Some questions on NetRexx open sourcing process

[David Requena]

Thanks for a great summary Kermit. I could never have expressed it better!

I would only add we're worried about being loosing valuable resources
previously available on the net and feel compelled to salvage what we can.

- I took over a piece of software which had been languishing for
years, unsupported by newer versions of its hosting environment. NetRexxJe
was a basic development tool for many (among the few) NetRexx developers.
- Recently Geocities closed doors. It was Kermit who prosecuted salvage of
'NetRexx at once' (thanks for this Kermit). That's a hell of a good gentle
intro for maybe-new-netrexx-programmers.
- AOL home town closed doors also. I've lost count of dead links now in my
bookmarks.
- NrxDoc latest version got lost long ago. Even for its own author!!

Just some sad examples...

PS: Kermit I've just become aware of how moot the RFE talk is right now.
According to proceedings of Ren?'s presentation at JVM Language Summit 2008
there already existed a NetRexx v. 3.00 back in September 2008 (which wasn't
out yet ;-) ). There is even an interesting 'recent additions section'!

In the meantime they could even have included planned dynamic features!
We'll have to wait till a release is made to see if any RFE's still apply.

2010/2/5 Kermit Kiser <[hidden email]>

> Chip & all ;
>
> As Jeff just pointed out in a separate message, when you go to the RexxLA
> web site, you see a lot of "Under Construction" messages and very little in
> the way of resources. I am not saying that RexxLA is a total failure at
> providing Rexx info, just that it does not look very successful yet. And
> when it comes to NetRexx info, there is virtually zero. Hopefully that will
> improve when NetRexx is finally open source. But that project is almost two
> years underway now (officially) and we still don't know when it will
> complete.
>
> Meanwhile, some of us are doing stuff with NetRexx right now. That includes
> a desire to see NetRexx succeed and grow to remain a viable option for our
> projects long term. When Ren? said we should start getting stuff ready, I
> admit I misinterpreted that to mean that RFE discussion was now fair game,
> but even with that on hold along with some of our projects like my scripting
> modules which are waiting on OSS NetRexx, there is still a lot we can do. It
> is great if RexxLA wants to provide the central portal for NetRexx and
> related resources, but that does not mean that we should put everything on
> hold indefinitely. On the contrary, we should be getting ready now as Ren?
> said. And since Rexx projects are mostly separate items on Sourceforge now,
> a consolidated repository would put NetRexx way ahead, IMHO.
>
> Thanks to RexxLA, IBM (Mike & Ren??) found a way to open source NetRexx.
> But most of RexxLA is probably not interested in NetRexx at this point, so
> WE are the NetRexx community right now.
>
> Just my opinion.
>
> -- Kermit
>
>
>
> Chip Davis wrote:
>
>> Thank you for that clarification, David.  I agree that it is crucial to a
>> productive discussion to establish exactly what we are talking about. :-)
>>
>> Let me first establish that I do not speak for the Open Source NetRexx
>> Project, nor in any official capacity for the RexxLA.  _I_ have not signed
>> any NDAs, and am privy to no more than anyone else on this discussion list.
>>  I speak from experience, and My Humble Opinions are exactly that.  But I
>> took the Rexx Red Pill a long time ago.
>>
>> If I may paraphrase, you have made a distinction between "NetRexx the
>> Language" and "NetRexx the Community".  IMHO, if NetRexx is to succeed, we
>> need to do everything in our power to erase that distinction and avoid
>> creating any mechanism that would tend to draw such a distinction.  That is
>> not only the "Rexx Way", that's the "FOSS Way".
>>
>> In my experience, a single point of contact (especially for nascent
>> efforts like FOSS NetRexx) is absolutely critical to wider visibility and
>> acceptance.  This was the primary raison d'?tre for the creation of RexxLA
>> in the first place.
>> There were many, many sites, vendors, and disassociated users of Rexx, on
>> many different platforms, who needed a single place to go for information,
>> help, news, etc.
>>
>> Sure, there was comp.lang.rexx but it had become obvious by the late
>> Eighties that, due to the "open" nature of newsgroups, its signal-to-noise
>> ratio had dropped almost to the point of unintelligibility.  Sadly, things
>> have not gotten significantly better in the intervening twenty years and I
>> don't expect them to, when the cost to participate is so low.  One sees the
>> exact same phenomenon when attending a symphony concert in the park, versus
>> the hall.  Same symphony, substantially different audience behavior,
>> considerably poorer musical experience.
>>
>> In short, I see no advantage, and many pitfalls in making a distinction
>> between "that which is RexxLA" and "that which is the NetRexx Community".
>>  RexxLA is merely the umbrella organization with a legal name and address
>> and incorporation papers that can be responsible for the resources that IBM
>> is willing to convey.
>>
>> As for technical issues regarding where certain files should be
>> maintained, or which server should host which discussion group, or other
>> such minor details, I see no reason why the model (if not the specifics)
>> used in the ooRexx Project would not be a good place to start.  (I will
>> admit that I'm not a big fan of SourceForge any more, but that itself is a
>> technical detail.)
>>
>> In closing, I'd like to address the subtle yet pervasive idea implicit in
>> your "ivory tower" references.  In the "Cathedral and Bazaar" model, the
>> NetRexx Project is most decidedly _not_ a cathedral, nor a tower (of
>> whatever material).
>>
>> NetRexx has always been product of a single, extraordinarily talented
>> Craftsman working at his shop in the bazaar, building tools for us peasants
>> and giving them away for free.  During the week, he is indeed a "Toolmaker
>> to the King" but on weekends he's in the bazaar working for free.  However,
>> the King always makes the rules and, while willing to release the rights to
>> one of the Craftsman's tools, the Royal Lawyers must be appeased.  Thus, and
>> very temporarily, the King's henchmen are guarding the Craftsman's door
>> while an inventory is taken. They will be gone very soon and the sun will
>> again stream into the Craftsman's shoppe.
>>
>> It is RexxLA's best interest, no to mention a legal requirement, that it
>> support NetRexx in the Open Source Way.
>>
>> -Chip-
>>
>> On 2/5/10 12:57 David Requena said:
>>
>>> Chip,
>>>
>>> Yours is more or less my own point. I'd like to express a couple slight
>>> precisions though.
>>>
>>> First, I believe where mixing 2 very distinct concepts here:
>>>
>>> - Open source NetRexx. This is RexxLA's concern and I think it will be
>>> obviously benefiting from "the REXX way".
>>> - Open source NetRexx user code/resources. This is the NetRexx language
>>> users community concern and is benefiting from "a bookmark file full of
>>> broken or obsolete links" right now.
>>>
>>> I for one have been talking about the second point the whole time. We, as
>>> the user base of the language, would greatly benefit from a consolidation of
>>> the current array of dispersed (and rapidly vanishing into oblivion) of
>>> netrexx resources on the net.
>>>
>>> Second, just because a given language's user community doesn't has a
>>> direct role in the language design, implementation, etc. doesn't mean it may
>>> have an important influence on that language's ultimate fate. Any effort
>>> made to help growing a strong, enthusiastic, user community is just another
>>> brick thrown at the foundation of the tower (pun intended ;-) ) we all are
>>> trying to build. That would be "the current times way"...
>>>
>>>
>>> 2010/2/5 Chip Davis <[hidden email] <mailto:[hidden email]>>
>>>
>>>    Tom,
>>>
>>>    I'm sure Ren? and those working diligently to effect the transfer of
>>>    NetRexx from IBM to RexxLA appreciate the enthusiasm that you,
>>>    David, Kermit, and others have demonstrated for this project, but I
>>>    think we all need to slow down and take a breath.
>>>
>>>    Our baby is not yet walking and you are already discussing which
>>>    motorcycle to buy for him. :-)
>>>
>>>    I think a couple of things should be self-evident: we have some
>>>    world-class intellectual resources within RexxLA (Ren? and Mike, for
>>>    starters), we have the experience of having done the same process
>>>    with ooRexx, and we have a passionate (rabid?) group of NetRexx
>>>    enthusiasts trying to wait for Christmas morning.
>>>
>>>    There are many fundamental issues that need to be addressed by the
>>>    NetRexx team and the RexxLA Board and Officers.  Believe me, we have
>>>    a lot more questions than you do, many of which are quite prosaic
>>>    and unglamorous, but crucial to the success of this project.  We
>>>    learned a lot from the ooRexx project and will incorporate those
>>>    lessons into the organization and charter of the NetRexx project.
>>>
>>>    At this point, fixed written-in-stone specifics are hard to come by,
>>>    so I hope you (and everyone else chomping at the bit) will take
>>>    comfort in the philosophy to which all of us are committed, that
>>>    being the fundamental tenets of The Rexx Language, as laid out by
>>>    Mike in TRL and TNRL.  Ren? said as much in his answer to David's
>>>    fourth question, when he referred to following The Rexx Way (my
>>>    caps) of Discuss, Agree, Document, then Implement.
>>>
>>>    Thus, if you haven't seen any Discussion of a particular issue, it
>>>    probably means we haven't gotten to that stage yet.
>>>
>>>    I would suggest that you not make any decisions about various code
>>>    or doc repositories until we figure out where and how the NetRexx
>>>    code will be housed.  It will be much less confusing for users to
>>>    have only one location to find NetRexx resources, than a bookmark
>>>    file full of broken or obsolete links.
>>>
>>>    IMHO, of course.
>>>
>>>    -Chip-
>>>
>>>    On 2/5/10 05:22 Thomas Schneider said:
>>>     >   to follow UP the idea of putting *any and all* Utilities
>>>    (including mine)
>>>     > to www.NetRexx.org <http://www.NetRexx.org>:
>>>     >
>>>     > What does this mean for my (private) directory structure ?
>>>     >
>>>     > When you like, I can send you my current one, but I am not so much
>>>     > satisfied
>>>     > with what I do have, and will need (please)
>>>     >
>>>     > your advise:
>>>     >
>>>     > How could/should we structure the whole project www.NetRexx.org
>>>    <http://www.NetRexx.org> ??
>>>     >
>>>     > Your advise will be welcome ...
>>>
>>>
>>>    On 2/5/10 03:48 Thomas Schneider said:
>>>
>>>        I am currently in the process to buildup the new
>>>        www.Rexx2Nrx.com <http://www.Rexx2Nrx.com> homepage.
>>>
>>>        Might it be that this would be a good instance to store all
>>>        those tools available in a central
>>>        place ?
>>>
>>>        Kermit Kiser schrieb:
>>>
>>>            I have seen a lot of RFEs floating around but no safe place
>>>            to store them all or discuss them. Could you do something
>>>            like that with your site?
>>>
>>>            David Requena wrote:
>>>
>>>                The thing is...  for the first time I got the feeling
>>>                this process is actually advancing in some direction.
>>>                We all, as NetRexx developers hoping to see the language
>>>                advance, are eagerly awaiting for the
>>>                culmination of a process about which we have near to no
>>>                information.
>>>
>>>
>>>
>>>
>>>    _______________________________________________
>>>    Ibm-netrexx mailing list
>>>    [hidden email] <mailto:[hidden email]>
>>>
>>>
>>>
>>>
>>> --
>>> Saludos / Regards,
>>> David Requena
>>>
>>>
>>> ------------------------------------------------------------------------
>>>
>>> _______________________________________________
>>> Ibm-netrexx mailing list
>>> [hidden email]
>>>
>>>
>>
>> _______________________________________________
>> Ibm-netrexx mailing list
>> [hidden email]
>>
>>
>>
>>  _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>

--
Saludos / Regards,
David Requena
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ns.hursley.ibm.com/pipermail/ibm-netrexx/attachments/20100205/ca6e4917/attachment-0001.html

[rvjansen]

Kermit,

let me first say that I fully agree with what Chip said. There should be a single point of contact and that point will be RexxLa - just because that is the best place where you can also find the best Rexx people.

I have, however, an elegant solution to the problem. www.netrexx.org is mine now, and points to one of my servers (that is not running right now in case you want to check, but that is beside the point that I am going to make).

There is nothing to stop us from organizing public knowledge about NetRexx and ~resources on this site - that eventually will be donated to RexxLA. Actually, it should have been done a long time ago. When I said that we should start doing things I meant it and I think you misinterpreted that a lot less than you thought when writing that email. The point that Chip is making about publicy and unmoderated mailing list is valid though, and I am not going to entertain or participate a lot in discussion how things should be. I will have a look at what I like about other websites of programming languages and learn from that.

There is a reason for the 'work in progress; ' signs on the RexxLA site, and that is because the people with the knowledge to do webdesign and maintenance are also working on implementations of Rexx, and humans need to sleep. No-one is paid for this; the fee structure does not allow this.

Please send me your stuff, links, ideas and things. We'll see how it goes. When it goes well, I'll switch it on. When the time comes, it will transition to RexxLA. I think this model will accomodate all previously mentioned points of view.

best regards,

Ren?.


On 5 feb 2010, at 22:45, Kermit Kiser wrote:

> Chip & all ;
>
> As Jeff just pointed out in a separate message, when you go to the RexxLA web site, you see a lot of "Under Construction" messages and very little in the way of resources. I am not saying that RexxLA is a total failure at providing Rexx info, just that it does not look very successful yet. And when it comes to NetRexx info, there is virtually zero. Hopefully that will improve when NetRexx is finally open source. But that project is almost two years underway now (officially) and we still don't know when it will complete.
>
> Meanwhile, some of us are doing stuff with NetRexx right now. That includes a desire to see NetRexx succeed and grow to remain a viable option for our projects long term. When Ren? said we should start getting stuff ready, I admit I misinterpreted that to mean that RFE discussion was now fair game, but even with that on hold along with some of our projects like my scripting modules which are waiting on OSS NetRexx, there is still a lot we can do. It is great if RexxLA wants to provide the central portal for NetRexx and related resources, but that does not mean that we should put everything on hold indefinitely. On the contrary, we should be getting ready now as Ren? said. And since Rexx projects are mostly separate items on Sourceforge now, a consolidated repository would put NetRexx way ahead, IMHO.
>
> Thanks to RexxLA, IBM (Mike & Ren??) found a way to open source NetRexx. But most of RexxLA is probably not interested in NetRexx at this point, so WE are the NetRexx community right now.
>
> Just my opinion.
>
> -- Kermit
>
>
> Chip Davis wrote:
>> Thank you for that clarification, David.  I agree that it is crucial to a productive discussion to establish exactly what we are talking about. :-)
>>
>> Let me first establish that I do not speak for the Open Source NetRexx Project, nor in any official capacity for the RexxLA.  _I_ have not signed any NDAs, and am privy to no more than anyone else on this discussion list.  I speak from experience, and My Humble Opinions are exactly that.  But I took the Rexx Red Pill a long time ago.
>>
>> If I may paraphrase, you have made a distinction between "NetRexx the Language" and "NetRexx the Community".  IMHO, if NetRexx is to succeed, we need to do everything in our power to erase that distinction and avoid creating any mechanism that would tend to draw such a distinction.  That is not only the "Rexx Way", that's the "FOSS Way".
>>
>> In my experience, a single point of contact (especially for nascent efforts like FOSS NetRexx) is absolutely critical to wider visibility and acceptance.  This was the primary raison d'?tre for the creation of RexxLA in the first place.
>> There were many, many sites, vendors, and disassociated users of Rexx, on many different platforms, who needed a single place to go for information, help, news, etc.
>>
>> Sure, there was comp.lang.rexx but it had become obvious by the late Eighties that, due to the "open" nature of newsgroups, its signal-to-noise ratio had dropped almost to the point of unintelligibility.  Sadly, things have not gotten significantly better in the intervening twenty years and I don't expect them to, when the cost to participate is so low.  One sees the exact same phenomenon when attending a symphony concert in the park, versus the hall.  Same symphony, substantially different audience behavior, considerably poorer musical experience.
>>
>> In short, I see no advantage, and many pitfalls in making a distinction between "that which is RexxLA" and "that which is the NetRexx Community".  RexxLA is merely the umbrella organization with a legal name and address and incorporation papers that can be responsible for the resources that IBM is willing to convey.
>>
>> As for technical issues regarding where certain files should be maintained, or which server should host which discussion group, or other such minor details, I see no reason why the model (if not the specifics) used in the ooRexx Project would not be a good place to start.  (I will admit that I'm not a big fan of SourceForge any more, but that itself is a technical detail.)
>>
>> In closing, I'd like to address the subtle yet pervasive idea implicit in your "ivory tower" references.  In the "Cathedral and Bazaar" model, the NetRexx Project is most decidedly _not_ a cathedral, nor a tower (of whatever material).
>>
>> NetRexx has always been product of a single, extraordinarily talented Craftsman working at his shop in the bazaar, building tools for us peasants and giving them away for free.  During the week, he is indeed a "Toolmaker to the King" but on weekends he's in the bazaar working for free.  However, the King always makes the rules and, while willing to release the rights to one of the Craftsman's tools, the Royal Lawyers must be appeased.  Thus, and very temporarily, the King's henchmen are guarding the Craftsman's door while an inventory is taken. They will be gone very soon and the sun will again stream into the Craftsman's shoppe.
>>
>> It is RexxLA's best interest, no to mention a legal requirement, that it support NetRexx in the Open Source Way.
>>
>> -Chip-
>>
>> On 2/5/10 12:57 David Requena said:
>>> Chip,
>>>
>>> Yours is more or less my own point. I'd like to express a couple slight precisions though.
>>>
>>> First, I believe where mixing 2 very distinct concepts here:
>>>
>>> - Open source NetRexx. This is RexxLA's concern and I think it will be obviously benefiting from "the REXX way".
>>> - Open source NetRexx user code/resources. This is the NetRexx language users community concern and is benefiting from "a bookmark file full of broken or obsolete links" right now.
>>>
>>> I for one have been talking about the second point the whole time. We, as the user base of the language, would greatly benefit from a consolidation of the current array of dispersed (and rapidly vanishing into oblivion) of netrexx resources on the net.
>>>
>>> Second, just because a given language's user community doesn't has a direct role in the language design, implementation, etc. doesn't mean it may have an important influence on that language's ultimate fate. Any effort made to help growing a strong, enthusiastic, user community is just another brick thrown at the foundation of the tower (pun intended ;-) ) we all are trying to build. That would be "the current times way"...
>>>
>>>
>>> 2010/2/5 Chip Davis <[hidden email] <mailto:[hidden email]>>
>>>
>>>    Tom,
>>>
>>>    I'm sure Ren? and those working diligently to effect the transfer of
>>>    NetRexx from IBM to RexxLA appreciate the enthusiasm that you,
>>>    David, Kermit, and others have demonstrated for this project, but I
>>>    think we all need to slow down and take a breath.
>>>
>>>    Our baby is not yet walking and you are already discussing which
>>>    motorcycle to buy for him. :-)
>>>
>>>    I think a couple of things should be self-evident: we have some
>>>    world-class intellectual resources within RexxLA (Ren? and Mike, for
>>>    starters), we have the experience of having done the same process
>>>    with ooRexx, and we have a passionate (rabid?) group of NetRexx
>>>    enthusiasts trying to wait for Christmas morning.
>>>
>>>    There are many fundamental issues that need to be addressed by the
>>>    NetRexx team and the RexxLA Board and Officers.  Believe me, we have
>>>    a lot more questions than you do, many of which are quite prosaic
>>>    and unglamorous, but crucial to the success of this project.  We
>>>    learned a lot from the ooRexx project and will incorporate those
>>>    lessons into the organization and charter of the NetRexx project.
>>>
>>>    At this point, fixed written-in-stone specifics are hard to come by,
>>>    so I hope you (and everyone else chomping at the bit) will take
>>>    comfort in the philosophy to which all of us are committed, that
>>>    being the fundamental tenets of The Rexx Language, as laid out by
>>>    Mike in TRL and TNRL.  Ren? said as much in his answer to David's
>>>    fourth question, when he referred to following The Rexx Way (my
>>>    caps) of Discuss, Agree, Document, then Implement.
>>>
>>>    Thus, if you haven't seen any Discussion of a particular issue, it
>>>    probably means we haven't gotten to that stage yet.
>>>
>>>    I would suggest that you not make any decisions about various code
>>>    or doc repositories until we figure out where and how the NetRexx
>>>    code will be housed.  It will be much less confusing for users to
>>>    have only one location to find NetRexx resources, than a bookmark
>>>    file full of broken or obsolete links.
>>>
>>>    IMHO, of course.
>>>
>>>    -Chip-
>>>
>>>    On 2/5/10 05:22 Thomas Schneider said:
>>>     >   to follow UP the idea of putting *any and all* Utilities
>>>    (including mine)
>>>     > to www.NetRexx.org <http://www.NetRexx.org>:
>>>     >
>>>     > What does this mean for my (private) directory structure ?
>>>     >
>>>     > When you like, I can send you my current one, but I am not so much
>>>     > satisfied
>>>     > with what I do have, and will need (please)
>>>     >
>>>     > your advise:
>>>     >
>>>     > How could/should we structure the whole project www.NetRexx.org
>>>    <http://www.NetRexx.org> ??
>>>     >
>>>     > Your advise will be welcome ...
>>>
>>>
>>>    On 2/5/10 03:48 Thomas Schneider said:
>>>
>>>        I am currently in the process to buildup the new
>>>        www.Rexx2Nrx.com <http://www.Rexx2Nrx.com> homepage.
>>>
>>>        Might it be that this would be a good instance to store all
>>>        those tools available in a central
>>>        place ?
>>>
>>>        Kermit Kiser schrieb:
>>>
>>>            I have seen a lot of RFEs floating around but no safe place
>>>            to store them all or discuss them. Could you do something
>>>            like that with your site?
>>>
>>>            David Requena wrote:
>>>
>>>                The thing is...  for the first time I got the feeling
>>>                this process is actually advancing in some direction.
>>>                We all, as NetRexx developers hoping to see the language
>>>                advance, are eagerly awaiting for the
>>>                culmination of a process about which we have near to no
>>>                information.
>>>
>>>
>>>
>>>
>>>    _______________________________________________
>>>    Ibm-netrexx mailing list
>>>    [hidden email] <mailto:[hidden email]>
>>>
>>>
>>>
>>>
>>> --
>>> Saludos / Regards,
>>> David Requena
>>>
>>>
>>> ------------------------------------------------------------------------
>>>
>>> _______________________________________________
>>> Ibm-netrexx mailing list
>>> [hidden email]
>>>
>>
>>
>> _______________________________________________
>> Ibm-netrexx mailing list
>> [hidden email]
>>
>>
>>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>

[rvjansen]

David,

when I spoke to Mike the last Rexx Language Symposium last year, we briefly discussed dynamic features, and he mentioned that in his opinion the interpreter api in the current public version would already allow for this. I tried it some time ago, and this is what I came up with:


options binary
import COM.ibm.netrexx.process.NetRexxA
/**
 * Class Dynamic implements...
 * <BR>
 * Created on: di, 16, jun 2009 22:37:53 +0200
 */
class Dynamic extends SourceTransformer  -- take that out later

  interpreter = NetRexxA()
  /**
   * Default constructor
   */
  method Dynamic()

  method interpret(file=String) returns Class
    files=[file]
    classname = Rexx file
    classname = classname.substr(1,classname.pos('.')-1)
    say classname
    flags=['nocrossref', 'verbose0']
    interpreter.parse(files, flags)
    return interpreter.getClassObject(null, classname) -- find the hello Class
   
  method main(args=String[]) static
    t = Dynamic()
    source = Arraylist()
    source.add(Rexx "options binary")
    source.add(Rexx "class TestDyn")
    source.add(Rexx "method main(args = String[]) static")
    source.add(Rexx "say 'Hello Dynamic NetRexx World'")

    writeFileFromArrayList('TestDyn.nrx', source)
   
    dynamicClass = t.interpret('TestDyn.nrx')

    say dynamicClass


note that I never took out SourceTransformer, but that class is only used to call writeFileFromArrayList - I am sure you can write a file yourself.

Just as an illustration that I went complaining about the static nature of NetRexx for years without understanding the wealth that is already in there. One could make an API out of that without the need for changing the core functionality.

best regards,

Ren?.











>
> PS: Kermit I've just become aware of how moot the RFE talk is right now. According to proceedings of Ren?'s presentation at JVM Language Summit 2008 there already existed a NetRexx v. 3.00 back in September 2008 (which wasn't out yet ;-) ). There is even an interesting 'recent additions section'!
>
> In the meantime they could even have included planned dynamic features! We'll have to wait till a release is made to see if any RFE's still apply.

[David Requena]

Chip,

I'd like to make clear that at no point in time I've criticized how the open
sourcing process is taking place or how RexxLA is going to manage the
language at a latter time. To be honest, until very recently I didn't
believe said process was going anywhere. At some point I got convinced this
is really going to happen eventually and kindly made some obvious questions
just out of curiosity. RexxLA is by express desire of its members a closed
community. That is a fact. It is so for the very same reasons you state
below. So I had to ask and hope for someone responding.

Of course I have my own opinions on these "non-criticized" topics above but
I'm not interested on discussing them here at all. May be one day at
RexxLA's list.. ;-)
What I really *do* be interested on is in NetRexx itself, in my ability to
go on indefinitely using it, and maybe one day beholding it going mainstream
allowing me to earn a living with it instead of java, c++ and c# (not to
mention lotusscript and vb...)

I think the whole debate over websites got a little bit out of hand. In my
view the open source netrexx part of the discussion just came up at nearly
the same time. I've been the whole time saying opensourced netrexx does not
yet exist. When it becomes a reality of course a single point of contact (or
at least a main central one) is the way to go. Hell, I even presented the
current ooREXX model as how I envision the whole thing at the end of the
road. But the fact still is: *** the time has not come yet ***

What we (at least me!) were discussing is: look, here we are; 3 individuals
with netrexx centric projects we are actually working on. We share a passion
(I lack a better word, really) for the language. We have our projects hosted
at 3 different sites. Thanks to this very same list we're already
collaborating ("hey would you opensource your parser so I can use it? - yes,
what would be a best license", "could you help me with that jEdit
idiosincracy?" - "dont' really know, I'll do my best"). At some point one
proposes more or less informally we could put our projects together. That
would benefit each others projects bringing more attention to them (note
we're already cross linking/mentioning). That would benefit potential users
too. Ok, we start talking about how such an endeavour would best be
approached.

See? This is community building up. Maybe it will catch and attract more
individuals until a certain critical mass is built up, maybe not and then
the trend will vanish. Nobody knows at this time. You can not discuss,
design, document, implement, then release a community the REXX way (or any
other way for that matter) it just happens when the right individuals
coincide at the right time at the right place.

Look at this very same list. I've been following it for years now. There has
been possibly more traffic in the last two weeks that in the whole preceding
two years. And this is good indeed. At least we're holding discussions
as gratifying as this very same one!

The distinction over language user community and language owner entity is
not one I made. It just exists right now, I now for sure I'm not part of the
latter at the moment. A (nearly dead) community of  netrexx users has
existed all the way long while the language itself was in limbo. An owner
has existed too. And the thing is, this is not bad in itself. We would have
gone on using netrexx as long as we could was it not to be open sourced.
I've been using gcc many years without ever having a look at its source
code.

Admittedly at the FOSS world this distinction is often moot, mainly for
languages that were born open source to start with. In other cases start
mingling to the point you cannot really tell which is what. That's what I
would like to see with netrexx in the end, when it's ripe as an OSS product.
Again the fact is: *** we're not still there ***

So let the community grow if that's what's gona happen. I community needs to
build around the language. Someone must be there, waiting, when OSS NetRexx
finally arrives. We are definitely a bit enthusiastic at the moment, I know.
We're relieved, we've been worried about our language
of choice disappearing, seeing our tools going unsupported for years. Now we
*do know* the language is gona be open sourced. Hurray! NetRexx is gona
live! We wont be depending on some 3 letter company shutting down the server
holding the downloads of NetRexxC and NetRexxD, nor the netrexx-list (our
only point of contact) being eventually withdrawn.

PS: From now on I'll be trying not to use the word 'community'
on further discussion, I'm already tired of it. Besides, hell, there is only
five indiciduals involved in the thread!!

PS2: I'll be surely involved in the discussion of technical details related
to the upcoming release of oss netrexx when the time has come. For the time
being I'll stick to the technical requirements the software I maintain
imposes on me.

PS3: No words I could possibly utter but convey praise enough for the work
of that craftsman under the heavy hand of the king. Noww. being under that
reign gave him the opportunity to raise general awareness of his own craft.
Life is made of these incongruences... Thanks Mike!!!

PS4: I'm sorry, no nice words for the king's lawyers. Maybe at a latter
time.


2010/2/5 Chip Davis <[hidden email]>

> Thank you for that clarification, David.  I agree that it is crucial to a
> productive discussion to establish exactly what we are talking about. :-)
>
> Let me first establish that I do not speak for the Open Source NetRexx
> Project, nor in any official capacity for the RexxLA.  _I_ have not signed
> any NDAs, and am privy to no more than anyone else on this discussion list.
>  I speak from experience, and My Humble Opinions are exactly that.  But I
> took the Rexx Red Pill a long time ago.
>
> If I may paraphrase, you have made a distinction between "NetRexx the
> Language" and "NetRexx the Community".  IMHO, if NetRexx is to succeed, we
> need to do everything in our power to erase that distinction and avoid
> creating any mechanism that would tend to draw such a distinction.  That is
> not only the "Rexx Way", that's the "FOSS Way".
>
> In my experience, a single point of contact (especially for nascent efforts
> like FOSS NetRexx) is absolutely critical to wider visibility and
> acceptance.  This was the primary raison d'?tre for the creation of RexxLA
> in the first place.
> There were many, many sites, vendors, and disassociated users of Rexx, on
> many different platforms, who needed a single place to go for information,
> help, news, etc.
>
> Sure, there was comp.lang.rexx but it had become obvious by the late
> Eighties that, due to the "open" nature of newsgroups, its signal-to-noise
> ratio had dropped almost to the point of unintelligibility.  Sadly, things
> have not gotten significantly better in the intervening twenty years and I
> don't expect them to, when the cost to participate is so low.  One sees the
> exact same phenomenon when attending a symphony concert in the park, versus
> the hall.  Same symphony, substantially different audience behavior,
> considerably poorer musical experience.
>
> In short, I see no advantage, and many pitfalls in making a distinction
> between "that which is RexxLA" and "that which is the NetRexx Community".
>  RexxLA is merely the umbrella organization with a legal name and address
> and incorporation papers that can be responsible for the resources that IBM
> is willing to convey.
>
> As for technical issues regarding where certain files should be maintained,
> or which server should host which discussion group, or other such minor
> details, I see no reason why the model (if not the specifics) used in the
> ooRexx Project would not be a good place to start.  (I will admit that I'm
> not a big fan of SourceForge any more, but that itself is a technical
> detail.)
>
> In closing, I'd like to address the subtle yet pervasive idea implicit in
> your "ivory tower" references.  In the "Cathedral and Bazaar" model, the
> NetRexx Project is most decidedly _not_ a cathedral, nor a tower (of
> whatever material).
>
> NetRexx has always been product of a single, extraordinarily talented
> Craftsman working at his shop in the bazaar, building tools for us peasants
> and giving them away for free.  During the week, he is indeed a "Toolmaker
> to the King" but on weekends he's in the bazaar working for free.  However,
> the King always makes the rules and, while willing to release the rights to
> one of the Craftsman's tools, the Royal Lawyers must be appeased.  Thus, and
> very temporarily, the King's henchmen are guarding the Craftsman's door
> while an inventory is taken. They will be gone very soon and the sun will
> again stream into the Craftsman's shoppe.
>
> It is RexxLA's best interest, no to mention a legal requirement, that it
> support NetRexx in the Open Source Way.
>
> -Chip-
>
>
> On 2/5/10 12:57 David Requena said:
>
>> Chip,
>>
>> Yours is more or less my own point. I'd like to express a couple slight
>> precisions though.
>>
>> First, I believe where mixing 2 very distinct concepts here:
>>
>> - Open source NetRexx. This is RexxLA's concern and I think it will be
>> obviously benefiting from "the REXX way".
>> - Open source NetRexx user code/resources. This is the NetRexx language
>> users community concern and is benefiting from "a bookmark file full of
>> broken or obsolete links" right now.
>>
>> I for one have been talking about the second point the whole time. We, as
>> the user base of the language, would greatly benefit from a consolidation of
>> the current array of dispersed (and rapidly vanishing into oblivion) of
>> netrexx resources on the net.
>>
>> Second, just because a given language's user community doesn't has a
>> direct role in the language design, implementation, etc. doesn't mean it may
>> have an important influence on that language's ultimate fate. Any effort
>> made to help growing a strong, enthusiastic, user community is just another
>> brick thrown at the foundation of the tower (pun intended ;-) ) we all are
>> trying to build. That would be "the current times way"...
>>
>>
>> 2010/2/5 Chip Davis <[hidden email] <mailto:[hidden email]>>
>>
>>
>>    Tom,
>>
>>    I'm sure Ren? and those working diligently to effect the transfer of
>>    NetRexx from IBM to RexxLA appreciate the enthusiasm that you,
>>    David, Kermit, and others have demonstrated for this project, but I
>>    think we all need to slow down and take a breath.
>>
>>    Our baby is not yet walking and you are already discussing which
>>    motorcycle to buy for him. :-)
>>
>>    I think a couple of things should be self-evident: we have some
>>    world-class intellectual resources within RexxLA (Ren? and Mike, for
>>    starters), we have the experience of having done the same process
>>    with ooRexx, and we have a passionate (rabid?) group of NetRexx
>>    enthusiasts trying to wait for Christmas morning.
>>
>>    There are many fundamental issues that need to be addressed by the
>>    NetRexx team and the RexxLA Board and Officers.  Believe me, we have
>>    a lot more questions than you do, many of which are quite prosaic
>>    and unglamorous, but crucial to the success of this project.  We
>>    learned a lot from the ooRexx project and will incorporate those
>>    lessons into the organization and charter of the NetRexx project.
>>
>>    At this point, fixed written-in-stone specifics are hard to come by,
>>    so I hope you (and everyone else chomping at the bit) will take
>>    comfort in the philosophy to which all of us are committed, that
>>    being the fundamental tenets of The Rexx Language, as laid out by
>>    Mike in TRL and TNRL.  Ren? said as much in his answer to David's
>>    fourth question, when he referred to following The Rexx Way (my
>>    caps) of Discuss, Agree, Document, then Implement.
>>
>>    Thus, if you haven't seen any Discussion of a particular issue, it
>>    probably means we haven't gotten to that stage yet.
>>
>>    I would suggest that you not make any decisions about various code
>>    or doc repositories until we figure out where and how the NetRexx
>>    code will be housed.  It will be much less confusing for users to
>>    have only one location to find NetRexx resources, than a bookmark
>>    file full of broken or obsolete links.
>>
>>    IMHO, of course.
>>
>>    -Chip-
>>
>>    On 2/5/10 05:22 Thomas Schneider said:
>>     >   to follow UP the idea of putting *any and all* Utilities
>>    (including mine)
>>     > to www.NetRexx.org <http://www.NetRexx.org>:
>>
>>     >
>>     > What does this mean for my (private) directory structure ?
>>     >
>>     > When you like, I can send you my current one, but I am not so much
>>     > satisfied
>>     > with what I do have, and will need (please)
>>     >
>>     > your advise:
>>     >
>>     > How could/should we structure the whole project www.NetRexx.org
>>    <http://www.NetRexx.org> ??
>>
>>     >
>>     > Your advise will be welcome ...
>>
>>
>>    On 2/5/10 03:48 Thomas Schneider said:
>>
>>        I am currently in the process to buildup the new
>>        www.Rexx2Nrx.com <http://www.Rexx2Nrx.com> homepage.
>>
>>
>>        Might it be that this would be a good instance to store all
>>        those tools available in a central
>>        place ?
>>
>>        Kermit Kiser schrieb:
>>
>>            I have seen a lot of RFEs floating around but no safe place
>>            to store them all or discuss them. Could you do something
>>            like that with your site?
>>
>>            David Requena wrote:
>>
>>                The thing is...  for the first time I got the feeling
>>                this process is actually advancing in some direction.
>>                We all, as NetRexx developers hoping to see the language
>>                advance, are eagerly awaiting for the
>>                culmination of a process about which we have near to no
>>                information.
>>
>>
>>
>>
>>    _______________________________________________
>>    Ibm-netrexx mailing list
>>    [hidden email] <mailto:[hidden email]>
>>
>>
>>
>>
>> --
>> Saludos / Regards,
>> David Requena
>>
>>
>> ------------------------------------------------------------------------
>>
>>
>> _______________________________________________
>> Ibm-netrexx mailing list
>> [hidden email]
>>
>>
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>

--
Saludos / Regards,
David Requena
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ns.hursley.ibm.com/pipermail/ibm-netrexx/attachments/20100206/13129ea0/attachment-0001.html

[David Requena]

Nice trick Ren?!

I've not much experience with the interpreter but IIRC from de docs wouldn't
parsing the new class reinitialize the interpreter?
Would this code work would it be run from the interpreter itself? hmmm..
well.. that would be a separate interpreter instance...

but you couldn't re utilize a NetRexA instance without invalidating
references to a previously parsed class, right?
hmm.. feels a little bit clunky..

I've always wondered why Mike chosed to implement a filename based API for
the interpreter. It would have been so much more convenient should he
had chosen iostreams... we wouldn't have to go through the hops of temporary
files!

2010/2/6 Ren? Jansen <[hidden email]>

> David,
>
> when I spoke to Mike the last Rexx Language Symposium last year, we briefly
> discussed dynamic features, and he mentioned that in his opinion the
> interpreter api in the current public version would already allow for this.
> I tried it some time ago, and this is what I came up with:
>
>
> options binary
> import COM.ibm.netrexx.process.NetRexxA
> /**
>  * Class Dynamic implements...
>  * <BR>
>  * Created on: di, 16, jun 2009 22:37:53 +0200
>  */
> class Dynamic extends SourceTransformer  -- take that out later
>
>  interpreter = NetRexxA()
>  /**
>   * Default constructor
>   */
>  method Dynamic()
>
>  method interpret(file=String) returns Class
>    files=[file]
>    classname = Rexx file
>    classname = classname.substr(1,classname.pos('.')-1)
>    say classname
>    flags=['nocrossref', 'verbose0']
>    interpreter.parse(files, flags)
>    return interpreter.getClassObject(null, classname) -- find the hello
> Class
>
>  method main(args=String[]) static
>    t = Dynamic()
>    source = Arraylist()
>    source.add(Rexx "options binary")
>    source.add(Rexx "class TestDyn")
>    source.add(Rexx "method main(args = String[]) static")
>    source.add(Rexx "say 'Hello Dynamic NetRexx World'")
>
>    writeFileFromArrayList('TestDyn.nrx', source)
>
>    dynamicClass = t.interpret('TestDyn.nrx')
>
>    say dynamicClass
>
>
> note that I never took out SourceTransformer, but that class is only used
> to call writeFileFromArrayList - I am sure you can write a file yourself.
>
> Just as an illustration that I went complaining about the static nature of
> NetRexx for years without understanding the wealth that is already in there.
> One could make an API out of that without the need for changing the core
> functionality.
>
> best regards,
>
> Ren?.
>
>
>
>
>
>
>
>
>
>
>
> >
> > PS: Kermit I've just become aware of how moot the RFE talk is right now.
> According to proceedings of Ren?'s presentation at JVM Language Summit 2008
> there already existed a NetRexx v. 3.00 back in September 2008 (which wasn't
> out yet ;-) ). There is even an interesting 'recent additions section'!
> >
> > In the meantime they could even have included planned dynamic features!
> We'll have to wait till a release is made to see if any RFE's still apply.
>
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>

--
Saludos / Regards,
David Requena
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ns.hursley.ibm.com/pipermail/ibm-netrexx/attachments/20100206/ab4aa297/attachment.html

[David Requena]

Ren?,

The problem is we need some sourceforge-type infrastructure to support our
projects. At this time I think it does not make much sense to architect this
at netrexx.org just to host our little non-core netrexx projects.

That said, If you'd want to start the ball rolling with regard to the
upcoming release of open sourced NetRexx, I'll be of course contributing
with as much content, links, code, downloads, or whatever needed as I can.

Regard,
David

2010/2/6 Ren? Jansen <[hidden email]>

> Kermit,
>
> let me first say that I fully agree with what Chip said. There should be a
> single point of contact and that point will be RexxLa - just because that is
> the best place where you can also find the best Rexx people.
>
> I have, however, an elegant solution to the problem. www.netrexx.org is
> mine now, and points to one of my servers (that is not running right now in
> case you want to check, but that is beside the point that I am going to
> make).
>
> There is nothing to stop us from organizing public knowledge about NetRexx
> and ~resources on this site - that eventually will be donated to RexxLA.
> Actually, it should have been done a long time ago. When I said that we
> should start doing things I meant it and I think you misinterpreted that a
> lot less than you thought when writing that email. The point that Chip is
> making about publicy and unmoderated mailing list is valid though, and I am
> not going to entertain or participate a lot in discussion how things should
> be. I will have a look at what I like about other websites of programming
> languages and learn from that.
>
> There is a reason for the 'work in progress; ' signs on the RexxLA site,
> and that is because the people with the knowledge to do webdesign and
> maintenance are also working on implementations of Rexx, and humans need to
> sleep. No-one is paid for this; the fee structure does not allow this.
>
> Please send me your stuff, links, ideas and things. We'll see how it goes.
> When it goes well, I'll switch it on. When the time comes, it will
> transition to RexxLA. I think this model will accomodate all previously
> mentioned points of view.
>
> best regards,
>
> Ren?.
>
>
> On 5 feb 2010, at 22:45, Kermit Kiser wrote:
>
> > Chip & all ;
> >
> > As Jeff just pointed out in a separate message, when you go to the RexxLA
> web site, you see a lot of "Under Construction" messages and very little in
> the way of resources. I am not saying that RexxLA is a total failure at
> providing Rexx info, just that it does not look very successful yet. And
> when it comes to NetRexx info, there is virtually zero. Hopefully that will
> improve when NetRexx is finally open source. But that project is almost two
> years underway now (officially) and we still don't know when it will
> complete.
> >
> > Meanwhile, some of us are doing stuff with NetRexx right now. That
> includes a desire to see NetRexx succeed and grow to remain a viable option
> for our projects long term. When Ren? said we should start getting stuff
> ready, I admit I misinterpreted that to mean that RFE discussion was now
> fair game, but even with that on hold along with some of our projects like
> my scripting modules which are waiting on OSS NetRexx, there is still a lot
> we can do. It is great if RexxLA wants to provide the central portal for
> NetRexx and related resources, but that does not mean that we should put
> everything on hold indefinitely. On the contrary, we should be getting ready
> now as Ren? said. And since Rexx projects are mostly separate items on
> Sourceforge now, a consolidated repository would put NetRexx way ahead,
> IMHO.
> >
> > Thanks to RexxLA, IBM (Mike & Ren??) found a way to open source NetRexx.
> But most of RexxLA is probably not interested in NetRexx at this point, so
> WE are the NetRexx community right now.
> >
> > Just my opinion.
> >
> > -- Kermit
> >
> >
> > Chip Davis wrote:
> >> Thank you for that clarification, David.  I agree that it is crucial to
> a productive discussion to establish exactly what we are talking about. :-)
> >>
> >> Let me first establish that I do not speak for the Open Source NetRexx
> Project, nor in any official capacity for the RexxLA.  _I_ have not signed
> any NDAs, and am privy to no more than anyone else on this discussion list.
>  I speak from experience, and My Humble Opinions are exactly that.  But I
> took the Rexx Red Pill a long time ago.
> >>
> >> If I may paraphrase, you have made a distinction between "NetRexx the
> Language" and "NetRexx the Community".  IMHO, if NetRexx is to succeed, we
> need to do everything in our power to erase that distinction and avoid
> creating any mechanism that would tend to draw such a distinction.  That is
> not only the "Rexx Way", that's the "FOSS Way".
> >>
> >> In my experience, a single point of contact (especially for nascent
> efforts like FOSS NetRexx) is absolutely critical to wider visibility and
> acceptance.  This was the primary raison d'?tre for the creation of RexxLA
> in the first place.
> >> There were many, many sites, vendors, and disassociated users of Rexx,
> on many different platforms, who needed a single place to go for
> information, help, news, etc.
> >>
> >> Sure, there was comp.lang.rexx but it had become obvious by the late
> Eighties that, due to the "open" nature of newsgroups, its signal-to-noise
> ratio had dropped almost to the point of unintelligibility.  Sadly, things
> have not gotten significantly better in the intervening twenty years and I
> don't expect them to, when the cost to participate is so low.  One sees the
> exact same phenomenon when attending a symphony concert in the park, versus
> the hall.  Same symphony, substantially different audience behavior,
> considerably poorer musical experience.
> >>
> >> In short, I see no advantage, and many pitfalls in making a distinction
> between "that which is RexxLA" and "that which is the NetRexx Community".
>  RexxLA is merely the umbrella organization with a legal name and address
> and incorporation papers that can be responsible for the resources that IBM
> is willing to convey.
> >>
> >> As for technical issues regarding where certain files should be
> maintained, or which server should host which discussion group, or other
> such minor details, I see no reason why the model (if not the specifics)
> used in the ooRexx Project would not be a good place to start.  (I will
> admit that I'm not a big fan of SourceForge any more, but that itself is a
> technical detail.)
> >>
> >> In closing, I'd like to address the subtle yet pervasive idea implicit
> in your "ivory tower" references.  In the "Cathedral and Bazaar" model, the
> NetRexx Project is most decidedly _not_ a cathedral, nor a tower (of
> whatever material).
> >>
> >> NetRexx has always been product of a single, extraordinarily talented
> Craftsman working at his shop in the bazaar, building tools for us peasants
> and giving them away for free.  During the week, he is indeed a "Toolmaker
> to the King" but on weekends he's in the bazaar working for free.  However,
> the King always makes the rules and, while willing to release the rights to
> one of the Craftsman's tools, the Royal Lawyers must be appeased.  Thus, and
> very temporarily, the King's henchmen are guarding the Craftsman's door
> while an inventory is taken. They will be gone very soon and the sun will
> again stream into the Craftsman's shoppe.
> >>
> >> It is RexxLA's best interest, no to mention a legal requirement, that it
> support NetRexx in the Open Source Way.
> >>
> >> -Chip-
> >>
> >> On 2/5/10 12:57 David Requena said:
> >>> Chip,
> >>>
> >>> Yours is more or less my own point. I'd like to express a couple slight
> precisions though.
> >>>
> >>> First, I believe where mixing 2 very distinct concepts here:
> >>>
> >>> - Open source NetRexx. This is RexxLA's concern and I think it will be
> obviously benefiting from "the REXX way".
> >>> - Open source NetRexx user code/resources. This is the NetRexx language
> users community concern and is benefiting from "a bookmark file full of
> broken or obsolete links" right now.
> >>>
> >>> I for one have been talking about the second point the whole time. We,
> as the user base of the language, would greatly benefit from a consolidation
> of the current array of dispersed (and rapidly vanishing into oblivion) of
> netrexx resources on the net.
> >>>
> >>> Second, just because a given language's user community doesn't has a
> direct role in the language design, implementation, etc. doesn't mean it may
> have an important influence on that language's ultimate fate. Any effort
> made to help growing a strong, enthusiastic, user community is just another
> brick thrown at the foundation of the tower (pun intended ;-) ) we all are
> trying to build. That would be "the current times way"...
> >>>
> >>>
> >>> 2010/2/5 Chip Davis <[hidden email] <mailto:[hidden email]>>
> >>>
> >>>    Tom,
> >>>
> >>>    I'm sure Ren? and those working diligently to effect the transfer of
> >>>    NetRexx from IBM to RexxLA appreciate the enthusiasm that you,
> >>>    David, Kermit, and others have demonstrated for this project, but I
> >>>    think we all need to slow down and take a breath.
> >>>
> >>>    Our baby is not yet walking and you are already discussing which
> >>>    motorcycle to buy for him. :-)
> >>>
> >>>    I think a couple of things should be self-evident: we have some
> >>>    world-class intellectual resources within RexxLA (Ren? and Mike, for
> >>>    starters), we have the experience of having done the same process
> >>>    with ooRexx, and we have a passionate (rabid?) group of NetRexx
> >>>    enthusiasts trying to wait for Christmas morning.
> >>>
> >>>    There are many fundamental issues that need to be addressed by the
> >>>    NetRexx team and the RexxLA Board and Officers.  Believe me, we have
> >>>    a lot more questions than you do, many of which are quite prosaic
> >>>    and unglamorous, but crucial to the success of this project.  We
> >>>    learned a lot from the ooRexx project and will incorporate those
> >>>    lessons into the organization and charter of the NetRexx project.
> >>>
> >>>    At this point, fixed written-in-stone specifics are hard to come by,
> >>>    so I hope you (and everyone else chomping at the bit) will take
> >>>    comfort in the philosophy to which all of us are committed, that
> >>>    being the fundamental tenets of The Rexx Language, as laid out by
> >>>    Mike in TRL and TNRL.  Ren? said as much in his answer to David's
> >>>    fourth question, when he referred to following The Rexx Way (my
> >>>    caps) of Discuss, Agree, Document, then Implement.
> >>>
> >>>    Thus, if you haven't seen any Discussion of a particular issue, it
> >>>    probably means we haven't gotten to that stage yet.
> >>>
> >>>    I would suggest that you not make any decisions about various code
> >>>    or doc repositories until we figure out where and how the NetRexx
> >>>    code will be housed.  It will be much less confusing for users to
> >>>    have only one location to find NetRexx resources, than a bookmark
> >>>    file full of broken or obsolete links.
> >>>
> >>>    IMHO, of course.
> >>>
> >>>    -Chip-
> >>>
> >>>    On 2/5/10 05:22 Thomas Schneider said:
> >>>     >   to follow UP the idea of putting *any and all* Utilities
> >>>    (including mine)
> >>>     > to www.NetRexx.org <http://www.NetRexx.org>:
> >>>     >
> >>>     > What does this mean for my (private) directory structure ?
> >>>     >
> >>>     > When you like, I can send you my current one, but I am not so
> much
> >>>     > satisfied
> >>>     > with what I do have, and will need (please)
> >>>     >
> >>>     > your advise:
> >>>     >
> >>>     > How could/should we structure the whole project www.NetRexx.org
> >>>    <http://www.NetRexx.org> ??
> >>>     >
> >>>     > Your advise will be welcome ...
> >>>
> >>>
> >>>    On 2/5/10 03:48 Thomas Schneider said:
> >>>
> >>>        I am currently in the process to buildup the new
> >>>        www.Rexx2Nrx.com <http://www.Rexx2Nrx.com> homepage.
> >>>
> >>>        Might it be that this would be a good instance to store all
> >>>        those tools available in a central
> >>>        place ?
> >>>
> >>>        Kermit Kiser schrieb:
> >>>
> >>>            I have seen a lot of RFEs floating around but no safe place
> >>>            to store them all or discuss them. Could you do something
> >>>            like that with your site?
> >>>
> >>>            David Requena wrote:
> >>>
> >>>                The thing is...  for the first time I got the feeling
> >>>                this process is actually advancing in some direction.
> >>>                We all, as NetRexx developers hoping to see the language
> >>>                advance, are eagerly awaiting for the
> >>>                culmination of a process about which we have near to no
> >>>                information.
> >>>
> >>>
> >>>
> >>>
> >>>    _______________________________________________
> >>>    Ibm-netrexx mailing list
> >>>    [hidden email] <mailto:[hidden email]>
> >>>
> >>>
> >>>
> >>>
> >>> --
> >>> Saludos / Regards,
> >>> David Requena
> >>>
> >>>
> >>>
> ------------------------------------------------------------------------
> >>>
> >>> _______________________________________________
> >>> Ibm-netrexx mailing list
> >>> [hidden email]
> >>>
> >>
> >>
> >> _______________________________________________
> >> Ibm-netrexx mailing list
> >> [hidden email]
> >>
> >>
> >>
> > _______________________________________________
> > Ibm-netrexx mailing list
> > [hidden email]
> >
>
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>

--
Saludos / Regards,
David Requena
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ns.hursley.ibm.com/pipermail/ibm-netrexx/attachments/20100206/314f53e9/attachment-0001.html

[David Requena]

> That said, If you'd want to start the ball rolling with regard to the
> upcoming release of open sourced NetRexx, I'll be of course contributing
> with as much content, links, code, downloads, or whatever needed as I can.
>

Of course intended meaning was "contribute to www.netrexx.org"!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ns.hursley.ibm.com/pipermail/ibm-netrexx/attachments/20100206/6ddd7b7f/attachment.html

[Kermit Kiser]

Maybe I am a little dense but what do you mean by "dynamic features"?
This code is basically what I use in my scripting modules, NetRexxScript
and NetRexxJSR223 although it is missing a couple of steps that are
needed to locate the main method and invoke it. (More if you need to
isolate the dynamic code in a separate thread. I also transparently add
suitable environment variables while writing the code to the disk file.)
Is the objective to dynamically generate NetRexx code and execute it
from within a compiled program? That is basically what JSR223 does.

If the objective is to dynamically create Java byte code and add it to a
library of compiled classes in memory, then you would need to use the
NetRexx translator to produce Java code and pass it to the JSR199
compiler API to generate byte code. (I have not used JSR199 so I don't
know if you need a custom classloader also to do this, but having done a
classloader in NetRexx, I can verify that it is possible.) Am I even
close to what you guys are talking about?

-- Kermit


David Requena wrote:

> Nice trick Ren?!
>
> I've not much experience with the interpreter but IIRC from de docs
> wouldn't parsing the new class reinitialize the interpreter?
> Would this code work would it be run from the interpreter itself?
> hmmm.. well.. that would be a separate interpreter instance...
>
> but you couldn't re utilize a NetRexA instance without invalidating
> references to a previously parsed class, right?
> hmm.. feels a little bit clunky..
>
> I've always wondered why Mike chosed to implement a filename based API
> for the interpreter. It would have been so much more convenient should
> he had chosen iostreams... we wouldn't have to go through the hops of
> temporary files!
>
> 2010/2/6 Ren? Jansen <[hidden email] <mailto:[hidden email]>>
>
>     David,
>
>     when I spoke to Mike the last Rexx Language Symposium last year,
>     we briefly discussed dynamic features, and he mentioned that in
>     his opinion the interpreter api in the current public version
>     would already allow for this. I tried it some time ago, and this
>     is what I came up with:
>
>
>     options binary
>     import COM.ibm.netrexx.process.NetRexxA
>     /**
>      * Class Dynamic implements...
>      * <BR>
>      * Created on: di, 16, jun 2009 22:37:53 +0200
>      */
>     class Dynamic extends SourceTransformer  -- take that out later
>
>      interpreter = NetRexxA()
>      /**
>       * Default constructor
>       */
>      method Dynamic()
>
>      method interpret(file=String) returns Class
>        files=[file]
>        classname = Rexx file
>        classname = classname.substr(1,classname.pos('.')-1)
>        say classname
>        flags=['nocrossref', 'verbose0']
>        interpreter.parse(files, flags)
>        return interpreter.getClassObject(null, classname) -- find the
>     hello Class
>
>      method main(args=String[]) static
>        t = Dynamic()
>        source = Arraylist()
>        source.add(Rexx "options binary")
>        source.add(Rexx "class TestDyn")
>        source.add(Rexx "method main(args = String[]) static")
>        source.add(Rexx "say 'Hello Dynamic NetRexx World'")
>
>        writeFileFromArrayList('TestDyn.nrx', source)
>
>        dynamicClass = t.interpret('TestDyn.nrx')
>
>        say dynamicClass
>
>
>     note that I never took out SourceTransformer, but that class is
>     only used to call writeFileFromArrayList - I am sure you can write
>     a file yourself.
>
>     Just as an illustration that I went complaining about the static
>     nature of NetRexx for years without understanding the wealth that
>     is already in there. One could make an API out of that without the
>     need for changing the core functionality.
>
>     best regards,
>
>     Ren?.
>
>
>
>
>
>
>
>
>
>
>
>     >
>     > PS: Kermit I've just become aware of how moot the RFE talk is
>     right now. According to proceedings of Ren?'s presentation at JVM
>     Language Summit 2008 there already existed a NetRexx v. 3.00 back
>     in September 2008 (which wasn't out yet ;-) ). There is even an
>     interesting 'recent additions section'!
>     >
>     > In the meantime they could even have included planned dynamic
>     features! We'll have to wait till a release is made to see if any
>     RFE's still apply.
>
>
>     _______________________________________________
>     Ibm-netrexx mailing list
>     [hidden email] <mailto:[hidden email]>
>
>
>
>
> --
> Saludos / Regards,
> David Requena
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>  

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ns.hursley.ibm.com/pipermail/ibm-netrexx/attachments/20100205/6e23852f/attachment.html

[Aviatrexx]

On 2/5/10 21:03 Peter Rupp said:
> I have no experience moving products into the open-source world, so I
> was wondering how much assistance from Mike Cowlishaw will occur, or
> will he still be involved in the process and/or post-move involvement?

Well, there may be legal/administrative impediments to Mike's formal involvement
in the transfer (I don't know; see my previously noted lack of an NDA) but I
know that Ren? has Mike's email address... ;-)

> I'm assuming whoever makes modifications or enhancements to the language
> would require extensive knowledge of language processors, and be
> intimately familiar with Mike's Rexx philosophies?    That might be a
> tough act to follow :-).

Indeed.  I know of only one person that could follow that act, so we are hoping
that Mike himself will be able to carve out some time to mentor the Open NetRexx
project.  OTOH, at this point I think it is safe to say that, other than Mike,
the person who knows the most about the NetRexx internals is Ren?.  So I think
the Project is in the best possible hands.

But, as we've seen by this and other threads, the technical details are often
less of a problem than the organizational and interpersonal issues of creating a
community.

-Chip-

[Mike Cowlishaw-2]

Peter,

> I have no experience moving products into the open-source world, so I
> was wondering how much assistance from Mike Cowlishaw will occur, or
> will he still be involved in the process and/or post-move involvement?
> I'm assuming whoever makes modifications or enhancements to the language
> would require extensive knowledge of language processors, and be
> intimately familiar with Mike's Rexx philosophies?    That might be a
> tough act to follow :-).

I'll be happy to answer 'philosophy' questions, and 'why did he do that'
questions about the language.  I won't be able to contribute to the code
until after my retirement from IBM later this year, and even then I don't
really expect to do much with NetRexx (at least in the short term) because
I don't use Java or NetRexx for anything at the moment.

The code is actually fairly straightforward (and not as OO as it might be,
because it started off in plain Rexx).  It was evolving towards being able
to emit byte codes directly when I stopped working on it .. I think some
of the hooks are there already, but I forget.   In all, it should not need
an extensive knowledge of language processors.

For an overview of the internals, see:
http://speleotrove.com/misc/NetRexxI.pdf
And for the philosophies: http://speleotrove.com/misc/NetRexx2.pdf

It's good to see the activity on this list!

Mike






Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number
741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU





-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ns.hursley.ibm.com/pipermail/ibm-netrexx/attachments/20100206/32ce3780/attachment.html

[rvjansen]

Hi Kermit,

yes, code to locate the methods and invoke them is left as an excercise to the reader. The point here is that the translator already generates bytecode and has a classloader loading it into the running JVM; this without JSR199 or JSR223 and years before that. The translator stops short of writing bytecode to class files; also, the classes are not 'opened up' for us to be able to add methods and properties to their instances -afaik. JVM technology is not standing still however; the challenge obviously being, to do what JRuby can do but without the perfomance penalty. Not to say that JRuby is not moving in this area though.

I am looking forward to future discussions we will have in this area; and even if Mike C. will assist 'from a distance' this will be enormously beneficial to us.

best regards,

Ren? Jansen.

On 6 feb 2010, at 04:47, Kermit Kiser wrote:

> Maybe I am a little dense but what do you mean by "dynamic features"? This code is basically what I use in my scripting modules, NetRexxScript and NetRexxJSR223 although it is missing a couple of steps that are needed to locate the main method and invoke it. (More if you need to isolate the dynamic code in a separate thread. I also transparently add suitable environment variables while writing the code to the disk file.) Is the objective to dynamically generate NetRexx code and execute it from within a compiled program? That is basically what JSR223 does.
>
> If the objective is to dynamically create Java byte code and add it to a library of compiled classes in memory, then you would need to use the NetRexx translator to produce Java code and pass it to the JSR199 compiler API to generate byte code. (I have not used JSR199 so I don't know if you need a custom classloader also to do this, but having done a classloader in NetRexx, I can verify that it is possible.) Am I even close to what you guys are talking about?
>
> -- Kermit
>
>
> David Requena wrote:
>>
>> Nice trick Ren?!
>>
>> I've not much experience with the interpreter but IIRC from de docs wouldn't parsing the new class reinitialize the interpreter?
>> Would this code work would it be run from the interpreter itself? hmmm.. well.. that would be a separate interpreter instance...
>>
>> but you couldn't re utilize a NetRexA instance without invalidating references to a previously parsed class, right?
>> hmm.. feels a little bit clunky..
>>
>> I've always wondered why Mike chosed to implement a filename based API for the interpreter. It would have been so much more convenient should he had chosen iostreams... we wouldn't have to go through the hops of temporary files!
>>
>> 2010/2/6 Ren? Jansen <[hidden email]>
>> David,
>>
>> when I spoke to Mike the last Rexx Language Symposium last year, we briefly discussed dynamic features, and he mentioned that in his opinion the interpreter api in the current public version would already allow for this. I tried it some time ago, and this is what I came up with:
>>
>>
>> options binary
>> import COM.ibm.netrexx.process.NetRexxA
>> /**
>>  * Class Dynamic implements...
>>  * <BR>
>>  * Created on: di, 16, jun 2009 22:37:53 +0200
>>  */
>> class Dynamic extends SourceTransformer  -- take that out later
>>
>>  interpreter = NetRexxA()
>>  /**
>>   * Default constructor
>>   */
>>  method Dynamic()
>>
>>  method interpret(file=String) returns Class
>>    files=[file]
>>    classname = Rexx file
>>    classname = classname.substr(1,classname.pos('.')-1)
>>    say classname
>>    flags=['nocrossref', 'verbose0']
>>    interpreter.parse(files, flags)
>>    return interpreter.getClassObject(null, classname) -- find the hello Class
>>
>>  method main(args=String[]) static
>>    t = Dynamic()
>>    source = Arraylist()
>>    source.add(Rexx "options binary")
>>    source.add(Rexx "class TestDyn")
>>    source.add(Rexx "method main(args = String[]) static")
>>    source.add(Rexx "say 'Hello Dynamic NetRexx World'")
>>
>>    writeFileFromArrayList('TestDyn.nrx', source)
>>
>>    dynamicClass = t.interpret('TestDyn.nrx')
>>
>>    say dynamicClass
>>
>>
>> note that I never took out SourceTransformer, but that class is only used to call writeFileFromArrayList - I am sure you can write a file yourself.
>>
>> Just as an illustration that I went complaining about the static nature of NetRexx for years without understanding the wealth that is already in there. One could make an API out of that without the need for changing the core functionality.
>>
>> best regards,
>>
>> Ren?.
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> >
>> > PS: Kermit I've just become aware of how moot the RFE talk is right now. According to proceedings of Ren?'s presentation at JVM Language Summit 2008 there already existed a NetRexx v. 3.00 back in September 2008 (which wasn't out yet ;-) ). There is even an interesting 'recent additions section'!
>> >
>> > In the meantime they could even have included planned dynamic features! We'll have to wait till a release is made to see if any RFE's still apply.
>>
>>
>> _______________________________________________
>> Ibm-netrexx mailing list
>> [hidden email]
>>
>>
>>
>>
>> --
>> Saludos / Regards,
>> David Requena
>>
>>
>> _______________________________________________
>> Ibm-netrexx mailing list
>> [hidden email]
>>
>>  
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ns.hursley.ibm.com/pipermail/ibm-netrexx/attachments/20100206/90264976/attachment.html

[Aviatrexx]

Thank you David, for your thoughtful reply.  It is the nature of this medium
that a minor misunderstanding that would take seconds to clear up in a
face-to-face conversation, will take many KB to reconcile over the net.

I confess that the impression I got from your (and Kermit and Tom)'s enthusiasm
to set up a different Internet site for your products, was that you would then
be free to enhance NetRexx itself (with your existing list of RFEs) and offer an
alternate version to the world.  In large part, this impression was a result of
Tom's suggestion that he create a private discussion list for the four of us.  I
saw this as the imminent splintering of a not-yet-coalesced NetRexx community.

This is a sensitive issue for me personally, having been a member of the ANSI
Standard Committee for Rexx back in the Nineties, only to see Jeff Glatt pollute
the Rexx interpreter marketplace with his wildly-idiosyncratic and blatantly
non-standard Reginald interpreter.  And I find it illustrative that he did so
via comp.lang.rexx, not under the auspices of RexxLA (under which, it may be
noted, IBM REXX on its various platforms is the least ANSI-compliant of all the
different Rexx processors represented).

I have no doubt that one day such alternate NetRexx processors will exist (look
at how many Linux versions there are).  I just know in my bones that while the
FOSS NetRexx is scrambling for exposure and mind-share, we need to present a
single, unified platform to the world.

So I offer my apologies for misinterpreting the intent of this nascent
"developer group" (as opposed to the "support group" which will be responsible
for maintaining and enhancing NetRexx itself) and suggest that we keep this
discussion list as our primary point of communication.

I offered to host this listserver on one of my under-utilized virtual servers
before Ren? announced that he had a 'netrexx.org' server.  My offer still stands
but I think we would all agree that it would be best served on his site.  Ren?
has a lot on his plate at the moment however, so should IBM decide to pull the
plug on this list, we can quickly transfer this discussion to my server in the
interim.

And just FYI, the domains 'netrexx.biz' and 'netrexx.info' are still available
if someone wanted to create a site devoted to NetRexx-enabled products... :-)

-Chip-

On 2/6/10 02:00 David Requena said:

> Chip,
>
> I'd like to make clear that at no point in time I've criticized how the
> open sourcing process is taking place or how RexxLA is going to manage
> the language at a latter time. To be honest, until very recently I
> didn't believe said process was going anywhere. At some point I got
> convinced this is really going to happen eventually and kindly made some
> obvious questions just out of curiosity. RexxLA is by express desire of
> its members a closed community. That is a fact. It is so for the very
> same reasons you state below. So I had to ask and hope for someone
> responding.
>
> Of course I have my own opinions on these "non-criticized" topics above
> but I'm not interested on discussing them here at all. May be one day at
> RexxLA's list.. ;-)
> What I really *do* be interested on is in NetRexx itself, in
> my ability to go on indefinitely using it, and maybe one day beholding
> it going mainstream allowing me to earn a living with it instead of
> java, c++ and c# (not to mention lotusscript and vb...)
>
> I think the whole debate over websites got a little bit out of hand. In
> my view the open source netrexx part of the discussion just came up at
> nearly the same time. I've been the whole time saying opensourced
> netrexx does not yet exist. When it becomes a reality of course a single
> point of contact (or at least a main central one) is the way to go.
> Hell, I even presented the current ooREXX model as how I envision the
> whole thing at the end of the road. But the fact still is: *** the time
> has not come yet ***
>
> What we (at least me!) were discussing is: look, here we are; 3
> individuals with netrexx centric projects we are actually working on. We
> share a passion (I lack a better word, really) for the language. We have
> our projects hosted at 3 different sites. Thanks to this very same list
> we're already collaborating ("hey would you opensource your parser so I
> can use it? - yes, what would be a best license", "could you help me
> with that jEdit idiosincracy?" - "dont' really know, I'll do my best").
> At some point one proposes more or less informally we could put our
> projects together. That would benefit each others projects bringing more
> attention to them (note we're already cross linking/mentioning). That
> would benefit potential users too. Ok, we start talking about how such
> an endeavour would best be approached.
>
> See? This is community building up. Maybe it will catch and attract more
> individuals until a certain critical mass is built up, maybe not and
> then the trend will vanish. Nobody knows at this time. You can
> not discuss, design, document, implement, then release a community the
> REXX way (or any other way for that matter) it just happens when
> the right individuals coincide at the right time at the right place.
>
> Look at this very same list. I've been following it for years now. There
> has been possibly more traffic in the last two weeks that in the whole
> preceding two years. And this is good indeed. At least we're holding
> discussions as gratifying as this very same one!
>
> The distinction over language user community and language owner entity
> is not one I made. It just exists right now, I now for sure I'm not part
> of the latter at the moment. A (nearly dead) community of  netrexx users
> has existed all the way long while the language itself was in limbo. An
> owner has existed too. And the thing is, this is not bad in itself. We
> would have gone on using netrexx as long as we could was it not to be
> open sourced. I've been using gcc many years without ever having a look
> at its source code.
>
> Admittedly at the FOSS world this distinction is often moot, mainly for
> languages that were born open source to start with. In other cases start
> mingling to the point you cannot really tell which is what. That's what
> I would like to see with netrexx in the end, when it's ripe as an OSS
> product. Again the fact is: *** we're not still there ***
>
> So let the community grow if that's what's gona happen.
> I community needs to build around the language. Someone must be there,
> waiting, when OSS NetRexx finally arrives. We are definitely a
> bit enthusiastic at the moment, I know. We're relieved, we've been
> worried about our language of choice disappearing, seeing our tools
> going unsupported for years. Now we *do know* the language is gona be
> open sourced. Hurray! NetRexx is gona live! We wont be depending on some
> 3 letter company shutting down the server holding the downloads of
> NetRexxC and NetRexxD, nor the netrexx-list (our only point of contact)
> being eventually withdrawn.
>
> PS: From now on I'll be trying not to use the word 'community'
> on further discussion, I'm already tired of it. Besides, hell, there is
> only five indiciduals involved in the thread!!
>
> PS2: I'll be surely involved in the discussion of technical details
> related to the upcoming release of oss netrexx when the time has come.
> For the time being I'll stick to the technical requirements the software
> I maintain imposes on me.
>
> PS3: No words I could possibly utter but convey praise enough for the
> work of that craftsman under the heavy hand of the king. Noww. being
> under that reign gave him the opportunity to raise general awareness of
> his own craft. Life is made of these incongruences... Thanks Mike!!!
>
> PS4: I'm sorry, no nice words for the king's lawyers. Maybe at a latter
> time.
>
>
> 2010/2/5 Chip Davis <[hidden email] <mailto:[hidden email]>>
>
>     Thank you for that clarification, David.  I agree that it is crucial
>     to a productive discussion to establish exactly what we are talking
>     about. :-)
>
>     Let me first establish that I do not speak for the Open Source
>     NetRexx Project, nor in any official capacity for the RexxLA.  _I_
>     have not signed any NDAs, and am privy to no more than anyone else
>     on this discussion list.  I speak from experience, and My Humble
>     Opinions are exactly that.  But I took the Rexx Red Pill a long time
>     ago.
>
>     If I may paraphrase, you have made a distinction between "NetRexx
>     the Language" and "NetRexx the Community".  IMHO, if NetRexx is to
>     succeed, we need to do everything in our power to erase that
>     distinction and avoid creating any mechanism that would tend to draw
>     such a distinction.  That is not only the "Rexx Way", that's the
>     "FOSS Way".
>
>     In my experience, a single point of contact (especially for nascent
>     efforts like FOSS NetRexx) is absolutely critical to wider
>     visibility and acceptance.  This was the primary raison d'?tre for
>     the creation of RexxLA in the first place.
>     There were many, many sites, vendors, and disassociated users of
>     Rexx, on many different platforms, who needed a single place to go
>     for information, help, news, etc.
>
>     Sure, there was comp.lang.rexx but it had become obvious by the late
>     Eighties that, due to the "open" nature of newsgroups, its
>     signal-to-noise ratio had dropped almost to the point of
>     unintelligibility.  Sadly, things have not gotten significantly
>     better in the intervening twenty years and I don't expect them to,
>     when the cost to participate is so low.  One sees the exact same
>     phenomenon when attending a symphony concert in the park, versus the
>     hall.  Same symphony, substantially different audience behavior,
>     considerably poorer musical experience.
>
>     In short, I see no advantage, and many pitfalls in making a
>     distinction between "that which is RexxLA" and "that which is the
>     NetRexx Community".  RexxLA is merely the umbrella organization with
>     a legal name and address and incorporation papers that can be
>     responsible for the resources that IBM is willing to convey.
>
>     As for technical issues regarding where certain files should be
>     maintained, or which server should host which discussion group, or
>     other such minor details, I see no reason why the model (if not the
>     specifics) used in the ooRexx Project would not be a good place to
>     start.  (I will admit that I'm not a big fan of SourceForge any
>     more, but that itself is a technical detail.)
>
>     In closing, I'd like to address the subtle yet pervasive idea
>     implicit in your "ivory tower" references.  In the "Cathedral and
>     Bazaar" model, the NetRexx Project is most decidedly _not_ a
>     cathedral, nor a tower (of whatever material).
>
>     NetRexx has always been product of a single, extraordinarily
>     talented Craftsman working at his shop in the bazaar, building tools
>     for us peasants and giving them away for free.  During the week, he
>     is indeed a "Toolmaker to the King" but on weekends he's in the
>     bazaar working for free.  However, the King always makes the rules
>     and, while willing to release the rights to one of the Craftsman's
>     tools, the Royal Lawyers must be appeased.  Thus, and very
>     temporarily, the King's henchmen are guarding the Craftsman's door
>     while an inventory is taken. They will be gone very soon and the sun
>     will again stream into the Craftsman's shoppe.
>
>     It is RexxLA's best interest, no to mention a legal requirement,
>     that it support NetRexx in the Open Source Way.
>
>     -Chip-
>
>
>     On 2/5/10 12:57 David Requena said:
>
>         Chip,
>
>         Yours is more or less my own point. I'd like to express a couple
>         slight precisions though.
>
>         First, I believe where mixing 2 very distinct concepts here:
>
>         - Open source NetRexx. This is RexxLA's concern and I think it
>         will be obviously benefiting from "the REXX way".
>         - Open source NetRexx user code/resources. This is the NetRexx
>         language users community concern and is benefiting from "a
>         bookmark file full of broken or obsolete links" right now.
>
>         I for one have been talking about the second point the whole
>         time. We, as the user base of the language, would greatly
>         benefit from a consolidation of the current array of dispersed
>         (and rapidly vanishing into oblivion) of netrexx resources on
>         the net.
>
>         Second, just because a given language's user community doesn't
>         has a direct role in the language design, implementation, etc.
>         doesn't mean it may have an important influence on that
>         language's ultimate fate. Any effort made to help growing a
>         strong, enthusiastic, user community is just another brick
>         thrown at the foundation of the tower (pun intended ;-) ) we all
>         are trying to build. That would be "the current times way"...
>
>
>         2010/2/5 Chip Davis <[hidden email] <mailto:[hidden email]>
>         <mailto:[hidden email] <mailto:[hidden email]>>>
>
>
>            Tom,
>
>            I'm sure Ren? and those working diligently to effect the
>         transfer of
>            NetRexx from IBM to RexxLA appreciate the enthusiasm that you,
>            David, Kermit, and others have demonstrated for this project,
>         but I
>            think we all need to slow down and take a breath.
>
>            Our baby is not yet walking and you are already discussing which
>            motorcycle to buy for him. :-)
>
>            I think a couple of things should be self-evident: we have some
>            world-class intellectual resources within RexxLA (Ren? and
>         Mike, for
>            starters), we have the experience of having done the same process
>            with ooRexx, and we have a passionate (rabid?) group of NetRexx
>            enthusiasts trying to wait for Christmas morning.
>
>            There are many fundamental issues that need to be addressed
>         by the
>            NetRexx team and the RexxLA Board and Officers.  Believe me,
>         we have
>            a lot more questions than you do, many of which are quite prosaic
>            and unglamorous, but crucial to the success of this project.  We
>            learned a lot from the ooRexx project and will incorporate those
>            lessons into the organization and charter of the NetRexx project.
>
>            At this point, fixed written-in-stone specifics are hard to
>         come by,
>            so I hope you (and everyone else chomping at the bit) will take
>            comfort in the philosophy to which all of us are committed, that
>            being the fundamental tenets of The Rexx Language, as laid out by
>            Mike in TRL and TNRL.  Ren? said as much in his answer to David's
>            fourth question, when he referred to following The Rexx Way (my
>            caps) of Discuss, Agree, Document, then Implement.
>
>            Thus, if you haven't seen any Discussion of a particular
>         issue, it
>            probably means we haven't gotten to that stage yet.
>
>            I would suggest that you not make any decisions about various
>         code
>            or doc repositories until we figure out where and how the NetRexx
>            code will be housed.  It will be much less confusing for users to
>            have only one location to find NetRexx resources, than a bookmark
>            file full of broken or obsolete links.
>
>            IMHO, of course.
>
>            -Chip-
>
>            On 2/5/10 05:22 Thomas Schneider said:
>             >   to follow UP the idea of putting *any and all* Utilities
>            (including mine)
>             > to www.NetRexx.org <http://www.NetRexx.org>
>         <http://www.NetRexx.org>:
>
>             >
>             > What does this mean for my (private) directory structure ?
>             >
>             > When you like, I can send you my current one, but I am not
>         so much
>             > satisfied
>             > with what I do have, and will need (please)
>             >
>             > your advise:
>             >
>             > How could/should we structure the whole project
>         www.NetRexx.org <http://www.NetRexx.org>
>            <http://www.NetRexx.org> ??
>
>             >
>             > Your advise will be welcome ...
>
>
>            On 2/5/10 03:48 Thomas Schneider said:
>
>                I am currently in the process to buildup the new
>                www.Rexx2Nrx.com <http://www.Rexx2Nrx.com>
>         <http://www.Rexx2Nrx.com> homepage.
>
>
>                Might it be that this would be a good instance to store all
>                those tools available in a central
>                place ?
>
>                Kermit Kiser schrieb:
>
>                    I have seen a lot of RFEs floating around but no safe
>         place
>                    to store them all or discuss them. Could you do something
>                    like that with your site?
>
>                    David Requena wrote:
>
>                        The thing is...  for the first time I got the feeling
>                        this process is actually advancing in some direction.
>                        We all, as NetRexx developers hoping to see the
>         language
>                        advance, are eagerly awaiting for the
>                        culmination of a process about which we have near
>         to no
>                        information.
>
>
>
>
>            _______________________________________________
>            Ibm-netrexx mailing list
>            [hidden email]
>         <mailto:[hidden email]>
>         <mailto:[hidden email]
>         <mailto:[hidden email]>>
>
>
>
>
>         --
>         Saludos / Regards,
>         David Requena
>
>
>         ------------------------------------------------------------------------
>
>
>         _______________________________________________
>         Ibm-netrexx mailing list
>         [hidden email] <mailto:[hidden email]>
>
>
>
>     _______________________________________________
>     Ibm-netrexx mailing list
>     [hidden email] <mailto:[hidden email]>
>
>
>
>
> --
> Saludos / Regards,
> David Requena
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>

[Thomas.Schneider.Wien]

Hello Rene & all,
   first thanks for the follow up.

second, we *should* and *MUST* avoid, that everybody wanting to try &
USE one of the tools mentioned, *MUST be a member of REXXLA (with annual
fees).

This is archaic thinking.

When we are offering OPEN SOURCE Software, the dicussion list MUST be
OPEN as well,
e.g. accessible by any member. I, for instance, will use GOOGLE group
Rexx2Nrx for this
purpose,  where anybody canbecome a member without any fee. And I do NOT
like
to watch too many discusion groups as well. And I also do NOT like to
re-envent the wheel,
when not necessary.


Currently, REXXLA ist mostly ooRexx oriented.

And it might be well the case that we would need an ooRexx and NetRexx
discusion group.


Ren? Jansen schrieb:

> Kermit,
>
> let me first say that I fully agree with what Chip said. There should be a single point of contact and that point will be RexxLa - just because that is the best place where you can also find the best Rexx people.
>
> I have, however, an elegant solution to the problem. www.netrexx.org is mine now, and points to one of my servers (that is not running right now in case you want to check, but that is beside the point that I am going to make).
>
> There is nothing to stop us from organizing public knowledge about NetRexx and ~resources on this site - that eventually will be donated to RexxLA. Actually, it should have been done a long time ago. When I said that we should start doing things I meant it and I think you misinterpreted that a lot less than you thought when writing that email. The point that Chip is making about publicy and unmoderated mailing list is valid though, and I am not going to entertain or participate a lot in discussion how things should be. I will have a look at what I like about other websites of programming languages and learn from that.
>
> There is a reason for the 'work in progress; ' signs on the RexxLA site, and that is because the people with the knowledge to do webdesign and maintenance are also working on implementations of Rexx, and humans need to sleep. No-one is paid for this; the fee structure does not allow this.
>
> Please send me your stuff, links, ideas and things. We'll see how it goes. When it goes well, I'll switch it on. When the time comes, it will transition to RexxLA. I think this model will accomodate all previously mentioned points of view.
>
> best regards,
>
> Ren?.
>
>
> On 5 feb 2010, at 22:45, Kermit Kiser wrote:
>
>  
>> Chip & all ;
>>
>> As Jeff just pointed out in a separate message, when you go to the RexxLA web site, you see a lot of "Under Construction" messages and very little in the way of resources. I am not saying that RexxLA is a total failure at providing Rexx info, just that it does not look very successful yet. And when it comes to NetRexx info, there is virtually zero. Hopefully that will improve when NetRexx is finally open source. But that project is almost two years underway now (officially) and we still don't know when it will complete.
>>
>> Meanwhile, some of us are doing stuff with NetRexx right now. That includes a desire to see NetRexx succeed and grow to remain a viable option for our projects long term. When Ren? said we should start getting stuff ready, I admit I misinterpreted that to mean that RFE discussion was now fair game, but even with that on hold along with some of our projects like my scripting modules which are waiting on OSS NetRexx, there is still a lot we can do. It is great if RexxLA wants to provide the central portal for NetRexx and related resources, but that does not mean that we should put everything on hold indefinitely. On the contrary, we should be getting ready now as Ren? said. And since Rexx projects are mostly separate items on Sourceforge now, a consolidated repository would put NetRexx way ahead, IMHO.
>>
>> Thanks to RexxLA, IBM (Mike & Ren??) found a way to open source NetRexx. But most of RexxLA is probably not interested in NetRexx at this point, so WE are the NetRexx community right now.
>>
>> Just my opinion.
>>
>> -- Kermit
>>
>>
>> Chip Davis wrote:
>>    
>>> Thank you for that clarification, David.  I agree that it is crucial to a productive discussion to establish exactly what we are talking about. :-)
>>>
>>> Let me first establish that I do not speak for the Open Source NetRexx Project, nor in any official capacity for the RexxLA.  _I_ have not signed any NDAs, and am privy to no more than anyone else on this discussion list.  I speak from experience, and My Humble Opinions are exactly that.  But I took the Rexx Red Pill a long time ago.
>>>
>>> If I may paraphrase, you have made a distinction between "NetRexx the Language" and "NetRexx the Community".  IMHO, if NetRexx is to succeed, we need to do everything in our power to erase that distinction and avoid creating any mechanism that would tend to draw such a distinction.  That is not only the "Rexx Way", that's the "FOSS Way".
>>>
>>> In my experience, a single point of contact (especially for nascent efforts like FOSS NetRexx) is absolutely critical to wider visibility and acceptance.  This was the primary raison d'?tre for the creation of RexxLA in the first place.
>>> There were many, many sites, vendors, and disassociated users of Rexx, on many different platforms, who needed a single place to go for information, help, news, etc.
>>>
>>> Sure, there was comp.lang.rexx but it had become obvious by the late Eighties that, due to the "open" nature of newsgroups, its signal-to-noise ratio had dropped almost to the point of unintelligibility.  Sadly, things have not gotten significantly better in the intervening twenty years and I don't expect them to, when the cost to participate is so low.  One sees the exact same phenomenon when attending a symphony concert in the park, versus the hall.  Same symphony, substantially different audience behavior, considerably poorer musical experience.
>>>
>>> In short, I see no advantage, and many pitfalls in making a distinction between "that which is RexxLA" and "that which is the NetRexx Community".  RexxLA is merely the umbrella organization with a legal name and address and incorporation papers that can be responsible for the resources that IBM is willing to convey.
>>>
>>> As for technical issues regarding where certain files should be maintained, or which server should host which discussion group, or other such minor details, I see no reason why the model (if not the specifics) used in the ooRexx Project would not be a good place to start.  (I will admit that I'm not a big fan of SourceForge any more, but that itself is a technical detail.)
>>>
>>> In closing, I'd like to address the subtle yet pervasive idea implicit in your "ivory tower" references.  In the "Cathedral and Bazaar" model, the NetRexx Project is most decidedly _not_ a cathedral, nor a tower (of whatever material).
>>>
>>> NetRexx has always been product of a single, extraordinarily talented Craftsman working at his shop in the bazaar, building tools for us peasants and giving them away for free.  During the week, he is indeed a "Toolmaker to the King" but on weekends he's in the bazaar working for free.  However, the King always makes the rules and, while willing to release the rights to one of the Craftsman's tools, the Royal Lawyers must be appeased.  Thus, and very temporarily, the King's henchmen are guarding the Craftsman's door while an inventory is taken. They will be gone very soon and the sun will again stream into the Craftsman's shoppe.
>>>
>>> It is RexxLA's best interest, no to mention a legal requirement, that it support NetRexx in the Open Source Way.
>>>
>>> -Chip-
>>>
>>> On 2/5/10 12:57 David Requena said:
>>>      
>>>> Chip,
>>>>
>>>> Yours is more or less my own point. I'd like to express a couple slight precisions though.
>>>>
>>>> First, I believe where mixing 2 very distinct concepts here:
>>>>
>>>> - Open source NetRexx. This is RexxLA's concern and I think it will be obviously benefiting from "the REXX way".
>>>> - Open source NetRexx user code/resources. This is the NetRexx language users community concern and is benefiting from "a bookmark file full of broken or obsolete links" right now.
>>>>
>>>> I for one have been talking about the second point the whole time. We, as the user base of the language, would greatly benefit from a consolidation of the current array of dispersed (and rapidly vanishing into oblivion) of netrexx resources on the net.
>>>>
>>>> Second, just because a given language's user community doesn't has a direct role in the language design, implementation, etc. doesn't mean it may have an important influence on that language's ultimate fate. Any effort made to help growing a strong, enthusiastic, user community is just another brick thrown at the foundation of the tower (pun intended ;-) ) we all are trying to build. That would be "the current times way"...
>>>>
>>>>
>>>> 2010/2/5 Chip Davis <[hidden email] <mailto:[hidden email]>>
>>>>
>>>>    Tom,
>>>>
>>>>    I'm sure Ren? and those working diligently to effect the transfer of
>>>>    NetRexx from IBM to RexxLA appreciate the enthusiasm that you,
>>>>    David, Kermit, and others have demonstrated for this project, but I
>>>>    think we all need to slow down and take a breath.
>>>>
>>>>    Our baby is not yet walking and you are already discussing which
>>>>    motorcycle to buy for him. :-)
>>>>
>>>>    I think a couple of things should be self-evident: we have some
>>>>    world-class intellectual resources within RexxLA (Ren? and Mike, for
>>>>    starters), we have the experience of having done the same process
>>>>    with ooRexx, and we have a passionate (rabid?) group of NetRexx
>>>>    enthusiasts trying to wait for Christmas morning.
>>>>
>>>>    There are many fundamental issues that need to be addressed by the
>>>>    NetRexx team and the RexxLA Board and Officers.  Believe me, we have
>>>>    a lot more questions than you do, many of which are quite prosaic
>>>>    and unglamorous, but crucial to the success of this project.  We
>>>>    learned a lot from the ooRexx project and will incorporate those
>>>>    lessons into the organization and charter of the NetRexx project.
>>>>
>>>>    At this point, fixed written-in-stone specifics are hard to come by,
>>>>    so I hope you (and everyone else chomping at the bit) will take
>>>>    comfort in the philosophy to which all of us are committed, that
>>>>    being the fundamental tenets of The Rexx Language, as laid out by
>>>>    Mike in TRL and TNRL.  Ren? said as much in his answer to David's
>>>>    fourth question, when he referred to following The Rexx Way (my
>>>>    caps) of Discuss, Agree, Document, then Implement.
>>>>
>>>>    Thus, if you haven't seen any Discussion of a particular issue, it
>>>>    probably means we haven't gotten to that stage yet.
>>>>
>>>>    I would suggest that you not make any decisions about various code
>>>>    or doc repositories until we figure out where and how the NetRexx
>>>>    code will be housed.  It will be much less confusing for users to
>>>>    have only one location to find NetRexx resources, than a bookmark
>>>>    file full of broken or obsolete links.
>>>>
>>>>    IMHO, of course.
>>>>
>>>>    -Chip-
>>>>
>>>>    On 2/5/10 05:22 Thomas Schneider said:
>>>>     >   to follow UP the idea of putting *any and all* Utilities
>>>>    (including mine)
>>>>     > to www.NetRexx.org <http://www.NetRexx.org>:
>>>>     >
>>>>     > What does this mean for my (private) directory structure ?
>>>>     >
>>>>     > When you like, I can send you my current one, but I am not so much
>>>>     > satisfied
>>>>     > with what I do have, and will need (please)
>>>>     >
>>>>     > your advise:
>>>>     >
>>>>     > How could/should we structure the whole project www.NetRexx.org
>>>>    <http://www.NetRexx.org> ??
>>>>     >
>>>>     > Your advise will be welcome ...
>>>>
>>>>
>>>>    On 2/5/10 03:48 Thomas Schneider said:
>>>>
>>>>        I am currently in the process to buildup the new
>>>>        www.Rexx2Nrx.com <http://www.Rexx2Nrx.com> homepage.
>>>>
>>>>        Might it be that this would be a good instance to store all
>>>>        those tools available in a central
>>>>        place ?
>>>>
>>>>        Kermit Kiser schrieb:
>>>>
>>>>            I have seen a lot of RFEs floating around but no safe place
>>>>            to store them all or discuss them. Could you do something
>>>>            like that with your site?
>>>>
>>>>            David Requena wrote:
>>>>
>>>>                The thing is...  for the first time I got the feeling
>>>>                this process is actually advancing in some direction.
>>>>                We all, as NetRexx developers hoping to see the language
>>>>                advance, are eagerly awaiting for the
>>>>                culmination of a process about which we have near to no
>>>>                information.
>>>>
>>>>
>>>>
>>>>
>>>>    _______________________________________________
>>>>    Ibm-netrexx mailing list
>>>>    [hidden email] <mailto:[hidden email]>
>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> Saludos / Regards,
>>>> David Requena
>>>>
>>>>
>>>> ------------------------------------------------------------------------
>>>>
>>>> _______________________________________________
>>>> Ibm-netrexx mailing list
>>>> [hidden email]
>>>>
>>>>        
>>> _______________________________________________
>>> Ibm-netrexx mailing list
>>> [hidden email]
>>>
>>>
>>>
>>>      
>> _______________________________________________
>> Ibm-netrexx mailing list
>> [hidden email]
>>
>>    
>
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>
>  

[Thomas.Schneider.Wien]

Hello Rene.

First thanks for publishing this code snippet.

Second, having access to the OPEN NetRexx source would all of us allow a LOT
to LEARN from Mike's techniques and avoid duplicate code.

For me, of course, this all is a bit too late. I did it already my way,
with the same intent like Mike ones most probably, but using other
techniques already.

Tom.
 
Ren? Jansen schrieb:

> David,
>
> when I spoke to Mike the last Rexx Language Symposium last year, we briefly discussed dynamic features, and he mentioned that in his opinion the interpreter api in the current public version would already allow for this. I tried it some time ago, and this is what I came up with:
>
>
> options binary
> import COM.ibm.netrexx.process.NetRexxA
> /**
>  * Class Dynamic implements...
>  * <BR>
>  * Created on: di, 16, jun 2009 22:37:53 +0200
>  */
> class Dynamic extends SourceTransformer  -- take that out later
>
>   interpreter = NetRexxA()
>   /**
>    * Default constructor
>    */
>   method Dynamic()
>
>   method interpret(file=String) returns Class
>     files=[file]
>     classname = Rexx file
>     classname = classname.substr(1,classname.pos('.')-1)
>     say classname
>     flags=['nocrossref', 'verbose0']
>     interpreter.parse(files, flags)
>     return interpreter.getClassObject(null, classname) -- find the hello Class
>    
>   method main(args=String[]) static
>     t = Dynamic()
>     source = Arraylist()
>     source.add(Rexx "options binary")
>     source.add(Rexx "class TestDyn")
>     source.add(Rexx "method main(args = String[]) static")
>     source.add(Rexx "say 'Hello Dynamic NetRexx World'")
>
>     writeFileFromArrayList('TestDyn.nrx', source)
>    
>     dynamicClass = t.interpret('TestDyn.nrx')
>
>     say dynamicClass
>
>
> note that I never took out SourceTransformer, but that class is only used to call writeFileFromArrayList - I am sure you can write a file yourself.
>
> Just as an illustration that I went complaining about the static nature of NetRexx for years without understanding the wealth that is already in there. One could make an API out of that without the need for changing the core functionality.
>
> best regards,
>
> Ren?.
>
>
>
>
>
>
>
>
>
>
>
>  
>> PS: Kermit I've just become aware of how moot the RFE talk is right now. According to proceedings of Ren?'s presentation at JVM Language Summit 2008 there already existed a NetRexx v. 3.00 back in September 2008 (which wasn't out yet ;-) ). There is even an interesting 'recent additions section'!
>>
>> In the meantime they could even have included planned dynamic features! We'll have to wait till a release is made to see if any RFE's still apply.
>>    
>
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>
>  

[Thomas.Schneider.Wien]

Hello Chip,
   there should be ONE and ONLY one NetRexx Language and Compiler and
Interpreter.
And a central place (and one and only one) for NetRexxScript, NetrexxDE
(by the Way, David, I would name it NetRexxDev (for NetRexxDevelopment),
and, when wanted, my Rey (Rexx2Nrx) stuff.

I do hope I did initiate a useful discussion on this list.

I will now concentrate to Get Rey (Rexx for Java ::=== Rexx2Nrx release
7.00 out ASAP.)

And Rey is a FOSS project (just learned this acronym)...

Kind regards,
Tom.
==================================================================
Chip Davis schrieb:

> Thank you David, for your thoughtful reply.  It is the nature of this
> medium that a minor misunderstanding that would take seconds to clear
> up in a face-to-face conversation, will take many KB to reconcile over
> the net.
>
> I confess that the impression I got from your (and Kermit and Tom)'s
> enthusiasm to set up a different Internet site for your products, was
> that you would then be free to enhance NetRexx itself (with your
> existing list of RFEs) and offer an alternate version to the world.  
> In large part, this impression was a result of Tom's suggestion that
> he create a private discussion list for the four of us.  I saw this as
> the imminent splintering of a not-yet-coalesced NetRexx community.
>
> This is a sensitive issue for me personally, having been a member of
> the ANSI Standard Committee for Rexx back in the Nineties, only to see
> Jeff Glatt pollute the Rexx interpreter marketplace with his
> wildly-idiosyncratic and blatantly non-standard Reginald interpreter.  
> And I find it illustrative that he did so via comp.lang.rexx, not
> under the auspices of RexxLA (under which, it may be noted, IBM REXX
> on its various platforms is the least ANSI-compliant of all the
> different Rexx processors represented).
>
> I have no doubt that one day such alternate NetRexx processors will
> exist (look at how many Linux versions there are).  I just know in my
> bones that while the FOSS NetRexx is scrambling for exposure and
> mind-share, we need to present a single, unified platform to the world.
>
> So I offer my apologies for misinterpreting the intent of this nascent
> "developer group" (as opposed to the "support group" which will be
> responsible for maintaining and enhancing NetRexx itself) and suggest
> that we keep this discussion list as our primary point of communication.
>
> I offered to host this listserver on one of my under-utilized virtual
> servers before Ren? announced that he had a 'netrexx.org' server.  My
> offer still stands but I think we would all agree that it would be
> best served on his site.  Ren? has a lot on his plate at the moment
> however, so should IBM decide to pull the plug on this list, we can
> quickly transfer this discussion to my server in the interim.
>
> And just FYI, the domains 'netrexx.biz' and 'netrexx.info' are still
> available if someone wanted to create a site devoted to
> NetRexx-enabled products... :-)
>
> -Chip-
>
> On 2/6/10 02:00 David Requena said:
>> Chip,
>>
>> I'd like to make clear that at no point in time I've criticized how
>> the open sourcing process is taking place or how RexxLA is going to
>> manage the language at a latter time. To be honest, until very
>> recently I didn't believe said process was going anywhere. At some
>> point I got convinced this is really going to happen eventually and
>> kindly made some obvious questions just out of curiosity. RexxLA is
>> by express desire of its members a closed community. That is a fact.
>> It is so for the very same reasons you state below. So I had to ask
>> and hope for someone responding.
>>
>> Of course I have my own opinions on these "non-criticized" topics
>> above but I'm not interested on discussing them here at all. May be
>> one day at RexxLA's list.. ;-)
>> What I really *do* be interested on is in NetRexx itself, in my
>> ability to go on indefinitely using it, and maybe one day beholding
>> it going mainstream allowing me to earn a living with it instead of
>> java, c++ and c# (not to mention lotusscript and vb...)
>>
>> I think the whole debate over websites got a little bit out of hand.
>> In my view the open source netrexx part of the discussion just came
>> up at nearly the same time. I've been the whole time saying
>> opensourced netrexx does not yet exist. When it becomes a reality of
>> course a single point of contact (or at least a main central one) is
>> the way to go. Hell, I even presented the current ooREXX model as how
>> I envision the whole thing at the end of the road. But the fact still
>> is: *** the time has not come yet ***
>>
>> What we (at least me!) were discussing is: look, here we are; 3
>> individuals with netrexx centric projects we are actually working on.
>> We share a passion (I lack a better word, really) for the language.
>> We have our projects hosted at 3 different sites. Thanks to this very
>> same list we're already collaborating ("hey would you opensource your
>> parser so I can use it? - yes, what would be a best license", "could
>> you help me with that jEdit idiosincracy?" - "dont' really know, I'll
>> do my best"). At some point one proposes more or less informally we
>> could put our projects together. That would benefit each others
>> projects bringing more attention to them (note we're already cross
>> linking/mentioning). That would benefit potential users too. Ok, we
>> start talking about how such an endeavour would best be approached.
>>
>> See? This is community building up. Maybe it will catch and attract
>> more individuals until a certain critical mass is built up, maybe not
>> and then the trend will vanish. Nobody knows at this time. You can
>> not discuss, design, document, implement, then release a community
>> the REXX way (or any other way for that matter) it just happens when
>> the right individuals coincide at the right time at the right place.
>>
>> Look at this very same list. I've been following it for years now.
>> There has been possibly more traffic in the last two weeks that in
>> the whole preceding two years. And this is good indeed. At least
>> we're holding discussions as gratifying as this very same one!
>>
>> The distinction over language user community and language owner
>> entity is not one I made. It just exists right now, I now for sure
>> I'm not part of the latter at the moment. A (nearly dead) community
>> of  netrexx users has existed all the way long while the language
>> itself was in limbo. An owner has existed too. And the thing is, this
>> is not bad in itself. We would have gone on using netrexx as long as
>> we could was it not to be open sourced. I've been using gcc many
>> years without ever having a look at its source code.
>>
>> Admittedly at the FOSS world this distinction is often moot, mainly
>> for languages that were born open source to start with. In other
>> cases start mingling to the point you cannot really tell which is
>> what. That's what I would like to see with netrexx in the end, when
>> it's ripe as an OSS product. Again the fact is: *** we're not still
>> there ***
>>
>> So let the community grow if that's what's gona happen. I community
>> needs to build around the language. Someone must be there, waiting,
>> when OSS NetRexx finally arrives. We are definitely a bit
>> enthusiastic at the moment, I know. We're relieved, we've been
>> worried about our language of choice disappearing, seeing our tools
>> going unsupported for years. Now we *do know* the language is gona be
>> open sourced. Hurray! NetRexx is gona live! We wont be depending on
>> some 3 letter company shutting down the server holding the downloads
>> of NetRexxC and NetRexxD, nor the netrexx-list (our only point of
>> contact) being eventually withdrawn.
>>
>> PS: From now on I'll be trying not to use the word 'community' on
>> further discussion, I'm already tired of it. Besides, hell, there is
>> only five indiciduals involved in the thread!!
>> PS2: I'll be surely involved in the discussion of technical details
>> related to the upcoming release of oss netrexx when the time has
>> come. For the time being I'll stick to the technical requirements the
>> software I maintain imposes on me.
>>
>> PS3: No words I could possibly utter but convey praise enough for the
>> work of that craftsman under the heavy hand of the king. Noww. being
>> under that reign gave him the opportunity to raise general awareness
>> of his own craft. Life is made of these incongruences... Thanks Mike!!!
>>
>> PS4: I'm sorry, no nice words for the king's lawyers. Maybe at a
>> latter time.
>>
>> 2010/2/5 Chip Davis <[hidden email] <mailto:[hidden email]>>
>>
>>     Thank you for that clarification, David.  I agree that it is crucial
>>     to a productive discussion to establish exactly what we are talking
>>     about. :-)
>>
>>     Let me first establish that I do not speak for the Open Source
>>     NetRexx Project, nor in any official capacity for the RexxLA.  _I_
>>     have not signed any NDAs, and am privy to no more than anyone else
>>     on this discussion list.  I speak from experience, and My Humble
>>     Opinions are exactly that.  But I took the Rexx Red Pill a long time
>>     ago.
>>
>>     If I may paraphrase, you have made a distinction between "NetRexx
>>     the Language" and "NetRexx the Community".  IMHO, if NetRexx is to
>>     succeed, we need to do everything in our power to erase that
>>     distinction and avoid creating any mechanism that would tend to draw
>>     such a distinction.  That is not only the "Rexx Way", that's the
>>     "FOSS Way".
>>
>>     In my experience, a single point of contact (especially for nascent
>>     efforts like FOSS NetRexx) is absolutely critical to wider
>>     visibility and acceptance.  This was the primary raison d'?tre for
>>     the creation of RexxLA in the first place.
>>     There were many, many sites, vendors, and disassociated users of
>>     Rexx, on many different platforms, who needed a single place to go
>>     for information, help, news, etc.
>>
>>     Sure, there was comp.lang.rexx but it had become obvious by the late
>>     Eighties that, due to the "open" nature of newsgroups, its
>>     signal-to-noise ratio had dropped almost to the point of
>>     unintelligibility.  Sadly, things have not gotten significantly
>>     better in the intervening twenty years and I don't expect them to,
>>     when the cost to participate is so low.  One sees the exact same
>>     phenomenon when attending a symphony concert in the park, versus the
>>     hall.  Same symphony, substantially different audience behavior,
>>     considerably poorer musical experience.
>>
>>     In short, I see no advantage, and many pitfalls in making a
>>     distinction between "that which is RexxLA" and "that which is the
>>     NetRexx Community".  RexxLA is merely the umbrella organization with
>>     a legal name and address and incorporation papers that can be
>>     responsible for the resources that IBM is willing to convey.
>>
>>     As for technical issues regarding where certain files should be
>>     maintained, or which server should host which discussion group, or
>>     other such minor details, I see no reason why the model (if not the
>>     specifics) used in the ooRexx Project would not be a good place to
>>     start.  (I will admit that I'm not a big fan of SourceForge any
>>     more, but that itself is a technical detail.)
>>
>>     In closing, I'd like to address the subtle yet pervasive idea
>>     implicit in your "ivory tower" references.  In the "Cathedral and
>>     Bazaar" model, the NetRexx Project is most decidedly _not_ a
>>     cathedral, nor a tower (of whatever material).
>>
>>     NetRexx has always been product of a single, extraordinarily
>>     talented Craftsman working at his shop in the bazaar, building tools
>>     for us peasants and giving them away for free.  During the week, he
>>     is indeed a "Toolmaker to the King" but on weekends he's in the
>>     bazaar working for free.  However, the King always makes the rules
>>     and, while willing to release the rights to one of the Craftsman's
>>     tools, the Royal Lawyers must be appeased.  Thus, and very
>>     temporarily, the King's henchmen are guarding the Craftsman's door
>>     while an inventory is taken. They will be gone very soon and the sun
>>     will again stream into the Craftsman's shoppe.
>>
>>     It is RexxLA's best interest, no to mention a legal requirement,
>>     that it support NetRexx in the Open Source Way.
>>
>>     -Chip-
>>
>>
>>     On 2/5/10 12:57 David Requena said:
>>
>>         Chip,
>>
>>         Yours is more or less my own point. I'd like to express a couple
>>         slight precisions though.
>>
>>         First, I believe where mixing 2 very distinct concepts here:
>>
>>         - Open source NetRexx. This is RexxLA's concern and I think it
>>         will be obviously benefiting from "the REXX way".
>>         - Open source NetRexx user code/resources. This is the NetRexx
>>         language users community concern and is benefiting from "a
>>         bookmark file full of broken or obsolete links" right now.
>>
>>         I for one have been talking about the second point the whole
>>         time. We, as the user base of the language, would greatly
>>         benefit from a consolidation of the current array of dispersed
>>         (and rapidly vanishing into oblivion) of netrexx resources on
>>         the net.
>>
>>         Second, just because a given language's user community doesn't
>>         has a direct role in the language design, implementation, etc.
>>         doesn't mean it may have an important influence on that
>>         language's ultimate fate. Any effort made to help growing a
>>         strong, enthusiastic, user community is just another brick
>>         thrown at the foundation of the tower (pun intended ;-) ) we all
>>         are trying to build. That would be "the current times way"...
>>
>>
>>         2010/2/5 Chip Davis <[hidden email] <mailto:[hidden email]>
>>         <mailto:[hidden email] <mailto:[hidden email]>>>
>>
>>
>>            Tom,
>>
>>            I'm sure Ren? and those working diligently to effect the
>>         transfer of
>>            NetRexx from IBM to RexxLA appreciate the enthusiasm that
>> you,
>>            David, Kermit, and others have demonstrated for this project,
>>         but I
>>            think we all need to slow down and take a breath.
>>
>>            Our baby is not yet walking and you are already discussing
>> which
>>            motorcycle to buy for him. :-)
>>
>>            I think a couple of things should be self-evident: we have
>> some
>>            world-class intellectual resources within RexxLA (Ren? and
>>         Mike, for
>>            starters), we have the experience of having done the same
>> process
>>            with ooRexx, and we have a passionate (rabid?) group of
>> NetRexx
>>            enthusiasts trying to wait for Christmas morning.
>>
>>            There are many fundamental issues that need to be addressed
>>         by the
>>            NetRexx team and the RexxLA Board and Officers.  Believe me,
>>         we have
>>            a lot more questions than you do, many of which are quite
>> prosaic
>>            and unglamorous, but crucial to the success of this
>> project.  We
>>            learned a lot from the ooRexx project and will incorporate
>> those
>>            lessons into the organization and charter of the NetRexx
>> project.
>>
>>            At this point, fixed written-in-stone specifics are hard to
>>         come by,
>>            so I hope you (and everyone else chomping at the bit) will
>> take
>>            comfort in the philosophy to which all of us are
>> committed, that
>>            being the fundamental tenets of The Rexx Language, as laid
>> out by
>>            Mike in TRL and TNRL.  Ren? said as much in his answer to
>> David's
>>            fourth question, when he referred to following The Rexx
>> Way (my
>>            caps) of Discuss, Agree, Document, then Implement.
>>
>>            Thus, if you haven't seen any Discussion of a particular
>>         issue, it
>>            probably means we haven't gotten to that stage yet.
>>
>>            I would suggest that you not make any decisions about various
>>         code
>>            or doc repositories until we figure out where and how the
>> NetRexx
>>            code will be housed.  It will be much less confusing for
>> users to
>>            have only one location to find NetRexx resources, than a
>> bookmark
>>            file full of broken or obsolete links.
>>
>>            IMHO, of course.
>>
>>            -Chip-
>>
>>            On 2/5/10 05:22 Thomas Schneider said:
>>             >   to follow UP the idea of putting *any and all* Utilities
>>            (including mine)
>>             > to www.NetRexx.org <http://www.NetRexx.org>
>>         <http://www.NetRexx.org>:
>>
>>             >
>>             > What does this mean for my (private) directory structure ?
>>             >
>>             > When you like, I can send you my current one, but I am not
>>         so much
>>             > satisfied
>>             > with what I do have, and will need (please)
>>             >
>>             > your advise:
>>             >
>>             > How could/should we structure the whole project
>>         www.NetRexx.org <http://www.NetRexx.org>
>>            <http://www.NetRexx.org> ??
>>
>>             >
>>             > Your advise will be welcome ...
>>
>>
>>            On 2/5/10 03:48 Thomas Schneider said:
>>
>>                I am currently in the process to buildup the new
>>                www.Rexx2Nrx.com <http://www.Rexx2Nrx.com>
>>         <http://www.Rexx2Nrx.com> homepage.
>>
>>
>>                Might it be that this would be a good instance to
>> store all
>>                those tools available in a central
>>                place ?
>>
>>                Kermit Kiser schrieb:
>>
>>                    I have seen a lot of RFEs floating around but no safe
>>         place
>>                    to store them all or discuss them. Could you do
>> something
>>                    like that with your site?
>>
>>                    David Requena wrote:
>>
>>                        The thing is...  for the first time I got the
>> feeling
>>                        this process is actually advancing in some
>> direction.
>>                        We all, as NetRexx developers hoping to see the
>>         language
>>                        advance, are eagerly awaiting for the
>>                        culmination of a process about which we have near
>>         to no
>>                        information.
>>
>>
>>
>>
>>            _______________________________________________
>>            Ibm-netrexx mailing list
>>            [hidden email]
>>         <mailto:[hidden email]>
>>         <mailto:[hidden email]
>>         <mailto:[hidden email]>>
>>
>>
>>
>>
>>         --         Saludos / Regards,
>>         David Requena
>>
>>
>>        
>> ------------------------------------------------------------------------
>>
>>
>>         _______________________________________________
>>         Ibm-netrexx mailing list
>>         [hidden email] <mailto:[hidden email]>
>>
>>
>>
>>     _______________________________________________
>>     Ibm-netrexx mailing list
>>     [hidden email] <mailto:[hidden email]>
>>
>>
>>
>>
>> --
>> Saludos / Regards,
>> David Requena
>>
>>
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> Ibm-netrexx mailing list
>> [hidden email]
>>
>
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>

[rvjansen]

just to make the example standalone and show that you can access the methods and signatures using standard Java means, this more complete example, that is runnable by

nrc -exec -nojava Dynamic

--------------------------

options binary
import COM.ibm.netrexx.process.NetRexxA

class Dynamic

  interpreter = NetRexxA()
  /**
   * Default constructor
   */
  method Dynamic()

  method interpret(file=String) returns Class
    files=[file]
    classname = Rexx file
    classname = classname.substr(1,classname.pos('.')-1)
    say classname
    flags=['nocrossref', 'verbose0']
    interpreter.parse(files, flags)
    return interpreter.getClassObject(null, classname) -- find the hello Class
   
  method main(args=String[]) static
    t = Dynamic()
    source = Arraylist()
    source.add(Rexx "options binary")
    source.add(Rexx "class TestDyn")
    source.add(Rexx "method TestDyn(); say 'a new TestDyn'")
    source.add(Rexx "method SaySomethingElse(); say 'Hello New World'")
    source.add(Rexx "method main(args = String[]) static")
    source.add(Rexx "say 'Hello Dynamic NetRexx World'")

    writeFileFromArrayList('TestDyn.nrx', source)
   
    dynamicClass = t.interpret('TestDyn.nrx')

    -- write the methods and their signatures
    meths = Method[] dynamicClass.getDeclaredMethods()
    say "/* --------- methods ------------*/"
    loop i=0 to meths.length-1
      say "method" meths[i].getName()||'(\-'
      parms = Class[] meths[i].getParameterTypes()
      if parms.length = 0 then say ') \-'
      loop l=0 to parms.length-1
        say ' '||((Rexx l)+'a'.c2d()).d2c()||'='||parms[l].getName()||"\-"
        if l<parms.length-1 then say ', \-'
        else say ') \-'
      end -- loop
     
      mods = meths[i].getModifiers()
      say Modifier.toString(mods) '\-'
     
      say "returns" meths[i].getReturnType().getName()
     
      exeps = Class[] meths[i].getExceptionTypes()
      loop m=0 to exeps.length-1
        say "signals \-"
        say exeps[m] ',' "\-"
      end -- loop
     
    end
   
  method WriteFileFromArrayList(filename=Rexx, v=ArrayList) static
    do
      o = PrintWriter(BufferedWriter(FileWriter(filename)))
    loop i = 0  to v.size()-1
      line = rexx v.get(i)
      o.println(line)
    end
    o.close()
    catch IOException
      say 'exception in handling file' filename
    end

----------------------------

best regards,

Ren?.

[David Requena]

Kermit,

I wouldn't really know. When you read Ren?'s Document you find some hints
under the "Possibles" section:

Getting closer to Open Object Rexx
Especially the dynamic aspects - adding classes, methods at runtime
Might need to use the InvokeDynamic stuff here - this will bump up the
minimum Java version
req from 1.1.2 to 7!

The doc is at http://wiki.jvmlangsummit.com/pdf/33_Jansen_netrexx.pdf

I think Ren? did well moving this topic to a separate thread so I'll
elaborate there.

Anyway my point is that it's abvious now that NetRexx has continued to
evolve behind scenes. When it's finally released it will e already a
different product to what we have now.
So better we wait to see what the final outcome is prior to start proposing
language enhancement which may well be obsolete or irrelevant by that time.


2010/2/6 Kermit Kiser <[hidden email]>

>  Maybe I am a little dense but what do you mean by "dynamic features"? This
> code is basically what I use in my scripting modules, NetRexxScript and
> NetRexxJSR223 although it is missing a couple of steps that are needed to
> locate the main method and invoke it. (More if you need to isolate the
> dynamic code in a separate thread. I also transparently add suitable
> environment variables while writing the code to the disk file.) Is the
> objective to dynamically generate NetRexx code and execute it from within a
> compiled program? That is basically what JSR223 does.
>
> If the objective is to dynamically create Java byte code and add it to a
> library of compiled classes in memory, then you would need to use the
> NetRexx translator to produce Java code and pass it to the JSR199 compiler
> API to generate byte code. (I have not used JSR199 so I don't know if you
> need a custom classloader also to do this, but having done a classloader in
> NetRexx, I can verify that it is possible.) Am I even close to what you guys
> are talking about?
>
> -- Kermit
>
>
> David Requena wrote:
>
> Nice trick Ren?!
>
>  I've not much experience with the interpreter but IIRC from de docs
> wouldn't parsing the new class reinitialize the interpreter?
> Would this code work would it be run from the interpreter itself? hmmm..
> well.. that would be a separate interpreter instance...
>
>  but you couldn't re utilize a NetRexA instance without invalidating
> references to a previously parsed class, right?
> hmm.. feels a little bit clunky..
>
>  I've always wondered why Mike chosed to implement a filename based API
> for the interpreter. It would have been so much more convenient should he
> had chosen iostreams... we wouldn't have to go through the hops of temporary
> files!
>
> 2010/2/6 Ren? Jansen <[hidden email]>
>
>> David,
>>
>> when I spoke to Mike the last Rexx Language Symposium last year, we
>> briefly discussed dynamic features, and he mentioned that in his opinion the
>> interpreter api in the current public version would already allow for this.
>> I tried it some time ago, and this is what I came up with:
>>
>>
>> options binary
>> import COM.ibm.netrexx.process.NetRexxA
>> /**
>>  * Class Dynamic implements...
>>  * <BR>
>>  * Created on: di, 16, jun 2009 22:37:53 +0200
>>  */
>> class Dynamic extends SourceTransformer  -- take that out later
>>
>>  interpreter = NetRexxA()
>>  /**
>>   * Default constructor
>>   */
>>  method Dynamic()
>>
>>  method interpret(file=String) returns Class
>>    files=[file]
>>    classname = Rexx file
>>    classname = classname.substr(1,classname.pos('.')-1)
>>    say classname
>>    flags=['nocrossref', 'verbose0']
>>    interpreter.parse(files, flags)
>>    return interpreter.getClassObject(null, classname) -- find the hello
>> Class
>>
>>  method main(args=String[]) static
>>    t = Dynamic()
>>    source = Arraylist()
>>    source.add(Rexx "options binary")
>>    source.add(Rexx "class TestDyn")
>>    source.add(Rexx "method main(args = String[]) static")
>>    source.add(Rexx "say 'Hello Dynamic NetRexx World'")
>>
>>    writeFileFromArrayList('TestDyn.nrx', source)
>>
>>    dynamicClass = t.interpret('TestDyn.nrx')
>>
>>    say dynamicClass
>>
>>
>> note that I never took out SourceTransformer, but that class is only used
>> to call writeFileFromArrayList - I am sure you can write a file yourself.
>>
>> Just as an illustration that I went complaining about the static nature of
>> NetRexx for years without understanding the wealth that is already in there.
>> One could make an API out of that without the need for changing the core
>> functionality.
>>
>> best regards,
>>
>> Ren?.
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> >
>> > PS: Kermit I've just become aware of how moot the RFE talk is right now.
>> According to proceedings of Ren?'s presentation at JVM Language Summit 2008
>> there already existed a NetRexx v. 3.00 back in September 2008 (which wasn't
>> out yet ;-) ). There is even an interesting 'recent additions section'!
>> >
>> > In the meantime they could even have included planned dynamic features!
>> We'll have to wait till a release is made to see if any RFE's still apply.
>>
>>
>>  _______________________________________________
>> Ibm-netrexx mailing list
>> [hidden email]
>>
>>
>
>
> --
> Saludos / Regards,
> David Requena
>
>  ------------------------------
>
> _______________________________________________
> Ibm-netrexx mailing [hidden email]
>
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>
>

--
Saludos / Regards,
David Requena
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ns.hursley.ibm.com/pipermail/ibm-netrexx/attachments/20100208/d13e7d7d/attachment-0001.html

[David Requena]

Chip,

It's been a greatly rewarding experience reading your reply!

You're right in the nature of the medium (being a non native English speaker
helps a lot too) but then it also has its advantages. We can discuss
sensible topics at any needed length so we all get finally enriched.

I'm glad we now fully understand each other. Moreover, I'm happy we fully
agree!



2010/2/8 David Requena <[hidden email]>

> Kermit,
>
> I wouldn't really know. When you read Ren?'s Document you find some hints
> under the "Possibles" section:
>
>  Getting closer to Open Object Rexx
> Especially the dynamic aspects - adding classes, methods at runtime
> Might need to use the InvokeDynamic stuff here - this will bump up the
> minimum Java version
> req from 1.1.2 to 7!
>
> The doc is at http://wiki.jvmlangsummit.com/pdf/33_Jansen_netrexx.pdf
>
> I think Ren? did well moving this topic to a separate thread so I'll
> elaborate there.
>
> Anyway my point is that it's abvious now that NetRexx has continued to
> evolve behind scenes. When it's finally released it will e already a
> different product to what we have now.
> So better we wait to see what the final outcome is prior to start proposing
> language enhancement which may well be obsolete or irrelevant by that time.
>
>
> 2010/2/6 Kermit Kiser <[hidden email]>
>
>  Maybe I am a little dense but what do you mean by "dynamic features"?
>> This code is basically what I use in my scripting modules, NetRexxScript and
>> NetRexxJSR223 although it is missing a couple of steps that are needed to
>> locate the main method and invoke it. (More if you need to isolate the
>> dynamic code in a separate thread. I also transparently add suitable
>> environment variables while writing the code to the disk file.) Is the
>> objective to dynamically generate NetRexx code and execute it from within a
>> compiled program? That is basically what JSR223 does.
>>
>> If the objective is to dynamically create Java byte code and add it to a
>> library of compiled classes in memory, then you would need to use the
>> NetRexx translator to produce Java code and pass it to the JSR199 compiler
>> API to generate byte code. (I have not used JSR199 so I don't know if you
>> need a custom classloader also to do this, but having done a classloader in
>> NetRexx, I can verify that it is possible.) Am I even close to what you guys
>> are talking about?
>>
>> -- Kermit
>>
>>
>> David Requena wrote:
>>
>>  Nice trick Ren?!
>>
>> I've not much experience with the interpreter but IIRC from de docs
>> wouldn't parsing the new class reinitialize the interpreter?
>> Would this code work would it be run from the interpreter itself? hmmm..
>> well.. that would be a separate interpreter instance...
>>
>> but you couldn't re utilize a NetRexA instance without invalidating
>> references to a previously parsed class, right?
>> hmm.. feels a little bit clunky..
>>
>> I've always wondered why Mike chosed to implement a filename based API for
>> the interpreter. It would have been so much more convenient should he
>> had chosen iostreams... we wouldn't have to go through the hops of temporary
>> files!
>>
>> 2010/2/6 Ren? Jansen <[hidden email]>
>>
>>> David,
>>>
>>> when I spoke to Mike the last Rexx Language Symposium last year, we
>>> briefly discussed dynamic features, and he mentioned that in his opinion the
>>> interpreter api in the current public version would already allow for this.
>>> I tried it some time ago, and this is what I came up with:
>>>
>>>
>>> options binary
>>> import COM.ibm.netrexx.process.NetRexxA
>>> /**
>>>  * Class Dynamic implements...
>>>  * <BR>
>>>  * Created on: di, 16, jun 2009 22:37:53 +0200
>>>  */
>>> class Dynamic extends SourceTransformer  -- take that out later
>>>
>>>  interpreter = NetRexxA()
>>>  /**
>>>   * Default constructor
>>>   */
>>>  method Dynamic()
>>>
>>>  method interpret(file=String) returns Class
>>>    files=[file]
>>>    classname = Rexx file
>>>    classname = classname.substr(1,classname.pos('.')-1)
>>>    say classname
>>>    flags=['nocrossref', 'verbose0']
>>>    interpreter.parse(files, flags)
>>>    return interpreter.getClassObject(null, classname) -- find the hello
>>> Class
>>>
>>>  method main(args=String[]) static
>>>    t = Dynamic()
>>>    source = Arraylist()
>>>    source.add(Rexx "options binary")
>>>    source.add(Rexx "class TestDyn")
>>>    source.add(Rexx "method main(args = String[]) static")
>>>    source.add(Rexx "say 'Hello Dynamic NetRexx World'")
>>>
>>>    writeFileFromArrayList('TestDyn.nrx', source)
>>>
>>>    dynamicClass = t.interpret('TestDyn.nrx')
>>>
>>>    say dynamicClass
>>>
>>>
>>> note that I never took out SourceTransformer, but that class is only used
>>> to call writeFileFromArrayList - I am sure you can write a file yourself.
>>>
>>> Just as an illustration that I went complaining about the static nature
>>> of NetRexx for years without understanding the wealth that is already in
>>> there. One could make an API out of that without the need for changing the
>>> core functionality.
>>>
>>> best regards,
>>>
>>> Ren?.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> >
>>> > PS: Kermit I've just become aware of how moot the RFE talk is right
>>> now. According to proceedings of Ren?'s presentation at JVM Language Summit
>>> 2008 there already existed a NetRexx v. 3.00 back in September 2008 (which
>>> wasn't out yet ;-) ). There is even an interesting 'recent additions
>>> section'!
>>> >
>>> > In the meantime they could even have included planned dynamic features!
>>> We'll have to wait till a release is made to see if any RFE's still apply.
>>>
>>>
>>>  _______________________________________________
>>> Ibm-netrexx mailing list
>>> [hidden email]
>>>
>>>
>>
>>
>> --
>> Saludos / Regards,
>> David Requena
>>
>> ------------------------------
>>
>> _______________________________________________
>> Ibm-netrexx mailing [hidden email]
>>
>>
>> _______________________________________________
>> Ibm-netrexx mailing list
>> [hidden email]
>>
>>
>>
>
>
> --
> Saludos / Regards,
> David Requena
>
>

--
Saludos / Regards,
David Requena
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ns.hursley.ibm.com/pipermail/ibm-netrexx/attachments/20100208/2de34d60/attachment.html

[David Requena]

Ren?,

The intent was clear with the first instance of the example. Thanks for
supplying a readily runnable version.

Of course this is dynamic. It's dynamically creating a class on the fly then
accessing it. Some limitations aplly however if I understanf correctly
(please, correct me when not):

- You'd need to instantiate a NetRexxA object for each dynamic class if you
want refferences to the Class objects to remain valid.
- The NetRexx interpreter is not part of the NetRexx runtime package but
from the language processor one.
- You need to go through the hops of writting temp nrx files then having the
interpreter reading and parsing them. This one is really hindering tools
taking advantage of the interpreter such as Kermit's NetRexxScript. Surely
the interpreter is instantiating a FileInputStream to read the file. It
would have been so much more convenient had it oferred an InpuStream
interface.

The problem is the term 'dynamic' incarnates such a brad concept. When I
think of 'dynamic features of a language' I have in mind something more
along the lines of being able to access an existing class at runtime and
manipulating its methods and properties. Maybe even replacing existing
classes of already instantiated objects through some mechanism analoge to
objective-c's categories.

2010/2/6 Ren? Jansen <[hidden email]>

> just to make the example standalone and show that you can access the
> methods and signatures using standard Java means, this more complete
> example, that is runnable by
>
> nrc -exec -nojava Dynamic
>
> --------------------------
>
> options binary
> import COM.ibm.netrexx.process.NetRexxA
>
> class Dynamic
>
>  interpreter = NetRexxA()
>  /**
>   * Default constructor
>   */
>  method Dynamic()
>
>  method interpret(file=String) returns Class
>    files=[file]
>    classname = Rexx file
>    classname = classname.substr(1,classname.pos('.')-1)
>    say classname
>    flags=['nocrossref', 'verbose0']
>    interpreter.parse(files, flags)
>    return interpreter.getClassObject(null, classname) -- find the hello
> Class
>
>  method main(args=String[]) static
>    t = Dynamic()
>    source = Arraylist()
>    source.add(Rexx "options binary")
>    source.add(Rexx "class TestDyn")
>    source.add(Rexx "method TestDyn(); say 'a new TestDyn'")
>    source.add(Rexx "method SaySomethingElse(); say 'Hello New World'")
>    source.add(Rexx "method main(args = String[]) static")
>    source.add(Rexx "say 'Hello Dynamic NetRexx World'")
>
>    writeFileFromArrayList('TestDyn.nrx', source)
>
>    dynamicClass = t.interpret('TestDyn.nrx')
>
>    -- write the methods and their signatures
>    meths = Method[] dynamicClass.getDeclaredMethods()
>    say "/* --------- methods ------------*/"
>    loop i=0 to meths.length-1
>      say "method" meths[i].getName()||'(\-'
>      parms = Class[] meths[i].getParameterTypes()
>      if parms.length = 0 then say ') \-'
>      loop l=0 to parms.length-1
>        say ' '||((Rexx l)+'a'.c2d()).d2c()||'='||parms[l].getName()||"\-"
>        if l<parms.length-1 then say ', \-'
>        else say ') \-'
>      end -- loop
>
>      mods = meths[i].getModifiers()
>      say Modifier.toString(mods) '\-'
>
>      say "returns" meths[i].getReturnType().getName()
>
>      exeps = Class[] meths[i].getExceptionTypes()
>      loop m=0 to exeps.length-1
>        say "signals \-"
>        say exeps[m] ',' "\-"
>      end -- loop
>
>    end
>
>  method WriteFileFromArrayList(filename=Rexx, v=ArrayList) static
>    do
>      o = PrintWriter(BufferedWriter(FileWriter(filename)))
>    loop i = 0  to v.size()-1
>      line = rexx v.get(i)
>      o.println(line)
>    end
>    o.close()
>    catch IOException
>      say 'exception in handling file' filename
>    end
>
> ----------------------------
>
> best regards,
>
> Ren?.
>
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>

--
Saludos / Regards,
David Requena
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ns.hursley.ibm.com/pipermail/ibm-netrexx/attachments/20100208/4db024fc/attachment-0001.html

[Thomas.Schneider.Wien]

Hello Kermit & all,
   ... thank you *so much for forwarding Rene's document*.

I didn't know it, as I have been *not on this group* for a while  :-(

Tom.
=====================================================.
David Requena schrieb:

> Kermit,
>
> I wouldn't really know. When you read Ren?'s Document you find some
> hints under the "Possibles" section:
>
> Getting closer to Open Object Rexx
> Especially the dynamic aspects - adding classes, methods at runtime
> Might need to use the InvokeDynamic stuff here - this will bump up the
> minimum Java version
> req from 1.1.2 to 7!
>
> The doc is at http://wiki.jvmlangsummit.com/pdf/33_Jansen_netrexx.pdf
>
> I think Ren? did well moving this topic to a separate thread so I'll
> elaborate there.
>
> Anyway my point is that it's abvious now that NetRexx has continued to
> evolve behind scenes. When it's finally released it will e already a
> different product to what we have now.
> So better we wait to see what the final outcome is prior to start
> proposing language enhancement which may well be obsolete
> or irrelevant by that time.
>
>
> 2010/2/6 Kermit Kiser <[hidden email]
> <mailto:[hidden email]>>
>
>     Maybe I am a little dense but what do you mean by "dynamic
>     features"? This code is basically what I use in my scripting
>     modules, NetRexxScript and NetRexxJSR223 although it is missing a
>     couple of steps that are needed to locate the main method and
>     invoke it. (More if you need to isolate the dynamic code in a
>     separate thread. I also transparently add suitable environment
>     variables while writing the code to the disk file.) Is the
>     objective to dynamically generate NetRexx code and execute it from
>     within a compiled program? That is basically what JSR223 does.
>
>     If the objective is to dynamically create Java byte code and add
>     it to a library of compiled classes in memory, then you would need
>     to use the NetRexx translator to produce Java code and pass it to
>     the JSR199 compiler API to generate byte code. (I have not used
>     JSR199 so I don't know if you need a custom classloader also to do
>     this, but having done a classloader in NetRexx, I can verify that
>     it is possible.) Am I even close to what you guys are talking about?
>
>     -- Kermit
>
>
>     David Requena wrote:
>>     Nice trick Ren?!
>>
>>     I've not much experience with the interpreter but IIRC from de
>>     docs wouldn't parsing the new class reinitialize the interpreter?
>>     Would this code work would it be run from the interpreter itself?
>>     hmmm.. well.. that would be a separate interpreter instance...
>>
>>     but you couldn't re utilize a NetRexA instance without
>>     invalidating references to a previously parsed class, right?
>>     hmm.. feels a little bit clunky..
>>
>>     I've always wondered why Mike chosed to implement a filename
>>     based API for the interpreter. It would have been so much more
>>     convenient should he had chosen iostreams... we wouldn't have to
>>     go through the hops of temporary files!
>>
>>     2010/2/6 Ren? Jansen <[hidden email] <mailto:[hidden email]>>
>>
>>         David,
>>
>>         when I spoke to Mike the last Rexx Language Symposium last
>>         year, we briefly discussed dynamic features, and he mentioned
>>         that in his opinion the interpreter api in the current public
>>         version would already allow for this. I tried it some time
>>         ago, and this is what I came up with:
>>
>>
>>         options binary
>>         import COM.ibm.netrexx.process.NetRexxA
>>         /**
>>          * Class Dynamic implements...
>>          * <BR>
>>          * Created on: di, 16, jun 2009 22:37:53 +0200
>>          */
>>         class Dynamic extends SourceTransformer  -- take that out later
>>
>>          interpreter = NetRexxA()
>>          /**
>>           * Default constructor
>>           */
>>          method Dynamic()
>>
>>          method interpret(file=String) returns Class
>>            files=[file]
>>            classname = Rexx file
>>            classname = classname.substr(1,classname.pos('.')-1)
>>            say classname
>>            flags=['nocrossref', 'verbose0']
>>            interpreter.parse(files, flags)
>>            return interpreter.getClassObject(null, classname) -- find
>>         the hello Class
>>
>>          method main(args=String[]) static
>>            t = Dynamic()
>>            source = Arraylist()
>>            source.add(Rexx "options binary")
>>            source.add(Rexx "class TestDyn")
>>            source.add(Rexx "method main(args = String[]) static")
>>            source.add(Rexx "say 'Hello Dynamic NetRexx World'")
>>
>>            writeFileFromArrayList('TestDyn.nrx', source)
>>
>>            dynamicClass = t.interpret('TestDyn.nrx')
>>
>>            say dynamicClass
>>
>>
>>         note that I never took out SourceTransformer, but that class
>>         is only used to call writeFileFromArrayList - I am sure you
>>         can write a file yourself.
>>
>>         Just as an illustration that I went complaining about the
>>         static nature of NetRexx for years without understanding the
>>         wealth that is already in there. One could make an API out of
>>         that without the need for changing the core functionality.
>>
>>         best regards,
>>
>>         Ren?.
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>         >
>>         > PS: Kermit I've just become aware of how moot the RFE talk
>>         is right now. According to proceedings of Ren?'s presentation
>>         at JVM Language Summit 2008 there already existed a NetRexx
>>         v. 3.00 back in September 2008 (which wasn't out yet ;-) ).
>>         There is even an interesting 'recent additions section'!
>>         >
>>         > In the meantime they could even have included planned
>>         dynamic features! We'll have to wait till a release is made
>>         to see if any RFE's still apply.
>>
>>
>>         _______________________________________________
>>         Ibm-netrexx mailing list
>>         [hidden email] <mailto:[hidden email]>
>>
>>
>>
>>
>>     --
>>     Saludos / Regards,
>>     David Requena
>>
>>     ------------------------------------------------------------------------
>>     _______________________________________________ Ibm-netrexx
>>     mailing list [hidden email]
>>     <mailto:[hidden email]>
>
>     _______________________________________________
>     Ibm-netrexx mailing list
>     [hidden email] <mailto:[hidden email]>
>
>
>
>
>
> --
> Saludos / Regards,
> David Requena
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>