[Fwd: [Ibm-netrexx] Bug in NetRexx compiler (was: Problem extending PrintStream)]

[alansam]

IANAL.  However, if in legal fact IBM publicly announced in April 2008 their intention to open the source code for NetRexx, then aren't they accountable for their failure to do so in a timely fashion?

Certainly some of us have been harmed by that failure.

Considering the time delay and the probability that there is no question regarding the origin or ownership of the code, there is apparently a lack of motivation somewhere.

Not necessarily a lack of motivation; more likely just a lack of resources (and other priorities). I'm sure IBM Legal will have suffered from downsizing just as much as other parts of the corporation (and the economy in general).  In the scheme of things I doubt that NetRexx is a high-priority item.

I am curious though.  What harm has anyone suffered by the slowness of the process?

Alan.

--
Can't tweet, won't tweet!

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[billfen]

  On 8/18/2010 1:03 PM, Rick McGuire wrote:
> The announcement of intent was made by RexxLA, not IBM.  Any
> announcement of that sort coming directly from IBM would have carried
> all sorts of caveats about the possibility that it might not happen.
The RexxLA site says "NetRexx to be Open Source!  IBM announced in
April, 2008 their intention of making available their NetRexx product."

Certainly IBM was well aware of this statement, and by not refuting it
have made the announcement by default.

Should we write to the IBM Legal department and ask for clarification?  
They are the final word on what IBM has said publicly.

billfen

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[alansam]

On 18 August 2010 11:08, Bill Fenlason <[hidden email]> wrote:

Should we write to the IBM Legal department and ask for clarification?  They are the final word on what IBM has said publicly.

As Rene said way back in February:

"Gentlemen,

this is exactly what we don't need now. I strongly suggest you hold your horses - para el carro. Ahora."

Alan.

--

Can't tweet, won't tweet!

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[Robert L Hamilton]

There are 2 local colleges and 3 local businesses interested in nREXX/JAVA but are hung on the ownership situation.  They don't want to work with something owned by IBM and I don't blame them.

Good luck with the letter. . .


bobh

On Tue, Aug 17, 2010 at 3:38 PM, Bill Fenlason <[hidden email]> wrote:

 Bob,

I agree - my work on an Eclipse plugin is stalled as well.

Either IBM has turned the code over to RexxLA or they haven't.  It is either IBM  _or_  RexxLA that is holding it up, but not both.

In a private email on July 10th,  I asked the RexxLA president (Rene Jansen) which it was.  Unfortunately, I have not received a response yet.

I'm now drafting a formal (snail mail) letter to IBM Corporate requesting a public statement  - perhaps they will respond.  Since IBM owns the code, and as best I can tell IBM announced their intention to open source it in April 2008, they are primarily responsible for this situation.

billfen


On 8/17/2010 3:24 PM, Robert Hamilton wrote:

As an aside to these problems and concerns, I have two projects in limbo over the nRexx ownership issue. It would help if IBM would just publish an estimated date of ownership xfer. Or, even just state the open issues on the deal; just any kind of info from IBM.

bobh

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[rvjansen]

Alan,

thank you.

best regards,

René Vincent Jansen.

On 18 aug 2010, at 20:36, Alan Sampson wrote:

On 18 August 2010 11:08, Bill Fenlason <[hidden email]> wrote:

Should we write to the IBM Legal department and ask for clarification?  They are the final word on what IBM has said publicly.

As Rene said way back in February:

"Gentlemen,

this is exactly what we don't need now. I strongly suggest you hold your horses - para el carro. Ahora."

Alan.

--

Can't tweet, won't tweet!
_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[alansam]

On 18 August 2010 11:43, Robert Hamilton <[hidden email]> wrote:

There are 2 local colleges and 3 local businesses interested in nREXX/JAVA but are hung on the ownership situation.  They don't want to work with something owned by IBM and I don't blame them.

How do they feel about working with all the other IBM owned "intellectual property" that's embedded in every PC?  (Not to mention similar IBM intellectual property embedded in and licensed by software vendors who's software they do use.)  Do they hold the same reservations with Microsoft, Oracle etc. intellectual property?  Note that Java itself is proprietary too.

NetRexx is free to download and use and the license is liberal.  This sounds like an excuse rather than a reason.

Alan.
--
Can't tweet, won't tweet!

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[Bruce Skelly]

The problem should not be that IBM owns it, but that it is currently unsupported.

On Aug 18, 2010, at 11:43 AM, Robert Hamilton wrote:

There are 2 local colleges and 3 local businesses interested in nREXX/JAVA but are hung on the ownership situation.  They don't want to work with something owned by IBM and I don't blame them.

Good luck with the letter. . .


bobh

On Tue, Aug 17, 2010 at 3:38 PM, Bill Fenlason <[hidden email]> wrote:

 Bob,

I agree - my work on an Eclipse plugin is stalled as well.

Either IBM has turned the code over to RexxLA or they haven't.  It is either IBM  _or_  RexxLA that is holding it up, but not both.

In a private email on July 10th,  I asked the RexxLA president (Rene Jansen) which it was.  Unfortunately, I have not received a response yet.

I'm now drafting a formal (snail mail) letter to IBM Corporate requesting a public statement  - perhaps they will respond.  Since IBM owns the code, and as best I can tell IBM announced their intention to open source it in April 2008, they are primarily responsible for this situation.

billfen


On 8/17/2010 3:24 PM, Robert Hamilton wrote:

As an aside to these problems and concerns, I have two projects in limbo over the nRexx ownership issue. It would help if IBM would just publish an estimated date of ownership xfer. Or, even just state the open issues on the deal; just any kind of info from IBM.

bobh

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[Aviatrexx]

[Possibly futile attempt to close this thread, or at least extract it
from a hopelessly convoluted cross-threaded Subject: line.]

I'm with Mike, Rick, and the rest of the responders who have intimate
experience in large corporations.  If any of the IBM parties involved
in this effort wanted to issue any sort of official progress report to
the outside world (us), such a statement could be interpreted as a
binding offer and used as the basis for legal action.  They would then
be required to surmount numerous hurdles the legal department would
throw in their way.

The effect of the above is that usually there are only two forms of
such an announcement: "We have done this" or "We are not going to do
this".

Frankly, this is an excellent example of "No news is good news".  As
much as we ALL would like to know when something is going to happen
with NetRexx, I'm not willing to risk annoying the IBMPTB to the point
that the easy answer is "No".

I don't know the situation in Bob's, Bill's, or Peter's corporate
world, but everyone I know is also doing the work of their two or
three colleagues that have been laid off.  I have to believe that, for
whoever is working on it, the open-sourcing of NetRexx is on a
time-available basis.

Remember: Patience is a Virtue... :-)

-Chip-

On 8/18/10 17:03 Rick McGuire said:
 > On Wed, Aug 18, 2010 at 12:54 PM, Bill Fenlason
<[hidden email]> wrote:
 >> On 8/18/2010 9:17 AM, Rick McGuire wrote:
 >>> Strangely enough, the silence is actually a good sign.  It means the
 >>> process is proceeding.  If this was not going to happen, there would
 >>> be a clear statement from IBM to have effect.  While the legal
 >>> clearances are taking place, details of the progress will be IBM
 >>> confidential.  No statement will come from IBM until the release is
 >>> certain...and that determination can only be made at the very end of
 >>> the process.
 >> IANAL.  However, if in legal fact IBM publicly announced in April
2008 their
 >> intention to open the source code for NetRexx, then aren't they
accountable
 >> for their failure to do so in a timely fashion?
 >
 > The announcement of intent was made by RexxLA, not IBM.  Any
 > announcement of that sort coming directly from IBM would have carried
 > all sorts of caveats about the possibility that it might not happen.
 >
 >> Certainly some of us have been harmed by that failure.
 >>
 >> Considering the time delay and the probability that there is no
question
 >> regarding the origin or ownership of the code, there is apparently
a lack of
 >> motivation somewhere.
 >
 > Even lawyers are a finite resource within IBM, and the particular
 > attorneys that would handle this type of effort have as their first
 > priority managing requests for products that actually make IBM money.
 > The easy answer for something like this is to just say NO.  The fact
 > this has taken this long means that a real due-diligence effort is
 > taking place to ensure that this code can be cleanly released.

On 8/17/10 20:38 Bill Fenlason said:
  > Either IBM has turned the code over to RexxLA or they haven't.  It is

> either IBM  _or_  RexxLA that is holding it up, but not both.
>
> In a private email on July 10th,  I asked the RexxLA president (Rene
> Jansen) which it was.  Unfortunately, I have not received a response yet.
>
> I'm now drafting a formal (snail mail) letter to IBM Corporate
> requesting a public statement  - perhaps they will respond.  Since IBM
> owns the code, and as best I can tell IBM announced their intention to
> open source it in April 2008, they are primarily responsible for this
> situation.
>
> On 8/17/2010 3:24 PM, Robert Hamilton wrote:
>> As an aside to these problems and concerns, I have two projects in
>> limbo over the nRexx ownership issue. It would help if IBM would just
>> publish an estimated date of ownership xfer. Or, even just state the
>> open issues on the deal; just any kind of info from IBM.

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[billfen]

On 8/18/2010 2:36 PM, Alan Sampson wrote:

On 18 August 2010 11:08, Bill Fenlason <[hidden email]> wrote:

Should we write to the IBM Legal department and ask for clarification?  They are the final word on what IBM has said publicly.

As Rene said way back in February:

"Gentlemen,

this is exactly what we don't need now. I strongly suggest you hold your horses - para el carro. Ahora."

February was SIX MONTHS AGO!  April 2008 was TWENTY-EIGHT months ago!  If not now, when?

The seemingly endless calls for patience remind me of a fable involving a wolf....


_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[KP Kirchdörfer]

Am Mittwoch, 18. August 2010, 22:26:45 schrieben Sie:

>   On 8/18/2010 2:36 PM, Alan Sampson wrote:
> > On 18 August 2010 11:08, Bill Fenlason <[hidden email]
> >
> > <mailto:[hidden email]>> wrote:
> >     Should we write to the IBM Legal department and ask for
> >     clarification?  They are the final word on what IBM has said
> >     publicly.
> >
> > As Rene said way back in February:
> >
> > "Gentlemen,
> >
> > this is exactly what we don't need now. I strongly suggest you hold your
> > horses - para el carro. Ahora."
>
> February was SIX MONTHS AGO!  April 2008 was TWENTY-EIGHT months ago!
> If not now, when?
>
> The seemingly endless calls for patience remind me of a fable involving
> a wolf....

Or the Debian mantra "it's released, when it's ready".

Either you do it yourself (resolve the legal issues in this case) or you have
to be patient (like I have to, for example, with official Debian support for an
apache server with name based certificates support).
(And yes, I've spent years and still do, to build a router distro that fit my
needs, and I had a lot to learn  - I guess as much as you'd have to to solve
the outstanding legal issues for NetRexx becoming open source.)

Just complaining annoys everyone involved working on OSS.

I agree there is little space to add something to Netrexx right now; but at
least there are David's pages on kenai.com as an anchor, if you have to add
something for the time being.

kp
_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[alansam]

February was SIX MONTHS AGO!  April 2008 was TWENTY-EIGHT months ago!  If not now, when?

The seemingly endless calls for patience remind me of a fable involving a wolf....

I cannot speak for Rene and/or IBM but based on his latest reply (above) I assume that the process is moving along.  The problem with stirring the pot is that it's really easy for the folks in IBM Legal to just say "no" and walk away from the process.  Consider that it's much easier for them to do that than to proceed anyway.  If they perceive that the community are causing them problems, particularly with upper management, they are well within their rights to do just that.

While the squeaky wheel oft times gets the grease it can also lead to the whole cart being scrapped too.  As Rene hasn't indicated that there are problems with the process I for one am content to wait and not rock the boat.  Observe that back in February Rene indicated that "we don't need" interference.  There has to be a very good reason for that.

Personally, I'd hate to be responsible for the project being canned just because I'm getting impatient with IBM's process.  Are you prepared to be that person?

Alan.
--
Can't tweet, won't tweet!

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[George Hovey-2]

Amen.
George

On Wed, Aug 18, 2010 at 5:51 PM, Alan Sampson <[hidden email]> wrote:

February was SIX MONTHS AGO!  April 2008 was TWENTY-EIGHT months ago!  If not now, when?

The seemingly endless calls for patience remind me of a fable involving a wolf....

I cannot speak for Rene and/or IBM but based on his latest reply (above) I assume that the process is moving along.  The problem with stirring the pot is that it's really easy for the folks in IBM Legal to just say "no" and walk away from the process.  Consider that it's much easier for them to do that than to proceed anyway.  If they perceive that the community are causing them problems, particularly with upper management, they are well within their rights to do just that.

While the squeaky wheel oft times gets the grease it can also lead to the whole cart being scrapped too.  As Rene hasn't indicated that there are problems with the process I for one am content to wait and not rock the boat.  Observe that back in February Rene indicated that "we don't need" interference.  There has to be a very good reason for that.

Personally, I'd hate to be responsible for the project being canned just because I'm getting impatient with IBM's process.  Are you prepared to be that person?

Alan.
--
Can't tweet, won't tweet!

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[Rupp Peter - prupp]

I agree with Alan, so thank you for your words of wisdom.  Since, I’m the one who instigated this discussion, I probably should have have kept my personal thoughts to myself…instead of ruffling a lot of feathers.  The  last thing I want to do is interfere with IBM’s processes.     

In the meantime, netRexx (current) works well for me.  (We use it extensively here at Acxiom)

Thanks for your patience with me.

 

From: [hidden email] [mailto:[hidden email]] On Behalf Of George Hovey
Sent: Wednesday, August 18, 2010 7:43 PM
To: IBM Netrexx
Subject: Re: [Fwd: [Ibm-netrexx] Bug in NetRexx compiler(was:ProblemextendingPrintStream)]

 

Amen.
George

On Wed, Aug 18, 2010 at 5:51 PM, Alan Sampson <[hidden email]> wrote:

February was SIX MONTHS AGO!  April 2008 was TWENTY-EIGHT months ago!  If not now, when?

The seemingly endless calls for patience remind me of a fable involving a wolf....

 

I cannot speak for Rene and/or IBM but based on his latest reply (above) I assume that the process is moving along.  The problem with stirring the pot is that it's really easy for the folks in IBM Legal to just say "no" and walk away from the process.  Consider that it's much easier for them to do that than to proceed anyway.  If they perceive that the community are causing them problems, particularly with upper management, they are well within their rights to do just that.

While the squeaky wheel oft times gets the grease it can also lead to the whole cart being scrapped too.  As Rene hasn't indicated that there are problems with the process I for one am content to wait and not rock the boat.  Observe that back in February Rene indicated that "we don't need" interference.  There has to be a very good reason for that.

Personally, I'd hate to be responsible for the project being canned just because I'm getting impatient with IBM's process.  Are you prepared to be that person?

Alan.
--
Can't tweet, won't tweet!

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

 

***************************************************************************
The information contained in this communication is confidential, is
intended only for the use of the recipient named above, and may be legally
privileged.

If the reader of this message is not the intended recipient, you are
hereby notified that any dissemination, distribution or copying of this
communication is strictly prohibited.

If you have received this communication in error, please resend this
communication to the sender and delete the original message or any copy
of it from your computer system.

Thank You.
****************************************************************************
_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[billfen]

  To those of you in RexxLA and elsewhere who want to paint me as "the
bad guy", so be it, and I'll say no more.

But before jumping to conclusions, you might try a little digging for
the actual truth about who has (and hasn't) done what, when and why.

As the saying goes, don't believe anything you read or hear and only
half of what you see.

billfen
_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[Bruce Skelly]

Or how about Hope for the best, and prepare for the worst.

I don't think anybody is painting you as "the bad guy."

I think we are just saying that there is more to gain by waiting  
patiently for developments, than storming the castle walls.

Since the RexxLA group has succeeded in getting IBM to open source the  
bulk of Object Rexx, I am content to follow their lead on the proper  
method to getting NetRexx open sourced.

Remember, IBM is under no obligation to open source NetRexx.  
Companies often announce intentions to do something, only later to  
retract them.

Bruce


On Aug 19, 2010, at 9:44 AM, Bill Fenlason wrote:

> To those of you in RexxLA and elsewhere who want to paint me as "the  
> bad guy", so be it, and I'll say no more.
>
> But before jumping to conclusions, you might try a little digging  
> for the actual truth about who has (and hasn't) done what, when and  
> why.
>
> As the saying goes, don't believe anything you read or hear and only  
> half of what you see.
>
> billfen
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[Kermit Kiser]

I am puzzled Bill - last year you told us that you could add NetRexx
support to Eclipse but this year you said there are problems that
require the open source code to resolve.

I have not seen any messages describing the nature of those problems.

Granted that David Requena and I were not able to do as much as we
wanted with the NetRexxDE and NetRexxScript IDEs in jEdit due to lack of
NetRexx source code access, we did ask help here and managed to work
around the problems enough to make substantial progress.

I am curious about why you feel that the delay in NetRexx open source
has delayed Eclipse support for NetRexx.

-- Kermit

BTW: Is anyone working on NetBeans support for NetRexx?


On 8/19/2010 9:44 AM, Bill Fenlason wrote:

>  To those of you in RexxLA and elsewhere who want to paint me as "the
> bad guy", so be it, and I'll say no more.
>
> But before jumping to conclusions, you might try a little digging for
> the actual truth about who has (and hasn't) done what, when and why.
>
> As the saying goes, don't believe anything you read or hear and only
> half of what you see.
>
> billfen
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>
>

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

El 19/08/2010 22:26, Kermit Kiser escribió:

BTW: Is anyone working on NetBeans support for NetRexx?

I did look into it time ago.

It seems to be quite feasible once one has an antlr grammar for the NetRexx language.

An almost-complete such thing is said to exist. Unfortunately I never could get my hands on that antlr file.

Not that I'm an antlr expert either..


_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[Thomas.Schneider.Wien]

Hello David & all,
   I'm currently *not* working on netbeans support, as I'm currently too much busy to support classic Rexx, COBOL, and the PL/I.

When this is finished, I will, for sure, try to develop a NetBeans support, howeverm and that's the reason why I did subscribe [nbdev], the NetNeans
development group, some time ago...

But first, I'll have to complete my current queue ...
Sorry that I cannot help you immediately. :-(
Tom.
===========================================================================================================
Am 19.08.2010 23:00, schrieb David Requena:

El 19/08/2010 22:26, Kermit Kiser escribió:

BTW: Is anyone working on NetBeans support for NetRexx?

I did look into it time ago.

It seems to be quite feasible once one has an antlr grammar for the NetRexx language.

An almost-complete such thing is said to exist. Unfortunately I never could get my hands on that antlr file.

Not that I'm an antlr expert either..

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[billfen]

  On 8/19/2010 4:26 PM, Kermit Kiser wrote:

> I am puzzled Bill - last year you told us that you could add NetRexx
> support to Eclipse but this year you said there are problems that
> require the open source code to resolve.
>
> I have not seen any messages describing the nature of those problems.
>
> Granted that David Requena and I were not able to do as much as we
> wanted with the NetRexxDE and NetRexxScript IDEs in jEdit due to lack
> of NetRexx source code access, we did ask help here and managed to
> work around the problems enough to make substantial progress.
>
> I am curious about why you feel that the delay in NetRexx open source
> has delayed Eclipse support for NetRexx.
>
> -- Kermit
>
> BTW: Is anyone working on NetBeans support for NetRexx?

Kermit,

The following is probably more than you want to know, and I assume many
readers will tl;dr.  But since you asked, here is a brief version of my
NetRexx Eclipse project description and status report.

 From a design standpoint, because of the sophistication of the Eclipse
Java IDE, I felt that more than a "color the keywords, click to compile
for error messages" approach was necessary.  Users of the Eclipse Java
IDE are used to instant (while typing) error feedback and the illusion
of an instant incremental compile, and NetRexx users would expect no
less, including the full "look and feel" of Eclipse.  For example, using
the excellent Eclipse Java debug facility on a combination of NetRexx
(converted to Java) and native Java files (with Trace as a bonus) is
certainly appealing, and there are numerous other desirable Eclipse
capabilities.  I'm not saying that a simple Eclipse plugin editor that
provides the basics of keyword coloring and a batch compile is not
useful - it just wasn't my goal.  Certainly a plugin like that could
(and probably should) have been done for Eclipse by now.  Hopefully
someone else interested in Eclipse will take on that type of project.

Building  an Eclipse NetRexx IDE that looks and acts similar to (and is
reasonably integrated with) the Eclipse Java IDE is a very ambitious
task (at least for me).  From day one I wanted the NetRexx source to see
what was usable and adaptable, and I counted on basing design decisions
on it.  There is no sense in badly imitating a real thing when the real
thing is available.  But because the source was unavailable initially, I
thought the best approach at the time was a traditional token based
scanner / parser operating incrementally in the Eclipse framework.  In
theory I could have tried to adapt the JDE, but just starting to
understand all that is a huge task.  The JFlex scanner generator /
JavaCC parser generator combination had worked for me in the past and I
used that.  I find JavaCC more straight forward than Antlr, and JFlex
matched my experience with Flex.  (Also, I had a JFlex scanner and
JavaCC grammar for PL/I as a start.  The JFlex and JavaCC Eclipse
plugins work and are well supported).

GUI performance is very, very important, and the error checking and
syntax/semantic coloring have to be incremental.  Compiling with every
document change event could be a real dog - the Eclipse background
reconciler is bad enough as is.  Without the NetRexx source code, making
the grammar match the reference implementation is difficult, and it's
all about the details.  To be effective, the match needs to be nearly
perfect since the illusion has to be of immediate incremental compiling,
just like the Java IDE.  When I last worked on it, the scanner and
grammar definitions were in reasonable shape and essentially complete,  
but the grammar still had a few glitches, and needs to be carefully
verified by comparison with the NetRexx source.  Incremental scanning
and parsing (statement level, with problem marker error messages) work,
although I still had some AST related stuff to iron out . It would
certainly have been useful to see exactly how Mike handled some aspects
of that, although I don't believe he used a formal AST.  Possibly I
could have used the NetRexx internal structure and code and actually
done an incremental compile (really slick if possible) but that is
unlikely now without major redesign.  The editor actions of "Compile to
Java" and "Interpret" are a bit of a kludge and have some problems, and
revising and adding some new options and other interface code to NetRexx
would certainly help there.  Lots of the Eclipse bells and whistles are
remaining and in various stages of completion - outline page, content
assist, hover, double click, folding, help, etc., some of which might be
helped by the NetRexx code.  Of course there is still lots of debugging
to do and things I need to learn about inside Eclipse.   Obviously the
project a big job, and it is currently incomplete with significant
holes.  The NetRexx source, if available, could potentially save a lot
of effort.  My current opinion that it is unlikely that I can
successfully implement the kind of Eclipse plugin I envision without the
NetRexx source code.  (Not saying it can't be done - and hats off to
anyone who does it!  Any experienced Eclipse user who is also familiar
with NetRexx knows almost exactly what is wanted and needed.)

I stopped active work on the project some months ago when I realized the
futility of waiting for the NetRexx source.  I started the project over
a year ago and the NetRexx source had been promised for well over a year
even at that time.  Had I foreseen the fact that IBM and RexxLA would
make absolutely no progress toward releasing the NetRexx source code in
all the time since then, it's unlikely that I would have even started
the effort.

  When (IF?) the source code finally becomes available, I'll read it
carefully and decide if it would be best to totally start over by
integrating that code into a new Eclipse editor, to continue on the
current path, or to attempt some combination of the two.  I don't want
to spend any more time and effort possibly going further down the wrong
road, so until then, I'm working on other projects and trying to
restrain my frustration with the NetRexx scene (obviously not very
successfully lately - my apologies as appropriate :).

billfen
_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[Kermit Kiser]

Bill ;

I am not an Eclipse user much less an Eclipse plugin developer. The
reason for that is because I program only in NetRexx and I don't want to
deal with Java syntax any more than I have to in order to use the Java
libraries. I know very little about Eclipse except that it is about 10
years overdue for supporting NetRexx from my point of view! I did look
into writing a NetRexx plugin for Eclipse last year, but as a non user
it was quite overwhelming and I quickly gave up that idea.

I appreciate the work you have done in the Eclipse area. If it ever
reaches usability I would help test. It might even replace the kludgey
Android NetRexx IDE that I built in jEdit early this year!

-- Kermit


On 8/20/2010 2:02 AM, Bill Fenlason wrote:

>  On 8/19/2010 4:26 PM, Kermit Kiser wrote:
>> I am puzzled Bill - last year you told us that you could add NetRexx
>> support to Eclipse but this year you said there are problems that
>> require the open source code to resolve.
>>
>> I have not seen any messages describing the nature of those problems.
>>
>> Granted that David Requena and I were not able to do as much as we
>> wanted with the NetRexxDE and NetRexxScript IDEs in jEdit due to lack
>> of NetRexx source code access, we did ask help here and managed to
>> work around the problems enough to make substantial progress.
>>
>> I am curious about why you feel that the delay in NetRexx open source
>> has delayed Eclipse support for NetRexx.
>>
>> -- Kermit
>>
>> BTW: Is anyone working on NetBeans support for NetRexx?
> Kermit,
>
> The following is probably more than you want to know, and I assume
> many readers will tl;dr.  But since you asked, here is a brief version
> of my NetRexx Eclipse project description and status report.
>
> From a design standpoint, because of the sophistication of the Eclipse
> Java IDE, I felt that more than a "color the keywords, click to
> compile for error messages" approach was necessary.  Users of the
> Eclipse Java IDE are used to instant (while typing) error feedback and
> the illusion of an instant incremental compile, and NetRexx users
> would expect no less, including the full "look and feel" of Eclipse.  
> For example, using the excellent Eclipse Java debug facility on a
> combination of NetRexx (converted to Java) and native Java files (with
> Trace as a bonus) is certainly appealing, and there are numerous other
> desirable Eclipse capabilities.  I'm not saying that a simple Eclipse
> plugin editor that provides the basics of keyword coloring and a batch
> compile is not useful - it just wasn't my goal.  Certainly a plugin
> like that could (and probably should) have been done for Eclipse by
> now.  Hopefully someone else interested in Eclipse will take on that
> type of project.
>
> Building  an Eclipse NetRexx IDE that looks and acts similar to (and
> is reasonably integrated with) the Eclipse Java IDE is a very
> ambitious task (at least for me).  From day one I wanted the NetRexx
> source to see what was usable and adaptable, and I counted on basing
> design decisions on it.  There is no sense in badly imitating a real
> thing when the real thing is available.  But because the source was
> unavailable initially, I thought the best approach at the time was a
> traditional token based scanner / parser operating incrementally in
> the Eclipse framework.  In theory I could have tried to adapt the JDE,
> but just starting to understand all that is a huge task.  The JFlex
> scanner generator / JavaCC parser generator combination had worked for
> me in the past and I used that.  I find JavaCC more straight forward
> than Antlr, and JFlex matched my experience with Flex.  (Also, I had a
> JFlex scanner and JavaCC grammar for PL/I as a start.  The JFlex and
> JavaCC Eclipse plugins work and are well supported).
>
> GUI performance is very, very important, and the error checking and
> syntax/semantic coloring have to be incremental.  Compiling with every
> document change event could be a real dog - the Eclipse background
> reconciler is bad enough as is.  Without the NetRexx source code,
> making the grammar match the reference implementation is difficult,
> and it's all about the details.  To be effective, the match needs to
> be nearly perfect since the illusion has to be of immediate
> incremental compiling, just like the Java IDE.  When I last worked on
> it, the scanner and grammar definitions were in reasonable shape and
> essentially complete,  but the grammar still had a few glitches, and
> needs to be carefully verified by comparison with the NetRexx source.  
> Incremental scanning and parsing (statement level, with problem marker
> error messages) work, although I still had some AST related stuff to
> iron out . It would certainly have been useful to see exactly how Mike
> handled some aspects of that, although I don't believe he used a
> formal AST.  Possibly I could have used the NetRexx internal structure
> and code and actually done an incremental compile (really slick if
> possible) but that is unlikely now without major redesign.  The editor
> actions of "Compile to Java" and "Interpret" are a bit of a kludge and
> have some problems, and revising and adding some new options and other
> interface code to NetRexx would certainly help there.  Lots of the
> Eclipse bells and whistles are remaining and in various stages of
> completion - outline page, content assist, hover, double click,
> folding, help, etc., some of which might be helped by the NetRexx
> code.  Of course there is still lots of debugging to do and things I
> need to learn about inside Eclipse.   Obviously the project a big job,
> and it is currently incomplete with significant holes.  The NetRexx
> source, if available, could potentially save a lot of effort.  My
> current opinion that it is unlikely that I can successfully implement
> the kind of Eclipse plugin I envision without the NetRexx source
> code.  (Not saying it can't be done - and hats off to anyone who does
> it!  Any experienced Eclipse user who is also familiar with NetRexx
> knows almost exactly what is wanted and needed.)
>
> I stopped active work on the project some months ago when I realized
> the futility of waiting for the NetRexx source.  I started the project
> over a year ago and the NetRexx source had been promised for well over
> a year even at that time.  Had I foreseen the fact that IBM and RexxLA
> would make absolutely no progress toward releasing the NetRexx source
> code in all the time since then, it's unlikely that I would have even
> started the effort.
>
>  When (IF?) the source code finally becomes available, I'll read it
> carefully and decide if it would be best to totally start over by
> integrating that code into a new Eclipse editor, to continue on the
> current path, or to attempt some combination of the two.  I don't want
> to spend any more time and effort possibly going further down the
> wrong road, so until then, I'm working on other projects and trying to
> restrain my frustration with the NetRexx scene (obviously not very
> successfully lately - my apologies as appropriate :).
>
> billfen
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
>
>
>

_______________________________________________
Ibm-netrexx mailing list
[hidden email]