NetRexx enhancements -- reply

[Fernando Cassia-2]

On Sat, Sep 25, 2010 at 2:07 PM, Robert Hamilton <[hidden email]> wrote:
> PLEASE DON'T MESS WITH NETREXX OR TEXAS.
>
> Bobh

HA-HA. Who said this list was dead?. It can even be funny. :)

FC
_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[alansam]

On 25 September 2010 10:07, Robert Hamilton <[hidden email]> wrote:

PLEASE DON'T MESS WITH NETREXX OR TEXAS.

Bobh

Do what you like with Texas; leave NetRexx alone! :-D
 

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

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

Robert,

You got me here. I've no idea of what iinbdfi might mean :-)


El 24/09/2010 19:06, Robert Hamilton escribió:

My vote goes for iinbdfi -- the old reliable

bobh

On Fri, Sep 24, 2010 at 9:58 AM, David Requena <[hidden email]> wrote:

El 24/09/2010 15:06, Mike Cowlishaw escribió:

Hey guys, we're talking about a language which incorporates (of all things!!) '&&' as a (i)logical 'XOR' operator.  

 

What would be the logical symbol/character for XOR?  '⊕' would be just as incomprehensible to the average programmer...  [but yes, '||' would be what I might have chosen, had that not been widely in use for 'concatenate'].

While I admit that *any* chosen symbol would be completely arbitrary, this particular case is specially unfortunate in my view.

- For any computer illiterate or novice programmer "&&" means "some form of 'AND'". That is because the symbol '&' has a very precise meaning in plain English (which happens to be 'AND').
- For any experienced programmer coming from C or C++ (and many others) '&&' is "logical AND".
- For programmers coming from java '&&' is "short-circuiting logical AND".

Given that 2nd and 3rd groups comprise a big chunk of the programmers population today, by choosing '&&' as an XOR operator NetRexx has gotten to alienate most of its actual and potential users. The problem is not alleviated by the fact that a single '&' means indeed "logical AND" in ´NetRexx.

A language may of course choose '=' as an non-equal logical operator (after all some arbitrary symbol needs to be chosen). The thing is that such an election would be very unintuitive for *most* people.

---
Saludos / Kind regards.
David Requena

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

hmmm... aside of the fact that maths symbols are just as arbitrary as any other set, you seem to consider that any programmer has a wide maths background.
Although that was true once upon a time, it's not so any longer I'm afraid.

El 24/09/2010 20:14, Connor Birch escribió:

If we are going to change these symbols, lets not just plump for some other random symbol.   Using ∧ for exclusive OR would be as crazy as using = for not equals, since ∧ is the mathematical symbol for AND!   

If we change the symbols, lets be consistent and virtuous and use the proper mathematical ones.   They may seem arbitrary to someone who does not know them, but the same could be said of any mathematical symbol; if we went down that route it would be a license to use all sorts of random symbols due only to personal preference, and some one would come up with a different sort of OR and we'd have ||| or &&& or &&| or something terrible.   

If, for consistency with our mathematics mates, we use = for equals, + for addition, then lets also use ∧ for AND, ∨ for OR, and ⊕ for EXCLUSIVE OR.   It turns out that they are just as easy to remember as any other symbol and actually look nicer in expressions I think since they don't get mistaken for anything else.

After all | does not really mean OR any more than any other symbol, and EXCLUSIVE OR is neither OR OR (||) or AND AND (&&).   If we use the proper mathematics symbols, we'll not end up having to invent new compound symbols every time we include a new operation. 

I think some of the confusion stems from the words used, and not using the correct words when talking about the operators.   When we say AND, what we really mean is BOTH.   When we say OR, what we really mean is EITHER, and when we say EXCLUSIVE OR, what we really mean is ONE OF.   When we think of it like this, the temptation to use & for BOTH is less attractive, and using the established mathematical symbols is easier to accept.   In todays world, there is no difficulty in using the correct symbols; they have been provided by fellow human beings so that we can use them, lets not insult these good people by ignoring their efforts while making do with something far less adequate.

If there is not much appetite for NetRexx to evolve, then this is the wrong forum to discuss these things.   If the approach taken by Tebetha (which has no qualms about using the most appropriate symbols, or moving on from what other languages have done) makes sense to anyone, and they want to learn more about Tabetha (which has syntactic elegance for human usability as one of its key aims), I'd be happy to contribute to a separate forum (when I'm not so busy).   Let me know, Connor at JoinedUpBusiness dot com.

Connor.

  

On 24 Sep 2010, at 17:37, Mike Cowlishaw wrote:

   What would be the logical symbol/character for XOR?  

   '⊕' would be

   just as incomprehensible to the average programmer...

Cool!  Where'd you find that glyph, Mike?  The Wiccan 

codepage?  The Double-Byte Vampire Character Set?  ;-))

Google (for example): HTML exclusive or

   [but yes,

   '||' would be what I might have chosen, had that not 

   been widely in

   use for 'concatenate'].

I always wondered why the compound operator '/|' wasn't chosen.

Visually it says "not-or" which isn't too big a conceptual 

leap to 'exclusive-or' for most people.

I'm not sure, either (see post of a few minutes ago).  There was precedent for
&&, I'm sure, so I followed that.

Mike


_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

Yeah Chip,

I remember that 80's books which described in full detail how in year 2000, cars would drive us home themselves, trains would travel airborne, and countless other technological marvels.

Instead we had the we suffered the Y2K panic issue because we couldn't be sure that we had eradicated the use of two digits for year representation.

I'm afraid we all are set for a long wait before "the whole keyboard/glyph problem evaporates" :-(

El 24/09/2010 21:54, Chip Davis escribió:

Frankly, if second-guessing Mike about his choice of glyph for an relatively obscure operator is the best we can do, I'd say there are better uses for our time.  He was working with a 3270 keyboard and an expected initial programmer base familiar with EXEC2 and PL/I.  And I suspect that when he started the design, he had no idea it was going to be publicly available in the first place.  NetRexx inherited those choices honestly.

Before long, we'll have voice-dictated programs and the whole keyboard/glyph problem evaporates.  That will spawn a movement to make programming terms more euphonic (we can say "substring" now!).  With the development of RMM technology, we'll find out how faulty our cognitive processes are, and people will wax nostalgic for the good old days of striking a specific key to get a specific character.

As for adding syntactic sugar to NetRexx, I find it hard on my programming pancreas.  There is an elegance in the conceptual simplicity of the Rexx design that such accretions mar.  Sort of like putting spinner hubcaps on a Roll-Royce, imho.

I wish I could look back at when I was his age and say that a sub-optimal choice for a mathematical symbol was one of my most enduring regrets... :-/

-Chip-


_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

El 25/09/2010 2:43, George Hovey escribió:

I support evolution for sound reasons, but not revolution.  NetRexx isn't mostly wrong, it's hugely right.

Amen brother :-)


_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

El 25/09/2010 15:34, Mike Cowlishaw escribió:

Yes, this isn't an option now -- I was just saying it *was* a nice idea.

 

And none of this is about 'saving keystrokes'.   What the A+=expression saves is the duplication of the name, which is a source of errors when names are changed (forgetting to change one or other of them).   The same saving as not repeating 'BigDecimal' in:

 

  foo=BigDecimal(5)

 

rather than: 

 

  BigDecimal foo=new BigDecimal(5);

That's indeed my own argument for supporting the += operators.
I was just comparing a+1 to a+=1 when talking aboy typesaving :-)

 

[of course there is other unnecessary syntax that NetRexx avoids, there. :-)]

 

Mike

---

From: [hidden email] [[hidden email]] On Behalf Of David Requena
Sent: 25 September 2010 11:19
To: IBM Netrexx
Subject: Re: [Ibm-netrexx] NetRexx enhancements -- reply

El 24/09/2010 14:51, Mike Cowlishaw escribió:

And I still like:  a+expression  (in general:  symbol operation expression) as a shorter way of saying the same thing (does not work in classic or oo Rexx, of course, because a clause that is an expression is a command).

This particular one I don't like very much as 'a+1' is currently an expression yielding a result leaving 'a' unmodified.

Would that new behaviour be only applied when that expression were to be found itself alone in a clause?
What about b=a+1? Would that also increment 'a'?
And how about 'a=1;a=a+1'? Would 'a' be assigned the value '2' or '3'? (OK, '2', but that is not evident at first sight)

Admittedly those questions might be answered (for example if first question answer was 'yes'). But frankly it seems to me that such a syntax (having the very same expression resulting on different behaviour depending on context) would make the 'astonishing factor' to sky-rocket just to save *ONE* keystroke over 'a += 1'

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

Well, last news available to me:

- The only existent implementation will at some point be handed to RexxLA.
- Several comités (steering?, technical?) will be formed inside RexxLA in order to make the language progress in "The REXX way"(TM)
- At a latter time further iterations of said implementations will be made available to general public.

Time schedule? Who nows!
In the mean time some people here are trying to somewhat influence potential future events with their on ideas :-)

I think I've stated before here my believing that we've already failed as a community.
Nothing prevented anyone to develop an alternative implementation. That would have been the natural reaction of a live user community to having NetRexxC inert, unchanged, in the hands  of a company which obviously never was really interested in it, for that many years. That didn't happen.

Failed we have.
Wait for open-sourcing we do,
as the language won't die,
we'd like to be sure.
 

El 25/09/2010 15:36, Mike Cowlishaw escribió:

Sounds like everyone likes the += enhancement.   Looking forward to seeing it in NetRexx 3!  When are you guys actually going to ship that?

 

Mike

---

From: [hidden email] [[hidden email]] On Behalf Of David Requena
Sent: 25 September 2010 11:46
To: IBM Netrexx
Subject: Re: [Ibm-netrexx] NetRexx enhancements -- reply

Mike,

In general all my my remarks in this list have not been in any way criticisms about the language itself or your long time away taken design decisions.
I may keep proposing enhancements but that is because the world kept revolving while NetRexx remained frozen in time (e.g. NetRexx is not fully compatible with the java library any more).

That seems to be the wall we all are stumbling time and again.

I'm not even complaining about '&&' which I went on using after a mild surprise when I saw it first time. My original point was that I don't see the '+=' operators family as such evil complexity risers given that we have things as && --> XOR no one seems to have an issue with.

In general, I favour '+=' adoption as I try to always choose got variable names.
Which in turn implies these get rather longish and that I often change variable names.
Which in turn calls for 'search and replace'.
'search and replace' *IS TRUE EVIL* if you don't learn the particular regex flavour supported by your editor.
Any regex flavour is abstruse by definition

Under this light, any syntax which helps reducing the number of instances I need to change is welcome.


El 24/09/2010 18:34, Mike Cowlishaw escribió:

But David ....

El 24/09/2010 15:06, Mike Cowlishaw escribió:

Hey guys, we're talking about a language which incorporates (of all things!!) '&&' as a (i)logical 'XOR' operator.

 

What would be the logical symbol/character for XOR?  '⊕' would be just as incomprehensible to the average programmer...  [but yes, '||' would be what I might have chosen, had that not been widely in use for 'concatenate'].

While I admit that *any* chosen symbol would be completely arbitrary, this particular case is specially unfortunate in my view.

- For any computer illiterate or novice programmer "&&" means "some form of 'AND'". That is because the symbol '&' has a very precise meaning in plain English (which happens to be 'AND'). 

 

Agreed.  But || was already taken by PL/I to mean concatenate, and && was already used in several languages for 'exclusive or'.   I wish I had chosen the PL/I ^ (possibly it wasn't using that then, or the character was not on keyboards, or other languages were using it for the power operator; I forget the reason).

- For any experienced programmer coming from C or C++ (and many others) '&&' is "logical AND". 

 

That wasn't an issue when Rexx was designed .. hardly anyone used C in 1979 (K&R was published in late 1978, and the other documents about C that I had seen all seemed very fluid -- in fact I chose the % operator from early C documents -- and then it changed (in C) to mean something different).

- For programmers coming from java '&&' is "short-circuiting logical AND". 

 

Quite.  So  NetRexx could add the ^ operator to mean XOR, and phase out &&, leaving it there just for old-timer Rexx folk. 

Given that 2nd and 3rd groups comprise a big chunk of the programmers population today, by choosing '&&' as an XOR operator NetRexx has gotten to alienate most of its actual and potential users. The problem is not alleviated by the fact that a single '&' means indeed "logical AND" in ´NetRexx.

A language may of course choose '=' as an non-equal logical operator (after all some arbitrary symbol needs to be chosen). The thing is that such an election would be very unintuitive for *most* people. 

Almost anything based on C is unintuitive for most people :-)).  That's why I did NetRexx.  However, basing it on Rexx did imply trying to stay as close to Rexx as possible -- including the operators.  But I entirely agree that && for XOR was not a good choice.  || would have been better.    History ...

 

Mike 

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

El 25/09/2010 16:02, Chip Davis escribió:

If, as has been asserted, "search and replace are evil" (with requisite apologies to Dijkstra) why not have a token that means "whatever is on the left of the '='"?  Something like:

  aVeryLongAndComplicatedVariableName=~+1

That seems a weird argument to fend the use of:

    aVeryLongAndComplicatedVariableName += 1

In which way an arbitrary 3 symbol sequence would present any advantage over a 2 symbol one? One that is already familiar for a large number of people

And yes, 'search and replace' *is* harmful :-)


_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

The fact that I cannot see the symbols as entered by you in my BlackBerry, but see them ok on thunderbird, would wouldn't on epoc is very real...

Even the glyphs you take for granted as easily typed on a US keyboard (\, |, #, [, ], {, }) require a 2 key combo in my Spanish dictionary. One involving a key that US keyboards lack (Alt-Gr). Without Alt-Gr I'd need Ctrl + Alt + Symbol-key.

And that is just for the labelled glyphs. Entering 'Σ' (do you see a summatory?) required me to open windows chrmap.exe, selecting it and then copy and paste here.



El 25/09/2010 16:11, Connor Birch escribió:

On 25 Sep 2010, at 01:43, George Hovey wrote:

It seems to me that we must limit NetRexx's character set (for all elements except identifiers and strings) to ASCII which is a subset of at least CP1252, ISO-8859-1 and UTF8 encodings.  Using other characters, no matter how attractive, will get us into deep trouble with respect to keyboard entry

Any 'problem' with entering the characters is more imagined that real.   For at least a decade, people wanting to enter accents for roman letters have been able to use 'Dead Keys'.   They press a Dead key that represents the family of accents (nothing appears to happen) then they press the roman letter, and the accented character appears.  We could easily use a dead key (implemented in software), or even use an ALT key to allow programming symbols to be typed (A produces the AND symbol etc, very easy to remember).   And there is an alternative to using the keyboard, in the same way as we add special symbols to a Word document for example, or into an email like this ↵

Can't see a problem at all, they are there for us to use.   

As an analogy, if we have all the correct sizes of screwdriver provided for us in our toolbox, who would choose not to use some of them?

Kind regards,

Connor.   

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

George,

I've been a RexxLA member for about 6 months and all I can say is that any new developments on the subject are not public there either.
There's not much NetRexx talk on the association members list (mainly REXX/ooRexx centric) but rest assured there's no secret conspiracy there.
At least not one accessible to regular full-voting members ;-)

El 25/09/2010 19:31, George Hovey escribió:

You remark 'As for "NetRexx 3", maybe there should be community agreement well before anything is implemented?'  But it isn't exactly clear who the community is.  RexxLA's homepage seems to suggest that they will become the proprietors of the NetRexx brand.  How their development process works is a mystery since the bylaws section of their site seems to be permanently "under construction"; and it isn't clear that outsiders have any voice whatsoever.

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

El 25/09/2010 19:40, George Hovey escribió:

David,
Would you mind elaborating on your remark "NetRexx is not fully compatible with the java library any more"?

Of course. I've already discussed this extensively at the list.

Let me first state that these are not actual shortcomings in NetRexx itself but merely a consequence of being immutable for so long while sitting on top of such a rapidly evolving platform as java. To summarize, the 2 cases I've spotted (there're probable more):

Some standard library classes can no longer be extended from NetRexx. Java nowadays supports different return types when overriding methods while NetRexx does not. Of course the feature has been used in the standard library. See http://ibm-netrexx.215625.n3.nabble.com/Another-issue-with-current-NetRexxC-sorry-to-say-tt1187437.html#a1192691 for the discussion on the topic which includes a previous report by Kermit Kisses as well as a (nasty) workaround.

A couple months ago there was a longish discussion here about NetRexx minor classes versus java inner classes. Although it was agreed that inner classes was a trully ugly addition to java the fact remains that Sun tightly wrapped some areas of the library (swing, multi-threading model and sockets come to mind) around them. NetRexx non-inner, kind-of-limited-inheritance, minor classes force the programmer to expose private state/behaviour out of the classes in some instances. In a few classes it's just not possible to use these library classes in the intended way. Search for 'dependent classes' and 'minor classes' at http://ibm-netrexx.215625.n3.nabble.com/ for the discussion. The subject was altered several times so I cannot provide a direct link.




_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

El 25/09/2010 20:07, Fernando Cassia escribió:

On Sat, Sep 25, 2010 at 1:26 PM, George Hovey [hidden email] wrote:

But we don't have the correct tools in our toolbox.  Your ideas on entering
special characters strike me as horrific.

+1 indeed.

In fact, I´ll vote twice against horror.

+1+1

Let me support that too:

numberOfVotesAgainstHorror = numberOfVotesAgainstHorror + 1

Funny enough I actually copy-and-pasted the name of the variable!
 :-D

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

Wouldn't that be PLEASE MESS WITH ((NOT NETREXX) AND (NOT TEXXAS)) ??

:-)

El 25/09/2010 22:55, Alan Sampson escribió:

On 25 September 2010 10:07, Robert Hamilton <[hidden email]> wrote:

PLEASE DON'T MESS WITH NETREXX OR TEXAS.

Bobh

Do what you like with Texas; leave NetRexx alone! :-D
 

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

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[David Requena]

El 25/09/2010 20:09, Fernando Cassia escribió:

HA-HA. Who said this list was dead?. It can even be funny. :)

Well it was mostly dead for a few years but fortunately  it seems to be enjoying a 2nd youth now :-)

Wow!! Now that I've been looking at the Nabble archives I see I'm the 2nd most prolific poster ever (226 posts). Only second to (of course) Thomas.Schneider.Wien (360)  ;-)

To be honest there's a certain individual which seems to be cheating at the ranking by sharing his posts among 3 personalities: Mike Cowlishaw (83) , Mike Cowlishaw-2 (146) and Mike Cowlishaw-3 (57)

I'm closing the gap Mr. MFC...  ;-)

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[alansam]

On 25 September 2010 14:07, David Requena <[hidden email]> wrote:

Robert,

You got me here. I've no idea of what iinbdfi might mean :-)

Really old but I don't think I've seen it expressed as a TLA before.  Expressed in PERL however:
use constant iinbdfi => q/If it's not broken don't fix it/;
 :-D

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

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[Aviatrexx]

While I agree with the sentiment of your acronym, I also have a SHARE
button that says "Perl is Just Rexx with Bad Syntax".

-Chip-

On 9/26/10 00:44 Alan Sampson said:

>
>
> On 25 September 2010 14:07, David Requena <[hidden email]
> <mailto:[hidden email]>> wrote:
>
>     Robert,
>
>     You got me here. I've no idea of what iinbdfi might mean :-)
>
>
> Really old but I don't think I've seen it expressed as a TLA before.  
> Expressed in PERL however:
> *use constant iinbdfi => q/If it's not broken don't fix it/;*
>  :-D
>
> Alan.

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[alansam]

On 26 September 2010 08:30, Chip Davis <[hidden email]> wrote:

While I agree with the sentiment of your acronym, I also have a SHARE button that says "Perl is Just Rexx with Bad Syntax".

-Chip-

you should add "and CPAN".  It's the CPAN library that makes PERL a winner.  They have a huge user community focused on delivering quality solutions and extensions to the base libraries that makes PERL a world beater.  Would that the same approach was taken by RexxLA and the NetRexx community.  There are some small Rexx/NetRexx contributions that have been made and posted for public consumption but there's no central repository where programmers can go to reference them; eventually the links go stale so the knowledge and tools are lost to the community.

Alan.

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

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[Bruce Skelly]

I agree with you Alan, not just for NetRexx but for ooRexx too.

On another topic, have any of you tried Processing <http://www.processing.org/>.  I wish that NetRexx would be as easy to download and get running.  I took a little online class on processing, and in just a few hours we had a neat little animation program up and running.  Also, with a single click, executables were produced for Windows, Mac OS X, and Linux.

Bruce

On Sep 26, 2010, at 9:12 AM, Alan Sampson wrote:

On 26 September 2010 08:30, Chip Davis <[hidden email]> wrote:

While I agree with the sentiment of your acronym, I also have a SHARE button that says "Perl is Just Rexx with Bad Syntax".

-Chip-

you should add "and CPAN".  It's the CPAN library that makes PERL a winner.  They have a huge user community focused on delivering quality solutions and extensions to the base libraries that makes PERL a world beater.  Would that the same approach was taken by RexxLA and the NetRexx community.  There are some small Rexx/NetRexx contributions that have been made and posted for public consumption but there's no central repository where programmers can go to reference them; eventually the links go stale so the knowledge and tools are lost to the community.

Alan.

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

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

[Robert L Hamilton]

IF IT AIN'T BROKE DON'T FIX IT. . .

BOBH

On Sun, Sep 26, 2010 at 10:30 AM, Chip Davis <[hidden email]> wrote:

While I agree with the sentiment of your acronym, I also have a SHARE button that says "Perl is Just Rexx with Bad Syntax".

-Chip-

On 9/26/10 00:44 Alan Sampson said:

On 25 September 2010 14:07, David Requena <[hidden email] <mailto:[hidden email]>> wrote:

   Robert,

   You got me here. I've no idea of what iinbdfi might mean :-)


Really old but I don't think I've seen it expressed as a TLA before.  Expressed in PERL however:
*use constant iinbdfi => q/If it's not broken don't fix it/;*
 :-D

Alan.

_______________________________________________
Ibm-netrexx mailing list
[hidden email]

_______________________________________________
Ibm-netrexx mailing list
[hidden email]