Delphi, Pascal, delphi-mode, pascal-mode, and free software

Delphi, Pascal, delphi-mode, pascal-mode, and free software

[Stefan Monnier]

Can someone enlighten me on the status of Delphi w.r.t Free Software?
AFAICT it's proprietary (and there's a Free Software alternative called
Lazarus which is IIUC a sort of library for Free Pascal).

Also, what's the relationship between pascal-mode and delphi-mode?
Is there some overlap between the two, should there be more overlap,
or less?


        Stefan

Re: Delphi, Pascal, delphi-mode, pascal-mode, and free software

[Stefan Monnier]

Ping?

>>>>> "Stefan" == Stefan Monnier <monnier@iro.umontreal.ca> writes:

> Can someone enlighten me on the status of Delphi w.r.t Free Software?
> AFAICT it's proprietary (and there's a Free Software alternative called
> Lazarus which is IIUC a sort of library for Free Pascal).

> Also, what's the relationship between pascal-mode and delphi-mode?
> Is there some overlap between the two, should there be more overlap,
> or less?


>         Stefan

Re: Delphi, Pascal, delphi-mode, pascal-mode, and free software

[Simon South]

On Fri, 2010-09-10 at 13:05 +0200, Stefan Monnier wrote:
> > Can someone enlighten me on the status of Delphi w.r.t Free Software?
> > AFAICT it's proprietary (and there's a Free Software alternative called
> > Lazarus which is IIUC a sort of library for Free Pascal).

Yes, that's correct.  (Well, I believe Lazarus is more an
IDE-slash-RAD-tool than a set of libraries.  But yes, the idea is that
Lazarus + Free Pascal would be a free replacement for Delphi.)

> > Also, what's the relationship between pascal-mode and delphi-mode?
> > Is there some overlap between the two, should there be more overlap,
> > or less?

Each of these targets a different major dialect of Pascal and is
completely separate from the other.  Having a single mode might be nice
but I think writing a parser that would (neatly) handle both dialects
would be a big challenge.  Plus I seem to recall the two modes are quite
different architecturally at the moment, so at least one would need to
be rewritten.

-- 
Simon South
ssouth@simonsouth.ca

Re: Delphi, Pascal, delphi-mode, pascal-mode, and free software

[Stefan Monnier]

>> > Can someone enlighten me on the status of Delphi w.r.t Free Software?
>> > AFAICT it's proprietary (and there's a Free Software alternative called
>> > Lazarus which is IIUC a sort of library for Free Pascal).
> Yes, that's correct.  (Well, I believe Lazarus is more an
> IDE-slash-RAD-tool than a set of libraries.  But yes, the idea is that
> Lazarus + Free Pascal would be a free replacement for Delphi.)

Hmm... so we should rename delphi.el to something else, since we don't
like packages that are dedicated to supporting proprietary software.
From the above I gather than the language part of Delphi is basically
the same as Free Pascal, so maybe we could rename it free-pascal.el?

>> > Also, what's the relationship between pascal-mode and delphi-mode?
>> > Is there some overlap between the two, should there be more overlap,
>> > or less?
> Each of these targets a different major dialect of Pascal and is
> completely separate from the other.

OK.  Indeed from looking at them there didn't seem to be much shared
ancestry.  Do you happen to know what major dialect of Pascal is
supported by pascal-mode?

> Having a single mode might be nice but I think writing a parser that
> would (neatly) handle both dialects would be a big challenge.
> Plus I seem to recall the two modes are quite different
> architecturally at the moment, so at least one would need to
> be rewritten.

Merging seems difficult, indeed.  But at least from what you say, it
seems like it would be desirable to try and shape future changes so as
to reduce the differences rather than increase them.


        Stefan

Re: Delphi, Pascal, delphi-mode, pascal-mode, and free software

[Simon South]

On Fri, 2010-09-10 at 14:48 +0200, Stefan Monnier wrote:
> From the above I gather than the language part of Delphi is basically
> the same as Free Pascal, so maybe we could rename it free-pascal.el?

Well, it's only one of several dialects Free Pascal understands (and not
the default).  The actual name of the language is Object Pascal, so how
about object-pascal.el?  

> OK.  Indeed from looking at them there didn't seem to be much shared
> ancestry.  Do you happen to know what major dialect of Pascal is
> supported by pascal-mode?

No... I'd assume it's standard Pascal (ISO 7185) but I don't know for
sure.  (Actually, if it isn't, it probably should be.)

> Merging seems difficult, indeed.  But at least from what you say, it
> seems like it would be desirable to try and shape future changes so as
> to reduce the differences rather than increase them.

Agreed!

-- 
Simon South
ssouth@simonsouth.ca

Re: Delphi, Pascal, delphi-mode, pascal-mode, and free software

[Richard Stallman]

The issue of editing modes for languages whose compilers are nonfree
is a difficult borderline case.  If we strictly interpret the rule
of not supporting nonfree software, we should reject those modes.

However, I think it is ok to include such modes when the language is
widely used.  In such cases, Emacs support gives little encouragement
to use the language.  For instance, Emacs has had a Java mode since
long before there was a free Java platform, and I think it was good
to have this mode.

On the other hand, if there is an obscure language whose only
implementation is proprietary, it would be good to refuse to let
Emacs publicize its existence.

    >> > Also, what's the relationship between pascal-mode and delphi-mode?
    >> > Is there some overlap between the two, should there be more overlap,
    >> > or less?
    > Each of these targets a different major dialect of Pascal and is
    > completely separate from the other.

The two Lisp programs may be completely separate, but a different
question is how different these Pascal dialects are.  Would it be
easy to modify one of these modes to handle all the Pascal dialects
we want to handle?

Re: Delphi, Pascal, delphi-mode, pascal-mode, and free software

[Simon South]

On Sat, 2010-09-11 at 01:30 -0400, Richard Stallman wrote:
> Would it be easy to modify one of these modes to handle all the Pascal
> dialects we want to handle?

This should be possible with delphi-mode if we limit ourselves to
ISO-standard and Object Pascal, which I'm learning are the only two
dialects still commonly used anyway.  That would cover source code for
GNU Pascal and Free Pascal, the two major free implementations, and
Delphi as well.

In fact this is probably already true to a large extent and it would
mostly be a matter of testing and rounding out delphi-mode's
functionality.

-- 
Simon South
ssouth@simonsouth.ca

Re: Delphi, Pascal, delphi-mode, pascal-mode, and free software

[Stefan Monnier]

>> From the above I gather than the language part of Delphi is basically
>> the same as Free Pascal, so maybe we could rename it free-pascal.el?
> Well, it's only one of several dialects Free Pascal understands (and not
> the default).  The actual name of the language is Object Pascal, so how
> about object-pascal.el?  

Sounds good.  Of course, we'd need to rename vars and functions inside
(and add a few defalias/defvaralias to maintain some amount of backward
compatibility, of course) and "object-pascal-" sounds like too long
a prefix.  How 'bout "opascal-"?


        Stefan

Re: Delphi, Pascal, delphi-mode, pascal-mode, and free software

[Stefan Monnier]

>> Would it be easy to modify one of these modes to handle all the Pascal
>> dialects we want to handle?

> This should be possible with delphi-mode if we limit ourselves to
> ISO-standard and Object Pascal, which I'm learning are the only two
> dialects still commonly used anyway.  That would cover source code for
> GNU Pascal and Free Pascal, the two major free implementations, and
> Delphi as well.

> In fact this is probably already true to a large extent and it would
> mostly be a matter of testing and rounding out delphi-mode's
> functionality.

One thing I'd like to see with delphi.el (other than renaming it) is to
use "standard" facilities like the syntax-table, rather than do all the
parsing by hand.

Do you have any idea what that would take?


        Stefan

Re: Delphi, Pascal, delphi-mode, pascal-mode, and free software

[Simon South]

On Sat, 2010-09-11 at 11:56 +0200, Stefan Monnier wrote:
> Of course, we'd need to rename vars and functions inside (and add a few
> defalias/defvaralias to maintain some amount of backward
> compatibility, of course) and "object-pascal-" sounds like too long a
> prefix.  How 'bout "opascal-"?

Sounds good.  I'll add this to my list.

-- 
Simon South
Independent Software Developer
Phone: +1 647 258 5662
Email: ssouth@simonsouth.ca

Re: Delphi, Pascal, delphi-mode, pascal-mode, and free software

[Simon South]

On Sat, 2010-09-11 at 12:52 +0200, Stefan Monnier wrote:
> One thing I'd like to see with delphi.el (other than renaming it) is to
> use "standard" facilities like the syntax-table, rather than do all the
> parsing by hand.
> 
> Do you have any idea what that would take?

No, this is all still pretty new to me.  Is there a guide to best
practices for major modes anywhere, or could you suggest a "good" mode I
could use as an example?

-- 
Simon South
Independent Software Developer
Phone: +1 647 258 5662
Email: ssouth@simonsouth.ca

Re: Delphi, Pascal, delphi-mode, pascal-mode, and free software

[Stefan Monnier]

>> One thing I'd like to see with delphi.el (other than renaming it) is to
>> use "standard" facilities like the syntax-table, rather than do all the
>> parsing by hand.
>> Do you have any idea what that would take?

> No, this is all still pretty new to me.  Is there a guide to best
> practices for major modes anywhere, or could you suggest a "good" mode I
> could use as an example?

I've been working on octave-mod.el to make it use generic facilities
(some of them brand new), so you could use it as a guide.
I don't vouch for the whole octave-mod.el file, but at least the
syntax-table, buffer-navigation, and indentation should be a reasonably
good example.  You can try it on test/indent/octave.m.

If you have a good Object Pascal file to use, we could add it as
test/indent/opascal.pas which could also be used for regression tests
(e.g. of the indentation code).


        Stefan