🚀 go-pugleaf

In article <MARCUS.94Nov1172625@jetsam.cs.pdx.edu>, marcus@jetsam.cs.pdx.edu (Marcus Daniels) writes:
>>>>>> "TC" == Tom Christiansen <tchrist@mox.perl.com> writes:
> In article <395p66$2j4@csnews.cs.Colorado.EDU> Tom Christiansen <tchrist@mox.perl.com> writes:

TC> At the risk of repeating myself, I'll state that in this
TC> community, only technical criteria matter.  Politics do not drive
TC> the technical, but vice versa.  To see it as otherwise is to fight
TC> a losing battle.  Until gnu script appears, it's vaporware of a
TC> purely political nature, and just so much hype and hormones.  Once
TC> it does, it has the potential of being a technical wonder and
TC> filling some gaping niche (which most of us can't see), at which
TC> point it will be judged on its technical merits.

> * The hero of Perl, the advocate of the `anti-elitist' language, is
>   now upset that GNU would dare propose an alternative substrate
>   for extention languages. 

A slight clarification: While Tom Christiansen has done a lot for Perl and is
definitely its most prominent evangelist, Perl was created by Larry Wall, as
opposed to what your phrase "the hero of Perl" would seem to imply.

> * Okay, now we see this is confusing, so the
>   only thing left to do is throw up some FUD.  Did you know that 
>   Gel is out to stop you from implementing and using languages
>   like Perl and Python?!  Yes it is true!

I respect the effort that Tom Lord et al. will put into Gel, but I can't avoid
the impression that Gel was an afterthought to a mostly politically motivated
campaign against Tcl:

 - rms started the discussion with an article entitled "Why you should not use
   Tcl" (Not: "Why we will not use Tcl for project GNU" or "The wonderful
   language we will create").

 - The alternative he proposed was at that moment pure vaporware; it had
   neither a name nor people assigned to it, nor any clear design principles
   except to interface to Tk using as many parentheses as possible :-)

 - After initial public discussion, rms posted another article showing that
   he at least had done some research afterwards to back up his initial claims.
   It was only at that point that anything remotely resembling a specific
   plan for an extension language emerged.

Now Gel at least has a name and personnel, but I still don't think it will be
much of a success:

 - While TCL, Perl, and probably also Python, are mature now, Gel will take at
   least half a year to implement, and probably a lot more until a significant
   GNU program uses it.
 - Even what little is known about Gel now is far less than
   uncontroversial. Even the Scheme community seems to be rather unenthusiastic
   about some of the proposed extensions to Scheme.
 - And finally, as an *extension* language, Gel will be unuseable for
   commercial programs. This point seems to have been frequently misunderstood:
   the question is not whether *user scripts* written in Gel will be
   automatically copylefted: they won't be. The problem is that the Gel
   *system* will have to be linked into the program to be scripted, and if Gel
   is copylefted, every scriptable program using it would have to be
   copylefted. This situation is significantly different from language
   implementations like gcc, where copylefting has not hindered commercial use.

So in summary, Gel is likely to be late, to offer no significant improvements,
and to impose crippling restrictions on non-GNU use. Considering that, Tom
Christiansen's assessment in the Subject: line seems to be on the mark.

Matthias

-----
Matthias Neeracher <neeri@iis.ee.ethz.ch> http://err.ethz.ch/members/neeri.html
   "There once was an Age of Reason, but we've progressed beyond it."
                                   -- Ayn Rand, _Atlas Shrugged_