Two Flavors of ZSH ?

Two Flavors of ZSH ?

[Eskandar Ensafi]

Hello,

I used to be involved in ZSH development, but I have been away for
about two years.

I recall that around the time that I was still involved (ZSH 2.6
was the official release and talk about ZSH 3.0 had just started),
there were two flavors of ZSH: the official distribution and the
Zefram-patched distribution.

It was my understanding that when Zoltan Hidvegi took over the list
from Richard Coleman (is Zoltan still in charge of things?), he was
going to incorporate Zefram's patches into the official ZSH release
in the interest of preventing ZSH from proliferating into several
different and potentially incompatible products.

After I rejoined this mailing list, I noticed a zsh-3.1.2-zefram3
patch.  Does this mean that we have more than one flavor of ZSH in
the world?

Regards,
Eskandar

P.S.  Is dynamic loading available in the current ZSH, and if so,
      is it stable on supported platforms?  A few years ago, I was
      going to add a module called Objective-Z to add Objective-C
      support to ZSH.  I gave up on it because it required custom
      patches to ZSH and worked only on systems with GNU C (in fact,
      my prototype only ran under NEXTSTEP, which is now OPENSTEP).
      Perhaps integrated dynamic loading supprt will make my task
      simpler and more portable (the idea was to dynamically load
      Objective-C classes to extend the shell's functionality).

--
Eskandar Ensafi               <esky@cs.ucla.edu>
Space Computer Corporation    <ensafi@spacecomputer.com>
http://www.spacecomputer.com

Re: Two Flavors of ZSH ?

[Andrew Main]

Eskandar Ensafi wrote:
>I recall that around the time that I was still involved (ZSH 2.6
>was the official release and talk about ZSH 3.0 had just started),
>there were two flavors of ZSH: the official distribution and the
>Zefram-patched distribution.

That would be the Zoltan-patched distribution.  When Zoltan took over,
his patches went into the baseline.

>After I rejoined this mailing list, I noticed a zsh-3.1.2-zefram3
>patch.  Does this mean that we have more than one flavor of ZSH in
>the world?

That's a very recent development.  Zoltan's too busy to make releases
at the moment, so I made an unofficial release with current patches,
in order to provide a base for further development.  This shouldn't be
a significant branch, but we're still waiting for Zoltan.

-zefram

Re: Two Flavors of ZSH ?

[Bruce Stephens]

esky@CS.UCLA.EDU said:
> It was my understanding that when Zoltan Hidvegi took over the list
> from Richard Coleman (is Zoltan still in charge of things?), he was
> going to incorporate Zefram's patches into the official ZSH release in
> the interest of preventing ZSH from proliferating into several
> different and potentially incompatible products. 

I think that's still the plan.  It's just that there hasn't been a release 
from Zoltan for a long time, for some reason.  I'd hope that everything that 
Zefram's version has will be in the official release; I don't recall anything 
controversial (other than the bashisms in the compilation scripts, which were 
a oversight on Zefram's part).

esky@CS.UCLA.EDU said:
> P.S.  Is dynamic loading available in the current ZSH, and if so,
>       is it stable on supported platforms? 

I think so.  I only have access to Solaris2.5.1 and Linux (elf, 
2.0.something). On those platforms where it's supported, I think zle and other 
pretty important things are modules (typically), so I'd expect more complaints 
if it didn't work.

Re: Two Flavors of ZSH ?

[Andrej Borsenkow]

On Wed, 28 Jan 1998, Eskandar Ensafi wrote:

> 
> P.S.  Is dynamic loading available in the current ZSH, and if so,
>       is it stable on supported platforms?  A few years ago, I was
>       going to add a module called Objective-Z to add Objective-C
>       support to ZSH.  I gave up on it because it required custom
>       patches to ZSH and worked only on systems with GNU C (in fact,
>       my prototype only ran under NEXTSTEP, which is now OPENSTEP).
>       Perhaps integrated dynamic loading supprt will make my task
>       simpler and more portable (the idea was to dynamically load
>       Objective-C classes to extend the shell's functionality).
> 

I am working with dynaload since it first appeared (zsh-3.1.0) and haven't
got any noticable problems (apart from initial build quirks - but they
should be sorted now).

As for your question ... A dynamically loaded module can currently provide
new builtin and zle widget. It cannot e.g. add new option, new completion
rule or like. And it definitely cannot change/extend syntax and/or
grammar. So, if you Objetive-Z fits in additional builtin paradigm - yes,
it is doable now. But if you planned to extend grammmar to add Objective C
- I am afraid, not.


-------------------------------------------------------------------------
Andrej Borsenkow 		Fax:   +7 (095) 252 01 05
SNI ITS Moscow			Tel:   +7 (095) 252 13 88

NERV:  borsenkow.msk		E-Mail: borsenkow.msk@sni.de
-------------------------------------------------------------------------