any news on 1.6 yet?

any news on 1.6 yet?

[Han-Wen Nienhuys]

Sorry for pestering you again about the guile release, but I'm
starting to get a little impatient. The mailing list has been quite
dead for months, and it seems that development is being halted until
the new stable release is out of the door.

Also, my friend Jan submitted a patch for building GUILE crosscompiled
to a cygwin target,

  http://mail.gnu.org/pipermail/guile-devel/2002-February/004562.html

which was met with

>Thanks very much.  I'd like to wait until the people more familiar
>with this particular issue can check the patch out before we apply it.
>So this fix may not make it into 1.5.7, but presuming everything's
>fine, it should make it into 1.5.8.
>
>Thanks again.

I'm under the distinct impression that "the people more familiar" are
more busy with putting various air castles into workbook files instead
of evaluating and writing code.

And if this sounds bitter, then you're darn right.  We spend a lot of
time with juggling around with various GUILE versions to find out what
bugs are coming from what combinations, and it would make our life a
lot easier if we would have one single, standardized, up-to-date GUILE
version.

But perhaps we're just whining, since we're the only guys that seem to
actually _use_ GUILE.

ObConstructive

Can someone puleaze put in the following remark into the GUILE
documentation?


  Be warned that the thread library (pthreads) will impose a 512k
  limit on the stack size, when compiling GUILE with thread
  support. 

We spent a _long_ time being flabbergasted at why our program dumped
core on perfectly valid data and innocent instructions.

--
Han-Wen Nienhuys   |   hanwen@cs.uu.nl    | http://www.cs.uu.nl/~hanwen/


_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel

Re: any news on 1.6 yet?

[Rob Browning]

Han-Wen Nienhuys <hanwen@cs.uu.nl> writes:

> Sorry for pestering you again about the guile release, but I'm
> starting to get a little impatient. The mailing list has been quite
> dead for months, and it seems that development is being halted until
> the new stable release is out of the door.

Speaking for myself, that's true -- I'm not really working on HEAD
ATM, but I do have a number of things that should go into HEAD once
I'm ready -- i.e a month or two ago I got GMP working for bignums.  If
we can decide how to handle the licensing, those changes can go in.

In any case, now that various issues in my real-life have settled
down.  I'm now starting back up work on the release, FWIW.  I'd like
to give you a schedule, but I can't yet.  However, I should have a new
status update posted in less than a week.

> Also, my friend Jan submitted a patch for building GUILE crosscompiled
> to a cygwin target,
>
>   http://mail.gnu.org/pipermail/guile-devel/2002-February/004562.html
>
> which was met with
>
>>Thanks very much.  I'd like to wait until the people more familiar
>>with this particular issue can check the patch out before we apply it.
>>So this fix may not make it into 1.5.7, but presuming everything's
>>fine, it should make it into 1.5.8.
>
> I'm under the distinct impression that "the people more familiar" are
> more busy with putting various air castles into workbook files instead
> of evaluating and writing code.

I will look back at this patch.  I'd also like to hear if Marius (or
anyone else) has any comments.

> Can someone puleaze put in the following remark into the GUILE
> documentation?
>
>
>   Be warned that the thread library (pthreads) will impose a 512k
>   limit on the stack size, when compiling GUILE with thread
>   support. 
>
> We spent a _long_ time being flabbergasted at why our program dumped
> core on perfectly valid data and innocent instructions.

Wow, that's annoying -- do you know if this is an arch dependency, OS
dependency, etc?

-- 
Rob Browning
rlb @defaultvalue.org, @linuxdevel.com, and @debian.org
Previously @cs.utexas.edu
GPG=1C58 8B2C FB5E 3F64 EA5C  64AE 78FE E5FE F0CB A0AD

_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel

Re: any news on 1.6 yet?

[Han-Wen]

rlb@defaultvalue.org writes:
> > Can someone puleaze put in the following remark into the GUILE
> > documentation?
> >
> >
> >   Be warned that the thread library (pthreads) will impose a 512k
> >   limit on the stack size, when compiling GUILE with thread
> >   support. 
> >
> > We spent a _long_ time being flabbergasted at why our program dumped
> > core on perfectly valid data and innocent instructions.
> 
> Wow, that's annoying -- do you know if this is an arch dependency, OS
> dependency, etc?

You're f*cked once you link with libpthreads (I could produce a crash
linking to guile --without-threads and libpthreads). Apparently the
thread package does include some functions to set the stackspace, but
those must be called before a thread is spawned. I know zilch about
threads, so I don't know the details behind this.

We managed to get the stack usage of our functions and GC mark
functions down (which is probably a good idea performance wise), but
it was annoying nevertheless.

-- 

Han-Wen Nienhuys   |   hanwen@cs.uu.nl   |   http://www.cs.uu.nl/~hanwen 

_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel