From: Johan Parin
Subject: Problem w/ ILISP 5.8 + XEmacs 20.4 + Allegro 4.3.1
Date:
Message-ID: <kviu7gahce.fsf@uab.ericsson.se>
Hello everybody,
After switching to XEmacs 20.4 I have problems getting ILISP to
work. The supplied ILISP version is 5.8.
What happens is that when I try to do e.g. C-c l (to load a file), it
displays
Started initializing ILISP
and nothing more happens. If I try to interact with the lisp after
that, I get the error message:
As of 20.3, `concat' no longer works with compiled-function objects
...
What could be the problem? Solutions?
Any help is appreciated.
BTW, ILISP seems to work fine with CMU CL, it's only with Allegro I
have this problem.
---------------------------------------------------------------------------
Johan Parin | phone: +46 (8) 7274036
Verification Engineer | fax: +46 (8) 7273830
AXE Control System Design Group | email: ···········@uab.ericsson.se
Ericsson AXE Research & Development |
G�talandsv�gen 230 |
Box 1505, 125 25 �LVSJ� |
SWEDEN |
----------------------------------------------------------------------------
From: William Deakin
Subject: Re: Problem w/ ILISP 5.8 + XEmacs 20.4 + Allegro 4.3.1
Date:
Message-ID: <379339E8.7CCEC054@pindar.com>
Johan Parin wrote:
> After switching to XEmacs 20.4 I have problems getting ILISP to
> work. The supplied ILISP version is 5.8.
Having installed the inferior lisp package with xemacs 21.1-p3, I cannot get
ilisp to work with CLISP.
I have posted this to comp.emacs.xemacs and gnu.emacs.help and thought that this
was a CLISP problem, but I also have observed the "concat is no longer
supported with compiled code after version 20.3." There are 129 concat's in the
ilisp code.
I'm sorry for cross posting.
:-) will
William Deakin <·····@pindar.com> writes:
> Johan Parin wrote:
>
> > After switching to XEmacs 20.4 I have problems getting ILISP to
> > work. The supplied ILISP version is 5.8.
>
> Having installed the inferior lisp package with xemacs 21.1-p3, I cannot get
> ilisp to work with CLISP.
>
> I have posted this to comp.emacs.xemacs and gnu.emacs.help and
> thought that this
> was a CLISP problem, but I also have observed the "concat is no longer
> supported with compiled code after version 20.3." There are 129
> concat's in the
> ilisp code.
>
From what I see, there have been a lot of changes in the XEmacs
distribution, which ended up breaking the ILISP code. The "concat"
problem seems very ugly, just to mention one, but I believe that
something more fundamental is going on.
Since I am not a XEmacs user, I cannot really help you right away.
If anybody has any clue about what is happening with XEmacs/ILISP,
please let me know or write to ILISP mailing list. (for the time
being, but not much longer, <·····@naggum.no>).
Cheers
--
Marco Antoniotti ===========================================
PARADES, Via San Pantaleo 66, I-00186 Rome, ITALY
tel. +39 - 06 68 10 03 17, fax. +39 - 06 68 80 79 26
http://www.parades.rm.cnr.it/~marcoxa
* Marco Antoniotti wrote:
> From what I see, there have been a lot of changes in the XEmacs
> distribution, which ended up breaking the ILISP code. The "concat"
> problem seems very ugly, just to mention one, but I believe that
> something more fundamental is going on.
concat has change (is going to change from my point of view) in at
least a couple of ways:
(concat 1 2 3)
used to produce
"123"
and does something else now.
I think it also used to work to concatenate byte-compiled functions
because they were just strings. I don't think that works now, and it
shouldn't work in any post-v19 emacs (whether or not it is XEmacs).
Perhaps the right thing is to send a backtrace (turn debug-on-error
on, and see where it barfs) to the list.
--tim
(posted here because I'm not on the ilisp list any more)
From: William Deakin
Subject: Re: Problem w/ ILISP 5.8 + XEmacs 20.4 + Allegro 4.3.1
Date:
Message-ID: <37958850.5CDC33AD@pindar.com>
Tim Bradshaw wrote:
> Perhaps the right thing is to send a backtrace (turn debug-on-error on,
> and see where it barfs) to the list.
Hmmm. I've just tried this and now suspect this is a collusion between my
incompetence and my inexperience of using clisp and xemacs, but I cannot
get xemacs running clisp to barf, I can only get it to hang. For example:
First deleted all compiled ilisp files and then:
1) started xemacs and in the *scratch* buffer and set debug on error in
the lisp-interaction menu.
2) run ilisp (typing ESC-x load-library and ilisp at the corresponding
prompt)
3) run clisp (typing ESC-x run-ilisp and clisp-hs at the corresponding
prompt)
clisp in the clisp buffer then works quite happily but if I try and get a
command from another buffer to execute:
1) open a file foo.lisp
2) type something in the file like (setf foo "bar")
3) type ESC-C-x to execute this command in clisp
xemacs then hangs. But no barf. The command line say 'Started
initialising ILISP' and the lisp buffers say (ILISP :interupt). If I
repeat this with debug-on-signal and debug-on-warning switched on I still
cannot get any barf. If I break out the clisp process is knackered
reporting a 'wrong type argument: characterp lambda'.
Sorry about this. BUT IT IS DRIVING ME INSANE. My wife has already has
had words with me about being bad tempered about the house....
:-| Will
William Deakin <·····@pindar.com> writes:
> Tim Bradshaw wrote:
>
> > Perhaps the right thing is to send a backtrace (turn debug-on-error on,
> > and see where it barfs) to the list.
>
> Hmmm. I've just tried this and now suspect this is a collusion between my
> incompetence and my inexperience of using clisp and xemacs, but I cannot
> get xemacs running clisp to barf, I can only get it to hang. For example:
>
> First deleted all compiled ilisp files and then:
> 1) started xemacs and in the *scratch* buffer and set debug on error in
> the lisp-interaction menu.
> 2) run ilisp (typing ESC-x load-library and ilisp at the corresponding
> prompt)
> 3) run clisp (typing ESC-x run-ilisp and clisp-hs at the corresponding
> prompt)
>
> clisp in the clisp buffer then works quite happily but if I try and get a
> command from another buffer to execute:
>
> 1) open a file foo.lisp
> 2) type something in the file like (setf foo "bar")
> 3) type ESC-C-x to execute this command in clisp
>
> xemacs then hangs. But no barf. The command line say 'Started
> initialising ILISP' and the lisp buffers say (ILISP :interupt). If I
> repeat this with debug-on-signal and debug-on-warning switched on I still
> cannot get any barf. If I break out the clisp process is knackered
> reporting a 'wrong type argument: characterp lambda'.
>
> Sorry about this. BUT IT IS DRIVING ME INSANE. My wife has already has
> had words with me about being bad tempered about the house....
This has (may have) been fixed in the upcoming 5.9.
Put
(setf clisp-hs-program "clisp -I")
somewhere in your init files.
Marco
--
Marco Antoniotti ===========================================
PARADES, Via San Pantaleo 66, I-00186 Rome, ITALY
tel. +39 - 06 68 10 03 17, fax. +39 - 06 68 80 79 26
http://www.parades.rm.cnr.it/~marcoxa
From: William Deakin
Subject: Re: Problem w/ ILISP 5.8 + XEmacs 20.4 + Allegro 4.3.1
Date:
Message-ID: <37959502.899DAB8A@pindar.com>
Marco Antoniotti wrote:
> This has (may have) been fixed in the upcoming 5.9.
Thank you. This sounds like music to my ears! Do you need anybody to test this?
Or is there anyway I can help you with your excellent work?
> Put (setf clisp-hs-program "clisp -I") somewhere in your init files.
I am apologise that I omitted to say that I have run clisp with all
permutations of -I (and -a) and that this seemed to make no difference :-(
Best Regards,
:-) Will
From: Tim Bradshaw
Subject: Re: Problem w/ ILISP 5.8 + XEmacs 20.4 + Allegro 4.3.1
Date:
Message-ID: <nkjn1wqb8ux.fsf@tfeb.org>
William Deakin <·····@pindar.com> writes:
> 1) open a file foo.lisp
> 2) type something in the file like (setf foo "bar")
> 3) type ESC-C-x to execute this command in clisp
>
> xemacs then hangs. But no barf. The command line say 'Started
> initialising ILISP' and the lisp buffers say (ILISP :interupt). If I
> repeat this with debug-on-signal and debug-on-warning switched on I still
> cannot get any barf. If I break out the clisp process is knackered
> reporting a 'wrong type argument: characterp lambda'.
This kind of lossage is really common. ilisp does all sorts of
subprocess fiddlery which can just leave it wedged. It's a lot better
now (at least for me) but it used to be just hopeless. You can
sometimes fix it by making sure the inits are compiled
(ilisp-compile-inits?) because some of it is races I think, and you
can also recover, sometimes, by aborting and then doing a repair-ilisp
which tries to get things back to a sane state. Failing that you need
to ask an ilisp guru I think.
--tim
From: Johan Parin
Subject: Re: Problem w/ ILISP 5.8 + XEmacs 20.4 + Allegro 4.3.1
Date:
Message-ID: <kvso6i59w4.fsf@uab.ericsson.se>
Tim Bradshaw writes:
Tim> William Deakin <·····@pindar.com> writes:
<snip>
>>
>> xemacs then hangs. But no barf. The command line say 'Started
>> initialising ILISP' and the lisp buffers say (ILISP :interupt). If I
>> repeat this with debug-on-signal and debug-on-warning switched on I still
>> cannot get any barf. If I break out the clisp process is knackered
>> reporting a 'wrong type argument: characterp lambda'.
Tim> This kind of lossage is really common. ilisp does all sorts of
Tim> subprocess fiddlery which can just leave it wedged. It's a lot
Tim> better now (at least for me) but it used to be just hopeless.
Tim> You can sometimes fix it by making sure the inits are compiled
Tim> (ilisp-compile-inits?) because some of it is races I think, and
Tim> you can also recover, sometimes, by aborting and then doing a
Tim> repair-ilisp which tries to get things back to a sane state.
Tim> Failing that you need to ask an ilisp guru I think.
I think I may have failed to clarify in my original message that the
concat problem only occurs when I interrupt it in the "Initializing
ILISP" state, so the concat problem is really only part of the
issue. The main problem is the hung state.
I tried to compile the Allegro init file, but this did not help.
I also tried to do repair-ilisp in the concat error state, this only
got me out of the concat error state, I still got the hung problem
when I tried to load the file again.
BTW Is there a current ILISP WWW site - the link at the XEmacs home
page points to nothing. I would like to subscribe to the mailing list
and retrieve an archive, if possible.
Thanks.
- Johan Parin
Tim Bradshaw <···@tfeb.org> writes:
> * Marco Antoniotti wrote:
>
> > From what I see, there have been a lot of changes in the XEmacs
> > distribution, which ended up breaking the ILISP code. The "concat"
> > problem seems very ugly, just to mention one, but I believe that
> > something more fundamental is going on.
>
> concat has change (is going to change from my point of view) in at
> least a couple of ways:
>
> (concat 1 2 3)
>
> used to produce
>
> "123"
>
> and does something else now.
Al right! That's it. I will check the ILISP sources and change all
the (CONCAT ...) with (CONCATENATE 'STRING ...) if possible. :)
>
> I think it also used to work to concatenate byte-compiled functions
> because they were just strings. I don't think that works now, and it
> shouldn't work in any post-v19 emacs (whether or not it is XEmacs).
>
> Perhaps the right thing is to send a backtrace (turn debug-on-error
> on, and see where it barfs) to the list.
Cheers
--
Marco Antoniotti ===========================================
PARADES, Via San Pantaleo 66, I-00186 Rome, ITALY
tel. +39 - 06 68 10 03 17, fax. +39 - 06 68 80 79 26
http://www.parades.rm.cnr.it/~marcoxa