Guile.el

Guile.el

[tbrooke]

I've been trying to load guile.el from 1.90 without any luck. I'm not sure
it adds much to the existing scheme modes but I wanted to try it. It asks
for channels.scm and guile-scheme.scm but still doesn't seem to work. I'm
using Emacs 23.1.50 

Any thought

Tom
-- 
View this message in context: http://www.nabble.com/Guile.el-tp24253470p24253470.html
Sent from the Gnu - Guile - User mailing list archive at Nabble.com.

Re: Guile.el

[Neil Jerram]

tbrooke <tom.brooke@gmail.com> writes:

> I've been trying to load guile.el from 1.90 without any luck. I'm not sure
> it adds much to the existing scheme modes but I wanted to try it. It asks
> for channels.scm and guile-scheme.scm but still doesn't seem to work. I'm
> using Emacs 23.1.50 
>
> Any thought

What are you trying to do?  guile.el is pretty old, and I wouldn't be
surprised if it had bitrotted.

    Neil

Re: Guile.el

[Tom Brooke]

[-- Attachment #1: Type: text/plain, Size: 675 bytes --]

No Problem the scheme modes work fine - I just thought I read somewhere
about new improved debugger support that I thought was incorporated in the
.el

On Mon, Jun 29, 2009 at 2:40 PM, Neil Jerram <neil@ossau.uklinux.net> wrote:

> tbrooke <tom.brooke@gmail.com> writes:
>
> > I've been trying to load guile.el from 1.90 without any luck. I'm not
> sure
> > it adds much to the existing scheme modes but I wanted to try it. It asks
> > for channels.scm and guile-scheme.scm but still doesn't seem to work. I'm
> > using Emacs 23.1.50
> >
> > Any thought
>
> What are you trying to do?  guile.el is pretty old, and I wouldn't be
> surprised if it had bitrotted.
>
>    Neil
>

[-- Attachment #2: Type: text/html, Size: 1091 bytes --]

Re: Guile.el

[Andrew Bagdanov]

On Mon, Jun 29, 2009 at 9:59 PM, Tom Brooke<tom.brooke@gmail.com> wrote:
> No Problem the scheme modes work fine - I just thought I read somewhere
> about new improved debugger support that I thought was incorporated in the
> .el
>

For what it's worth, I have been fiddling with guile 1.90 recently and
Geiser works pretty damn well for me:

 http://programming-musings.org/2009/05/14/geiser/

Best,

-Andy

> On Mon, Jun 29, 2009 at 2:40 PM, Neil Jerram <neil@ossau.uklinux.net> wrote:
>>
>> tbrooke <tom.brooke@gmail.com> writes:
>>
>> > I've been trying to load guile.el from 1.90 without any luck. I'm not
>> > sure
>> > it adds much to the existing scheme modes but I wanted to try it. It
>> > asks
>> > for channels.scm and guile-scheme.scm but still doesn't seem to work.
>> > I'm
>> > using Emacs 23.1.50
>> >
>> > Any thought
>>
>> What are you trying to do?  guile.el is pretty old, and I wouldn't be
>> surprised if it had bitrotted.
>>
>>    Neil
>
>

Re: Guile.el

[Neil Jerram]

Andrew Bagdanov <bagdanov@gmail.com> writes:

> On Mon, Jun 29, 2009 at 9:59 PM, Tom Brooke<tom.brooke@gmail.com> wrote:
>> No Problem the scheme modes work fine - I just thought I read somewhere
>> about new improved debugger support that I thought was incorporated in the
>> .el
>>
>
> For what it's worth, I have been fiddling with guile 1.90 recently and
> Geiser works pretty damn well for me:
>
>  http://programming-musings.org/2009/05/14/geiser/

There's also GDS, which is bundled in 1.9.0 and documented in the
Guile manual.  A big caveat though: I haven't actually tried it since
all the new compilation support has gone in; so it may or may not
work.  If you try it, please let us know!

Regards,
        Neil

Re: Guile.el

[Ludovic Courtès]

Hi,

Tom Brooke <tom.brooke@gmail.com> writes:

> No Problem the scheme modes work fine - I just thought I read somewhere about
> new improved debugger support that I thought was incorporated in the .el

The "new debugging support" mentioned in 1.8.5's `NEWS' is GDS (info
"(guile) Using Guile in Emacs"), which allows users to evaluate Scheme
code, single-step through it, display backtraces, etc.

You may also want to look at Geiser, as was already mentioned.  It
provides a REPL, autodoc (documentation of the object at point), smart
completion, and more.

Thanks,
Ludo'.

Re: Guile.el

[Dmitry Dzhus]

Ludovic Courtès wrote:

> The "new debugging support" mentioned in 1.8.5's `NEWS' is GDS (info
> "(guile) Using Guile in Emacs"), which allows users to evaluate Scheme
> code, single-step through it, display backtraces, etc.

> You may also want to look at Geiser, as was already mentioned.  It
> provides a REPL, autodoc (documentation of the object at point), smart
> completion, and more.

I wonder if something can be hacked for Guile to make it work with
SLIME. Somehow this GDS thing seems to resemble SLIME to me. Amirite?
-- 
Happy Hacking.

http://sphinx.net.ru
む

Re: Guile.el

[Neil Jerram]

Dmitry Dzhus <dima@sphinx.net.ru> writes:

> Ludovic Courtès wrote:
>
>> The "new debugging support" mentioned in 1.8.5's `NEWS' is GDS (info
>> "(guile) Using Guile in Emacs"), which allows users to evaluate Scheme
>> code, single-step through it, display backtraces, etc.
>
>> You may also want to look at Geiser, as was already mentioned.  It
>> provides a REPL, autodoc (documentation of the object at point), smart
>> completion, and more.
>
> I wonder if something can be hacked for Guile to make it work with
> SLIME. Somehow this GDS thing seems to resemble SLIME to me. Amirite?

In general terms, yes.  GDS, SLIME and Geiser are all in the same
ballpark.

I don't like NIH syndrome, so I'd happily allow GDS to wither if there
was a generic Scheme/Emacs debugging and autodoc etc. environment that
could do all the same things.

I'm afraid I haven't looked at SLIME or Geiser enough in detail yet,
but there are two things about GDS that I think are significantly
different.

1. Using the trap callouts from Guile's evaluator.  As opposed - I
guess - to instrumenting and reevaluating the code that you want to
step through.  Given the move towards the VM though, this point is
very much in flux now.

2. Not relying on standard input and output for communication with the
Guile process that is being debugged.  I think this is important,
because apps often use standard input and output for other things!

Regards,
        Neil

Re: Guile.el

[Jose A. Ortega Ruiz]

Neil Jerram <neil@ossau.uklinux.net> writes:

> 2. Not relying on standard input and output for communication with the
> Guile process that is being debugged.  I think this is important,
> because apps often use standard input and output for other things!

geiser is based on comint-mode, meaning (as i'm sure you know) that
reliance on input/ouput is very easy to get rid of: comint can handle
telnet connections transparently, so it's just a matter of launching a
remote repl on guile's side, without any change to the elisp code,
geiser's scheme support libraries. i've just been lazy writing the repl
server thing :)

jao