emacs recoded in guile

I heard there was a project to convert emacs code from emacs lisp to guile. I think the project has stalled but I think this should be made a priority by FSF to go through with this because emacs lisp is an antiquated dialect of lisp that lacks lexical scoping. If emacs was coded in guile I think that would attract a lot of people to emacs as it would be a lot better to do configuration scripting in a popular lisp dialect rather than have to learn a dialect that you will never use for anything else. In fact people might start learning emacs just to be able to use guile. I think guile is moderately fast so it might be possible to replace the small amount of C code in emacs and make it a pure lisp app

People learn Emacs in order to use ANY Lisp, because it's the only editor that doesn't suck for writing Lisp.  Emacs Lisp is going to stay until every. last. .el script. has been converted to Guile, not just the main ones but the 20 years worth of add-ons.

>>2
People learn Emacs in order to use ANY Lisp, because it's the only editor that doesn't suck for writing Lisp.
This is why I switched to vi.

>>1
You just have no idea how big a job that would be, do you?

>>4
actually I do, its not just a matter of translating from one language to another or even refactoring code to fit a new language. Emacs would have to be redesigned from the ground up by experts not only in guile/scheme but by programmers who are experienced in designing complex systems. And all the scripts would have to be redesigned to interface with this new system.

tldr: its a job much bigger than the freetards at FSF who copied/rebranded unix utilities code could handle

>>5
Implying that emacs was originally a unix utility.

Implying that one can't buy boring programming work

>>1,5
Nobody is going to replace all elisp code in emacs. It was tried with MIT/GNU Scheme and Edwin and that didn't go anywhere. The idea now is to use the Guile VM to run elisp, scheme and javascript in unison.

http://lists.gnu.org/archive/html/emacs-devel/2010-04/msg00665.html

The idea now is to use the Guile VM to run elisp, scheme and javascript in unison.
that plan sucks

rather than properly rebuild the app they just want to do the easier job of giving it scriptability in more popular languages and hope the community will translate the emacs lisp scripts on their own. This might add new users in the short term but will make emacs more of a incomprehensible mess in the long term

emacs is an important enough editor that it deserves a complete rewrite that would keep all existing functionality. Its a job that cant be done by volunteers but would would require major investment by corporations whos employees use emacs. Perhaps much of the heavy work could be outsourced

anyone know of any statistics released by major corporations on the editors their employees use? I think people would be surprised to know how common emacs is used in the IT industry

>>2
Needing an IDE to write in a language indicates a seriously flawed language.

Whats wrong with elisp? or emacs?
As far as i know guile sucks.

>Whats wrong with elisp?
I already said, but I'll say again, its a very old and obsolete dialect of lisp that no one uses except for emacs. A person should not have to learn a programming language they will only use for one app.

>or emacs?
emacs is an ancient app that was very limited by the hardware it originally ran on. it cant continue to just be added on to, at some point the design has to be organized into one unified design. when emacs was designed lisp could not even run on a regular computer because it was too memory intensive and specially computers had to be designed just to run lisp (look up lisp machines). We live in an age now where Lisp can run on anything, emacs needs to be redesigned to be a lisp only app

>As far as i know guile sucks.
please tell me why guile sucks, because guile is nothing more than an implementation of scheme, so are you saying scheme sucks? or is guile just a bad implementation of scheme?

>>12
I want to further mention that lisp and functional programming languages in general are very mainstream now. It is very rare that you will meet a CS student or IT professional who has not at least tried such a language. So it should really be no problem hiring qualified professionals to work on redesigning emacs

And I wish more Emacs code would be written in CL(actual CL, not just CL on top of Emacs), but that's just not going to happen (outside of all the Emacs clones like Climacs, Zmacs(OpenGenera, etc), Hemlock, etc).

The reality is that Emacs has a large legacy of code written for it, and you would either need to replace its core (with a functional equivalent), while making sure compatibility is preserved and it's not buggy. In the end, you might just see that it's a lot of work, especially in the light of all the existing non-core code, and the benefit of rewriting isn't all that huge, especially when you can still use non-elisp dialects (such as the incomplete cl mode).

>>12
please tell me why guile sucks
I have had bad experiences with guile, its a pain in the ass to debug.

I agree with the rest tough.
But to be honest most of these are not very usefull in practice and dangerous. Redesigning a such large programm such as emacs is very risky.
Also emacs runs just fine the way i see it and elisp is not really good, but neither is it bad.

>>1,5,9,14
Guile already supports elisp the language, just type ,language elisp at the prompt( I'm sure there is/was a plan to get lua running on it too). All that you should need to do (although I'd guess there is plenty more) is to implement all the elisp primitives for editing (drawing the screen etc.) and you should be able to drop the existing elisp on top of it.

>>8
Edwins development hasn't kept up with emacs trunk, but it's still a working editor and mail client, and if you're using MIT scheme it's passable. There is a separate effort to update Edwin (edwin48?), but I never kept up the news on it.

>>15
Guile has changed quite a lot over the past few years, so it may be worth another look if this wasn't recently. My own judgement of Guile was based on the 1.7.* series, and I was pleasantly surprised by 1.9.*.

Guile is bloat.
http://code.google.com/p/femtolisp/

And I wish more Emacs code would be written in CL
scheme (and its embedded guile implementation) is a much simpler dialect of lisp than CL and so would be much more suitable as a scripting/configuration language for emacs

>>18
Why do you think programming for the Emacs plaform should be done in a "simpler" language than programming for other platforms?

>>19
same reason that most all embedded languages are simple, it makes things easier to extend and build on. the basic rule of any API is to make functionality as simple as possible because its very easy to add complexity whereas its almost impossible to take away complexity....case in point being emacs, simple restructuring will not suffice, but rather a redesign by someone who understands emacs and what major changes have to be made while keeping as much of the original functionality as possible

now that QT is GPLed it would be nice to see Scheme/Guile added to the list of languages with QT bindings, CL is already on the list
http://en.wikipedia.org/wiki/Qt_%28framework%29#Bindings

>>20
Why does that apply less to programming for Windows?

>>22
Away foul troll, and return hence to thine bridge.

>>21
IIRC chicken has QT bindings