Lightweight, C-only implementation of Emacs

Lightweight, C-only implementation of Emacs

[Jeffrey Walton]

Hi Everyone,

I'm trying to find a lightweight C-only implementation of Emacs for
Linux. I need it for situations like this:
https://redmine.pfsense.org/issues/9682 .

Searching is pretty much useless due to irrelevant results. I found
Zile at https://www.gnu.org/software/software.html#allgnupkgs , but it
fails to build (it configures OK).

I understand I will probably loose some functionality due to decoupling Lisp.

Does anyone have a list of Emacs replacements?

Thanks in advance.

Jeff

Re: Lightweight, C-only implementation of Emacs

[Bob Newell]

> I'm trying to find a lightweight C-only implementation of Emacs

You might give 'fe' a try. It's a little-known editor that, while very
far from being a full Emacs, has just enough Emacs functionality to be
highly useful for many editing tasks. It has the advantage of being
very small and very fast. And, as I've used it for many years now and
it's become an essential tool for me, I've learned to compile it in
lots of environments. Maybe it will work for you (if all you want to
do is edit text). Find it here:

http://www.moria.de/~michael/fe/

-- 
Bob Newell
Honolulu, Hawai`i

Via Linux/Emacs/Gnus/BBDB.

Re: Lightweight, C-only implementation of Emacs

[Noam Postavsky]

On Sat, 17 Aug 2019 at 21:26, Jeffrey Walton <noloader@gmail.com> wrote:

> I'm trying to find a lightweight C-only implementation of Emacs for
> Linux. I need it for situations like this:
> https://redmine.pfsense.org/issues/9682 .

There are few alternatives listed here:
https://www.emacswiki.org/emacs/ErsatzEmacs

You can also use TRAMP for this kind of thing.

Re: Lightweight, C-only implementation of Emacs

[Stefan Monnier]

> Searching is pretty much useless due to irrelevant results. I found
> Zile at https://www.gnu.org/software/software.html#allgnupkgs , but it
> fails to build (it configures OK).

I believe nowadays Zile is the most common answer to your question.
It builds well enough for Debian AFAICT, so you can likely fix the build
(e.g. by contacting the maintainer).


        Stefan

Re: Lightweight, C-only implementation of Emacs

[Ulrich Mueller]

>>>>> On Sat, 17 Aug 2019, Jeffrey Walton wrote:

> I'm trying to find a lightweight C-only implementation of Emacs for
> Linux. I need it for situations like this:
> https://redmine.pfsense.org/issues/9682 .

> Searching is pretty much useless due to irrelevant results. I found
> Zile at https://www.gnu.org/software/software.html#allgnupkgs , but it
> fails to build (it configures OK).

> I understand I will probably loose some functionality due to decoupling Lisp.

> Does anyone have a list of Emacs replacements?

There is a list of Emacs implementations in the Gentoo wiki:
https://wiki.gentoo.org/wiki/Project:Emacs/Emacs-like_editors

I also did a comparision of small Emacsen several years ago:
https://bugs.gentoo.org/384451 (especially, see the table attached at
https://384451.bugs.gentoo.org/attachment.cgi?id=287721).

Re: Lightweight, C-only implementation of Emacs

[Jeffrey Walton]

On Mon, Aug 19, 2019 at 1:36 AM Ulrich Mueller <ulm@gentoo.org> wrote:
> ...
> > I understand I will probably loose some functionality due to decoupling Lisp.
>
> > Does anyone have a list of Emacs replacements?
>
> There is a list of Emacs implementations in the Gentoo wiki:
> https://wiki.gentoo.org/wiki/Project:Emacs/Emacs-like_editors

You should probably take Zile off the list or make a note it is not
recommended for use. According to Reuben Thomas, one of the Zile
maintainers (from Zile Issue 58880):

    I consider Zile to be in maintenance mode. I do not
    recommend it use, I will not be adding features ...

Jeff

Re: Lightweight, C-only implementation of Emacs

[Ulrich Mueller]

>>>>> On Mon, 03 Aug 2020, Jeffrey Walton wrote:

>> > Does anyone have a list of Emacs replacements?
>> 
>> There is a list of Emacs implementations in the Gentoo wiki:
>> https://wiki.gentoo.org/wiki/Project:Emacs/Emacs-like_editors

> You should probably take Zile off the list or make a note it is not
> recommended for use. According to Reuben Thomas, one of the Zile
> maintainers (from Zile Issue 58880):

>     I consider Zile to be in maintenance mode. I do not
>     recommend it use, I will not be adding features ...

Well, it is feature complete then.

The same is true for most of the lightweight implementations in the
list. Why would that be a problem?

Re: Lightweight, C-only implementation of Emacs

[Jeffrey Walton]

On Mon, Aug 3, 2020 at 3:37 PM Ulrich Mueller <ulm@gentoo.org> wrote:
>
> >>>>> On Mon, 03 Aug 2020, Jeffrey Walton wrote:
>
> >> > Does anyone have a list of Emacs replacements?
> >>
> >> There is a list of Emacs implementations in the Gentoo wiki:
> >> https://wiki.gentoo.org/wiki/Project:Emacs/Emacs-like_editors
>
> > You should probably take Zile off the list or make a note it is not
> > recommended for use. According to Reuben Thomas, one of the Zile
> > maintainers (from Zile Issue 58880):
>
> >     I consider Zile to be in maintenance mode. I do not
> >     recommend it use, I will not be adding features ...
>
> Well, it is feature complete then.
>
> The same is true for most of the lightweight implementations in the
> list. Why would that be a problem?

"I do not recommend it use."

Jeff

Re: Lightweight, C-only implementation of Emacs

[Ulrich Mueller]

>>>>> On Mon, 03 Aug 2020, Jeffrey Walton wrote:

>> Well, it is feature complete then.
>> 
>> The same is true for most of the lightweight implementations in the
>> list. Why would that be a problem?

> "I do not recommend it use."

*shrug* It compiles and works, and upstream appears to be responsive.
No good reason for dropping it from a distro (and breaking existing
users' workflows).

Re: Lightweight, C-only implementation of Emacs

[Jeffrey Walton]

On Mon, Aug 3, 2020 at 3:56 PM Ulrich Mueller <ulm@gentoo.org> wrote:
>
> >>>>> On Mon, 03 Aug 2020, Jeffrey Walton wrote:
>
> >> Well, it is feature complete then.
> >>
> >> The same is true for most of the lightweight implementations in the
> >> list. Why would that be a problem?
>
> > "I do not recommend it use."
>
> *shrug* It compiles and works, and upstream appears to be responsive.
> No good reason for dropping it from a distro (and breaking existing
> users' workflows).

*shrug* The author told you not to use it. What don't you understand
about the statement?

Re: Lightweight, C-only implementation of Emacs

[Stefan Monnier]

> *shrug* The author told you not to use it.  What don't you understand
> about the statement?

FWIW, I use it very happily.  It might be worthwhile trying to find out
why he doesn't recommend its use and what he recommends instead.
Maybe the issue is simply that he lost interest and we should be looking
for another maintainer.


        Stefan

Re: Lightweight, C-only implementation of Emacs

[Gregory Heytings via Emacs development discussions.]

>
>> *shrug* The author told you not to use it.  What don't you understand 
>> about the statement?
>
> FWIW, I use it very happily.  It might be worthwhile trying to find out 
> why he doesn't recommend its use and what he recommends instead. Maybe 
> the issue is simply that he lost interest and we should be looking for 
> another maintainer.
>

The main (and serious) limitation of Zile is that it has no support for 
Unicode.  In fact it is in practice limited to ASCII characters.  There 
should apparently be a "displayable-characters" configuration variable to 
use 8-bit character sets, but it is nowhere documented how it should be 
set.

Gregory

Re: Lightweight, C-only implementation of Emacs

[Amin Bandali]

[-- Attachment #1: Type: text/plain, Size: 902 bytes --]

Jeffrey Walton <noloader@gmail.com> writes:

> On Mon, Aug 3, 2020 at 3:56 PM Ulrich Mueller <ulm@gentoo.org> wrote:
>>
>> >>>>> On Mon, 03 Aug 2020, Jeffrey Walton wrote:
>>
>> >> Well, it is feature complete then.
>> >>
>> >> The same is true for most of the lightweight implementations in the
>> >> list. Why would that be a problem?
>>
>> > "I do not recommend it use."
>>
>> *shrug* It compiles and works, and upstream appears to be responsive.
>> No good reason for dropping it from a distro (and breaking existing
>> users' workflows).
>
> *shrug* The author told you not to use it. What don't you understand
> about the statement?

That's not how free software works.  The author is of course free to
make recommendations about their software, and the community is free to
decide to whether take the author up on that recommendation or not.

See <https://www.fsfla.org/blogs/lxo/pub/new-dawn>.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 857 bytes --]

Re: Lightweight, C-only implementation of Emacs

[Daniele Nicolodi]

On 03/08/2020 15:52, Amin Bandali wrote:
> Jeffrey Walton <noloader@gmail.com> writes:
> 
>> On Mon, Aug 3, 2020 at 3:56 PM Ulrich Mueller <ulm@gentoo.org> wrote:
>>>
>>>>>>>> On Mon, 03 Aug 2020, Jeffrey Walton wrote:
>>>
>>>>> Well, it is feature complete then.
>>>>>
>>>>> The same is true for most of the lightweight implementations in the
>>>>> list. Why would that be a problem?
>>>
>>>> "I do not recommend it use."
>>>
>>> *shrug* It compiles and works, and upstream appears to be responsive.
>>> No good reason for dropping it from a distro (and breaking existing
>>> users' workflows).
>>
>> *shrug* The author told you not to use it. What don't you understand
>> about the statement?
> 
> That's not how free software works.  The author is of course free to
> make recommendations about their software, and the community is free to
> decide to whether take the author up on that recommendation or not.

"I do not recommend its use" from an author abut the free software they
wrote has many possible meanings. Among those: "use it but please don't
bother me if it breaks and eats your data", "I know it has (serious)
flaws that I don't want even think about", "I am very tired of
supporting this piece of software, but I continue doing so for spirit of
service toward those (hopefully few) that still depend on it. Please
stop using it so I can enjoy doing something else".

Of course no one can forbid you to use the software the author himself
deprecated, but I think it would be nice to the author to follow their
recommendation and look for alternatives, or pick up or share the burden
of maintaining it.

I don't know if any of this applies to Zile, but maybe inquiring,
preferably in private, with the author and maintainer would be a good
idea before putting it in a list of suggested software (I admit I
haven't looked at the list debated here, thus I don't know if this applies).

Cheers,
Dan

Re: Lightweight, C-only implementation of Emacs

[Amin Bandali]

[-- Attachment #1: Type: text/plain, Size: 1981 bytes --]

Daniele Nicolodi writes:

> On 03/08/2020 15:52, Amin Bandali wrote:
[...]
>
> "I do not recommend its use" from an author abut the free software they
> wrote has many possible meanings. Among those: "use it but please don't
> bother me if it breaks and eats your data", "I know it has (serious)
> flaws that I don't want even think about", "I am very tired of
> supporting this piece of software, but I continue doing so for spirit of
> service toward those (hopefully few) that still depend on it. Please
> stop using it so I can enjoy doing something else".
>
> Of course no one can forbid you to use the software the author himself
> deprecated, but I think it would be nice to the author to follow their
> recommendation and look for alternatives, or pick up or share the burden
> of maintaining it.
>
> I don't know if any of this applies to Zile, but maybe inquiring,
> preferably in private, with the author and maintainer would be a good
> idea before putting it in a list of suggested software (I admit I
> haven't looked at the list debated here, thus I don't know if this applies).
>
> Cheers,
> Dan

Right, I'd generally agree.  I just wanted to clarify that we are not
under any obligation to (blindly) obey recommendations about using a
certain piece of free software, not in the way noloader makes it sound
like we must.  But of course it's often a good idea to consider these
recommendations, especially when they come from the author/maintainer of
the software in question.

In the specific case of Zile, having now looked through the issue
referenced earlier, <https://savannah.gnu.org/bugs/?58880>, I see the
maintainer mentioning that "Zile still has users, so I maintain it for
them", which is great news :-).  Earlier they mention that they will not
be adding new features at the time being, but will continue to fix bugs.
Which is certainly not what I'd have guessed if I'd only read noloader's
messages of heavily questioning/discouraging any use of Zile.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 857 bytes --]

Re: Lightweight, C-only implementation of Emacs

[Daniele Nicolodi]

On 03/08/2020 17:56, Amin Bandali wrote:
> In the specific case of Zile, having now looked through the issue
> referenced earlier, <https://savannah.gnu.org/bugs/?58880>, I see the
> maintainer mentioning that "Zile still has users, so I maintain it for
> them", which is great news :-).

Are you sure this does not fall into the "I am very tired of supporting
this piece of software, but I continue doing so for spirit of service
toward those (hopefully few) that still depend on it" category?

Cheers,
Dan

Re: Lightweight, C-only implementation of Emacs

[andres.ramirez]

Hi. Ulrich.

>>>>> "Ulrich" == Ulrich Mueller <ulm@gentoo.org> writes:


    Ulrich> *shrug* It compiles and works, and upstream appears to be responsive.  No good reason
    Ulrich> for dropping it from a distro (and breaking existing users' workflows).


qemacs could be an alternative. qemacs has basic suppor for org-mode as
a bonus.

Best Regards

Re: Lightweight, C-only implementation of Emacs

[Ergus]

On Mon, Aug 03, 2020 at 09:32:26PM +0000, andres.ramirez wrote:
>Hi. Ulrich.
>
>>>>>> "Ulrich" == Ulrich Mueller <ulm@gentoo.org> writes:
>
>
>    Ulrich> *shrug* It compiles and works, and upstream appears to be responsive.  No good reason
>    Ulrich> for dropping it from a distro (and breaking existing users' workflows).
>
>
>qemacs could be an alternative. qemacs has basic suppor for org-mode as
>a bonus.
>
>Best Regards
>

Some time ago I came with this same question. After some tests and forks
I found that qemacs was a better alternative compared to Zile for
different reasons:

1) The code was cleaner and simpler to extend, understand and maintain.

2) It does not depend on gnulib or any external tool, which considering
I wanted to port it to cmake was a key plus. (gnulib only supports
autotools)

3) The support for color and syntax highlight was finished and fully
working with extension possibilities.

4) It had all the infrastructure to create extension in C (with dynamic
libraries and an exposed API), like the emacs C-modules.

5) Basic support implemented for some major modes like org-mode, and
others. (this can be extended with modules as mentioned in point 4)

6) gui interface with xlibs and terminal version.

In general I made a request to add some functionalities to Zile and
Qemacs and I never received any answer from any of them so I just made
my personal forks:

https://github.com/Ergus/Zile

https://github.com/Ergus/Qemacs

Re: Lightweight, C-only implementation of Emacs

[chad]

[-- Attachment #1: Type: text/plain, Size: 812 bytes --]

Linus Torvalds supposedly uses a version of mg (formerly Micro GNU Emacs)
that is maintained along with the Linux kernel:

  https://git.kernel.org/pub/scm/editors/uemacs/uemacs.git

Hope that helps,
~Chad

On Sat, Aug 17, 2019 at 6:27 PM Jeffrey Walton <noloader@gmail.com> wrote:

> Hi Everyone,
>
> I'm trying to find a lightweight C-only implementation of Emacs for
> Linux. I need it for situations like this:
> https://redmine.pfsense.org/issues/9682 .
>
> Searching is pretty much useless due to irrelevant results. I found
> Zile at https://www.gnu.org/software/software.html#allgnupkgs , but it
> fails to build (it configures OK).
>
> I understand I will probably loose some functionality due to decoupling
> Lisp.
>
> Does anyone have a list of Emacs replacements?
>
> Thanks in advance.
>
> Jeff
>
>

[-- Attachment #2: Type: text/html, Size: 1474 bytes --]

Re: Lightweight, C-only implementation of Emacs

[Ergus via Emacs development discussions.]

Sorry for necrobumping this thread, but after some time I have
discovered Qemacs and tried for a while (a week).

https://savannah.nongnu.org/projects/qemacs

I need to say that it is a more functional C-only alternative to emacs
because unlike zile:

- Does not depend of gnulib, gc, or external non-standard libraries or
  emacs itself. (zile needs emacs to run the tests)
- Not even a complex autotools file is needed. (gnulib integration is
  very hard and for some reason they don't support anything else than
  autotools... which sometimes is too complex for small projects and
  limits integration with other tools like CMake and enforces the use
  and knowledge of autotools/autoconf and so on)
- It have syntax-highlight for some of the common languages like
  C/Rust/swift/perl and others.
- It has search candidates highlight.
- It seems to be very modular to implement modules and languages modes
  (In C of course).
- It has some simple support for org-mode
- Allows vertical splits not only horizontal.
- It has gui and tui interfaces (with -nw as usual, zile does not have
- gui version)
- Multiplatform support (win32 included, and specialization for some
  terminals like xterm and so on.)
- Provides and API to create plugins in C without recompiling the
  editor, as it loads .so files.
- No garbage collection is needed.
- Line numbers functionality supported.
- utf-8 support and bidirectional editing.

I should say that I added some primitive support for linum, colors, fill
 column-indicator and mouse interaction to zile in my personal fork on
 github... but couldn't ever commit it as I never got a reply to join to
 the project. So I freely added some non-gnu changes like migrate it to
 CMake But if anyone is interested:

https://github.com/Ergus/Zile

Limitations:

- No lisp machine at all (configuration is in a file with a C-like
  syntax) which from some points of view is not a limitation. (
  performance, no gc, no C-to-lisp code api needed, no
  byte-compiler/lisp interpreter to maintain)
- Extensibility is more complex as it requires compiling the plugin.
- Have had only 2 contributors in it's history, which is good because
  all the code is very organized and standard. So there is not community
  there. (but actually is the same with Zile)
- Still developed with CVS, I asket to move it to git... but who knows
  if I even get a reply.
- It is in savannah but it is a nongnu project.
- The documentation is very simple... maybe too short.

Hope this helps.

On Wed, Aug 21, 2019 at 03:58:27PM -0700, chad wrote:
>Linus Torvalds supposedly uses a version of mg (formerly Micro GNU Emacs)
>that is maintained along with the Linux kernel:
>
>  https://git.kernel.org/pub/scm/editors/uemacs/uemacs.git
>
>Hope that helps,
>~Chad
>
>On Sat, Aug 17, 2019 at 6:27 PM Jeffrey Walton <noloader@gmail.com> wrote:
>
>> Hi Everyone,
>>
>> I'm trying to find a lightweight C-only implementation of Emacs for
>> Linux. I need it for situations like this:
>> https://redmine.pfsense.org/issues/9682 .
>>
>> Searching is pretty much useless due to irrelevant results. I found
>> Zile at https://www.gnu.org/software/software.html#allgnupkgs , but it
>> fails to build (it configures OK).
>>
>> I understand I will probably loose some functionality due to decoupling
>> Lisp.
>>
>> Does anyone have a list of Emacs replacements?
>>
>> Thanks in advance.
>>
>> Jeff
>>
>>