GNU Emacs as a lightweight IDE (seeking volunteers)

GNU Emacs as a lightweight IDE (seeking volunteers)

[Anand Tamariya]

[-- Attachment #1: Type: text/plain, Size: 2375 bytes --]

*TLDR* I'm seeking volunteers for merging my changes to Emacs. Reach out to
me if you're interested.

Hi,
I've been working on some improvements over CEDET to turn GNU Emacs into a
lightweight IDE (limited or no debugging). If you have a small project in
any of the supported languages, you can get IDE like features without any
additional installation. Given the framework, supporting more languages
boils down to simply adding a new grammar and overloading few functions.

Presently, it works for the following languages (demo links in brackets):
- C ( https://youtu.be/dUEHo5nrnc8 , https://youtu.be/04HUN52M8Sc )
- LaTex ( https://youtu.be/KL0KNa6Nh4M )
- SQL ( https://youtu.be/dUbqdY0KU0U , https://youtu.be/yJ55cQez5Tc )
- HTML + JS + CSS ( https://youtu.be/jJZMRZINAL8 ,
https://youtu.be/FkOYUPjdID8 )
- Java ( https://youtu.be/MWpvKSTz4qc , https://youtu.be/zmRac_BNbv0 ,
https://youtu.be/U2L0iLZI_cE , https://youtu.be/4UcPIdTJLOM )
- Elisp ( https://youtu.be/i-qwqKbedCA , https://youtu.be/DMg2rDn4LzQ )

Complete playlist is here
https://youtube.com/playlist?list=PLW9poAEUvGDAMYvvznljaNtvooaJZxsFQ

I believe it's now stable enough for a merge with main branch. However,
when I look back, the changes have been huge (87 changed files). You can
see the changes here
https://gitlab.com/atamariya/emacs/-/compare/master...dev (scroll down the
page). Hence, I intend to do this in following stages:


*Stage one (Testing)*Though I've been using this for over a year, I believe
it can do with some testing by others.


*Stage two (Minor changes)*I'll be accepting PRs for these to my github
repo:

   - If you want to see additional comments in code
   - If you have suggestions for performance improvements
   - If you have suggestions for code improvement in general
   - If you dislike code formatting



*Stage three (Major changes)*This should be final merge with main. We would
need to do it in small chunks over emails in this mailing list. However, we
can't get to this stage before the other two stages as that might mean
breaking the stability.

It's not a separate package as I believe this should be part of base emacs.
So I developed it as such. Given that CEDET is part of Emacs, if you have
ideas on packaging it as a separate package, I can use your help.

If this sounds interesting and you'd like to help, I'd like to hear from
you.

Regards,
Anand

[-- Attachment #2: Type: text/html, Size: 3359 bytes --]

Re: GNU Emacs as a lightweight IDE (seeking volunteers)

[Richard Stallman]

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > I've been working on some improvements over CEDET to turn GNU Emacs into a
  > lightweight IDE (limited or no debugging).

That is a very good project.  But why not add debugging support?

-- 
Dr Richard Stallman
Chief GNUisance of the GNU Project (https://gnu.org)
Founder, Free Software Foundation (https://fsf.org)
Internet Hall-of-Famer (https://internethalloffame.org)

Re: GNU Emacs as a lightweight IDE (seeking volunteers)

[Anand Tamariya]

[-- Attachment #1: Type: text/plain, Size: 161 bytes --]

> That is a very good project.  But why not add debugging support?
>
I'd love to. But an individual can only do so much. Hence, the call for
volunteers.

-Anand

[-- Attachment #2: Type: text/html, Size: 397 bytes --]

Re: GNU Emacs as a lightweight IDE (seeking volunteers)

[T.V Raman]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset=gb18030, Size: 502 bytes --]

Anand Tamariya <atamariya@gmail.com> writes:

"light weight IDE" sounds like a contradiction in terms -- Most "IDE's"
start off at least as a bantam weight and progress rapidly to the
heavy-weight category  before becoming over-weight and retiring.


>     That is a very good project.  But why not add debugging support?
>
> I'd love to. But an individual can only do so much. Hence, the call
> for volunteers.
>
> -Anand
>

-- 

Thanks,

--Raman
7©4 Id: kg:/m/0285kf1  •0Ü8

Re: GNU Emacs as a lightweight IDE (seeking volunteers)

[Richard Stallman]

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > I'd love to. But an individual can only do so much. Hence, the call for
  > volunteers.

That's a fine approach.  I misunderstood your initial message as saying
that debugging support was excluded from the scope of the project.  I think
you mean, rather, that debugging is not implemented _yet_.
-- 
Dr Richard Stallman
Chief GNUisance of the GNU Project (https://gnu.org)
Founder, Free Software Foundation (https://fsf.org)
Internet Hall-of-Famer (https://internethalloffame.org)

Re: GNU Emacs as a lightweight IDE (seeking volunteers)

[Anand Tamariya]

[-- Attachment #1: Type: text/plain, Size: 321 bytes --]

> "light weight IDE" sounds like a contradiction in terms -- Most "IDE's"
> start off at least as a bantam weight and progress rapidly to the
> heavy-weight category  before becoming over-weight and retiring.
>
Sorry, I don't get your point. Do you have a problem with the naming or are
you saying this effort is futile?

[-- Attachment #2: Type: text/html, Size: 577 bytes --]

Re: GNU Emacs as a lightweight IDE (seeking volunteers)

[Pankaj Jangid]

"T.V Raman" <raman@google.com> writes:

> "light weight IDE" sounds like a contradiction in terms -- Most "IDE's"
> start off at least as a bantam weight and progress rapidly to the
> heavy-weight category  before becoming over-weight and retiring.

Emerging trend is to delegate most of the IDE functions to external
processes. May be that, that will make IDE’s light-weight some day.

I am using Eglot package. Very neatly written code. Indeed light-weight.

--
    Pankaj

Re: GNU Emacs as a lightweight IDE (seeking volunteers)

[Dmitry Gutov]

Hi Anand,

I think you will find that anybody on emacs-devel who has time and 
energy to work on CEDET commits straight to master.

Your work sounds interesting, and we will probably want to see it in the 
core, but the onus on doing the necessary effort should be on you.

In the end it will come down to submitting a number of patches for 
proper review, as well as good descriptions for those patches.

Before that, it might be helpful to post them in a branch for anyone who 
might want to take a look, to give a pre-review, so to speak.

Relatedly, while looking through your /compare/master...dev link, I 
noticed some questionable changes, though. Like some commented-out code, 
or checked in company-semantic.el (did you add some extra features to 
it?), or changes to a number of files in /progmodes/ (!). These should 
have to go, unless accompanied with real good justifications.

For our patch submissions rules, you can read the file CONTRIBUTE.

Some new automated tests to go with the changes would also be welcome.

On 04.02.2021 09:00, Anand Tamariya wrote:
> *TLDR* I'm seeking volunteers for merging my changes to Emacs. Reach out 
> to me if you're interested.
> 
> Hi,
> I've been working on some improvements over CEDET to turn GNU Emacs into 
> a lightweight IDE (limited or no debugging). If you have a small project 
> in any of the supported languages, you can get IDE like features without 
> any additional installation. Given the framework, supporting more 
> languages boils down to simply adding a new grammar and overloading few 
> functions.
> 
> Presently, it works for the following languages (demo links in brackets):
> - C ( https://youtu.be/dUEHo5nrnc8 <https://youtu.be/dUEHo5nrnc8> , 
> https://youtu.be/04HUN52M8Sc <https://youtu.be/04HUN52M8Sc> )
> - LaTex ( https://youtu.be/KL0KNa6Nh4M <https://youtu.be/KL0KNa6Nh4M> )
> - SQL ( https://youtu.be/dUbqdY0KU0U <https://youtu.be/dUbqdY0KU0U> , 
> https://youtu.be/yJ55cQez5Tc <https://youtu.be/yJ55cQez5Tc> )
> - HTML + JS + CSS ( https://youtu.be/jJZMRZINAL8 
> <https://youtu.be/jJZMRZINAL8> , https://youtu.be/FkOYUPjdID8 
> <https://youtu.be/FkOYUPjdID8> )
> - Java ( https://youtu.be/MWpvKSTz4qc <https://youtu.be/MWpvKSTz4qc> , 
> https://youtu.be/zmRac_BNbv0 <https://youtu.be/zmRac_BNbv0> , 
> https://youtu.be/U2L0iLZI_cE <https://youtu.be/U2L0iLZI_cE> , 
> https://youtu.be/4UcPIdTJLOM <https://youtu.be/4UcPIdTJLOM> )
> - Elisp ( https://youtu.be/i-qwqKbedCA <https://youtu.be/i-qwqKbedCA> , 
> https://youtu.be/DMg2rDn4LzQ <https://youtu.be/DMg2rDn4LzQ> )
> 
> Complete playlist is here 
> https://youtube.com/playlist?list=PLW9poAEUvGDAMYvvznljaNtvooaJZxsFQ 
> <https://youtube.com/playlist?list=PLW9poAEUvGDAMYvvznljaNtvooaJZxsFQ>
> 
> I believe it's now stable enough for a merge with main branch. However, 
> when I look back, the changes have been huge (87 changed files). You can 
> see the changes here 
> https://gitlab.com/atamariya/emacs/-/compare/master...dev 
> <https://gitlab.com/atamariya/emacs/-/compare/master...dev> (scroll down 
> the page). Hence, I intend to do this in following stages:
> 
> _*Stage one (Testing)
> *_Though I've been using this for over a year, I believe it can do with 
> some testing by others.
> 
> _*Stage two (Minor changes)
> *_I'll be accepting PRs for these to my github repo:
> 
>   * If you want to see additional comments in code
>   * If you have suggestions for performance improvements
>   * If you have suggestions for code improvement in general
>   * If you dislike code formatting
> 
> 
> _*Stage three (Major changes)
> *_This should be final merge with main. We would need to do it in small 
> chunks over emails in this mailing list. However, we can't get to this 
> stage before the other two stages as that might mean breaking the stability.
> 
> It's not a separate package as I believe this should be part of base 
> emacs. So I developed it as such. Given that CEDET is part of Emacs, if 
> you have ideas on packaging it as a separate package, I can use your help.
> 
> If this sounds interesting and you'd like to help, I'd like to hear from 
> you.
> 
> Regards,
> Anand

Re: GNU Emacs as a lightweight IDE (seeking volunteers)

[Stefan Monnier]

Anand Tamariya [2021-02-04 12:30:08] wrote:
> *TLDR* I'm seeking volunteers for merging my changes to Emacs.

I'd be quite happy to see your new features merged into Emacs, indeed.
That'll take work, but I'd really like to see it happen.

> *Stage one (Testing)*Though I've been using this for over a year, I believe
> it can do with some testing by others.
> *Stage two (Minor changes)*I'll be accepting PRs for these to my github
> repo:
> *Stage three (Major changes)*This should be final merge with main. We would
> need to do it in small chunks over emails in this mailing list. However, we
> can't get to this stage before the other two stages as that might mean
> breaking the stability.

I don't understand what kind of "breaking the stability" worries on in
this regard.  Usually there aren't any serious conflicts between merging
patches into an upstream project and continuing development&refinement
of a downstream project.

IIUC you've been working on this for a while already, so I suspect you
have a good idea of which parts of your changes are "pretty stable" and
which parts aren't.

I suggest you start sending some small patches for inclusion right away.
I suggest you start small with obvious fixes, additional tests, code
clarifications, addition of helpful comments, docstring fixes, those
kind of things.

Most of us here are not very familiar with CEDET's code (I'm a bit more
familiar with it now, since I've been working on recently to convert it
to use `lexical-binding`, but its overall structure is still far from
clear to me), so it'll take time for us to understand your code enough
to review it, and I believe you're not very familiar with our habits
either, so it'll take some time for you to learn how to present your
patches in a way that we can accept them.  Those things work best when
starting with "easy stuff" and then work our way up from there.

> It's not a separate package as I believe this should be part of base emacs.
> So I developed it as such. Given that CEDET is part of Emacs, if you have
> ideas on packaging it as a separate package, I can use your help.

As you may know, I'm a big believer in separate packages, so I may end
up suggesting to make some things into separate packages, but I suspect
some parts of your changes will belong in Emacs itself anyway.

BTW, beside those small&simple patches, a good way to start might be
also to provide an overview of what the "total patch set" contains
(in terms of code more than in terms of user-level features).


        Stefan

Re: GNU Emacs as a lightweight IDE (seeking volunteers)

[Anand Tamariya]

[-- Attachment #1: Type: text/plain, Size: 1159 bytes --]

> Before that, it might be helpful to post them in a branch for anyone who
> might want to take a look, to give a pre-review, so to speak.
>
It's available in gitlab branch
https://gitlab.com/atamariya/emacs/-/tree/dev . If you want it in some
other repo, can you provide me access to that repo?

Relatedly, while looking through your /compare/master...dev link, I
> noticed some questionable changes, though. Like some commented-out code,
> or checked in company-semantic.el (did you add some extra features to
> it?), or changes to a number of files in /progmodes/ (!). These should
> have to go, unless accompanied with real good justifications.
>
Changes to company-semantic.el will of course go into company package. It's
presently in my repo for my convenience. I've added some post-completion
and template changes (eg SQL).
Changes to some files (e.g. in progmodes) is the reason why I want atleast
couple of more users to test the functionality. I don't want to push some
changes that might have wide impact on the basis of my individual testing
alone. What if some existing functionality is breaking that I'm unaware of
(something that I don't use)?

[-- Attachment #2: Type: text/html, Size: 1670 bytes --]

Re: GNU Emacs as a lightweight IDE (seeking volunteers)

[Dmitry Gutov]

On 09.03.2021 07:28, Anand Tamariya wrote:
> 
>     Before that, it might be helpful to post them in a branch for anyone
>     who
>     might want to take a look, to give a pre-review, so to speak.
> 
> It's available in gitlab branch 
> https://gitlab.com/atamariya/emacs/-/tree/dev 
> <https://gitlab.com/atamariya/emacs/-/tree/dev> . If you want it in some 
> other repo, can you provide me access to that repo?

Gitlab looks good enough for me.

Regarding your access to the Emacs repository, it's probably a 
discussion for later.

>     Relatedly, while looking through your /compare/master...dev link, I
>     noticed some questionable changes, though. Like some commented-out
>     code,
>     or checked in company-semantic.el (did you add some extra features to
>     it?), or changes to a number of files in /progmodes/ (!). These should
>     have to go, unless accompanied with real good justifications.
> 
> Changes to company-semantic.el will of course go into company package. 
> It's presently in my repo for my convenience. I've added some 
> post-completion and template changes (eg SQL).

We want to obsolete company-semantic sooner or later because it 
duplicates CEDET's completion-at-point-functions function, hopefully 
migrating whatever features are missing.

> Changes to some files (e.g. in progmodes) is the reason why I want 
> atleast couple of more users to test the functionality. I don't want to 
> push some changes that might have wide impact on the basis of my 
> individual testing alone. What if some existing functionality is 
> breaking that I'm unaware of (something that I don't use)?

Even if nobody complains, you will still have to explain/justify every 
such change.