Summer of Code 2009

Summer of Code 2009

[Nick Roberts]

I suggested earlier that Emacs should get involved in the Summer of Code.  See
http://www.gnu.org/software/soc-projects/ideas.html for the GNU Project's
involvement last year.  Google have announced Summer of Code 2009 and I presume
GNU will be involved again, although I have not seen this stated yet.  Below is
what I would like to send to summer-of-code@gnu.org.  This could be just one
suggestion for Emacs, although if there are many others it might need a
separate website, and I think, Google allocates a fixed number of projects to
each organisation.  I have put emacs-devel@gnu.org as a contact point, which
seems right to me, but might cause some annoyance.

Stephan/Yidong - as Emacs maintainers are you agreeable to this?  If yes, are
there any changes you would wish me to make to the description?  Each project
require a mentor (estimated effort five hours per week) and I would volunteer
for this project.

Any other ideas?

-- 
Nick                                           http://www.inet.net.nz/~nickrob


Emacs
----- 

To migrate the GDB graphical interface in Emacs to completely use GDB/MI (GDB's
Machine Interface).  Currently the underlying lisp package, gdb-ui.el, uses a
mixture of annotations and GDB/MI to interact with GDB.  Annotations are being
deprecated and GDB/MI provides a more robust interface, as well as supporting
the new features of GDB.  There is currently a package called gdb-mi available
from the Emacs Lisp Package Archive (http://tromey.com/elpa) that fully uses
GDB/MI but does not have all the features of gdb-ui.el.  It could be used as a
starting point and, if the project is successful, it is anticipated that the
resulting code would replace gdb-ui.el after the release of Emacs 23.1.  See
http://users.snap.net.nz/~nickrob/ or post to emacs-devel@gnu.org for more
information.

Re: Summer of Code 2009

[Chong Yidong]

Nick Roberts <nickrob@snap.net.nz> writes:

> Stephan/Yidong - as Emacs maintainers are you agreeable to this?  If
> yes, are there any changes you would wish me to make to the
> description?  Each project require a mentor (estimated effort five
> hours per week) and I would volunteer for this project.

I am agreeable, and the description looks fine to me.  Bear in mind that
if you manage to get someone to work on this, that person needs to
complete a copyright assignment for the completed project to be included
in Emacs.

Re: Summer of Code 2009

[Nick Roberts]

 > I am agreeable, and the description looks fine to me.  Bear in mind that
 > if you manage to get someone to work on this, that person needs to
 > complete a copyright assignment for the completed project to be included
 > in Emacs.

Thanks.  I think Google decide how many students are allocated to GNU, then
someone from the GNU Project decides which individual projects are allocated
a students, based on submissions.  So I'm guessing that copyright assignment
is dealt with by them before work starts.  In any case, I am aware that
assignment is necessary.


-- 
Nick                                           http://www.inet.net.nz/~nickrob

Re: Summer of Code 2009

[Daniel Clemente]

Hi,
  I think this is a very good opportunity to tackle the big projects which have been around for a long time and which would improve Emacs considerably.


  As examples:

1. Integrate CEDET into Emacs
-----------------------------
  This will ease support for other languages and integrate features like code completion, syntax checking, project handling, ... Some other tools like JDEE (Java support) or ECB could also be included.


2. Good multithreading support for Emacs
----------------------------------------
  This is much wanted, specially for background processes like checking mail. It is necessary to profit from multicore processors, very common nowadays.
  There were already attempts to do this.


3. Version control system and bug tracker
-----------------------------------------
  This would include:
- setting the Bazaar repository, migrating from CVS to Bazaar
- web interface to the bug tracker, similar to the other projects like Bugzilla and Launchpad, where you could see reports, send a new bug, query for bugs, ...
- and at the same time, mail interface, where you can issue all commands just via mail
- integration between VC and bug tracker; similar to what Bundle Buggy (in Launchpad) does: patch tracking, asking for automatic merges from the mailing list, patch approval, ...
- some other infrastructure tasks could be added, like automatic building (continuous integration), breakage detection (get noticed when code doesn't compile), validation at commit time, ...



  There are for sure more ambitious projects; and I'm sure that many students would apply for them. Emacs needs mentors first.


  Greetings,

Daniel

Re: Summer of Code 2009

[Richard M Stallman]

Your proposals are good ones.

Another change we should ask Google to support
is to eliminate the requirement for window and frame sizes
to be integral numbers of characters and lines.

Re: Summer of Code 2009

[Nick Roberts]

Daniel Clemente writes:
 > 
 > 
 > Hi,
 >   I think this is a very good opportunity to tackle the big projects which
 >   have been around for a long time and which would improve Emacs
 >   considerably.
 > 
 > 
 >   As examples:
 > 
 > 1. Integrate CEDET into Emacs
 > -----------------------------
 >   This will ease support for other languages and integrate features like
 >   code completion, syntax checking, project handling, ... Some other tools
 >   like JDEE (Java support) or ECB could also be included.
 > 
 > 
 > 2. Good multithreading support for Emacs
 > ----------------------------------------
 >   This is much wanted, specially for background processes like checking
 >   mail. It is necessary to profit from multicore processors, very common
 >   nowadays.  There were already attempts to do this.
 >  ...


One SoC project involves a student working for about about twelve weeks.  I'm
not sure how much progress they could make with these tasks in this time.  I
would think it is better to have smaller clearly defined ones.

-- 
Nick                                           http://www.inet.net.nz/~nickrob

Re: Summer of Code 2009

[Daniel Clemente]

> One SoC project involves a student working for about about twelve weeks.  I'm
> not sure how much progress they could make with these tasks in this time.  I
> would think it is better to have smaller clearly defined ones.

  Yes, of course; projects should be adapted to those 12 weeks. But they still can be the first steps of a bigger project.


  We could track all project ideas on EmacsWiki; there is also a wish list there.


-- Daniel

Re: Summer of Code 2009

[Lennart Borgman]

On Fri, Jan 16, 2009 at 3:18 PM, Daniel Clemente <dcl441-bugs@yahoo.com> wrote:
>
>> One SoC project involves a student working for about about twelve weeks.  I'm
>> not sure how much progress they could make with these tasks in this time.  I
>> would think it is better to have smaller clearly defined ones.
>
>  Yes, of course; projects should be adapted to those 12 weeks. But they still can be the first steps of a bigger project.
>
>
>  We could track all project ideas on EmacsWiki; there is also a wish list there.


That sounds good to me. Maybe you could open and take responsibility
for a page there about this? (I think some management of the page will
be needed.)

Re: Summer of Code 2009

[Will Farrington]

[-- Attachment #1: Type: text/plain, Size: 1644 bytes --]


On Jan 16, 2009, at 2:10 AM, Nick Roberts wrote:

> Daniel Clemente writes:
>>
>>
>> Hi,
>>  I think this is a very good opportunity to tackle the big projects  
>> which
>>  have been around for a long time and which would improve Emacs
>>  considerably.
>>
>>
>>  As examples:
>>
>> 1. Integrate CEDET into Emacs
>> -----------------------------
>>  This will ease support for other languages and integrate features  
>> like
>>  code completion, syntax checking, project handling, ... Some other  
>> tools
>>  like JDEE (Java support) or ECB could also be included.
>>
>>
>> 2. Good multithreading support for Emacs
>> ----------------------------------------
>>  This is much wanted, specially for background processes like  
>> checking
>>  mail. It is necessary to profit from multicore processors, very  
>> common
>>  nowadays.  There were already attempts to do this.
>> ...
>
>
> One SoC project involves a student working for about about twelve  
> weeks.  I'm
> not sure how much progress they could make with these tasks in this  
> time.  I
> would think it is better to have smaller clearly defined ones.
>
> -- 
> Nick                                           http://www.inet.net.nz/~nickrob

As a student who participated in GSoC last year, I'd be thrilled to do  
work on Emacs.

Personally, as far as more-accomplishable tasks go, I'd recommend  
having a couple geared towards bringing various ports (namely the  
Cocoa one) up to feature parity with the others (namely the Gtk one).  
These would go a long way towards improving Emacs while being simple  
enough to work on to be accomplished within the 12-week limit.


[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 2427 bytes --]

Re: Summer of Code 2009

[Nick Roberts]

 > As a student who participated in GSoC last year, I'd be thrilled to do  
 > work on Emacs.
 > 
 > Personally, as far as more-accomplishable tasks go, I'd recommend  
 > having a couple geared towards bringing various ports (namely the  
 > Cocoa one) up to feature parity with the others (namely the Gtk one).  
 > These would go a long way towards improving Emacs while being simple  
 > enough to work on to be accomplished within the 12-week limit.

It might be acheivable but I doubt GNU would agree to a student from
their allocated number working on a port of Emacs to a proprietary platform.

-- 
Nick                                           http://www.inet.net.nz/~nickrob

Re: Summer of Code 2009

[Alex Ott]

Hello

may be it will good idea to implement general-purpose multi major mode
support for emacs? As i remember most of implementations (except the mmm)
have some hacks for concrete modes?

-- 
With best wishes, Alex Ott, MBA
http://alexott.blogspot.com/        http://xtalk.msk.su/~ott/
http://alexott-ru.blogspot.com/

Re: Summer of Code 2009

[Chong Yidong]

Alex Ott <alexott@gmail.com> writes:

> may be it will good idea to implement general-purpose multi major mode
> support for emacs? As i remember most of implementations (except the
> mmm) have some hacks for concrete modes?

My impression is that this may be too intricate for a GSoC, but maybe
someone who has worked on this (e.g. on of the mumamo devs) can weigh
in.

Re: Summer of Code 2009

[Glenn Morris]

Nick Roberts wrote:

>  > bringing various ports (namely the Cocoa one) up to feature
>  > parity with the others
[...]
> It might be acheivable but I doubt GNU would agree to a student from
> their allocated number working on a port of Emacs to a proprietary
> platform.

s/Cocoa/GNUstep - problem solved.

Re: Summer of Code 2009

[Stefan Monnier]

>> > bringing various ports (namely the Cocoa one) up to feature
>> > parity with the others
> [...]
>> It might be acheivable but I doubt GNU would agree to a student from
>> their allocated number working on a port of Emacs to a proprietary
>> platform.

> s/Cocoa/GNUstep - problem solved.

Actually, the GNUstep port would indeed be very valuable.  The main part
would be to get the dump to work.


        Stefan

Re: Summer of Code 2009

[Alex Ott]

Hello

>>>>> "SM" == Stefan Monnier writes:
 > bringing various ports (namely the Cocoa one) up to feature parity with
 > the others
 >> [...]
 > It might be acheivable but I doubt GNU would agree to a student from
 > their allocated number working on a port of Emacs to a proprietary
 > platform.

 >> s/Cocoa/GNUstep - problem solved.

 SM> Actually, the GNUstep port would indeed be very valuable.  The main
 SM> part would be to get the dump to work.

With current build of gnu emacs on mac, there is also problem with
incorrect font displaying - for example, it use different fonts for
russian/english words, etc.


-- 
With best wishes, Alex Ott, MBA
http://alexott.blogspot.com/        http://xtalk.msk.su/~ott/
http://alexott-ru.blogspot.com/

Re: Summer of Code 2009

[Will Farrington]

[-- Attachment #1: Type: text/plain, Size: 861 bytes --]


On Jan 17, 2009, at 7:53 PM, Nick Roberts wrote:

>> As a student who participated in GSoC last year, I'd be thrilled to  
>> do
>> work on Emacs.
>>
>> Personally, as far as more-accomplishable tasks go, I'd recommend
>> having a couple geared towards bringing various ports (namely the
>> Cocoa one) up to feature parity with the others (namely the Gtk one).
>> These would go a long way towards improving Emacs while being simple
>> enough to work on to be accomplished within the 12-week limit.
>
> It might be acheivable but I doubt GNU would agree to a student from
> their allocated number working on a port of Emacs to a proprietary  
> platform.

Eliminating bugs from Emacs is *always* a good thing, even for  
proprietary platforms. I think discouraging people from trying to  
improve Emacs based upon the platform they use sends the wrong message.

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 2427 bytes --]

Re: Summer of Code 2009

[Chong Yidong]

Will Farrington <wcfarrington@gmail.com> writes:

> Eliminating bugs from Emacs is *always* a good thing, even for
> proprietary platforms. I think discouraging people from trying to
> improve Emacs based upon the platform they use sends the wrong
> message.

That's not the point.  Certainly, contributions are welcome for all
platforms.  However, in the case of GSoC, we're trying to select a small
number of projects to be worked on.  In this case, it makes sense to
steer people towards efforts that do not narrowly benefit non-free
platforms.

Re: Summer of Code 2009

[Richard M Stallman]

I think MuMaMo does the job well enough.  All that's needed now
is to integrate it.

Re: Summer of Code 2009

[Nick Roberts]

 > > It might be acheivable but I doubt GNU would agree to a student from
 > > their allocated number working on a port of Emacs to a proprietary  
 > > platform.
 > 
 > Eliminating bugs from Emacs is *always* a good thing, even for  
 > proprietary platforms. I think discouraging people from trying to  
 > improve Emacs based upon the platform they use sends the wrong message.

I think some perspective is needed here: last year GNU, as an organisation, was
allocated 9 students which meant under one student per project from those that
had applied.  Unless GNUstep is a get out of jail clause, I think giving people
false hope sends out the wrong message.  Of course, all these ideas are to be
encouraged and should get the support of the mailing list, but probably only
one, at most, will receive a grant.

-- 
Nick                                           http://www.inet.net.nz/~nickrob

Re: Summer of Code 2009

[Will Farrington]

[-- Attachment #1: Type: text/plain, Size: 1369 bytes --]

On Jan 18, 2009, at 10:02 PM, Chong Yidong wrote:

> Will Farrington <wcfarrington@gmail.com> writes:
>
>> Eliminating bugs from Emacs is *always* a good thing, even for
>> proprietary platforms. I think discouraging people from trying to
>> improve Emacs based upon the platform they use sends the wrong
>> message.
>
> That's not the point.  Certainly, contributions are welcome for all
> platforms.  However, in the case of GSoC, we're trying to select a  
> small
> number of projects to be worked on.  In this case, it makes sense to
> steer people towards efforts that do not narrowly benefit non-free
> platforms.

I agree that steering students in this direction of course is a good  
thing, but I also think that students shouldn't be expected to adopt a  
new platform with which they should focus themselves unless they are  
already interested in doing so. Obviously, there are more than likely  
some students who are interested in more open platforms, and more than  
likely GNU will favor those who do so -- which is understandable. But  
at the same time, I think it's important to note that by supporting  
the idea of improving Emacs on *all* platforms, more students are  
consequently encouraged to apply to work on GNU projects (which is  
arguably beneficial with regards to the number of spots GNU is  
allotted during the course of the program).

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 2427 bytes --]

Re: Summer of Code 2009

[Will Farrington]

[-- Attachment #1: Type: text/plain, Size: 1175 bytes --]


On Jan 19, 2009, at 1:45 AM, Nick Roberts wrote:

>>> It might be acheivable but I doubt GNU would agree to a student from
>>> their allocated number working on a port of Emacs to a proprietary
>>> platform.
>>
>> Eliminating bugs from Emacs is *always* a good thing, even for
>> proprietary platforms. I think discouraging people from trying to
>> improve Emacs based upon the platform they use sends the wrong  
>> message.
>
> I think some perspective is needed here: last year GNU, as an  
> organisation, was
> allocated 9 students which meant under one student per project from  
> those that
> had applied.  Unless GNUstep is a get out of jail clause, I think  
> giving people
> false hope sends out the wrong message.  Of course, all these ideas  
> are to be
> encouraged and should get the support of the mailing list, but  
> probably only
> one, at most, will receive a grant.

At the same time, GNU is allotted "slots" based upon the number of  
interested parties, among other things, so I think encouraging work on  
all platforms (due note that an emphasis on open platforms is more  
than acceptable) would result in more interest among applicants overall.

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 2427 bytes --]

Re: Summer of Code 2009

[Will Farrington]

[-- Attachment #1: Type: text/plain, Size: 1099 bytes --]

On Jan 18, 2009, at 10:02 PM, Chong Yidong wrote:

> Will Farrington <wcfarrington@gmail.com> writes:
>
>> Eliminating bugs from Emacs is *always* a good thing, even for
>> proprietary platforms. I think discouraging people from trying to
>> improve Emacs based upon the platform they use sends the wrong
>> message.
>
> That's not the point.  Certainly, contributions are welcome for all
> platforms.  However, in the case of GSoC, we're trying to select a  
> small
> number of projects to be worked on.  In this case, it makes sense to
> steer people towards efforts that do not narrowly benefit non-free
> platforms.

I agree that steering students in this direction of course is a good  
thing, but I also think that students shouldn't be expected to adopt a  
new platform with which they should focus themselves unless they are  
already interested in doing so. Obviously, there are more than likely  
some students who are interested in more open platforms, and more than  
likely GNU will favor those who do so -- which is understandable. But  
at the same time, I think it's important that

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 2427 bytes --]

Re: Summer of Code 2009

[Richard M Stallman]

    But  
    at the same time, I think it's important to note that by supporting  
    the idea of improving Emacs on *all* platforms, more students are  
    consequently encouraged to apply to work on GNU projects (which is  
    arguably beneficial with regards to the number of spots GNU is  
    allotted during the course of the program).

Even if that's true, it counts for little when set against our
principal mission: to teach people that proprietary software is unjust.

It would be foolish and self defeating to blatantly ignore our ethical
principles just to attract more developers.

Re: Summer of Code 2009

[Dan Nicolaescu]

Chong Yidong <cyd@stupidchicken.com> writes:

  > Nick Roberts <nickrob@snap.net.nz> writes:
  > 
  > > Stephan/Yidong - as Emacs maintainers are you agreeable to this?  If
  > > yes, are there any changes you would wish me to make to the
  > > description?  Each project require a mentor (estimated effort five
  > > hours per week) and I would volunteer for this project.
  > 
  > I am agreeable, and the description looks fine to me.  Bear in mind that
  > if you manage to get someone to work on this, that person needs to
  > complete a copyright assignment for the completed project to be included
  > in Emacs.

How about also having a proposal for improving VC?
- implement vc-pull / vc-push
- improve dealing with branches
- add other features that are needed for modern version control systems.

Re: Summer of Code 2009

[Stefan Monnier]

>> > Stephan/Yidong - as Emacs maintainers are you agreeable to this?  If
>> > yes, are there any changes you would wish me to make to the
>> > description?  Each project require a mentor (estimated effort five
>> > hours per week) and I would volunteer for this project.
>> 
>> I am agreeable, and the description looks fine to me.  Bear in mind that
>> if you manage to get someone to work on this, that person needs to
>> complete a copyright assignment for the completed project to be included
>> in Emacs.

> How about also having a proposal for improving VC?
> - implement vc-pull / vc-push
> - improve dealing with branches
> - add other features that are needed for modern version control systems.

That sounds very good, especially because the code is fairly simple.


        Stefan

Re: Summer of Code 2009

[Daniel Clemente]

Hi, I wrote in EmacsWiki the list of ideas which have appeared on the mailing list with relation to Summer of Code 2009:
  http://www.emacswiki.org/emacs/SummerOfCode2009

  Anyone can add more; and please restructure the page if you want. I will also try to keep the list consistent and update it.
  Please add some deadlines to your org-modes so that you don't miss the dates to submit proposals!  :-) We can collect ideas until March, so there's still time to discuss and choose the project.

  Remember: we don't only need ideas, we need also mentors (people who can help the students and track their progress).

  Greetings,
Daniel