From kdevelop-devel  Wed Apr 19 08:55:42 2006
From: Hamish Rodda <rodda () kde ! org>
Date: Wed, 19 Apr 2006 08:55:42 +0000
To: kdevelop-devel
Subject: Re: Google Summer of Code 2006 Project proposals
Message-Id: <200604191855.45302.rodda () kde ! org>
X-MARC-Message: https://marc.info/?l=kdevelop-devel&m=114543709625012
MIME-Version: 1
Content-Type: multipart/mixed; boundary="--nextPart2554091.XlYElihUD3"

--nextPart2554091.XlYElihUD3
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

On Wednesday 19 April 2006 06:37, Alexander Neundorf wrote:
> Hi,
>
> On Tuesday 18 April 2006 01:48, Matt Rogers wrote:
> > Hi,
> >
> > As you may have heard/read, Google will be doing Summer of Code again f=
or
> > 2006. I'd like those of you interested to think of ideas for projects
> > that could be done for KDevelop trunk. Depending on the project, some
> > projects could also be for KDevelop 3.4 as well, although I don't as of
> > yet know when we'll do the final release for that.
> >
> > Initially, we'll have to track projects here on the mailing list, and
> > I'll hopefully be able to move them to a wiki page soon.
>
> Since I won't have time for mentoring and I am too little involved in
> kdevelop development, I post it here:
>
> Currently KDevelop doesn't have reliable C++ code
> navigation/autocompletion/refactoring support.
>
> We have our own C/C++ parser with r++ (AFAIK). There are also other optio=
ns
> to implement this, each with different pros and cons.
> As there are:
> -cscope/kscope
> -ctags
> -gcc-xml

While we appreciate the suggestion, these possibilities have been discussed=
=20
before and it all boils down to one point: code.

The approach being taken by active developers is our internal parser, and w=
e=20
are confident that with time it will get up to scratch.  Currently I'm=20
refactoring it from stdlib to qt, to make it more accessibile to myself and=
=20
other developers for hacking.

So unless there's someone willing to mentor a potentially competing method =
for=20
parsing support (and kdevelop-pg will cover not just c++ but as many other=
=20
languages as parser definitions are written) then the only SoC project on=20
parsing + related features that makes sense to me is working on the interna=
l=20
stuff.

Cheers,
Hamish.

--nextPart2554091.XlYElihUD3
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (GNU/Linux)

iD8DBQBERfsRH8BtnSmIlUYRAoDGAKDsS+WTB29hpMIoVYVZlLXEBtpV+wCgv3wZ
xkjjUwbM//0YKMop9ZW9RdI=
=scNQ
-----END PGP SIGNATURE-----

--nextPart2554091.XlYElihUD3--

_______________________________________________
KDevelop-devel mailing list
KDevelop-devel@barney.cs.uni-potsdam.de
http://barney.cs.uni-potsdam.de/mailman/listinfo/kdevelop-devel