GNOME.org
 

Re: Is GTK+ a cross-platform toolkit ?

hi;

you're operating under a bit of a misunderstanding, so I hope I can
clarify it for you: there is no way, for any of the people listed on
the website, to stop or prevent other developers to work on GTK on
different platforms. you also don't need any validation from any of
the people on that list if you want to package GTK on different
platforms — including Linux (it's not like packagers for Linux
distributions ask for our permission either).

the only thing that's required, by the license we use, is that any
change done to the source code is published under the same terms as
GTK. common courtesy (and what we obviously prefer so that everyone
benefits from those changes) would be to work upstream, if you make
changes that allow porting the toolkit and its stack, though it's not
strictly necessary. the MacOS port work has been recently folded back
upstream, instead of being hosted as a set of random patches; the work
for Windows has pretty much always been done upstream. these are
changes to the code, though; packaging is different, and usually
should not involve patches.

the Windows port discussion has always been derailed by the "we need
an installer" argument; while me *may* need a reference one, the whole
issue is simply orthogonal to the fact that we need reliable builds of
GTK and its stack, complete with some form of QA or smoketesting,
otherwise there won't ever be anything to distribute in the first
place. we can only fix the Windows code inside GLib and GTK (and
Pango, and Cairo, and Clutter, and whatever) if we actually can get it
tested first — otherwise we get notifications of bugs 9/12 months down
the line, when somebody decides to update their ad hoc build. for this
reason, ensuring that cross-compilation keeps working may be a
solution: at least, it would allow us to do continuous integration
like we do on Linux, these days. if people want Windows to be a
supported platform, QA should probably take the precedence on shipping
installer templates; on top of that, shipping DLLs that can be easily
dropped in an existing application in order to check for regressions
is a good step forward.

finally, and as an aside, the gtk website is tracked in bugzilla: if
you want something hosted on it, as well as changing its contents, you
should open a bug; we can obviously discuss it on the mailing list
first, but that's the process we use.

ciao,
 Emmanuele.

On 5 March 2013 11:46, Bálint Réczey <balint balintreczey hu> wrote:

To clarify the audience of the question in the subject line even
further, the current
core team per http://www.gtk.org/development.php is the following:

Name    Affiliation
Matthias Clasen         Red Hat
Behdad Esfahbod         Google
Benjamin Otte   Red Hat
Federico Mena Quintero  Novell
Alexander Larsson       Red Hat
Tristan Van Berkom      Openismus
Carlos Garnacho         Lanedo GmbH
Kristian Rietveld       Lanedo GmbH
Michael Natterer        Lanedo GmbH
Ryan Lortie     Codethink Ltd
Emmanuele Bassi         Intel Open Source Technology Center

AFAIK they can officially accept Tarnynko's offer, turn it down or
implicitly declarevGTK+ officially dead as a cross-platform toolkit
by not responding to the question.
They know about the thread here, since one of them directed us here
with the question.

To add my personal opinion I prefer GTK+ over Qt and I would be very
sad if I had to say good bye to GTK+.

Cheers,
Balint

2013/3/5  <tarnyko tarnyko net>:

Paul Davis writes:


On Mon, Mar 4, 2013 at 6:56 PM, <tarnyko tarnyko net> wrote:


Hi folks,
The subject of this mail is taken from the corresponding thread on
"gtk-list". The thread itself (a few days old) can be read here :

https://mail.gnome.org/**archives/gtk-list/2013-**February/msg00055.html<https://mail.gnome.org/archives/gtk-list/2013-February/msg00055.html>

I think the question is relevant to gtk-devel.
In short, we are questioning the lack of an official GTK+3 Bundle on
Win32
; such bundles exist in the wild, but we think GTK+ should better
advertise
its cross-platform nature.
Another problem is the difficulty to bring the topic to someone "in
charge" (for what it stands here). Hoping to have some feedback guys.



give it up. or at least *read* the archives.
summary:
   * GTK+ is cross-platform to the extent that there are people willing to
step up and help maintain ports to various platforms.
   * if by cross-platform you mean "there are binary releases for platform
X", then no, you're not likely to see that for several platform
   * binary releases for non-Linux platforms are not necessarily a good
idea
   * there are very few people around interested in and with the time to
maintain the windows or OS X ports
this is an open source, under-funded, under-manned project. if you want to
change that, then finding actual man-hours to put into it (yours or
someone
else's) will go a LOT further than mailing about it.
oh, and there is no-one "in charge" when it comes to non-Linux releases.
--p



Hi Paul,
"this is an open source, under-funded, under-manned project."
I understand that fully.
"if you want to change that, then finding actual man-hours to put into it
(yours or someone else's) will go a LOT further than mailing about it."
Well here, we have some people already having done actual work, and wanting
to pull it. Me for example.
You are saying that the subject has already been debated, and that "binary
releases for non-Linux platforms are not necessarily a good idea" (I'm
digging the archives right now). Why ? There have been GTK+2 bundles
available for download on http://www.gtk.org for some time now.
(We're speaking policy here, but to support my point in technical terms :
- my reusable GTK+3 build environment :
http://www.tarnyko.net/repo/gtk3_build_system/ ;
- sample bundle I have produced : http://www.tarnyko.net/en/?q=node/20)
Regards,
Tarnyko

_______________________________________________
gtk-devel-list mailing list
gtk-devel-list gnome org
https://mail.gnome.org/mailman/listinfo/gtk-devel-list




--
W: http://www.emmanuelebassi.name
B: http://blogs.gnome.org/ebassi/
[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]