Re: GNOME: lack of strategic roadmap

[Philip Van Hoof <pvanhoof gnome org>]

I'll be replying to both Martyn and Alberto here.

> On 23/02/10 12:36, Alberto Ruiz wrote:

[CUT]

> > As promising as the situation was, I don't think they seriously
> > invested in the toolkit itself AFACT, during all this years RedHat
> > (through mclasen and alexl) and individual contributors on their spare
> > time have been the only ones doing a serious investment in the
> > toolkit. There was never a full time maintainer dedicated to make sure
> > that GTK+ was moving forward to support those mobile platforms (and to
> > help mclasen on the hard task of reviewing patches and making releases
> > for both GLib and GTK+). To be honest, I don't think that's the kind
> > of interest we expected.

It's churlish or unrefined to claim that ie. Nokia didn't seriously
invest in Gtk+ during its N770, n800, Diablo and Fremantle years (let's
not hide the company names you're targeting here).

Via Imendio's services, and later Lanedo (and I'm sure Martyn can tell
you a lot about Lanedo), and through the other smaller companies, they
actually did. 

This is also the model that was heavily advised to them, by many of the
people in the GNOME community during the Villanova and Stuttgart years.

Acknowledging the help we receive from companies still seems to be a
difficult task for some people in our community. It's unhelpful for our
goals, and we should all consider being more careful here, imo.


On Tue, 2010-02-23 at 14:12 +0000, Martyn Russell wrote:
> Actually, Nokia invested quite heavily in GTK+. Imendio/Lanedo had more 
> developers than Red Hat working on it full time over the past years (I 
> could be wrong here). We had Mitch, Kris, Tim, Sven and some work from 
> others at times in the company (this doesn't include personal time 
> involvement which we have recently seen a great deal of from people like 
> Carlos Garnacho on the MPX branch).

This is also how I perceived it over the years, and still do.

> On 23/02/10 12:36, Alberto Ruiz wrote:
> > I often hear complaints about how the RedHat guys turn down patches
> > from other contributors (mostly from members of companies competing
> > with them),

Well if that's the case, then that's something to be madly angry about.

I don't know nor do I claim this to be the case. I have not seen this
being the case, not for Gtk+ (I'm not involved in its development).

Having a person in the position of maintainer while being pushed by his
company to reject patches from competitors, is more or less the kind of
situation that I'd even propose to actively ban from GNOME.

It's most damaging for our goals, it's viral and it instantly murders
any interest in our development platform by the (other) mobile vendors.

> On 23/02/10 12:36, Alberto Ruiz wrote:

> > but I yet have to see any of those companies investing
> > some of their resources on helping to review all those pending patches
> > waiting in bugzilla and making sure they have a way to get their own
> > patches upstream.

You must be looking at a different bugzilla and git repo than I am.

I do such community patch reviews and other community work while being
paid for contracting, frequently. And I'm not the only one, at all.

It's often a bit of negotiating and balancing with prioritized work
within the sprint that got planned for us. But that's fair: they're the
paying customer. You ought to have some respect for your customer.

It's quite standard here in Western philosophy.

To be fair is the company that I do contracting for being very fair in
this regard. I couldn't ask for more.

They call it "we work upstream". Perhaps we should all give some credit
for that? They did and do listen, at a high level of management.

On Tue, 2010-02-23 at 14:12 +0000, Martyn Russell wrote:
> Actually, I think that the Red Hat maintainers of the toolkit had an 
> interest in stability (for ISVs) and that stifled development. As such 
> developing anything in GTK+ takes a lot longer than it should and that's 
> why it is always hard to get into development there or to fix something. 
> This has long been the internal politic of GTK+.

Right (that's how I perceived it too)

On Tue, 2010-02-23 at 14:12 +0000, Martyn Russell wrote:
> I am perhaps not the best person to comment here, Tim for example, has 
> had much more personal and professional involvement in the toolkit and 
> is much better to make comment on this. My view here is just from a very 
> casual contributor watching over a number of years in a company that has 
> GTK+ expertise.
> 
> > My bottom line is that I don't think that in reality the MeeGo news
> > are going to make any difference to GTK+ (I do wonder, however, what
> > are Intel plans on Clutter long term wise)
> 
> I don't either. There are a lot of companies using it internally that 
> never tell public communities about its use of GTK+. British Telecom is 
> one of them. I remember when Owen was setting up the projects page for 
> gtk.org and I wanted to submit our use cases back then, but internally 
> they didn't want to make it public in case customers were worried about 
> the fact that we were using "open source".
> 
> There are thousands of applications using GTK+ too, let's not forget 
> that, what are they going to move to instead if they don't use GTK+?
> 
> I also thought that Andrew Savory's point was incredibly pertinent.
> 
> As a company we do receive requests for GTK+ support, so I don't think 
> it is fair to say that GTK+ is dead.

No, it's not dead. But I wouldn't say that it's the leading toolkit in
the spectrum of the mobile at this moment. This arguably used to be the
case a few years ago.

I think it helped a lot to get all the middleware being considered by
the mobile companies. For example GStreamer, Telepathy, Pulseaudio and
recently Tracker. Even GConf, can you believe that??

We'll probably see less young guys who are doing awesome new middleware
at this moment having their projects picked up by the mobile companies
in future, unless we deliver a compelling "pushing them" product soon.

So I'm still worried. But I also know many people are working hard on
different concepts and ideas. Clutter, for example, was and is a very
good try. Vala and gobject-introspection (the javascript stuff, for
example) too.

We'll need more. Much more. And better. It's fun.



Cheers,

Philip 



-- 
Philip Van Hoof, freelance software developer
home: me at pvanhoof dot be 
gnome: pvanhoof at gnome dot org 
http://pvanhoof.be/blog
http://codeminded.be