my blog

Wednesday 6 April 2011

Enjoy GNOME 3 today!

GNOME 3.0

GNOME 3 banner by Jakub Steiner & Lapo Calamandrei

After nine years in the world of GNOME 2, our community is now moving one step further with the availability of the first GNOME 3 release. We're writing an important new line in the history of our project.

It's hard to ignore the feelings this event creates.

I remember the long discussion Lucas and I had about the future of GNOME, in front of his hotel at FOSDEM 2008.
I remember how, at GUADEC 2008 in Istanbul, the release team discussed the idea of finally doing GNOME 3.
I remember the Boston hackfest in October 2008, and all the ideas that were generated there.
I remember writing the first draft of our plan for GNOME 3.0, late at night at the end of March 2009, while being hosted by Christophe.
I remember meeting with as many teams as possible for two days at GUADEC 2010 in the Hague, to decide if we should delay the release.
And I obviously remember the Bangalore Hackfest we did last week to work on the final details on the release.

I love our project and what we achieved. You will love it too!

Getting ready to celebrate GNOME 3? Maybe in Lyon?

Today, we'll release GNOME 3 to the world.

We still need to finalize a few things here and there, but the release is mostly ready. While every .0 release generates some stress (especially when it's one starting a new major cycle), there has been much confidence about how smooth this would all go. This is mostly thanks to the Bangalore Hackfest, where we were able to prepare for this big step.

I actually came back from India on Monday around noon and I immediately started working on the release. We nearly got all tarballs in time (thanks!), and even though we had to re-roll a few tarballs to fix some extremely visible issues, everything went fine so far. I was able to quickly publish the 3.0.0 modulesets so that anybody could help with smoketesting and I built everything, at least twice. The few visible issues I've seen in my tests are either already fixed in re-rolled tarballs or in git, or with a patch in bugzilla. Since those issues are not necessarily considered as blockers (it depends on how severe they are), some of them might stay for 3.0.0; but they'll certainly be fixed for 3.0.1, due in one month. So all in all, things are good there. There is still some coordination to be done for the website update, though.

The bad news is that my body apparently doesn't want to keep working like this for long: what was supposed to be a short break at 6PM yesterday ended up being the end of the day, as I fell asleep for the whole night. I guess considerably eating on my sleep time in the last two weeks and falling sick during the last days in India didn't help :-)

It's slowly getting hard to keep focusing on just the release in this last day, since I already find myself thinking about all the release celebration! There's much excitement all around, and all the launch parties around the world make me think we'll all have fun for the celebration. I'll go to the Lyon party tomorrow (Thursday), and I certainly hope to meet many people there. So don't hesitate to pass by if you're near Lyon tomorrow!

Now, going back to release mode :-)

Friday 1 April 2011

Delaying GNOME 3.0, again

Update: in case people had any doubt, this was an April 1st joke :-)

The Bangalore Hackfest was really useful for the release team to evaluate the status of GNOME 3. We really want GNOME 3 to be amazing, and various recent events lead us to wonder if doing the release next week is a good thing; we had a lot of discussion and meetings, and we even had a call with the Board to evaluate different options.

It was not an easy decision, but after announcing that GNOME 3 would occur in September 2010 instead of March 2010 as originally planned, and then pushing it back to March 2011, we have to announce another delay: GNOME 3.0 is now scheduled for September 2011.

The announcement linked above explains the various reasons for this decision, which was really hard to take. You can also go read the next mail I sent afterwards to the same mailing list for more information. I really want to thank all the community for their hard work, and everybody should just keep rocking. Because what we achieved so far is still pretty impressive.

Of course, this changes a bit the topic for the talks we're going to deliver to the GNOME.Asia conference, especially my keynote scheduled tomorrow on April 2nd...

Wednesday 9 March 2011

WebKit-powered gnome-web-photo

A few weeks ago, I was looking at what was still using an old version of XULRunner in openSUSE, and one of the few applications that was left with no port to XULRunner 2 (the one that will be out with Firefox 4) was gnome-web-photo. In case you don't know gnome-web-photo, it's a small utility originally written by Christian Persch, to capture screenshots of a web page, which can also be used to generate a thumbnail for web pages. It's actually the main reason why we care about it in openSUSE: having a thumbnailer for HTML files is a nice little touch.

So I look at porting gnome-web-photo to XULRunner 2, and... I felt some pain. I asked Christian about his plans and if he thought porting this utility to WebKit could be a good idea. As Christian was not working on it actively, he was open to the idea. After a few hours of coding (WebKitGTK+ is really cool, with its easy-to-use API), a rewrite was ready.

I did some more cleanup and added the option to print directly a web page to a printer, and then released gnome-web-photo 0.10. The only real regression, as far as I know, is that it won't handle correctly very tall web pages (like this one): it will simply cut the web page, because we're hitting some limits in some libraries below us. This is probably fixable by scrolling the page and writing the resulting image ourselves, but I really wanted to get the rewrite out and I'm not even sure people would notice this regression ;-)

Enjoy!

Monday 7 March 2011

Google Summer of Code 2011 for GNOME & openSUSE

Every once in a while, I fall into a trap that causes me to care about a specific topic. Last week, this happened for the Google Summer of Code 2011.

Twice.

I'm apparently going to be a GSoC co-admin for both GNOME and openSUSE, assuming the two organizations get accepted. But I'm not unhappy about that, since GSoC is one really amazing opportunity for free software projects to get useful contributions, but more importantly, to introduce new people to the projects. People who will stay as contributors later on, if we do a good job at making them feel welcome.

GNOME

GNOME has participated to all GSoC, and that's something we're very proud of. The best part is that past students have become highly involved in GSoC in later years, with some of them being the main admins for GNOME. We usually have a team of several admins (at least four), and everybody has experience of GSoC, so organizing our participation is probably easier than for many projects. We've put all of our GSoC documentation online, and that's really the place to visit if you want to be a mentor or student for GSoC on a GNOME-related project.

We've just started collecting project ideas. If you work on something GNOME-related (or a cross-desktop technology), don't hesitate to add your project ideas there. The admin team will triage the list of ideas later on, so don't worry if your idea seems to be lost in a big list of ideas :-) Christophe will send a proper request for ideas in the next few days (if he hasn't already, I haven't read all my mails).

With GNOME 3 just around the corner, there is without a doubt a good opportunity to attract students: those are exciting times for GNOME where a student could make a big difference for 3.2/3.4 with a single project, and become a core actor of the GNOME 3 development, and therefore of the GNOME community!

openSUSE

For openSUSE, things are different: we participated three times (2006, 2008 and 2009), and our application was unfortunately rejected last year. So a few people worked hard in the past few weeks to increase our chances to participate this year (special thanks to Manu who did a good part of the job).

We gathered all the relevant information on the wiki and our community already offered many different project ideas (roughly 40, as of right now). What's exciting is with openSUSE, we offer topics ranging from low-level C code to Ruby on Rails, from infrastructure tools to end-user features, from openSUSE-specific topics to cross-distribution ones, etc. Our list of ideas is extremely broad, and we believe the technologies we cover are exciting for students. Thanks to this preliminary work, we've already got several students contacting us about the projects. Isn't that cool?

As the openSUSE Foundation might not be setup in time to receive GSoC money, we're considering various options as to what to do with the money. The two main contenders are leaving the money to Google for future similar initiatives (GSoC or Code-In, for example), and giving the money to another non-profit organization that we believe is important. We welcome feedback on this, so raise your voice if you have an opinion :-)

- page 5 of 119 -

by Vincent