Hacker’s den

You might have (or not) noticed that I haven’t really posted anything in the last month or so. That is because my uni exams are still on the agenda. What happens with GSoC? Well, I’ve got some things done, but certainly not to the point that I’m satisfied with what I did, and not enough to pass midterm. Can I catch up? Perhaps … let the race against time start

There happens to be something called DVCS. People said its gonna solve all our problems. But then, different people had different view on our problems, and they had different ideas how to solve them. Out of the flux came to life generation one, followed by generation two and finally generation three. Evolution, they say. So we have Mercurial, Bazaar, Git as representatives from the latest DVCS evolutionary branch.

Now, there is this project that we’ll call GNOME for needs of this little essay. There are a lot of smart folks there doing some amazing work, pushing boundaries of what can be done with every release at least a little bit further. They’ve been using CVS for quite some time, and altough they didn’t really like its deficiencies, they always somehow managed to chum along with it. The “In good and evil” kind of relationship, you know. And then one day, enough! Decision was made to make a migration to Subversion, since it was supposed to make things better. As it happens to be, like with everything else, it brought good and bad things on the table. But eventually most of the things were polished by the rocking GNOME sysadmins, even if there is only a few of them.

Today, we’ve got some other winds blowing. There is an ongoing campaign to make a switch to a DVCS. Here is where things become a little more trickier. People usually discuss things before doing them, and that is actually good! But it becomes quite useless when all you’re doing is throwing balls at each other, and not actually listening to what others have to say. Personally, I am not a fanboy of any DVCS, am quite involved with both bzr and git communities, and have at least tried other DVCSs like Mercurial and Darcs. So before you even mention it, I’m not promoting any of those. What would *really* rock, and I know that upstream developers would appreciate it, is pointing out all those use-cases where you’ve experienced problems with a certain DVCS. If we can do that, upstream developers could work on improving their tools where needed. Perhaps, just perhaps, we can even give them a hand? At least one developer to the developers pool of some DVCS project would really make a difference, because those projects are quite young and don’t have many hands behind them.

Our administrators are worn out from SVN migration, and they are cautious of yet another migration just for the fun of it. If you really want to migrate to your favorite DVCS, write a plan and develop tools needed for the proper migration, do the required testing, and collaborate with upstream developers to improve their DVCSs in needed areas. Let us talk. For quite some time now I’ve been working with KDE to prepare a GitoriousKDE specification which could be further discussed at Akademy. Basically the specification defines a variation of Gitorious project specifically tailored to the needs of KDE project. Perhaps we can do something similar for the Gnome? Then, a moment before you start thinking about it, forget that it supports only git and let your imagination work. Perhaps we can let the developer of individual module to choose desired VCS from range of {bzr,git,hg} trio? Adding others would be fairly trivial anyway, since VCS stuff would be abstracted through not-yet-existant rvcs.

If you tell me that a common base for one of the key infrastructure pieces for KDE and GNOME is not useful, just stop for a moment, get a glass of water, and read this post again. Lets show FOSS collaboration in all its glory with this effort.

I am far from done with expressing my opinion on this matter, but I challenge you now …  I challenge you to communicate and collaborate, I challenge you to make things happen.

Another two weeks have passed, and during that time I have been really busy. Regardless, I have managed to get some things done, and have more in preparation.

• played with BuildService web client
• discussed various Gnome client stuff with my mentor
• made some notes about UI for the client
• helped people with BS web client bugs

Today is actually the last day before hacking officially starts, and I have been pretty busy with real life stuff, so other then a little bit of hacking not much has been done this week. I will write another post on that topic in form of Weekly report anyway.

I would like to thank everyone who shared their suggestions for the applications name. In the end it was decided that the application will be named “obsc”, standing for “OBS Spec Client”, but I’ll definitely keep all other names in mind for some other projects in the future.

It has been quite a long time since Nokia announced accepted applicants to the N810 maemo device program. I’ve had various troubles in getting it, mostly related to the fact that the code was valid only in Austria, and I don’t live there. But anyway, it’s here now and I am sure it will be fun to hack on it.

I’ve got a bunch of neat ideas scattered in my head, patiently waiting to come to fruition. Big thanks to Nokia and the Maemo community for this great opportunity.

Its about time I start writing reports so you can see what am I working on. First in the series is a bi-weekly one due to community bonding and learning period, but I intend to switch to regular weekly cycles once things settle down a bit.

• played around with BuildService
• tried playing around with openSUSE Beta1
• learnt LaTeX
• participated in various BuildService-related discussions
• explored BuildService API
• familiarized myself with osc
• got to know various community members
• helped some people with openSUSE problems

That would be all for now. Although according to GSoC timeline hacking starts on May 26, I hope to start writing some prototype code sooner. Hacking is fun

You remember that Gnome client for the openSUSE Build Service that I am supposed to be working on? Well, I am looking for the right name, so suggestions are more then welcome. It would be great if it could be without the leading “G”, as it will be backend <-> frontend, so I’d rather have a name for backend, and then just have frontends named after it plus graphical toolkit name. So far I have collected the following suggestions:

• Gosc
• Strawberry
• Oscberry

Rodrigo suggests that a name should have “osc” or “build” in its name making the tool’s goal more evident on first sight. And what name do you suggest?

So I believe it is the right time to say something about the project I will be working on this summer as part of Google Summer of Code program, and share my impressions about the people I met in the community.

As my mentor, Rodrigo, already briefly mentioned I will be working on Gnome client for the openSUSE Build Service. While the design and features-set is still to be discussed, initial thoughts are leading to a backend <-> frontend separation, thus allowing QT folks to easily write frontend on top of the created backend. We are aiming for a flexible approach where the applications features-set can be improved by various extensions. In general, the core application will probably consist of project manager and packages editor, initially supporting only rpm SPEC for openSUSE. Expect more of design talk in the coming week, as discussion progresses.

Everyone with whom I got in touch in the community has been extremely helpful, and my mentor is rocking! I have already got commit access to Build Service repository and my blog is aggregated on PlanetSuse. Thanks to everyone who helped to make it happen!

I have resisted opening a blog for quite a while now, even though various folks have been bugging me about it for some time now. So what is different this time? Well, the bugging combined with a strange turn in circumstances which lead to acceptance of my Google Summer of Code application to OpenSUSE project resulted in the decision to do it. So I will mostly be talking about various aspects of my work during Google Summer of Code, but don’t be surprised if I occasionally throw in bits and pieces from various other projects I am contributing to, some topics with regards to organization and management, or any other discipline I find interesting.

So to you all out there in the world, greetings, and welcome into the den!