Planet Collabora

Life: It's been a busy few days, and I should have been blogging every evening in order to keep up. (But I didn't, because I was busy.) I've been packing and getting ready for GCDS and trying to finish off some things before I leave. I did find time to go swimming with Rio one evening, and yesterday we all went to the fair. I won a fluffy penguin playing darts. (I was playing darts, not the penguin.) Thanks to Alex for the photo on the right.

The future of Metacity: It is fairly clear that Metacity will be replaced by its fork Mutter in the near future: Mutter is effectively Metacity 3. Although I have some loose ends to tie up in Metacity, it doesn't seem worth continuing hacking on Metacity 2 when the life is in the other fork. In addition, there are over five hundred bugs open against Metacity, more than I (as the only active maintainer) can humanly deal with. Mutter has far more contributors and the bugs will be far more easily dealt with.

CADT: However, this raises a problem. I can't just close the bugs because there's a new version: that would be repeating the GNOME 2.0 mistake which jwz called "cascade of attention-deficit teenagers". Therefore I will have to go through several hundred bugs and decide whether they are reproducible with Mutter, and if so reassign them. This will be a long and dreary job, and if anyone wants to help out I'd be happy to assign them a block.

Nargery: There is also a discussion about whether windows should be able to indicate to compositing managers that they are still working on drawing a window, to save the compositor diving in and drawing the existing pixmap, which may be uninitialised garbage. Some people question whether compositor-specific hints belong in the EWMH at all, or whether they belong in some separate spec.

Meme: Someone is asking "What was your first word?" Mine was "gone." My grandfather used to play a game with me when I was a baby. He would take an object, like a building block, and then hide it and say "Gone". Links:

• freedesktop.org is in bad shape. We can all help.
• The building that fell over.
• Giant dominoes: Symbolic toppling. I keep reading that as "symbolic topping", which must be good on representative cake.
• The longest known word in ancient Greek.

One of the highlights for me each GUADEC Gran Canaria Desktop Summit is FreeFA - the casual soccer tournament for geeks!

As of now, our roster looks a little short. If you're interested, log in, then register by adding yourself to the wiki page!

I'll see you on the field!

PS: we could probably use a couple volunteer photographers. I could only find about 10 photos on Flickr for 3 years of tournaments!

So the BBC is making a new drama series about the battle between the ZX spectrum and the BBC Micro. Currently codenamed ‘Syntax Era’. As it turns out Clive Sinclair, the creator of the ZX Spectrum, had his offices very close to the current Collabora office here in Cambridge. And due to our own Edward Hervey knowing some of the people involved, the production team behind this new series came by our offices to do some location scouting some weeks ago. It is a little bit up in the air if they are going to use our offices or not in the end, but there is a chance they will, and if that happens there is also chance you might catch some familiar faces as extras in this new series

O Firefox 3.5 foi lançado, e essa é uma boa notícia. Significa que os navegadores livres e/ou que respeitam padrões abertos continuam deixando comendo poeira os navegadores legados com o Internet Explorer (especialmente o 6, que ninguém merece, né?).

Entre outras coisas, o Firefox 3.5 tem performance de javascript muito melhor, parecida com a do Epiphany 2.27.3, e suporta bastante coisa de HTML5, incluindo as tags de audio e vídeo. Muito importante com relação a isso, é que ele suporta por padrão os formatos abertos (assim como a WebKitGTK+, que usa GStreamer, mas o suporte às tags ainda não funciona 100%). Tem algumas páginas muito interessantes para acompanhar a ‘adoção’ da nova release: http://downloadstats.mozilla.com/.

As pessoas devem saber da minha relação de amor é ódio com a Mozilla - mesmo hoje o navegador não se integra bem com meu GNOME, a API de embedding deixa muito a desejar, mas ninguém pode negar que o Firefox foi o que trouxe um clima de abertura para a Web, e exigiu que todos começassem a se preocupar com padrões, desempenho e qualidade. Se você usa um sistema operacional proprietário, largue logo os navegadores proprietários e use uma coisa que presta! =)

I think this is the complete schedule of talks that my colleagues are giving at the Desktop Summit this year.

Sat. 4 July
"QtScript bindings for Telepathy" - lightning talk by Ian Monroe, 15:30-16:30

Sun. 5 July
"The location-aware desktop" by Beaudoin, Pierre-Luc with Henri Bergius: 11:30, Room 2

"Profiling and Optimizing D-Bus APIs" by William Thompson: 12:30pm, Room 1

"Integrating VideoConferencing into Everyday Applications" by Olivier Crete: 12:30, Room 4

Tues. 7 July 2009
"Let's make GNOME a collaborative desktop" by Guillaume Desmottes: 11:00 - 11:45

"How to play libnice-ly with your NAT" by Youness El Alaoui: 15:00

"Pitivi Video Editor" by Edward Hervey: 15:45

Thurs. 9 July 2009
"Introduction to GStreamer development Tutorial" by Wim Taymans: 15:00

Fri. 10th July
"Tools for Authoring Awesome Docs" by Davyd Madeley: 11:00

I am bikeshedding my own yak-shaving. This should win an award.

I'm going to the Gran Canaria Desktop Summit next week. Developers, managers, and other free software enthusiasts in the GNOME and KDE communities get together on the Canary Islands, which are technically part of Spain but sit off the coast of Africa. Then I spend a week in Cambridge, England, working alongside my fellow Collaborans. Yup, it's all for work, and I won't even think about bringing a suit (other than a bathing suit).

Woke up at a good time, around seven.  Promptly and stupidly decided to go back to sleep to see what the end of the dream was; it turned out to be a nightmare.  Woke up again at about eleven and went to the gym.  Continued the run of stupid mistakes by forgetting to get lunch for Rio.  Sharon came by and brought her lunch instead.  I hate getting up late. :(

Later, went to the diner for dinner.  Talked to Alex about a shelving project he's working on.

Did a little tidying, but not very much.  But I've got some way towards Inbox Zero: I'm now down to four emails.

Today I learned that cd - changes to the directory you were in before the current one.

Fin gave me an old notebook of zirs to use as a logbook.  It's lovely.

It occurs to me that the simple system I built a while ago which mostly allows Ubuntu to come up in Shavian would also work to get Deseret, Unifon and Tengwar.  I wonder whether there's much of a market for Ubuntu in Tengwar.  Possibly good Slashdot fodder, anyway.

Joule-for-Dreamwidth is edging closer.  I also need to implement a per-day view with a paging system to get around this problem.

Five days until GCDS starts.

When it comes to security, I consider myself a moderate. On the modern scale from "I'll never regret posting these drunken photos of myself on Facebook, right?" to "I forge my own steel helmets because the government WANTS you to think tin foil is enough", I'm somewhere in the middle. I don't have a real problem with putting some personal information "in the cloud", since I trust the service providers to only display my phone number and mailing address to my friends (as I've requested). But I use different passwords everywhere and quickly replaced all my private SSH keys when I lost a USB drive that contained them.

(Though, admittedly, I'm fortunate in that nothing I'm interested in is illegal nor terribly controversial. So there isn't much beyond passwords, private keys, and private communications that I'd care about being out in the open.)

The other day, in an effort to reduce some clutter on my desk, I got some new, USB-powered hard disks for my personal back-ups to replace the big, loud, failing disks I currently use. So, knowing it's a privacy risk to dispose of hard disks without properly erasing them, I started some research into secure-deletion software. (Again, I'm not paranoid enough to incinerate the disks with thermite.)

The man page for wipe(1) brought up this little gem:

Be aware that harddisks are quite intelligent beasts those [sic] days.
...

I hereby speculate that harddisks can use the spare remapping area to secretly make copies of your data. Rising totalitarianism makes this almost a certitude. It is quite straightforward to implement some simple filtering schemes that would copy potentially interesting data. Better, a harddisk can probably detect that a given file is being wiped, and silently make a copy of it, while wiping the original as instructed.

(Emphasis mine)

Well, I hereby speculate that the author of that section has a very warped sense of other peoples' interest in their personal data. Unless you're a celebrity or involved in some very shady dealings (or your government is in terrible shape), I don't think anyone cares much about what's on your hard drive. And somehow I doubt any of those apply to the original author.

I really wonder what it is about the software field that makes some people go off the deep end with respect to privacy and security. Maybe the fact that we typically have a greater degree of control within our own domain than those in other fields? Or understanding that, after the locks are opened, it's possible to have your privacy invaded very rapidly nowadays (regardless of how likely it is, especially with the right locks and all but the worst locking practices)?

At any rate, your security is only compromised when someone cares more about copying your data than you care to secure it. A tiny bit of security goes a long way to locking out attackers. And beyond that point, you have to accept that you can't prevent this scenario:



CC-BY-NC 2.5, XKCD

Update: just to clarify, I don't think most people actually do live up to what I consider adequate/moderate security practices. Eg, I'll bet most used hard drives (inside a computer or otherwise) sold on eBay aren't even reformatted, much less properly erased.

So in preparation for heading of to Gran Canaria and GUADEC on Thursday I pushed a new Transmageddon release today. 0.11 is actually the first release I posted to sites such as Gnomefiles and Freshmeat so in some sense I guess I feel more confident about this version that earlier ones. A lot of new features included, like multipass encoding, videoflipping (so if your video is 90 degrees tilted you can correct that during transcoding) and better profiles. Still kinda rough at the edges though, and the very latest GStreamer releases are needed for everything to work. Still expecting quite a few bugs to be reported still.

In some sense development have stood still for a while as I have focused on testing various devices and GStreamer features. Remuxing is still on my todo list, but discovered that for instance an ac3parser is still needed in GStreamer to
do nice DVD conversions.

Was kinda nice to successfully use Transmageddon for a business related need at work, we had gotten a request at Collabora Multimedia which meant I needed to transcode a video into quicktime+h264+aac from a .vob file. Worked like a charm I mean while I have of course done hundred of transcodes as part of development, it was nice to do one for something ‘real’

Friendly reminder: the days left to vote to select foundation board members are running down!

Make it happen here: http://foundation.gnome.org/vote/vote.php?id=13

Photo (cc) by caribb: a typical sight during elections in Montréal, Québec.

It is my first participation in a such an online election, non national election.  Of course I am exited to be part of it.  Unfortunately, electronic voting makes a quite boring “Election’s night”.   I am used to the 3 or 4 hours of febrile waiting before the people’s choices are announced.  Where are the exit polls?

Really. I mean, really indeed.

Was talking about photos. Sometimes I do take photos that I like (sometimes they're not food). Here are seven:











Worked a lot of hours this week. Especially during the first half. Should make an effort at reading through my reading list, rather than staring at a screen this weekend.

After updating to Fedora Core 11 I noticed a new feature, the automatic font download system. Essentially it works like the automatic codec download system we have in GStreamer, but for fonts.

So judging by how often the font download box pops up the spam I am getting these days seems to be mostly in 3 languages Coptic, Syriac and N’Go I have to assume the spams are using random character sets to confuse spam filters, as I doubt that for instance either the ancient egyptians or their Coptic descendants of today are a big enough demographic for the spammers of the world

In the last weeks I have been asked several times to modify some components I’m working on to add the ability to split a full name in its components (first name, family name, etc.).
It looks like most people have great expectations about this working correctly but they get annoyed when it fails, and you can be sure it will fail. It will fail because it’s impossible to parse a name correctly, for instance:

Full name	First	Middle	Last
Barack Hussein Obama	Barack	Hussein	Obama
Pier Silvio Berlusconi	Pier Silvio		Berlusconi
José Rodríguez Zapatero	José		Rodríguez Zapatero

How can you do this automatically?

This becomes particularly silly if you cannot be sure that the string you are going to parse is actually a full name, for instance don’t try to parse a chat nickname. It’s true that gmail/gtalk uses your full name by default, but this is only a default and it’s true only for gmail.

To cut a long story short, please please please don’t try to parse names. You can see by yourself how hard it is, even if I’m just considering western-style names.
If you still don’t trust me here’s a quote from e-name-western.c, i.e. the file that does name parsing in libebook :

* <Nat> Jamie, do you know anything about name parsing?
* <jwz> Are you going down that rat hole? Bring a flashlight.

On a side note when you are trying to understand why some code is broken you can find some funny commits, like the great EDS purge

Update: I found this “serious” bug in e_name_western_parse .

I went to the dentist last night, specifically at the NYU College of Dentistry. I actually prefer the dental school experience to many private practice dentistries. The wait in the waiting room is shorter (2 hours per appointment actually spent in the chair, rather than an hour in intermittent waiting plus an hour in the chair), I get treated by eager-to-learn dentists in training rather than bored, laconic hygienists, and the student dentists are thorough and communicative. And they offer a 6pm-8pm slot. Very few private practices do.

Young student dentist Stringer was the one to phone me up to set up an appointment. He was more deft, gentle, and patient than several DDSes I've patronized. "Oh, you build up a lot of calculus here, because of your salivary gland. I have that too," he confided. He checked in with me about whether the ultrasonic cleaning dealie was running too hot and hurting me. "I don't like to use it, I don't think it's gentle enough," he said. He handed me the suction wand: "Raise your hand if you need me to stop so you can suction."

In further stereotype-demolishing, Stringer does not play World of Warcraft (nor does he wear Ira Glass glasses). My cousin-in-law-in-law Aaron, husband of Kristen, is on the road to full Dentistdom and enjoys WoW-style games. [pun about grinding omitted]

I told Stringer what his last name means in journalism; in retrospect, he has a new occupational surname, like Smith or Cooper.

I get curious about others' occupations. Firefighters, CAD designers, directors, transcriptionists, silversmiths, pastors, teachers, full-time caretakers, taxi drivers, deli owners, X-ray technicians, soldiers, construction workers, dentists. How does doing your job change the way you interact with others?

Last January, I announced Geolocation in Empathy.  All pending branches have now been merged and released in Empathy 2.27.3.

It took quite a lot of time to finalize it because we were quite busy and quite frankly while this is sexy, it isn’t a very important feature in Empathy :-).  In the following screenshots, you’ll discover that things have changed a lot since the original announcement.

First of all, the markers now include more information about the contact.  This uses the new markers in libchamplain.  It works nice for now (as I only have 3 or 4 contacts publishing their location), we’ll see with usage if the markers are just too big.

The map is now interactive: right clicking on a contact will bring up the same context menu you get on the contact list.

The Preferences UI was reworked to be simpler.  The previous UI left space for an hypothetical Manual address mode which was dropped.  The rationale is that Empathy shouldn’t have to care about addresses.  If you want to change the address, change it in Geoclue.

This is new since January: the tool tip now include your contact’s geolocation information.  This is the only part of all the geolocation changes that are present even if you don’t build with Geoclue or libchamplain.  It was impossible to display a map there as ClutterGtk doesn’t seem to like such windows hehe. We already know it is partly ugly and contains duplicate information It will be improved before final release.

Finally, the contact information dialog now displays a map and the detailed information about the contact’s location.

Don’t miss the FAQ that I populated with questions I was often asked during development.  Report bugs on the Geolocation component of Empathy (you can also see that we have work left to do).

I am not the only one who worked on this exiting feature, here are the details:

• Alban Crequy worked on the XMPP support and reviews;
• Dafydd Harries did the early work on the XMPP support;
• Guillaume Desmottes wrote the XMPP PEP code (the same used for OLPC) and reviewed the code;
• Pierre-Luc Beaudoin did the UIs, the libchamplain and geoclue integration and pursued the XMPP support;
• Xavier Claessens reviewed many times.

I can’t wait to see more people using this and show up on my map!

Long time without empathy blogpost. Lots of things happened in the Empathy/Telepathy world recently. Here I’ll present 4 new exciting feature of the upcoming Empathy 2.27.3. It will be an impressive release with not less than 49 bugs fixed!

Adium

The one year old adium branch got finally merged. This adds support for Adium Message Style themes in empathy. Themes are made of html and javascript, rendered using WebkitGtk. Of course webkit is optional dependency and old style themes can still be used.

Unfortunately all themes are not totally supported yet. Please check our wiki page for more information.

Geolocalisation

Another long standing branch got merged into Empathy master to add geoclue and libchamplain support. That means that you can now publish your position to your contacts. Your position can be calculater based on a connected GPS, based on your IP, or set manually to a postal address. You can of course tweak Empathy preferences to not publish the exact position for privacy.

If you have contacts publishing their position in your contact list, you’ll be able to see a world map with a marker at his position. The map is rendered thanks to libchamplain and animated using Clutter. Of course the map data is provided by OpenStreeMap.

This was done by Pierre-Luc Beaudoin and he will blog soon with more details and mandatory screenshots.

Desktop sharing

Work has been done to share your desktop with your contacts. This is done by creating a Telepathy StreamTube with your contact and passing the VNC protocol through it. Telepathy StreamTube spec is now undrafted and considered stable.

Guillaume will soon blog with more details about this.

File transfer

And last feature: The file transfer support in Empathy got large improvements. It now support checksum to make sure the transfer succeeded, and soon will support resume.

Some time ago I put out a request for someone to pull the AMR code in Android out and turn it into GStreamer plugins. Well Iago Toral Quiroga did exactly that and thus we now got support for encoding and decoding AMR-NB in GStreamer, combined with support for decoding AMR-WB. This turned out to be an indirect team effort as the current code is a combination of the original Android code, combined with some API glue code developed by Andres Mejia and Martin Storsjö, combined with the old AMR plugins in GStreamer which was meant to be used with some non-distributable reference code. These 3 code bases Iago massaged together into something pliable. The current plan is to merge the patch once bad and ugly unfreeze and we finally got some decent AMR plugins for GStreamer. A big thanks to Iago and everyone else involved.

This effort also fit in well with a bug hunt I have been on for a while to figure out why Transmageddon was not able to create files my phone could play back. After having spent a lot of time expanding on the transcoding engine in transmageddon and testing every possible variation I could come up with, it was finally Mark Nauwelaerts here at Collabora Multimedia who came to my rescue with this patch. Also that getting merged after the freeze.

Also been testing the ASF muxer that Thiago Sousa Santos is working on as part of this years Summer of Code. It is coming along very nicely, with WMA2 and WMV2 muxing working. Hopefully when I do my new Transmageddon release before heading down to Gran Canaria I can include ASF support too. Already got a nice batch of features in git master with multipass encoding, a lot of bugfixing to the profile support, h263 and video flipping. So with the efforts mentioned above I should also be able to add AMR-NB and ASF/Windows Media support to the mix too. I also hope it is a long time until the next time I need to think about stuff like pixel aspect ratio versus display aspect ratio

I still need an icon for Transmageddon if someone is up for the task.

LWN got a really interesting article discussing how Google have included H264 and AAC support in their Chrome browser using ffmpeg and the legal discussion that has come from that. It seems Chris DiBona and the Google lawyers have decided they can work around the LGPL by licensing patents for the ‘application’ instead of for the library implementing the functionality in question. Of course most of us would think that if you ship a library as part of your application, it is a part of your application, but Chris DiBona seems to feel that he licensed the H264 codec for use with the bookmarks list of Chrome and not the media engine

More seriously though DiBona tries to weasel out of the situation by claiming that the language of the LGPL saying ‘For example, if a patent license would not permit royalty-free redistribution of the Library by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Library. ‘ only applies if there is a specific language in the patent license hindering re-distribution of the code. He even manages to refer to the fact that the FSF made the language on this subject even clearer in the LGPLv3 as proof that his reading is correct. While at the same time claiming that what the FSF thinks about these issues is irrelevant.

Well I guess Chris is right about Google taking their responsibilities very seriously, at least as long as they say what he wants them to say Can’t help but feel that Google somewhere along the line went from ‘do no evil’ to ‘we are google, hence no evil can have been done’ which while sounding similar actually are very different.

Funniest part though is that another part of Google, Android, seems to think its not even legally fine to ship a LGPL media framework combined with Apache licensed codecs with their stack. But I guess the Android and Chrome departments have different lawyers, so if we are lucky maybe the Android department ends up suing the Chrome department Or maybe Chris DiBona wakes up and realize he could resolve this issue quickly by combining the Apache licensed H264 implementation in Android with his ffmpeg stack and thus resolve this issue.

While deploying Geoclue with friends, I came to the conclusion that we need a GUI tool for end users to setup Geoclue.  Geoclue-properties was born.  gstreamer-properties was an inspiration (for the name, and part of the visual aspect).

I started the project only last Monday night, but you can already do this with it:

• See your current address and postion according to Geoclue
• List installed providers and their provided services
• Set an address on the Manual provider
• Set the address for the current network on the Localnet provider
• List previously configured addresses in the Localnet provider

Just the last items save the user of having to use dbus-send incantations.   For the screenshot lovers, here’s your dose.

The project is still quite embryonic.  It is my first attempt at using Python for such a task and I lack knowledge (and quite frankly time) on how to create the project’s infrastructure (almost as if I am missing autotools — scary!).  If someone is willing to contribute that or point me the doc I’d appreciate!

Try it!

I just released the 0.6 version of my Elisa/Moovida RTBF plugin. The goal of this release is to integrate better with Moovida's user interface.

- Fix "Use this plugin" hook with Moovida.
- Use new, Moovida style, square icon.
- Move radios menu to "Internet -> Radio"

It should hopefully be available through Moovida's plugin section next week.

If you, like me, are ever looking for the source to Jana (incl. libjana, libjana-gtk and libjana-ecal), it has been moved to GNOME's git repo and can be found here.

For some reason Google doesn't already know that.

Well, not really my first patch. But the first thing I tried to mess with when I first started looking at WebKitGTK+ was the WebKitNetworkRequest object, because I was fancing the idea of writing stuff such as HTTP transactions monitoring, and things like that. So I wrote a big patch which exposed the internal WebCore object (ResourceRequest) fully through our own object. That was back in early 2008. We have come a long way since, and through all these months I got a broader perception of what kind of APIs we need, and how WebCore works. We also decided on going soup-only, which had a huge impact on what the final patch actually looks like.

The patch which finally got committed this week is, how can I put it, VERY different from what I had originally written. You can take a look at the long discussions about it in the bug report I used to track progress. I think I should point out that Marco Barisione and Christian Dywan were crucial in helping me get going with my contribution to WebKit at that time.

What this change gives us is basically the fact that a WebKitNetworkRequest now carries more than just the URI for the request (it actually carries with it a reference to the SoupMessage that will be used later in the request processing, which we are planning to expose in the near future), meaning that when WebKit API gives you a request, and you use it to cause a new load (for, say, opening in a new tab), you still get all the headers that were supposed to go with the request, so you don’t lose things such as, for instance, Referer. So, now, after more than 5 years, the bug that complained that Epiphany did not set Referer (and Galeon before that) for new tabs is finally closed.

By the way, this problem has been fixed for Mozilla’s browser back in 2002, but the embedding API is still buggy up to now. There is still hope, since there’s an attached patch that fixes the issue to be reviewed, and landed. If anyone is reading, it might be a good oportunity to get this fixed in there as well, so that users of applications that use Gecko’s embedding API can also benefit!

The PiTiVi team is very proud to announce the first release in the unstable 0.13 PiTiVi series.

This release is in memory of those who have lost their lives, friends, houses in the April 6th 2009 earthquake in l’Aquila, Italy.

You will find more info of the changes, bugfixes and so forth in the online release notes here .

To sum up this release :

• 3 developers working mostly full time : Alessandro Decina, Brandon Lewis and myself
• An insane QA/usability nazifreak who has overwhelmed the bugtracker with issues : Jean-Francois Fortin Tam
• Over 700 commits !
• Complete core rewrite, new multi-layered timeline, audio waveforms, video thumbnails
• Speed improvements

This is not the end, far from it. We’re already working on the next 0.13.2 release which should be available just before the Gran Canaria Desktop Summit in July with exciting improvements like : transitions, mixing, effects, more speed/memory improvements…

People say an image is worth a thousand words… then I guess a video is worth a million words. Jean-Francois did a screencast showing off the latest features of PiTiVi… and of course edited it with PiTiVi. You can find it in ogg or on youtube. Thanks Jean-Francois !

Enjoy the Show !

Seen quite a few very positive twitter feeds since Edward Hervey announced the 0.13.0.2 pre-release of the PiTiVi video editor. Also saw this blog entry by Bryan Lunduke praising our latest efforts around PiTiVi. So all Pitivi sceptics out there, this is the time to give PiTiVi another look!

Transmageddon
Nathan Willis wrote a nice article on lwn.net covering Transmageddon and Arista. If you are not a lwn.net subscriber you should seriously considering becoming so, or you just have to wait until the article becomes generally available next week.

To avoid scaring to many people away I also I finally updated the Transmageddon website today, so instead of looking like a 5 minute website it now looks like I at least gave it about an hour