This is an attempt at summarizing all the activity that happened in the last five months, though it’s far from being a short summary. If I had to pick a two words to describe my first five months at Canonical, it would be “Pure Awesomeness“. For a more detailed view, grab a cup of joe or your favorite other beverage and keep on reading.

Today is a special day. Exactly five months ago, on January 5th, I joined Canonical to work on the Landscape project.

It has been quite a ride so far, with two sprints with the Landscape team, AllHands and UDS in Barcelona just a week ago, and lots of excitement about the future. Saying that I’m completely stunned by the work everyone at Canonical has been putting together and how much the teams have grown in the last few months doesn’t do enough justice to it.

At AllHands and UDS we got a short preview of the things that are coming out in the next cycle and beyond. An example of that is the newly formed Design and User Experience (DUX) team which will not only be focusing on Ubuntu itself, but on many other areas across Canonical and the whole Open Source community in general.

At UDS, the DUX team had a special ‘booth’ where any person from any project could walk to them and get advice about their personal or favorite application. One person which has favorably used such advice already was Seif Lofty, from Gnome Zeitgeist. I met Seif during breakfast at the first day of UDS and I cannot describe how excited he was about simply *being* at UDS. And he was certainly twice as happy when he left.

Another person I had the joy to meet was David Siegel, of Gnome Do fame. We teamed during AllHands on the infamous “Fun In The Woods” activity, which had some people walking like zombies on the day after.

Even more importantly, I was able to meet many many more other colleagues from different teams and wrap up some loose ends from tasks that I got started during the first couple months. I’m definitely impressed by the amount of plain brilliant people that are part of Canonical as of today.

In a sense, being at a Canonical event is very much like being at a Plone conference. Everyone seems to be very receptive about new ideas and very friendly and laid back. And as a bonus, I was able to exercise my (not so secret now) power of throwing some crazy ideas around and see how they influence people. And man, I’m already impressed by the outcomes, just a week after the fact.

To me, this has been the most rewarding thing so far, to be able make big contributions not in lines of code, but in ideas that can make a concrete difference in the hands of the right people. This is something that can only be possible at a company the size Canonical is at the moment, where it’s just big enough that you can grab a mind or two to push an agenda without affecting the rest of the team and still small enough that you can influence decisions.

As my colleague Jamu would best describe, “I’m PUMPED!”. :)

But that’s not all. Software-wise, I was able to make some big contributions too. The Landscape team just finished a 6 month development cycle that brought many cool features to life. I’m really happy with that, and specially with the speed that this team can get features from the black board into reality. It’s also a much different environment than what I was used to, with very well-defined and refined processes for ensuring the overall quality of anything that is produced. One process that I’m specially enjoying is the requirement for two positive reviews before landing a branch. I hope to talk more about that soon (that is, sooner than 5 months from now *wink*).

As for my role in the team, it is quite different than what I’m used to. I’ve been focusing a lot on the UI aspects of Landscape, on ways to make things more obvious and more streamlined. I’ve been also writing a ton of Javascript, and collaborating with other teams to define better policies for Javascript testing in general. And finally, we will now have a person from the DUX team dedicated to working with us, which will push work on the Landscape UI even further.

I also had the chance of interacting with the Launchpad team, which has a much more refined process due to the size of their team. Over at Launchpad, I started a branch back in mid-December, even before starting at Canonical, to allow Launchpad to use the Chameleon Template Engine.

That was another wild ride, and during the course of this project I was able to contribute tons of fixes upstream to Malthe Borch to make Chameleon even more compatible with plain old ZPT. In fact, it is so compatible at the moment that due to the magic of z3c.ptcompat Launchpad will be able to run *both* Chameleon and ZPT with the flip of an environment variable. Even more stunning, the changes required to code were minimal, basically changing imports to use z3c.ptcompat, and in templates we’ve had to fix some non-XHTML compliant ones and remove unused i18n tags. I am happy to announce that this branch will soon be merged (it was submitted to PQM, successfully accepted and is waiting to land the buildbot queue). The bad news is that not all tests pass at the moment with Chameleon enabled, but we will be dogfooding and fixing those tests as we go. It was too much pain already to maintain a nearly 6 month old branch outside the main tree. ;)

I am really interested in many of the things that the Launchpad team is doing, process-wise. The PQM seems like a very nice idea for a bigger team like theirs, though it would probably be useful to our smaller team in Landscape too, and to others in general. Hopefully I will get a chance to explore it more and talk about it during the upcoming FISL 10, in Porto Alegre.

Lastly, but not least important, I’m also working on getting nightly builds of the Bzr Installer for Windows rolling, and a more streamlined process for the official builds. Karl Fogel, of Producing OSS fame, and our Launchpad Ombudsman is making sure I keep my promises about that, which is yet another great incentive.

All in all, there’s of course a ton of things I forgot to talk about and which happened in the last 5 months, but this post is already getting too long so I will stop right here and save some of the meat for a future one. Stay tuned!

Now with Google App Engine and Microsoft Azure, Python will be playing an important role in the two biggest Cloud Computing platforms out there. Guido should be feeling really proud of this.

These days, user-friendly distributions like Ubuntu make it easy for even a novice to get a desktop installation up and running smoothly. However, try to get an unusual piece of hardware working, or use a less stable distribution or piece of software, and you’ll probably face a few “challenges”. Now, someone who just want to do some word processing might refer to these as “problems” or “reasons to go back to Windows”, however, I personally find them one of the most exciting features of the operating system. Every time I have a problem, it takes some work and liberal use of everyone’s best friend Google to get over it, but each time I learn a little more about the software I’m using.

(from: 8 Reasons Why Every Computing Student Will Love Linux)

Update: You can also read the paper: Gestão de Pessoas no Contexto do Trabalho Colaborativo à Distância (yes, Portuguese-only too).

Besides that, I’ve recently discovered, after listening to Steve Gibson’s Security Now! podcast netcast that my development machine is actually 64-bit capable.

I had bought this machine back in May. It’s got a Pentium Core Duo, D 830 3ghz processor. And the motherboard is a ASUS P5ND2-SLI. It’s a very gaming-like machine, though I don’t do any gaming at all. I constantly run between 3-4 VMs on it though.

The important thing here, is something that had gone unnoticed when I built the machine. That the processor has this ‘feature’ called EM64T. And the motherboard supports it too. Well, the name is misleading part there. I thought it was just some fancy memory thingie. It means ‘Extended Memory 64 Technology’ (well, not anymore, now it has been renamed Intel 64).

Then I’ve learnt that this is actually Intel’s implementation of AMD64 technology. And that it can run Vista and XP 64.

So I’ve installed XP 64, which is also available as a trial (though I got mine from MSDN), and I must say that it feels much much faster. Whether that translates into real performance I can’t tell at the moment, but I plan to run some benchmarks. I still have the ‘old’ XP 32 installed on a separate disk.

Some interesting facts about XP 64:

• It’s based on Windows 2003 SP1 codebase
• It can run 32-bit applications just fine
  • This is because of a technology called WoW64 (Windows 32 on Windows 64). Zero noticeable speed impact.
• It installs 32-bit applications in ‘\Program Files (x86)’ by default.
• It installs registry keys on a ‘HKLM\Software\WoW6432Node’.

The last one is the most surprising. From what I can guess, when you run a 32-bit application it creates an environment that ‘hides’ the actual registry, and sets up some variables like %PROGRAMFILES% to point to different places. Very clever.

Now, a little confusion. When I went to download the 64-bit installer for Python 2.5, there are two options. ‘Itanium’ or ‘AMD64′. At that point I didn’t know that EM64T was equivalent to AMD64, so I tried the ‘Itanium’ installer. And to my surprise it did not install. But the AMD64 installer did work. I’ve reported that as a bug on the Python tracker. 

So, since I’m running 64-bit now, I hope I can help with the transition, specially on pywin32, so that we can have a native 64-bit Zope installer sometime over the next year.

Oh, and I can’t resist but post a screenshot of the 64-bit Firefox credits. It’s a trunk build, and the codename is ‘Minefield’. Wonder if there’s a more appropriate name for a ‘trunk’ version. I love the logo too. :)

My opinion is that they are doing a silent revolution. Instead of
going head to head with Linux and Apple they are making small steps
towards releasing significant parts of their software assets for free.

Today, it’s totally possible to be a Windows Developer and work with
several technologies without paying a single cent on development tools
licensing. You take the 2003 and Vista betas, a Express Edition compiler, and you
are writting your next application, ready for the to-be-released
versions of Windows.

I mean, seriously. Here’s some signs:

1. They released some major pieces of software as free downloads
   (Visual [you name it] Express, SQL Server Express). The Open Source
   community still has a ways to go to offer the same level of
   functionality as even a single one of those apps.
2. They have adopted the word ‘beta’ in some service/product
   announcements. :)
3. Recently, they released Microsoft Virtual PC and Virtual Server for
   free as well. This directly competes with the VMWare offerings.
4. Both Vista and Windows Server 2003 R2 are available as downloads
   with a simple registration and timed expiration from their site.
5. They have acquired SysInternals and WinInternals, adding Mark
   Russinovich and Bryce Cogswell to their operating system
   team. Those folks were known for providing low-level debugging
   tools for free.
6. They are working with the community in some fronts. For example,
   the WiX project on SourceForge, receives lots of community feedback
   and even patches. It’s a quite high traffic mailing list.
7. They have created a SourceForge-like website for hosting community
   projects centered around Microsoft products.
8. Have you looked at MSDN? Now that’s documentation. Ok, it lacks in
   quality in a few places. But can you name a single Open Source
   project that provides similar documentation? What’s the sound of
   one hand clapping?

How big of a dent that can make in the Open Source community? How many
people, frustrated by the lack of a good and free IDE will move back
to developing on Windows?

My advice: keep an eye on Microsoft. One day we might wake up and
sheesh they made the Windows source code available. Will Linux be
able to survive that?

./volunteers –all > gnome-br-list.

Between bot-building, test-fixing and birthday-partying,
December went by faster than the blink of an eye.

My biggest deception this year was the goldegg project. Lots of
traction happened in the month after launch, and then it got pretty
much abandoned. The milestone dates for Plone 2.5 quickly approach,
and no-one seems to care about it. I wouldn’t be surprised if the
whole work done got thrown out. :(

I’m leaving for a short vacation this sunday. Hope the world has
changed just a little (for better I hope) when I get back next year.