launchpadblog

Laura:  What do you do on the Launchpad team?
Vincent: Maintenance. Although I’m eagerly waiting for the sprint with gmb to get some hints on how to handle the beast 🙂  In the mean time, I’m focusing on fixing bugs and making the udd importer more testable.

Laura: Can we see something that you’ve worked on?
Vincent:  https://launchpad.net/bzr  and http://babune.ladeuil.net:24842/

Laura: Where do you work?
Vincent:  At home

Laura: What can you see from your office window? 
Vincent: The venerable Strasbourg post office, lovely old stones.

Laura:  What did you do before working on the Launchpad team?
Vincent:  Developing bzr.

Laura:  What did you do before working at Canonical?
Vincent:Various service/consulting work for > 20 years, including some episodes at software editors.

Laura: How did you get into free software?
Vincent:  With pleasure

I think the most important event was in 1993: I encountered a blocking bug in g++ related to C++ templates (way before it was standardized). That was a roadblock, no work-around and it was Friday afternoon. In despair, I posted a reproducing case in the related newsgroup. When I came back to work on Monday I got an email telling me the bug was known *and* fixed *and* where to get the patch for the compiler.

That was a light-bulb instant: free software support could be far superior to commercial software support !

One week later, I got a second email asking me if I was out of trouble… Amazing, not only did I get a fix faster than I could have dreamed, but the guy *came back* to ensure I got it…

I never looked back.

Laura:  What’s more important? Principle or pragmatism?
Vincent:  Both are important. If you forget one, be prepared to pay the cost. Both are dangerous too if you forget the other:

– being pragmatic only most often means you’re adding to your tech debt or rely on others to finish your work,

– respecting principles excessively means you never deliver anything.

Laura: Do you/have you contribute(d) to any free software projects?
Vincent: bzr is my most important contributions (including a few plugins). I’ve occasionally sent patches to gtk, perl modules and various other bzr upstream projects.

Laura:  Tell us something really cool about Launchpad that not enough people know about.
Vincent:  Pass  🙂

Laura:  Is there anything in particular that you want to change in Launchpad?
Vincent:  Make it easier to test against for all projects that rely on it (I’m probably biased here as the udd importer severely suffer from not being able to properly test interactions with launchpad (read *and* write (branch creation mainly)).

Laura: What do you do on the Launchpad team?
Jelmer: I’m one of the blue haired freaks on the Launchpad blue squad, although my hair isn’t actually blue – I’m sure we can fix this at the next squad sprint. At the moment, we are working on maintenance: fixing
critical bugs in Launchpad and dealing with incidents.

Laura: Can we see something that you’ve worked on?
Jelmer: I’ve contributed quite a bit to the code behind recipe builds. Most of my work has been on the backend though, not directly user-visible.

Laura: Where do you work?
Jelmer: Like most of us I work at home, which in my case is in Utrecht, the Netherlands. Occasionally I cowork with other teleworkers in Utrecht.

Laura: What can you see from your office window?
Jelmer: At the moment, I see just a big sad drapery made out of rain. On brighter days, I look out on a park and a canal.

Laura: What did you do before working on the Launchpad team?
Jelmer: The Blue squad, which I’m currently in, was originally the Bazaar team. Before that, I worked on the Launchpad team too. This was back in the days when there were no squads, but teams – I was in the Soyuz team. The inimitable Matt Revell interviewed me back in 2010:

Laura: How did you get into free software?
Jelmer: A long time ago, in high school, I ended up maintaining a few server machines running FreeBSD and Samba. After hitting some bugs, a dive into the source code followed to see what I could fix. I’ve been involved with various free software projects ever since.

Laura: What’s more important? Principle or pragmatism?
Jelmer: Do I really have to choose? That’s not very pragmatic.

Laura: Do you/have you contribute(d) to any free software projects?
Jelmer: Beside Launchpad, the main free software project I am involved in is Samba. There are several other projects that I have made major contributions to, such as Bazaar, CUPS, Wireshark, OpenChange, BitlBee.

I’m a Debian maintainer and Ubuntu uploader, mostly for projects I am involved in upstream. This knowledge comes in handy when working on the archive side of Launchpad.

Laura: Tell us something really cool about Launchpad that not enough people know about.
Jelmer: https://launchpad.net/builders lists all the Launchpad builders and
the mischief they are up to.

Laura: Is there anything in particular that you want to change in Launchpad?
Jelmer: It would still be really nice to have dashboards of some kind in
Launchpad. There is even a LEP.

Laura: What do you do on the Launchpad team?
John: I’m the team lead for the Blue squad. Right now our squad is on Maintenance, so I generally do the coordination work of our team with other teams and the system administrators.

Laura: Can we see something that you’ve worked on?
John: Before we switched to maintenance, our team was focusing on doing Bazaar work. Within the more recent time, we’ve done stuff like fixing up the Ubuntu Distributed Development package importer, and getting translations for Quantal started for Launchpad.

Laura:Where do you work?
John: I work from home in the Netherlands.

Laura: What can you see from your office window?
John: We have a forest near our house, and some of the other neighbors houses.

Laura: What did you do before working on the Launchpad team?
John: As mentioned, we were focused on development of Bazaar, though arguably that was still part of the Launchpad group (not officially, but in spirit).

Laura: What did you do before working at Canonical?
John: I worked for a company developing medical imaging algorithms, mostly for detection and visualization of disease.

Laura: How did you get into free software?
John: Our team wanted to use something better than CVS for development. At the time SVN was pretty hard to set up, and there was just the beginnings of a couple of tools for distributed version control. I got into tla at the beginning, and was happy when Canonical started Bazaar, and I was able to hack on it.

Laura: What’s more important? Principle or pragmatism?
John: I’m a fairly pragmatic engineer. I think it is good to use principle as a guideline, but in the end if the work isn’t in the hands of people using it, it is providing no benefit and is arguably wasted effort.

Laura: Do you/have you contribute(d) to any free software projects?
John: Well, Launchpad and Bazaar are both pretty clear things (and tla a little bit before that). I also developed some other tools while here at Canonical. Such as Meliae for profiling python memory.

Laura: Tell us something really cool about Launchpad that not enough people know about.
John: The UDD package importer turns the changes from debian packages into real VCS branches that you can do lots of nice stuff on. (annotate the history, log the history, see the graph over time, etc.)

I think we are at about 90% coverage, and you can do “bzr log ubuntu:package” to find out the recent history for a package in ubuntu.

Laura: Is there anything in particular that you want to change in Launchpad?
John: My own personal project with launchpad is improving the connection handling when accessing Bazaar branches. Right now it is approximately 3s just to do the ssh handshake and start talking to codehosting. I have some improvements that should decrease that significantly, but we encounter some strange hanging bugs that are only reproducible in production. And LP’s commitment to having minimal user-visible downtime is particularly problematic for SSH connections. A single HTTP request is less than 5s, but an SSH connection can legitimately be active for an hour if accessing a large project.