Version control discussion on the Python list

The Python developers have been discussing a migration off CVS on the python-dev mailing list. During the discussion, Bazaar-NG was mentioned. A few posts of note:

I’m going to have to play around with bzr a bit more, but it looks very nice (and should require less typing than baz …)

Version Control Workflow

Havoc: we are looking at ways to better integrate version control in Launchpad. There are many areas that could benefit from better use of version control, but I’ll focus on bug tracking since you mentioned it.

Take the attachment handling in Bugzilla, for instance. In non-ancient versions, you can attach statuses to attachments such as “obsolete” (which has some special handling in the UI — striking out obsolete attachments and making it easy to mark attachments as obsolete when uploading a new attachment). This makes it easy to track and manage a sequence of patches as a fix for a bug is developed (bug 118372 is a metacity bug with such a chain of patches).

If you look at this from a version control perspective, this sequence of patches forms a branch off the mainline of the software, where each newly attached patch is a new revision. The main differences being:

  • No explicit indication of what the patch was made against (code base or revision), or what options were used to create the patch.
  • No linkage between successive patches (can be a bit confusing if multiple patch series are attached to the same bug report).

So why not just use real version control to manage patches in the bug tracker? The big reason for projects using CVS or Subversion is that only authenticated users can create branches in the repository, and you don’t want to require contributors to ask permission before submitting fixes.

So this is an area where a distributed version control system can help: anyone can make a branch, so potential contributors don’t need permission to begin working on a bug. This also has the benefit that the contributors get access to the same tools as the developers (which is also helpful if they ever become a regular developer).

Now if you combine this with history sensitive merging and tell the bug tracker what the mainline branches of the products are, you can do some useful things:

  • Try and merge the changes from the bug fix branch onto the mainline, and see if it merges cleanly. This can tell a developer at a glance whether the patch has bitrotted. This could also be used to produce an up to date diff to the mainline, which can aid review of the changes.
  • Check if the bug fix branch has been merged into the mainline. No need for developers to manually flag the attachment as such.

We discussed some of these features in the context of Launchpad at the recent Brazil meeting.

Back from Brazil

  • Post author:
  • Post category:Uncategorized

I got back from the Launchpad sprint in São Carlos on Tuesday afternoon. It was hard work, but a lot of work got done. Launchpad is really coming together now, and will become even better as some of the things discussed at the sprint get implemented.

One of the things discussed was to formalise some of the development workflow we’ve been using to develop Launchpad inside Launchpad itself so that it will be usable by other projects.

I really enjoyed the time in Brazil. The food and fruit juices were great (especially the ones made from native fruits like Açaí).

At the end of the second week, Mark flew us up to Rio de Janeiro for the weekend on his jet:

Canonical One
Inside Canonical One

We took a cable car up Corcovado mountain to see the Cristo Redentor statue. There was a great view from the top.

Cristo Redentor
Rio de Janeiro from Corcovado

Since I was leaving that weekend, I didn’t fly back to São Carlos with everyone else. Instead Kiko got a local travel agent to book me a flight directly to São Paulo, so that I could catch my international flight.

Unfortunately, when I went to the Varig ticket counter to pay for the ticket there was no record of the booking, which was bad.

However, I was able to buy a ticket on the flight anyway (which was due to leave in an hour), which was good. I even ended up paying less for the ticket than I expected.

Once I got through security, I found the flight had been delayed, which was bad. After the departure time changing about 3 times, we ended up boarding about an hour after the original listed departure time. This happened to coincide with the listed departure time of the next Varig flight to São Paulo (which had the same gate listed too), causing some confusion.

They served chocolate fondue on the flight, which was nice.

When I reached São Paulo, it turned out that my checked luggage hadn’t, which was bad. I filled out a lost luggage form, and the staff said they’d try to get my bag to the international airport in time for my next flight if it turned up.

At the Buenos Aires and Auckland airports, I tried to find out whether my bags had made it onto the flight. The conversations would go something like this:

me: Hi, my bag got lost on a previous flight and I want to know whether it made it onto my current flight. Here is the lost luggage form with the bag tag number.
them: do you have a bag tag?
me: no, they took the tag when processing the lost luggage report.
them: well, I can’t track the bag without the bag tag. You should have kept the tag.
me:

It was almost the same in Sydney, except the guy at the desk took a look at the form and realised that it had a bag tag number on it (the thing they wanted the bag tag for), and found that the bag had been put on my flight. Sure enough, it turned up on the carousel.

The rest of the trip was fairly uneventful, which was nice.

HTTP resource watcher

  • Post author:
  • Post category:Uncategorized

I’ve got most of the features of my HTTP resource watching code I was working on for GWeather done. The main benefits over the existing gnome-vfs based code are:

  • Simpler API. Just connect to the updated signal on the resource object, and you get notified when the resource changes.
  • Supports gzip and deflate content encodings, to reduce bandwidth usage.
  • Keeps track of Last-Modified date and Etag value for the resource so that it can do conditional GETs of the resource for simple client side caching.
  • Supports the Expires header. If the update interval is set at 30 minutes but the web server says that the it won’t be updated for an hour, then use the longer timeout til the next check.
  • If a permanent redirect is received, then the new URI is used for future checks.
  • If a 410 Gone response is received, then future checks are not queued (they can be restarted with a refresh() call).

I’ve also got some code to watch the HTTP proxy settings in GConf, but that seems to trigger a hang in libsoup (bug 309867).

While I wrote the code for use in GWeather, it could be quite useful for other tasks that require watching an HTTP resource such as:

  • HTTP calendar backend of evolution-data-server.
  • A stock ticker applet like gtik.
  • Possibly an RSS reader.

The code is available in my Bazaar archive:

baz get http://www.gnome.org/~jamesh/arch/james@jamesh.id.au/http-resource--devel--0