Thu, 28 Aug 2008

Making Intrepid Solid

Making Intrepid Solid

With feature freeze now in effect the bulk of the big changes in Intrepid should now be done. There will still be new features entering the archive with the appropriate exceptions, but the rate will slow as we move forward.

Now we really need to focus on making Intrepid solid. We want to squash as many bugs as possible, so that when we deliver the final release it is something we can be proud of.

This is something that everyone can help with. There are plenty of ways to help out, so there should be something for everyone.

Testing

Simply running Intrepid and reporting bugs is a great start. It's still not recommended to run it if you aren't able to fix a system that doesn't boot, or where X doesn't start, but if you are then upgrading now will be a great help.

You can do more than just using the system though, pick an application and start testing all of the functionality, and report the bugs that you find. Some bugs only show up in certain locales, with certain hardware, or with certain combinations of packages, so try different things and look for serious problems.

Upgrade testing is an area that is under-tested until the last minute when floods of users upgrade. Also the easiest testing to do lots of is upgrading pretty standard installations, but this doesn't catch a lot of problems. So, when you are comfortable with running Intrepid upgrade and let apport file any upgrade problems that you find. You should also be on the lookout for unnecessary prompts that happen while upgrading, or packages that are left broken by the upgrade.

Even if you are not happy running Intrepid yet you can still potentially help upgrade testing, thanks to the unstoppable Michael Vogt. He has written a tool that will clone your system in to a kvm virtual machine, and then upgrade that. This means you can test a real world upgrade without risk to your system. If you do this a few times during the remaining time for Intrepid and file bugs, then you will have a much better chance of a hassle free upgrade to the final release. You can find more details on Michael's work here. (Not everyone has kvm capable hardware though unfortunately.)

Looking at bugs

As well as trying to find your own bugs you can look at the ones that other people have already found. There are several important things here.

The first is bug triage, trying to make sure that a bug report has all the information that it needs, and trying to set an appropriate priority. This is really important work, and we always need more help doing it, so consider joining the bugsquad and helping out.

At this time important bugs should also be milestoned so that they can be concentrated on for the release if possible. Deciding the different classes of bugs here is really tricky, and there can often be disagreements. It is important work though. If you see a bug that should probably become release critical then work with the bugsquad to triage it, and make sure to suggest that it is considered for release-critical status.

Developers can help by actually trying to fix these bugs. Some can be easy, for instance if they are known to be fixed elsewhere. Some can be really complex, and take a lot of effort. Fixing things from the release-critical bug list, and lists of other important bugs is always valuable.

Doing the easy things

There are some ways to improve the quality that are actually fairly easy. Though there is a freeze in effect in Ubuntu there is still loads of work going on elsewhere, and many, many bugs being fixed elsewhere. Pulling these fixes in to Ubuntu will improve the quality, while in theory being easier than coming up with a fix for a bug.

We should keep an eye on upstream projects and pick up bug fix point releases to the versions that we have. If the project doesn't do this then look out for important fixes going in to trunk and back-port them. If you are doing that then it can be worthwhile looking at the versions in other distributions that plan to release soon, and if they carry the same version suggesting that you share the workload of creating these point release, or at least collaborate on fixes and share them.

Adding external bug watches in launchpad is also a great way to help. Jorge explained this recently. This helps easily spot when there is a bug fix that we could pull in. When there is they will appear on harvest

Harvest (now with a new look) is another easy way to do things. It lists opportunities to fix things that should be fairly easy, such as bugs fixed elsewhere, or bugs with patches attached.

Making Intrepid+1 rock

In parallel with all the above now is the time to start thinking about what you want to achieve in the next cycle. If any of that requires changes to an upstream then speaking to them early can be a good idea, as you can get their feedback and see how it fits in to their plan. I'm sure everyone has loads of ideas, and a bit of preparation now can help you hit the ground running in the next cycle.

Posted at: 12:51 | category: /ubuntu | Comments (1)


Hello,

I fully agree with you.

Why invent the wheel twice I would say.

Actually I am very disappointed about the Intrepid-version. And to people who want may be change from Microsoft to Linux (Ubuntu) this is embarrassing.

They laugh.

The Hardy Heron I was and am very satisfied about that.

Because I am not an expert in software engineering I can not determine why it went so wrong with the "Intrepid".

Anyway if Ubuntu is going on like this way It will come to an end. And that is a pity. Maybe it is a good idea to edit only once a year a new version with LTS. In this way developers have the time to test the version thoroughly.

Let us hope for better times.

Greetings
Frans

Posted by Frans van der Star at Tue Dec 23 01:38:29 2008