Project Timelord Midterm Review

The new year saw the demise of the 10th Doctor. ;_;

Project Timelord, however, is rolling right along. Since we are right around alpha 3 and around half way through the cycle, I thought I’d take a moment to do a review on the progress of Project Timelord.

For your convenience, here are the goals that we had set out to accomplish (be)for(e) Kubuntu 10.04:

  • Announce Project Timelord to the world
  • Establish sensible bug triage policy
  • Re-evaluate support policies
  • Review all patches to core KDE modules
  • Fix translations, through whatever means necessary
  • Thoroughly test all Kubuntu utilities, and fix bugs when encountered
  • Replace gdebi-kde/install package use with KPackageKit
  • Replace update-notifier-kde with kubuntu-notification-helper
  • Port printer-applet-kde from KSystemTrayIcon to the new KStatusNotifierItem
  • Package and include kcm-touchpad, the Touchpad module for System Settings
  • Add KAuth integration to Userconfig

Quite a list, looking back on it. To start off the review:

Announce Project Timelord to the world

Status: Done

Yeah, this one was pretty much done from when we announced it. 😉

Establish sane bug triage policy

Status: Done

The new policy has been set in to place. The result has been a cleaner, more effective bug tracker at Launchpad. As a result, I personally have been able to spend a lot more time triaging bugs at myself now that the volume of bugs at Launchpad has decreased. Nobody’s complained either, so that’s a good sign. 🙂

Reevaluate support policy

Status: Todo

Not too much has been done here actually. Post-release support has always been a best-effort, uh, effort.  Volunteers welcome, etc, etc. Pretty much somebody just needs to write up a doc of some sort and propose it at a Kubuntu meeting, I suppose.

Review all patches to core KDE modules

Status: Done

The patch review got completed pretty early. Upstream KDE should now be aware of all the patches we are carrying. We were able to upstream a fair number of these patches, so that everybody can benefit from them. Overall, I’d say this is a success and something that we (upstream, us, everybody) can benefit from doing on a per-cycle basis.

In addition, Plasma desktop scripting has introduces a safer way to define our default desktop. We are now using this scripting API, rather than patching Plasma itself or providing pre-determined configuration files, to determine our default desktop configuration. Both of the previously-used methods have problems, and this scripting API leads to a much safer and more-powerful solution than either method we used in the past. See Harald’s blog for the nitty-gritty details.

Fix translations, through whatever means necessary

Status: Done

Right off the bat I’d like to say: Launchpad translations are still in use. Like it or not, the Launchpad Translations infrastructure is, well, an infrastructure, and as such moving away from using it is not trivial, and would have the possibility to introduce regressions (which would be especially bad for an LTS release). However, that being said there are no known regressions from the upstream KDE translations for core KDE modules. Multilingual testers, please keep on testing with Lucid to help ensure things remain that way. 🙂

Thoroughly test all Kubuntu utilities, and fix bugs when encountered

Status: Ongoing

This will be pretty much an ongoing task up until the final release of Kubuntu 10.04. That hasn’t stopped great progress from being made. I’d like to take this moment to recognize the work of Amichai Rothman, who has done a Herculean amount of work wrangling over a dozen bugs for the KDE frontends of Software-properties and Jockey, including some very annoying and old beasties that nobody had ever got around (or figured out) how to tame. In particular, I’d like to point out this massive changelog entry which closes a dozen bugs on its own. Amichai didn’t stop there, though, later that month he squashed a handful more.

I feel this portion of the project is going strong, and I think it highlights a success of the project in general. Success in this portion could not have been achieved without fresh contributors, and I would like to thank any and all new contributors for their work.

Replace gdebi-kde/install package use with KPackageKit

Status: Todo

This one is waiting on some improvements to be worked out with KPackageKit upstream to ensure that KPackageKit is a suitable replacement for our current batch installer install-package. This should be fairly trivial to work out before 10.04, so I’m still optimistic.

Replace update-notifier-kde with kubuntu-notification-helper

Status: Done

The replacement for update-notifier-kde, kubuntu-notification-helper, is now included with Kubuntu 10.04 as of alpha 2. It completely implements all of what update-notifier-kde did, and does it better. My blog post on it has details, but the main improvements are dramatically reduced RAM usage and tighter integration with the KDE notification system.

Port printer-applet-kde from KSystemTrayIcon to the new KStatusNotifierItem

Status: Done

Earlier I had posted that this was completed. (And even users of Kubuntu 9.10 + KDE 4.4 get this!) Unfortunately it didn’t make it in upstream KDE for 4.4, but a patch is available that separates this new feature from the other features/string changes in trunk. Feel free to bug your distro to include it, if you don’t use Kubuntu. 😉

Package and include kcm-touchpad, the Touchpad module for System Settings

Status: Done

The alphas of Kubuntu 10.04 now include the kcm-touchpad System Settings module, located in the Keyboard and Mouse module in System Settings. Testing is, as always, welcome.

Add KAuth integration to Userconfig

Status: Todo (postponed)

This one was already a bit of a stretch, and was noted as such in the original Timelord document. Nothing has been done on this yet. (Goes back to the only-recent availability of working KDE 4.4 python bindings) I’m less optimistic that this will make it for Kubuntu 10.04, but you never know. Perhaps a better goal for 10.10 would be to get this in kdeadmin or somesuch.

The Midterm Conclusion

Out of the 11 goals the original Timelord document set for 10.04:

  • 7 are totally finished
  • 1 is ongoing
  • 3 still need done

Not bad for halfway, especially when you consider that several of the “Todo” items that still need done are trivial. So far, I will personally consider the Timelord initiative a success. A lot of good has come from it; goals are being met, and we have gained several new contributors and friends because of it.

But it’s not too late to contribute! Contributing to Project Timelord is exactly like contributing to Kubuntu. If you would like to contribute to Kubuntu, you can always send us a line at the kubuntu-devel mailing list, or pop on by the #kubuntu-devel IRC channel on freenode.

Keep on rocking,

Jonathan, signing out.


20 Responses to Project Timelord Midterm Review

  1. sheytan says:

    Please make User config KCM more user friendly. I mean, something like in OS X. Where user can choose account picture, etc. Now it looks more advanced, and new users can be confused what is what. 🙂

    • Yuriy says:

      Userconfig is not for this. You can set your account picture in “About Me” on the main page of system settings. The one thing that really needs to be made more user friendly in userconfig is adding a new user, other than that it’s mostly for more advanced settings.

      • sheytan says:

        Well, i know i can do that. But why do we have to have 2 separated KCMs to do almost same thing? There should be just one.

      • Yuriy says:

        I agree that being able to edit information about your current user in two different places is a little confusing, but here is the general idea.

        You can set any personal information about *you* in About Me and any other settings for *you* in the other modules on the “General” page of system settings.

        Userconfig is for system administration — adding and removing users, changing settings of *other* users on the system (but including your own). If you are the only human user on the system, you should generally not need it. I think “About Me” may need to be improved a bit though for this to really be true though.

        Invariably since userconfig lets you change settings for a user, it’s tempting to start adding all kinds of settings in there for that user until it encompasses not only “About Me” but half of System Settings. That said, I think it makes sense to have all the basic information in one place when adding/modifying a user. It’s just a matter of defining that information and laying it out nicely. Account pictures are questionable here because on the one hand they are a simple piece of information for each user that is sometimes displayed on the log on screen, on the other hand they are really a personal setting rather than an administrative one.

        Finally, if you have any design input (mockups are good!) on how to make it more user friendly, please contact us on IRC or by email or use the wiki page. And of course, it’s python and a relatively simple app, so a great place to start for a new developer/contributor!

  2. binarylooks says:

    Great wotk on all sides, KDE and kubuntu. Lucid is coming along nicely. Now if we only could have the music store for amarok and now that I think of it, is ubutntu one on by default?

    Anayways. Thanks for the best kubuntu ever.

    • echidnaman says:

      Have you checked out the Magnatune store that Amarok comes with by default? I’ve not used either the Ubuntu One music store nor Magnatune myself, but I can’t imagine that Magnatune can be bad enough not to even compete with the Ubuntu music store. 😉

      • Yuriy says:

        Integrating the Ubuntu music store into Amarok would be pretty neat though, and can’t be all that hard with the framework there for the other services.

  3. nip says:

    Great work. Kubuntu deserves much more praise as a KDE distro, than it is currently getting. OpenSuse disappointed me, but Kubuntu saved the day.

    *also: the Ubuntu crowd deserve a better DE than Gnome, but that is a different matter*

  4. Zun says:

    Am I correct in assuming that kcm_touchpad is entirely distinct from synaptiks ( )?

  5. d2kx says:

    Kubuntu is making great progress, definitly. One major bug for me still is a big performance regression in the installer. The installer was already slow with Karmic, but it got seriously bad with Lucid. I have to minimize the installer (is it a fullscreen plasmoid?) between every step to make it go faster. It may be related to me using the opensource radeon driver, not sure, I’ll to do some more testing.

  6. anonim says:

    good job. I would personally concentrate efforts on getting kde better integrated than the atayana stuff, but well I guess this is really up to you & canonical people.

    Again, good job 😉

  7. jamboarder says:

    I’m a Kubuntu user that was on the verge of jumping ship when Timelord was announced and decided to stick around to see how the project went. I’m genuinely impressed and excited by the progress. It’s a great initiative and has kept this Kubuntu user in the fold.

    Inspiring stuff!!!

  8. Astreek says:

    Thank you for the hard work! Kubuntu 10.04 will be a nice release!

  9. David says:

    Great work! Kudos to you and the Kubuntu community!

  10. […] Project Timelord Midterm Review « Jonathan’s Blog […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: