Am 13.02.2013 10:47, schrieb Will Stephenson:
On 12/02/13 22:21, Marcus Moeller wrote:
Hi all.
Some of you might have followed the problem we had with KDM and domain logons.
This issue should be fixed in my last commit to the related BZ, but it was a real long and ugly way to get there.
I (as a designer) had to set up an Active Directory domain controller. I had to borrow another machine as my personal box is a low power Atom which is not capable for holding Virtual Machines.
After that I had to join a machine to that newly created domain to see how the theme behaves.
I spent a few sleepless hours on that, and felt forced to do so by Kenneth and Bruno.
I'm glad that you were able to get it set up and provide all the features our users require.
From my perspective we had never defined which function-set must be supported by the theme and just remembered that ppl called for a userlist in 12.1 which I have added. Just to let you know, only a very few of KDM themes support functionalities like userlist or domain logon, even not the official ones from the KDE project. And we had all kind of themes in the past.
I have not followed the 12.2 KDM design which was driven by Bruno, but this seems to have these functions. Sadly it was not very well received from the design point of view.
This leads me to my final summary: Functionality should never be driven by design. All other Display Managers I know, split these aspects. This is why I think KDM is broken and should be fixed.
I don't understand this part. Are you saying that it's possible to break KDM's functionality with the theming? How do other DMs separate these concerns so that even if the theming does not completely address all functions, they still work?
Yes, it's possible to break the functionality with theming. Functions are added through elements in the theme. Take a look at the source and you will see. Most of the other DEs always offer all functions and don't let them be switched on/off within the theme. GDM e.g. is not really themable at all.
This could be done by
replacing it with LightDM (and don't argue with missing systemd patches. Afaik these are there, they are just not accepted upstream, which we could not expect as they are using upstart and ignore the rest).
If we want to stay with KDM, we should clearly define which functions must be supported. And again: don't expect that a designer will set up an Active Directory domain controller! If we want to support that, we need backed help from a technical contact.
Perhaps adding a simulation mode in KDM so designers can test all aspects of a theme in a running system?
Yes, that would be an option, too. But it would also add even more complexibility. With the current situation, even if we adopt most of the parts from a previous theme, changes must be tested. And setting up a test-scenario as it was necessary to fix the recent bug, is way to much to expect from us. Greets Marcus -- To unsubscribe, e-mail: opensuse-artwork+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-artwork+owner@opensuse.org