Wednesday, April 27, 2011

Tokamak 5, Day 2

Yesterday, in defiance of the weather reports, the day was sunny and reasonably warm and set the stage for a very productive day 2 here at Tokamak in the Netherlands. We held four design sessions in the morning: 2 on libplasma2 (specifically the dual topics of isolating QGraphicsView from the core code and using Qt Components), one on plasma-desktop defaults (a button to show the activities, an auto-hiding pager when virtual desktops drop to one, some default launchers that track the default file manager and web browser, and much more) and one on a new first-boot screen designed with OEM style installs in mind.

Several dozen individual tasks were placed on our large kanban window before we were done and then we spent the afternoon working on implementations. Today we have a nice neat pile of sticky notes in a variety of colors in the "Done" column, and a few more in the In Progress from the night before.

Today, after everyone's finished with breakfast and we have a short status update, we will take on another set of topics into the design meetings and start the process of creating and migrating tasks anew. I expect the topics to include the tablet shell design, another libplasma2 topic and KActivityManager (which just moved into kdelibs/experimental!) and SLC.

Unfortunately the weather is not nearly as nice today for us: it is already heavily cloudy and there is a slight chill in the air. We also won't have the grace of Adriaan's pancake cooking mastery to fill our bellies this evening, so we'll have to improvise something new.

3 comments:

Ffejery said...

"...and KActivityManager (which just moved into kdelibs/experimental!)"
:D Please give Ivan a high-five for me.

toddrme2178 said...

Great to see activity support improving. I look forward to seeing more about what was discussed.

Regarding libplasma2, while you are in the process of modifying plasma, have you considered moving the settings for each plasmoid (location, dimensions, and plasmoid-specific settings) into separate files (one file for each plasmoid)? If this is possible, it would make it MUCH easier to remove plasmoids that are causing problems, as well as move them between systems. Currently if there is a plasmoid that is messing up your system, you need to hunt through the long single config file, and potentially have to determine which plasmoid to remove based on its screen position in pixels (if you have more than one of the same type). It makes fixing problems very difficult.

Further, if there is a problem in your config, with some typo or other random error breaking it, there is no choice but to remove the whole thing and start over from scratch. With the multiple-files approach, you could temporarily move each file out of confi directory to see if it is the one causing problems, and with the much smaller amount of text to look through spotting the error (for advanced users) would be easier.

Putting all the config files for plasmoids from a given activity in a single folder would also make it much easier to copy plasmoid settings between activities (just copy the file from one folder to another).

I know this may not be technically possible or may incurr too much of a performance penalty, but if it is feasible I think it would be very helpful for users. It is probably something that needs to be considered early on though.

josericardo said...

Regarding SLC, when can we expect the blog post you promised explaining what it is?
Hopefully it will be soon enough. Cheers.