PA4 tasks discovery process for Alarms application. Screenshots of the application can be found here.
Note: Running the application manually can be done with: plasma-windowed org.kde.active.alarms
One line per defect describing it. We will triage confirmed and unfixed to bugs.kde.org.
|TODO||Application components||In the org.kde.active.alarms package, there is a platformcomponents/application/ section which is no longer referencable. A solution for application components in this form needs to be found.||
|TODO||Date/Time settings are theme specific||The date and time setters only work when using the air-mobile svg theme as they reference throbber.svgz which is only found there. Problem: that element seems a bit mobile-specific: on the desktop the time settings should probably have a different look/behavior.||
|TODO||Sometimes alarms get duplicated||when running on a device (and only there) sometimes the akonadi alarm agent gets duplicated resulting in alarms tha ring twice. Probable issue: the order in which akonadi and the code using akonadi are started||
|TODO||Cannot enter message||When I enter the "Message" field, the keyboard pops up, but whenever I tap any key on it, it disappears immediately without entering anything||
|TODO||Deleting current alarm when viewed in editor||Press on an alarm, it comes up in the editor. Then delete it .. and the editor is still showing the now deleted alarm.||
Proposal: "Application" FormFactor in libplasma, the application specific components would be loaded in form factor constraint events.
One section per issue. User stories for workflow related issues. Include a problem statement and possible solution:
Problem: ... User story: ... Suggested Solution: ...
Jane sets up an alarm for taking a cake out of the oven in half an hour. At that point, her device is sleeping, but she wants to be reminded to take the cake out anyway.
Suggested solution: An alarm should wake up the device from sleep.
This is unlikely to be implemented unless we get some kernel hackers working on it. Once the system goes to sleep (suspend to ram, really) then it's up to the kernel to wake up. The other option is to not suspend to RAM at all but go into some very power-friendly mode where virtually everything else is suspended .. except something that watches for events like alarms. This brings up the related issue of applications being able to inhibit sleeping, which should already work but needs to be tested to confirm it is working on the PA images; in that case, alarms could also just inhibit auto-sleep. Not the most elegant, but possible. - aseigo