hein added a comment.

  I've tested now and things look good, the multi-activity-migration thing is 
solved. Here's my remaining concerns:
  
  - No Kickoff support: Personally I think we should actually migrate all 
applets at once. One of the main motivations for this is so users no longer 
unexpectedly lose favorites when swiching between applets using Alternatives. 
If Kickoff (the default applet) doesn't use it, that is not yet actually 
achieved. And since the missing support was based on a misunderstanding ...
  
  - I really dislike the submenu with the checkboxes, in particular for 
removing a favorite. It's just very complicated and hard to use, you need to 
uncheck a checkbox to remove. I know everybody is sick and tired of this merge 
taking so long, but I am really, really hesitant of shipping that to users. It 
sort-of is okay for windows because it's more like tagging there and remove is 
actually done by closing the window, but here it's incredibly awkward.
  
  - The initial sorting of the migrated superset seems random. How is it done?
  
  - In https://phabricator.kde.org/D7561 we adjusted libtaskmanager to use 
applications:<kservice menuId()> as URL whenever we're dealing with a KService 
that returns a non-empty menuId() instead of KService::entryPath(). The 
approach taken is to accept both as input (rc file) but resolve the latter to 
applications: when possible and then use that for the model data roles. As 
adding a new launcher is done from the data roles, newly-added launchers are 
stored as applications:. That makes it a non-absolute path, which fixes cases 
like (a) apps (or the menu editor) making a copy of a .desktop file in $HOME to 
add/update data there and the Task Manager not noticing or (b) having a 
launcher point at a .desktop file in $HOME that disappears later (e.g. menu 
editor reset). Kicker still uses entryPath so it suffers from both problems. 
I'd like this adjustment done here as well, using more or less the same 
approach, and I think it would make sense as part of this patch maybe because 
I'm not sure how it will affect your migration and so we don't add outdated 
stuff to the KAStats db in the first place that would need to be rewritten 
later otherwise?

REVISION DETAIL
  https://phabricator.kde.org/D3805

To: ivan, mart, hein
Cc: Zren, plasma-devel, ZrenBot, progwolff, lesliezhai, ali-mohamed, 
jensreuterberg, abetts, sebas, apol, mart, lukas

Reply via email to