Review Request 125198: by default do not restore the previous session on next login

[Harald Sitter]

> On Sept. 14, 2015, 6:05 a.m., Martin Gräßlin wrote:
> > +1 - as a note: in openSUSE users were quite angry that we had broken session management, they won't be glad about the choice. My personal opinion is (as many probably known) that I would drop all session management support as I consider it as a relict from pre-suspend times.
> 
> Marco Martin wrote:
>     I still tend to prefer it to suspend, starts things kindof where i was but slightly more clean slate than suspend (maybe is also due that i still do find suspend unreliable and horribly slow with 8 or more GB of ram).
>     It's fine to disable it by default, but the thing I'm afraid is that it may break again without being noticed, and i definitely want to keep it and keep it working for the time being.

Suspend to harddisk is sometimes weirdly broken. Suspend in general has some hardware detection problems (see steamos suspend disabled because usb devices do not get detected when coming out of suspend). At the same time as I mentioned yesterday on IRC session restore often requires explicit support to be implemented on an application level which also renders it part-defunct depending on the applications one uses as the application will often just come up with the default UI state which conceptually is not "restoring" but "reopening". All in all both options are meh at times.

So, I'd argue that suspend should be the go-to choice for the "restore" feature and we should demote session management as a "reopening"-only feature. Specifically that means changing the name and advertising it in that fashion which is line with what Kai suggested above as reopening is a much more immediate choice you want to make on logout actions. This also means that we'd consider it correct behavior if an application comes up in its default UI state. If an application chooses to implement restore that is cool as well but not required from a UX POV.


- Harald


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/125198/#review85343
-----------------------------------------------------------


On Sept. 12, 2015, 3:37 p.m., Harald Sitter wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/125198/
> -----------------------------------------------------------
> 
> (Updated Sept. 12, 2015, 3:37 p.m.)
> 
> 
> Review request for Plasma.
> 
> 
> Repository: plasma-workspace
> 
> 
> Description
> -------
> 
> As requested by the VDG.
> Rationale being that starting with an empty session all the time is
> substantially less aggressive than potentially starting a gazillion
> applications slowing down startup and greeting the user with a
> cluttered desktop.
> 
> 
> Diffs
> -----
> 
>   ksmserver/main.cpp 4808a80081c3f4322c0d1b3223fc65bcbfeb26c1 
>   ksmserver/shutdown.cpp 636ae66fcce1d5c39fd697925b9094abc44e4808 
> 
> Diff: https://git.reviewboard.kde.org/r/125198/diff/
> 
> 
> Testing
> -------
> 
> installed. wiped ksmserverrc. multiple logins always result in an empty session.
> 
> 
> Thanks,
> 
> Harald Sitter
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/plasma-devel/attachments/20150915/5382cff1/attachment.html>