> On April 12, 2011, 10:55 p.m., David Edmundson wrote:
> > account-filter-model.cpp, line 106
> > <http://git.reviewboard.kde.org/r/101108/diff/3/?file=14240#file14240line106>
> >
> >     I'm really not convinced this works, if you're going to treat them like 
> > ints then there's no point overriding this method at all.
> >     
> >     I'm pretty confident (though I may be wrong) this will put offline 
> > users at the top, then online, away, busy.
> >     
> >     I think  that we want to treat them as the enum values they are and 
> > need something more like:
> >     
> >     if (leftPresence == Tp::PresenceOnline) {
> >      return false;
> >     }
> >     if (rightPresence == Tp::PresenceOnline {
> >      return true;
> >     }
> >     
> >     if (leftPresence == Tp::PresenceAway) {
> >      return false;
> >     
> >     etc etc.
> >
> 
> Martin Klapetek wrote:
>     The purpose of lessThan(..) is to compare values to do the sort and it 
> seems like a good way to compare the uints (if they are ordered like 
> 0-online, 1-away...7-offline, otherwise it's no use). In your proposed way it 
> would have to compare all combinations of presences, no? Like if(leftPresence 
> == Online && rightPresence == Away) return true; if(leftPresence == Online && 
> rightPresence == Busy) return true;
>     
>     or am I getting this wrong? :)
>     
>     Also, Rémy, try to test with offline users shown.
> 
> David Edmundson wrote:
>     Your statement contains a massive "if" there.
>     I'm pretty sure the statuses are in this order:
>     
> http://telepathy.freedesktop.org/spec/Connection_Interface_Simple_Presence.html#Enum:Connection_Presence_Type
>     Which isn't a sensible order for listing.
>     
>     Also I don't think we need to do every possibly combo if it's written 
> like suggested, if either of them is "online" then that side 'wins' and you 
> return either true or false. Otherwise you continue onto the next presence, 
> if either of them are 'away' that way you only need each status once - though 
> it's still pretty lengthy.
>     
>     You may be right that it is easiest to turn the presence type into a rank 
> score, which does have 0 for online, 7 for offline, then compare those. I 
> would be happy with that. I just don't think it works as-is.

Ahh, now I'm getting your logic in the code above. Yes, this might be actually 
a way better solution...as always :P


- Martin


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://git.reviewboard.kde.org/r/101108/#review2599
-----------------------------------------------------------


On April 12, 2011, 9:05 p.m., Rémy Greinhofer wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://git.reviewboard.kde.org/r/101108/
> -----------------------------------------------------------
> 
> (Updated April 12, 2011, 9:05 p.m.)
> 
> 
> Review request for Telepathy.
> 
> 
> Summary
> -------
> 
> Sorts the contacts by name offers the possibility to sort it also by status.
> 
> 
> This addresses bug 246223.
>     http://bugs.kde.org/show_bug.cgi?id=246223
> 
> 
> Diffs
> -----
> 
>   account-filter-model.h d33231d 
>   account-filter-model.cpp 7182362 
>   main-widget.h 18f97ee 
>   main-widget.cpp a24ccf0 
>   main-widget.ui d000b9b 
> 
> Diff: http://git.reviewboard.kde.org/r/101108/diff
> 
> 
> Testing
> -------
> 
> 1. Open the contact list, the contacts are sorted by name.
> 2. Click on the "Sort by onlineness button", the contacts are sorted by 
> onlineness and by name.
> 
> 
> Thanks,
> 
> Rémy
> 
>

_______________________________________________
KDE-Telepathy mailing list
[email protected]
https://mail.kde.org/mailman/listinfo/kde-telepathy

Reply via email to