[Bugs] #1435 UNSP: Do no use random color if metadata color is not valid
Sugar Labs Bugs
bugtracker-noreply at sugarlabs.org
Wed Sep 30 11:19:16 EDT 2009
#1435: Do no use random color if metadata color is not valid
------------------------------------------+---------------------------------
Reporter: alsroot | Owner: erikos
Type: defect | Status: new
Priority: Unspecified by Maintainer | Milestone: Unspecified by Release Team
Component: sugar-toolkit | Version: 0.85.x
Severity: Unspecified | Keywords: r?
Distribution: Unspecified | Status_field: Needinfo
------------------------------------------+---------------------------------
Comment(by alsroot):
Replying to [comment:3 eben]:
> What does it mean to have an "invalid" entry? I understand the need to
have a sane fallback, of course, which merits a decision here, but
shouldn't this be an edge case?
In most cases, I think, it could be remains after broken ds, but activity
could create ds entries with missed icon-color property.
> That said, we shouldn't be using gray for this purpose, in my mind,
since that can potentially be confused with uninstantiated activities.
We're talking about entries for ongoing activities here, and those need to
be colored to signify that state. I think the only logical choice, when no
other identity is known, is to use the identity of the current owner. That
is, if it's in my Journal and my Journal isn't told its identity (its
colors), then it should assume my colors instead.
>
> This is likely to be the correct choice anyway in a number of cases,
such as files copied to the Journal from non-Sugar sources, files
downloaded by activities, etc. This application of identity should occur
at the time an entry is added to the Journal, not at the time it is
launched, so that all future actions (starting, sending, copying, etc.)
retain that identity going forward.
Ok, I'll post new patch.
--
Ticket URL: <http://bugs.sugarlabs.org/ticket/1435#comment:4>
Sugar Labs <http://sugarlabs.org/>
Sugar Labs bug tracking system
More information about the Bugs
mailing list