Re: Hard code freeze break request for nautilus

Have you tested this for a day? If so, here's 1 of 2 approvals.

> Hi,
> I'd like to request permission to break hard code freeze for the
> attached nautilus patch.  It has been approved by Alex, pending
> permission from the release team.
> The problem this patch fixes:
> if nautilus is launched without startup-notification (some examples of
> where this happens is using the "Desktop" entry in the "Places" panel
> menu or using bookmarks made in the filechooser; see bug 166242,
> comment 2), those windows have not been getting focus and have been
> stacked below the previous focus window.  Similarly, if the window was
> already open somewhere then metacity would get a request to show the
> existing window, but the request would appear to be very old and thus
> metacity wouldn't want to interrupt the users' work that it thought
> had occurred in the meantime by actually showing the window.
> The basic idea:
> gtk+ provides a default hint to the window manager about when a given
> window was launched.  This default is used by nautilus when
> startup-notification doesn't provide the exact launch time, but this
> default sucks for apps like nautilus that forward an
> "open-a-window-for-me" request to a previous instance.  This patch
> provides a pretty good approximation to launch time in those cases to
> use instead of the gtk+ default.
> Severity, trade-offs, etc.:
> This is basically a way to work around any regressions relative to
> Gnome 2.8 that bug 166722 is causing so that we can punt that bug
> until 2.12 (because fixing 166722, one of the showstoppers on Luis'
> list, would be far more involved).  The majority of the patch is a
> verbatim copy&paste of a single function from elsewhere in nautilus
> that has been used for months and thus should be perfectly safe.  The
> remainder of the patch is pretty short and I believe safe and
> straightforward.  It does introduce the possibility of nautilus
> windows stealing focus when they shouldn't (but only if the user is
> really fast and startup is really slow), but I don't think any better
> could be done when startup-notification isn't used and it makes the
> default behavior much saner.
> Thanks,
> Elijah
> --
> nautilus-list mailing list
> nautilus-list gnome org

Murray Cumming
murrayc murrayc com

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]