* Matt Zimmerman ([EMAIL PROTECTED]) wrote: > On Mon, Jan 29, 2007 at 07:51:31AM +0100, Mike Hommey wrote: > > On Sun, Jan 28, 2007 at 07:50:00PM -0500, Michael Gilbert <[EMAIL > > PROTECTED]> wrote: > > > severity 408460 wishlist > > > thanks > > > > > > On 1/26/07, Mike Hommey wrote: > > > >A workaround for this would be to set MOZ_DISABLE_PANGO. See the > > > >/usr/doc/share/iceweasel/README.Debian file. > > > > > > this behavior can be annoying to the user. would it be possible to > > > automatically set this option based on the language? for example, > > > only use pango for those languages that have rendering issues without > > > pango. > > > > Ubuntu does this, based on the user locale, and I think this is a very > > bad idea. The user locale is not necessarily related to the language > > used in the pages the user will look at. > > No, but it is our best approximation of whether that locale is especially > important to the user (which isn't an especially good one). It is trivial > for a power user to override this if it guesses incorrectly.
The other issue is one of surprise. A weird side effect of installing a locale will be that Firefox suddenly behaves differently and the connection will be completely non-obvious to the user. > > I also heard activating EXA instead of XAA in the X configuration > > improved the performance. > > Do you have a reference for this which we could follow up? I agree that the > current Ubuntu solution isn't ideal, but it was unacceptable for performance > to be as poor as it was for very common use cases. It really is a crappy situation, with both Pango and non-Pango renderers being broken in their own special ways. I wonder if any work is being done at speeding up Pango? Or is it Firefox's use of Pango that's to blame? -- Eric Dorland <[EMAIL PROTECTED]> ICQ: #61138586, Jabber: [EMAIL PROTECTED] 1024D/16D970C6 097C 4861 9934 27A0 8E1C 2B0A 61E9 8ECF 16D9 70C6
signature.asc
Description: Digital signature
