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

Re: Bug#401296: [g-i] The point about bugs related to keyboard handling



On Monday 04 December 2006 22:20, Attilio Fiandrotti wrote:
> > - fails to load on some machines with rgba pixel format
>
> I believe we've found the cause of this (fb0 and fb1 swapped): i'll
> reply to #400898 with a simple workaround for this bug tomorrow if i
> can get my hands on the syslog file

That would surprise me as I found out today how vesafb and vga16fb get 
loaded and AFAICT vesafb should _always_ get loaded first [1].
Either syslog or dmesg should be able to tell us for sure.

> looking at the newt frontend i see set_tile() is not implemented, while
> in the gtk frontend it is and window's title is updated *only* in
> set_title().
> I guess in the newt frontend the main title is updated outside
> set_title() too, is this correct?

No idea, you'll have to check the code or maybe try to catch Colin.

> Should the GTK frontend behave in a similar way?

Yes, that would probably be best.

[1] Turns out that vesafb is compiled into the kernel [2], but only 
activated if the boot param video=vesa is present as with g-i.
As vga16fb is a module, it will always be loaded later (by the 
debian-installer-startup.d script.
[2] This also means that the current code in the script that loads the 
framebuffer devices for i386 _is_ incorrect as it assumes that vesafb is 
modular.

Attachment: pgp8A8sSTHHMs.pgp
Description: PGP signature


Reply to: