Re: GTK frontend updated, please test unofficial miniiso
Eddy Petrişor wrote:
On 10/26/05, Attilio Fiandrotti <fiandro@tiscalinet.it> wrote:
Hi
Hello
(...omiitted..)
-In the partman's main menu partitions could be displayed as childs to
disks drives: this can be obtained by hacking a bit the frontend in
order not to have to change the partman templates
For this exact reason (the interface must know the backend) debconf's
specifications should be changed to support buffered questions.
JoeyH, any chance of that happening?
My target is doing this without having to touch partman, just adding
some "intelligence" to the frontend to let him understand what rows are
related to a drive and what rows are a partition inside the drive (like
i did for the question that asks to choose nation ("child"), where
continents are "fathers" )
-you can'tselect in lists by typing the first letter (as in newt
frontend) on GTKDFB (any version, on GTKX works) since this is an
unimplemented feature in GTKDFB (any version)
(Gdk-DirectFB-WARNING **: gdk_display_request_selection_notification
Unimplemented function)
this is a major setback related to what the newt frontend offers.
I'll report those bugs to Mike Emmel, but since he is working on GTKDFB
2.8 i think he won't have the time to correct them in GTKDFB 2.0.9 :(
-Testing on archs different from i386 : ATM this is the only arch where
miniiso images are knows to be built correctly.
EddyP is known to have made some experiments on PPC but it seems he's
stuck with segfaults that prevent the frontend from working properly.
Well, actually I managed to build an image without libary reduction
but it presented the flashing effect (FB is initiaited and then
crashes). I didn't had time to investigate in this matter to see what
happened (I ran the installer from the tmp directory of the build and
my system blocked - otoh, the mouse pointer was displayed and it moved
one pixel)
my 2 cents contribute on this: to state if the crashes are related to
GTKDFB libs or to the frontend only try to compile a simple "ghello.c"
application on your PPC and run it from inside the miniiso to see if it
crashes too.
When debugging under i386 tests like this helped a lot.
- label_string = malloc(strlen(q_get_description(q)) + 8 );
(...omitted...)
hpadbox = gtk_hbox_new (FALSE, 5);
I was always (since I grew some experience in programming) of the
oppinion that magic numbers are a thing to avoid as much possible.
Use #defines instead, it will make the code more lisible.
You're right, those parts of code are too hacky.
ciao
Attilio
Reply to: