Pango development

[Owen Taylor]

On Mon, 2004-03-01 at 10:32, Leonard den Ottolander wrote:
> Hello,
> 
> I am seeing crashes in gnome-terminal when I use the midnight commander.
> This issue seems to be pango related. I reported this issue 7 weeks ago.
> See http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=113284 . I think
> it might have to do something with the fact that I use ISO-8859-15 as my
> charset, since the crash seems to appear in a call to
> g_utf8_next_char().

> Should g_utf8_next_char() handle other charsets graciously? If not, then
> how can I use my system with another charset than UTF-8? Shouldn't pango
> handle other charsets as well? Has this changed in 1.3?

I'm somewhat at a loss as to why you think it is happening in
g_utf8_get_char(); all the backtraces on that bug-report show
a crash somewhere in pango-layout.c (without symbols, exactly
where is hard to tell.)

Pango checks UTF-8 validity on pango_layout_set_text() so generally
invalid UTF-8 can't crash Pango in strange locations. Though
this  crash is related to EelEllipsizingLabel, which a pretty awful
hack; so it's just barely possible it's stuffing bad text
into Pango in some fashion.

Like any bug, what is the most incredibly useful information to have is
the sequence of steps that is required to take a plain-vanilla system
and make it reproduce the crash.

> When studying the pango rpm I noticed a reference to
> http://bugzilla.gnome.org/show_bug.cgi?id=121488 . In that bug report
> Owen states: "Is it really worth rewriting stuff in the 1.2.x branch,
> which will shortly be dead." What does this comment mean with respect to
> the support of pango for Fedora 1 and RHEL 3?

For RHEL 3, I'll backport any fixes that are necessary; this is
generally a pretty small set; it would basically have to be a
user-triggable crash. (121488 actually was fixed in the
RHEL 3 SRPM with a local patch)

For Fedora 1; the same applies until we release FC2, at which point
it's Fedora Legacy's problem. :-)

Regards,
						Owen