Unfortunately Mozilla-1.0 remains as unstable as ever in face of
handling corrupted image files.  There don't seem to be as many
complaints about mozilla stability as I seem to be seeing.  One
particular site I use which auto-refreshes itself every hour and which
unfortunately often comes with a corrupted JPEG or two.  My Mozilla
sessions never run very long (only hours) if I leave that page on the
currently visible tab (regardless of whether the window is iconised or
not).

When I did an upgrade of gdk-pixbuf a few weeks ago I learned that it's
very "raw" code and does very little error checking.  To quote from the
TODO file:

* The loaders don't deal with exhausted memory conditions very
  gracefully.
[[....]]
  We do not have very good error reporting overall in the loading
  functions, either the synchronous file loaders or the progressive
  loaders.  We should figure out some common errors (out of memory,
  file not found, corrupted data, incomplete data, etc.) and add
  return values for them.  Yes, this means yet another API change.
  Live with it.

* We need a good stress-test suite!!!  Some nongraphical regression
  tests for `make check' would be nice.


Now if I'm not grossly mistaken gdk-pixbuf replaces Imlib.  The Mozilla
documentation says mozilla uses "Imagelib" to load images.  Is
graphics/imlib the same as what they call "Imagelib"?  Is Imagelib more
robust?

-- 
								Greg A. Woods

+1 416 218-0098;  <gwoods@acm.org>;  <g.a.woods@ieee.org>;  <woods@robohack.ca>
Planix, Inc. <woods@planix.com>; VE3TCP; Secrets of the Weird <woods@weird.com>