All changes that happened since the 0.93.2 release.

Please try out CVS if you can, we'll wait until Jan 18th to perform
the 0.93.3 release.  After that date, Jeroen will be back online and
will resume unstable developments on the CVS branch...

If there are issues, we would like to hear about them ASAP.

Raphael

- (2004.01.11 - RAM)
 * Fixed "fi->lifecount > 0" assertion failure when backing out data after a
   resuming mismatch.
 * When we run out of socket descriptors, try to close a banned fd, if any
   (both when making out a new connection or accepting one).
 * Made ban_reclaim_fd() a public function, that can be used when we run out
   of file descriptors.
 * Make sure pixmaps and other regular files are installed without
   the 'x' bit set.
 * Generalized call to ban_reclaim_fd() when opening files on out-of-fd
   condition by introducing the file_open() and file_create() calls.
 * Decreased node unstable cleanup timer from 1 day to 0.5 day.
 * New node info display configuration: show library size / # of files shared.
 * New properties: "show_gnet_info_shared_size", "show_gnet_info_shared_files".
 * When applying a patch message, make sure they did not already fill the
   QRP table before processing (should fix bug #762121).
 * Don't blindly assume pong information with hops=0 comes from the node,
   as LimeWire suffers from a bug breaking this assertion.
 * For prettyness, don't show a library size of 8K when the # of files shared
   is 0 (all UP will advertise 8K sharing so that this amount is a power of 2).
 * Remove -licuuc from the list of libs when it is present but does not
   contain the functions we want (fixes bug #859761).
 * Don't force bit 0 of amount of Kbytes shared to 1 if they share nothing.
 * Don't consider pongs with less than 8 kbytes shared as being ultra pongs.

- (2004.01.10 - RAM)
 * Fixed assert failure when all the entries from the cache are deleted.
 * When GTKG thinks it is offline, don't even try to do GWebCache updates.
 * [GTK1] Fixed titles[] array sizing using enum constants, avoiding too
   small sizing causing memory corruption [Patch from Thomas Schuerger].
 * [GTK2] Don't abort a download entry if it is a pseudo-header [Patch
   from Russell Francis] (fixes #874596).

- (2004.01.10 - Richard)
 * Moved generation of hostname for downloads pane to downloads_gui_common.c.
 * Fixed a memory leak in search_gui_common.c when search_hide_downloaded
   was activated.

Raphael


-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html
_______________________________________________
Gtk-gnutella-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/gtk-gnutella-devel

Reply via email to