From Bugzilla Helper: User-Agent: Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.4) Gecko/20030624 Netscape/7.1 (ax) Description of problem: I just tried to compile package oaf-0_6_10-6 from Fedora. The compiler said ac-query-expr.c(517): remark #592: variable "retval" is used before its value is set The source code is QueryExprConst retval; if (!fi->name) { g_warning ("Invalid function name '%s'", func_name); return retval; } Suggest initialise retval before use. Version-Release number of selected component (if applicable): oaf-0_6_10-6 How reproducible: Always Steps to Reproduce: 1. by inspection 2. 3. Additional info:
(Again, sorry about the inactivity) oaf is a package which has pretty much ceased development and is only retained for compatibility reasons only. For that reason, I'm don't think its worthwhile to release a new version of the package purely to fix a warning which won't actually cause a problem in practise. I've checked the latest version of the code (in libbbonobo/activation-server) and the bug has been fixed. Thanks.
>oaf is a package which has pretty much ceased development Strange. Current version is 0.6.10-12, which means six versions have elapsed since my bug report. >I'm don't think its worthwhile to release a new version Entirely your choice. A definate read of uninitialised memory in my code would have me re-coding it, but for oaf, that's up to you. >I've checked the latest version of the code and the bug has been fixed. Even stranger. oaf is still part of FC4 test1, and the bug report still exists in 0.6.10-12, dated around 27 April 2005.
(In reply to comment #2) > >oaf is a package which has pretty much ceased development > > Strange. > > Current version is 0.6.10-12, which means six versions have elapsed since my bug > report. I'm pretty sure I've explained this to you before - the "release" number is the one being bumped here, not the version number. So, what's happened since -6? : * Wed Mar 2 2005 Mark McLoughlin <markmc> 0.6.10-12 - Rebuild with gcc4 * Fri Aug 6 2004 Tim Waugh <twaugh> 0.6.10-11 - Fixed underquoted m4 definitions. * Tue Jun 15 2004 Elliot Lee <sopwith> - rebuilt * Tue Mar 02 2004 Elliot Lee <sopwith> - rebuilt * Fri Feb 13 2004 Elliot Lee <sopwith> - rebuilt * Mon Oct 27 2003 Havoc Pennington <hp> 0.6.10-8 - rebuild for exec shield * Mon Jun 23 2003 Havoc Pennington <hp> 0.6.10-7 - rebuild The package has been rebuilt a number of times with no changes apart from bumping the release number. The package has seen no real development at all and is only there for compatibility reasons. It is not worth time spending time fixing compiler warnings in such packages. > >I've checked the latest version of the code and the bug has been fixed. > > Even stranger. oaf is still part of FC4 test1, and the bug report still exists > in 0.6.10-12, dated around 27 April 2005. The "latest version of the code" I'm referring to is in the libbonobo package.