svn::talk 0.0.1
Filter: muesli at gmail dot com 

83 commits found for author ossi.
Mostly active for:   kdebase   kdebase 3.5   kdelibs   kdelibs 3.5   kdenetwork

TimeProjectRevisionAuthorRepliesMessageThursday, 31. August 2006
09:37kdebase 3.5579161ossi0put --no-in-notice in the makefile - the notice makes sense for manual invocations only.


TimeProjectRevisionAuthorRepliesMessageTuesday, 29. August 2006
20:15kdebase578633ossi0make pam check not bomb out if libdl was not found


TimeProjectRevisionAuthorRepliesMessageMonday, 28. August 2006
09:29kdelibs578040ossi6s_flags is an int, of course. thx ervin.
09:18kdebase578038ossi0svnrevertlast, now that the KCrash change is actually in. :}
09:16kdelibs578037ossi0change setSafer to a bit field: option not to try kdeinit and option to keep file descriptors open. kdm uses all three flags at once and kapp uses none of them, but what the heck. :)


TimeProjectRevisionAuthorRepliesMessageSunday, 27. August 2006
19:25kdebase577885ossi0make the clock qt4-clean
18:32kdebase577862ossi0- replace kapp with qapp + kinstance - some other setup tinkering kdm now "works" again.
18:23kdebase 3.5577860ossi0- safer file permissions (pretty uncritical, i think). - close file before anouncing it's name to avoid confusion
18:20kdebase577858ossi0- safer file permissions (pretty uncritical, i think). - close file before anouncing it's name to avoid confusion - clearer(?) error message
18:18kdebase577857ossi0make it "kdm-compatible" by changing kapp into qapp+kinst
18:17kdelibs577856ossi0make qt_arg[cv] public and add accessor for about[Data]. this is needed for kapp-less programs. this class needs some serious thinking ...
10:31kdebase577659ossi0(proper) coverity fixes. some reverts as well, cleared with coolo.


TimeProjectRevisionAuthorRepliesMessageTuesday, 22. August 2006
18:45kdebase575982ossi0now trying to clean up the getdomainname mess (which i sort of perfected) ...
18:18kdebase575979ossi1561ok, so this is what you get when you start off with someone else's buggy code and don't double-check. two others' guesswork already failed, so trying to do it right finally: clean up the sysinfo mess.


TimeProjectRevisionAuthorRepliesMessageSunday, 20. August 2006
20:21kdebase575125ossi5518make the setlogin() test more thorough - glibc provides a dummy
19:06kdelibs575090ossi0add CheckCSourceRuns from cmake cvs - remove after upgrade to cmake 2.5
15:10kdebase574925ossi1475make kdm & kcheckpass "cmake-clean" (i think). includes moving kcheckpass and some other stuff to workspace/. that wasn't fun ...
13:48kdelibs574886ossi0new macros macro_add_{compile,link}_flags(target "flags ...")


TimeProjectRevisionAuthorRepliesMessageSaturday, 19. August 2006
15:31kdebase 3.5574624ossi0fix unchecked malloc
13:45kdebase 3.5574588ossi0fix crash
12:24kdebase574560ossi0fix it the right way


TimeProjectRevisionAuthorRepliesMessageSunday, 04. June 2006
16:14kdenetwork548119ossi0irc.debian.org moved. backport on request.


TimeProjectRevisionAuthorRepliesMessageWednesday, 10. May 2006
20:39kdebase539526ossi0svnrevertlast


TimeProjectRevisionAuthorRepliesMessageMonday, 17. April 2006
09:53kdelibs530665ossi0lalala ... *of course* i had to forget to make -pedantic happy ...
09:48kdelibs530664ossi0forwardport (this time to the right branch) MergedStderr and CTtyOnly.
09:47kdelibs530663ossi0init d->utf8


TimeProjectRevisionAuthorRepliesMessageSaturday, 08. April 2006
08:58kdebase527443ossi1being anal about standards compliance
08:56kdebase 3.5527442ossi0unscrew font attribute evaluation (operator precedence!).
08:56kdebase527441ossi0unscrew font attribute evaluation (operator precedence!). i think this happened as a result of some stupid s&r.
08:55kdebase 3.5527439ossi0being anal about fallthrough comments
08:54kdebase527438ossi0being anal about fallthrough comments
08:42kdebase 3.5527432ossi0being anal about standards compliance


TimeProjectRevisionAuthorRepliesMessageWednesday, 15. March 2006
14:34kdebase518880ossi0overhaul of Xaccess stuff: - moved alias recursion check to config reader - more detailed error checking and reporting: - unresolved aliases are reported - aliases with unresolved hosts in denying rules invalidate the whole acl - support more features of the core in the config reader: - permitting rules can optionally start with =, so it is possible to use aliases in the permission part - aliases which are used only in the permimission parts of rules can contain wildcarded patterns
14:34kdebase 3.5518879ossi0overhaul of Xaccess stuff: - moved alias recursion check to config reader - more detailed error checking and reporting: - unresolved aliases are reported - aliases with unresolved hosts in denying rules invalidate the whole acl - support more features of the core in the config reader: - permitting rules can optionally start with =, so it is possible to use aliases in the permission part - aliases which are used only in the permimission parts of rules can contain wildcarded patterns
13:54kdebase518870ossi0introduce LogWarn() and adapt some Log* calls.
13:54kdebase 3.5518869ossi0introduce LogWarn() and adapt some Log* calls.


TimeProjectRevisionAuthorRepliesMessageThursday, 09. February 2006
20:11kdebase507711ossi0yes, it looks like calling pam_strerror with pamh == 0 is ok.
20:10kdebase 3.5507710ossi0thanks for the reminder ... as expected, (at least linux-)pam ignores the pamh argument to pam_strerror, so it can be safely set to null.
15:27kdebase507603ossi0don't recycle pam handles. there is just too much crap in the pam world out there for this to be reliable. as a side effect, all relevant error paths now do an immediate cleanup of the pam handle.
15:24kdebase 3.5507599ossi0don't recycle pam handles. there is just too much crap in the pam world out there for this to be reliable. as a side effect, all relevant error paths now do an immediate cleanup of the pam handle.
14:52kdebase507589ossi0Xsession: $DISPLAY is better than $PID for temp file name
14:51kdebase 3.5507588ossi0Xsession: $DISPLAY is better than $PID for temp file name


TimeProjectRevisionAuthorRepliesMessageFriday, 03. February 2006
19:34kdebase505422ossi0simpler, yet more effective Xauth writeout check.
19:34kdebase 3.5505421ossi0simpler, yet more effective Xauth writeout check.


TimeProjectRevisionAuthorRepliesMessageSunday, 29. January 2006
09:19kdelibs 3.5503454ossi0new communication flag CTtyOnly instructing setUsePty to create a controlling TTY without redirecting anything to it. yes, this really works without changing a single line in kprocess.cpp.
08:42kdelibs 3.5503447ossi0add new communication mode MergedStderr to avoid problem seen with KProcIO's includeStderr mode.


TimeProjectRevisionAuthorRepliesMessageFriday, 20. January 2006
20:51kdelibs 3.5500677ossi0blarhg, forgot to remove the logout() call.
20:48kdelibs 3.5500676ossi0revert r491786 (auto-logout() on master close): login() might be called from another process (child) than logout(), so the state variable simply does not work ... thanks to CCMAIL: Pavel Troller <patrol@sinus.cz> for the report.


TimeProjectRevisionAuthorRepliesMessageTuesday, 17. January 2006
05:19kdebase499146ossi0- don't assign potentially const * to a non-const *. - shuffle some headers.
05:16kdebase 3.5499144ossi0don't assign potentially const * to a non-const *.


TimeProjectRevisionAuthorRepliesMessageSunday, 15. January 2006
17:30kdebase498553ossi0add error handling to user Xauth writeout.
17:28kdebase 3.5498552ossi0add error handling to user Xauth writeout.
13:42kdebase498460ossi0don't crash if krb5 is compiled in but not used. don't leak string (ok, it's irrelevant, as the process will exit immediately afterwards anyway).
13:39kdebase 3.5498456ossi0don't crash if krb5 is compiled in but not used.


TimeProjectRevisionAuthorRepliesMessageSaturday, 14. January 2006
17:04kdebase 3.5498094ossi0another stab at the csh startup files hack: don't eval all output directly, but instead redirect only the output of 'export' into a temp file and source it.
17:04kdebase498093ossi0another stab at the csh startup files hack: don't eval all output directly, but instead redirect only the output of 'export' into a temp file and source it.


TimeProjectRevisionAuthorRepliesMessageFriday, 30. December 2005
19:44kdebase492709ossi0re-gen. a little late ...
19:35kdebase492705ossi0don't auto-regenerate here as well.


TimeProjectRevisionAuthorRepliesMessageTuesday, 27. December 2005
20:52kdebase 3.5491887ossi0create the right file ;)
14:51kdelibs 3.5491789ossi0initialize utf8 member
14:47kdelibs 3.5491786ossi0auto-logout() on master close (and consequently destruction). had this on disk for ages ...
13:58kdebase491779ossi0kwrited does not use tepty for ages


TimeProjectRevisionAuthorRepliesMessageSunday, 25. December 2005
18:23kdebase 3.5491352ossi0another indentation fix
18:21kdebase 3.5491350ossi0coolo's code is fine - after all i ok'd it - but he really needs to get his friggin' emacs under control. :)
10:40kdebase491297ossi0claim maintenance here as well
10:25kdebase491294ossi0don't remove xdmctl/ - it causes problems for SE Linux. xauth/ was never removed, either, so no loss.
10:24kdebase 3.5491293ossi0don't remove xdmctl/. it was meant to be nice ("clean up your crap, dude"), but it causes problems for SE Linux. xauth/ was never removed, either, so no loss. this fixes CCMAIL: 265920@bugs.debian.org


TimeProjectRevisionAuthorRepliesMessageWednesday, 21. December 2005
13:10kdebase490317ossi0and now fix it the right way.
13:10kdebase 3.5490316ossi0and now fix it the right way.
09:48kdebase490266ossi0and now finish the job ...
09:47kdebase 3.5490265ossi0and now finish the job ...
09:28kdebase490261ossi0absolute paths in "generated from ..." suck.
09:28kdebase 3.5490260ossi0absolute paths in "generated from ..." suck.


TimeProjectRevisionAuthorRepliesMessageThursday, 01. December 2005
17:30kdebase 3.5484737ossi0i suppose this compiles with in the other config ...


TimeProjectRevisionAuthorRepliesMessageThursday, 10. November 2005
14:12kdebase479451ossi0handle io error on listing lilo boot options
14:11kdebase 3.5479448ossi0handle io error on listing lilo boot options


TimeProjectRevisionAuthorRepliesMessageSaturday, 05. November 2005
14:14kdebase477985ossi0merge from 3.5 branch. krootimage remains to be done.
12:58kdebase 3.5477963ossi0finally stop the dockbar from virtually freely permutating my wmnet instances for the various interfaces: compare the command line as well.
12:19kdebase477955ossi0HAVE_LIBART => HAVE_LIBAGG => no SVG, because no libagg check in kdebase. will change significatly with qt 4.1 anyway.


TimeProjectRevisionAuthorRepliesMessageFriday, 04. November 2005
19:23kdebase 3.5477713ossi0disable user list view during the failed login delay - avoids DoS by clicking a user during the delay and is The Right Thing To Do anyway.


TimeProjectRevisionAuthorRepliesMessageTuesday, 01. November 2005
16:35kdebase 3.5476633ossi0fix popup position with xinerama (or was it multihead?)


TimeProjectRevisionAuthorRepliesMessageThursday, 27. October 2005
18:04kdebase 3.5474938ossi0nitpicks


TimeProjectRevisionAuthorRepliesMessageSaturday, 15. October 2005
09:02kdebase 3.5470796ossi0make diplays via xdmcp work again - read config here, too.