IRC Logs

Network: freenode Channel: #reactos

	Enter your search terms Submit search form
	Web www.irclog.org

[00:00]<fyzn2rff>greetings tamlin!
[00:00]<vrdfyg>Fireball! :-) I hope your "used bandwidth" problem has been solved.
[00:01]<emurgmcj>Cool, shutting down makes a nice B/S-SOD ;-)
[00:01]<vrdfyg>The reson I dropped by was a comment on the m/l, where I just couldn't believe this kind of crap still pop up - the *A vs. *W for (in this case) ScmStartSerive).
[00:01]<emurgmcj>errmm. B/B-SOD
[00:01]<fyzn2rff>tamlin - not really :(( I always had like 10Gb reserve, but recent time I exhausted everything with WDK, MSDN and stuff
[00:02]<fyzn2rff>tamlin - yep, I'm a bit surprised by Eric's commit
[00:02]<gnmmuzlxd>me too. I thought we decided A always calls W
[00:02]<vrdfyg>Fireball: Is it just me, or is it not obvious that ALL internal representation is *W, and the *A functions just mangle to/from A/W for strings?
[00:03]<zydvz2jz>I read it diff, maybe it should be A but? I cant read Eric mind.
[00:03]<vrdfyg>(and calling the *W functions, obviously)
[00:04]<vrdfyg>I mean, ROS isn't some kind of Win9x toy - it's an NT reimplementation, and in NT *ALL* strings are W (for good and bad), at least AFAIK.
[00:05]<fyzn2rff>well, it sounds obvious for long time already. I just tend to think Eric was busy with something else while coding this. I can't find a real reason to do 2 implementations
[00:06]<zydvz2jz>maybe hes going to make another change,,, will see.
[00:07]<fyzn2rff>I hope so
[00:09]<vrdfyg>Fireball: Have you had a chance to look at hpoussin's CreateProcess problem? Shooting frmo the hip, my initial guess is that it's a recent APC fix (!) that displays errors in other areas. Could anyone confirm this?
[00:10]<fyzn2rff>tamlin - I couldn't reproduce his problem
[00:10]<vrdfyg>Fireball: OK, let's delay that until he can verisy with HEAD.
[00:10]<vrdfyg>*verify
[00:12]<bynnnx>borland to bring back the "turbo" line
[00:12]<bynnnx>http://www.eweek.com/article2/0,1895,2000205,00.asp
[00:12]<vrdfyg>Though, the "FIX PS SDs!!" part seems... interesting.
[00:13]<fyzn2rff>:-)
[00:13]<wzwdzzrjg>(*#^(*#^%(*%#RE
[00:13]<fyzn2rff>WaxDragon ?
[00:13]<emurgmcj>wtf
[00:14]<wzwdzzrjg>I'm getting irritated.
[00:16]<wzwdzzrjg>I cannot edit a freaking text file with 0.3.0
[00:16]<wzwdzzrjg>I try to save it and it just hangs.
[00:16]<zydvz2jz>ok
[00:16]<vrdfyg>Sounds like I/O, and APC's are indeed involved there.
[00:17]<wzwdzzrjg>notepad is burning cpu
[00:17]<fyzn2rff>that's ugly
[00:18]<x-r-f-n000>sounds familiar
[00:18]<wzwdzzrjg>at least I can kill processes
[00:18]<vrdfyg>Question is, is it usermode or kernelmode CPU, and if usermode is it due to APC callbacks?
[00:18]<vrdfyg>Wait a second!
[00:18]<vrdfyg>APC *callbacks*!
[00:18]<zydvz2jz>did alex find the open/create file bug and fixed it?
[00:18]<wzwdzzrjg>well, at least I got the testing done.
[00:19]<zydvz2jz>I get a ntos kernel crash open lgpl.txt
[00:19]<zydvz2jz>in 030
[00:19]<vrdfyg>win32k has always been a bit shabby when it comes to callback from kmode to umode. Could recent APC fixes have displayed bugs in this area too?
[00:20]<vrdfyg>jimtabor: First I hear about it, but that sounds like an Io bug? Perhaps is relates to this (or worse, gives a feedback loop)?
[00:22]<cia-7>03amunger * r23529 10rosapps/packmgr/lib/ (options.cpp options.xml): Update the hardcoded ViewCVS URL. Now packmgr works on the bootcd.
[00:22]<vrdfyg>Coming to think of it, as this has been a source of errors, is thread affinity and qanta even implemented yet?
[00:23]<wzwdzzrjg>x
[00:23]<wzwdzzrjg>oops
[00:23]<zydvz2jz>bugcheck in ntoskrnl/mm/rpoolmgr.h:558 (RPoolRedZoneCheck)
[00:23]<vrdfyg>(not affinity really, but the previously missing quanta has displayed some... interesting phenomenas)
[00:23]<wzwdzzrjg>Now we just need to add the wallpaper.
[00:23]<efzjgm>Hi tamlin!
[00:24]<vrdfyg>jim: Excellent! Seems me convincing Royce a redzone was for good has again found a bug! :-)
[00:24]<vrdfyg>Elrond!
[00:24]<zydvz2jz>Assertion NextBlock->PrevSize == Block->Size failed at ./ntoskrnl/mm/rpoolmgr.h
[00:24]<vrdfyg>WD: Yuo talk about wallpapers when the discussion around you are about kernel crashes. Oh, the irony. :-)
[00:25]<vrdfyg>jim: Absolutely a memory overwite I'd say!
[00:25]<vrdfyg>(or under-write, even if less common)
[00:25]<zydvz2jz>yes
[00:26]<vrdfyg>jim: Couldn't the DR* registers be used to track this? Set "break on write of memory address x"?
[00:27]<vrdfyg>... assuming you can repro the problem.
[00:27]<zydvz2jz>oh~ it worked this time
[00:27]<vrdfyg>That's even more disturbing, that it might be seemingly stochastic.
[00:27]<zydvz2jz>Wax I can save and open,,, its funny~
[00:28]<zydvz2jz_w>tamlin over here
[00:28]<zydvz2jz_w>this works now
[00:28]<lyfylq106>yay. hi master
[00:28]<zydvz2jz_w>bye
[00:29]<lyfylq106>stochastic? ...
[00:29]<zydvz2jz>well~ dont know~
[00:29]<vrdfyg>Good evening (night even) disciple and grasshopper. I hope life treats you well.
[00:29]<zydvz2jz>thanks tamlin!
[00:29]<vrdfyg>filip: read "random"
[00:30]<lyfylq106>i read hackles the comics ;)
[00:30]<wzwdzzrjg>tamlin: I'm talking about getting 0.3.0 out the door, something which no one else seems to want to discuss.
[00:30]<zydvz2jz>B^]
[00:31]<wzwdzzrjg>;0)
[00:31]<lyfylq106>yeah, life treated me very well in the last days. i went for a lunch on sunday and returned home on mondey night.
[00:31]<lyfylq106>how about you?
[00:31]<vrdfyg>WD: Are you serious, or was that just a statement of frustration?
[00:31]<wzwdzzrjg>I'm serious.
[00:32]<lyfylq106>http://www.hackles.org/cgi-bin/archives.pl?request=66
[00:32]<wzwdzzrjg>save jimtabor ;0)
[00:32]<wzwdzzrjg>jimtabor: it was the options.xml that packmgr downloads. Maybe it kept an open handle after I killed it.
[00:32]<wzwdzzrjg>if you were able to edit files, good enough.
[00:33]<zydvz2jz>I had a kernel bug check now its gone,,, I can edit files now.
[00:33]<wzwdzzrjg>I'm gonna label 0.3.0 "Longest release evar"
[00:33]<vrdfyg>filip: Sounds like she was agreeable. :-) Me? I'll live. However strange it may seems, life might improve noticably shortly even, so I don't have any reason not to complain.
[00:34]<zydvz2jz>I wish I knew how to fix this ---> KeBugCheckEx at ./ntoskrnl/ke/process.c:315
[00:34]<vrdfyg>WD: Please don't. Labelling it "evar" will forever paint ROS as "the dudes that can't spell".
[00:35]<vrdfyg>;-)
[00:35]<vrdfyg>jim: Callstack?
[00:35]<bynnnx>ReactOS 0.3.0 "... and its about f**king time."
[00:35]<zydvz2jz>oh ~yeah~
[00:36]<zydvz2jz>where did they find that? it was in the asm code right?
[00:38]<bynnnx>on vc2003, should i use /G7 (opt for p4) for an amd64 athlon x2?
[00:38]<prdqjjt>Bizzeh: yes
[00:39]<bynnnx>just trying to see what i can squeze out of ffdshow :)
[00:39]<bynnnx>see if i can break the linker again

Back to #reactos
Or go to some related logs:
#macosx
#gentoo
#ubuntu
#blender
#gentoo