T2 IRC Log: 2009-11-05
This is the log as captured by an IRC bot in the channel. The statements are those of the individual people and might not neccessarily reflect the policy and legal rules as set forth by the T2 SDE Project.
--- Log opened Thu Nov 05 00:00:17 2009
04:27 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: mjungwirth__, koan
04:28 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: @ChanServ, mtr, mckoan|away
04:28 -!- Netsplit over, joins: @ChanServ
04:28 -!- Netsplit over, joins: mckoan|away, mtr
04:29 -!- Netsplit over, joins: koan, mjungwirth__
08:10 -!- capey [n=mp@telakka.wapice.com] has joined #t2
08:22 < capey> hmm... should there be another qt package that also builds qtdesigner et al or should the current qt4 package to be changed to use the "all of qt" package? seems a bit overkill to change the current package since mostly only need the qt libs not the dev stuff
08:40 -!- mckoan|away is now known as mckoan
09:13 < rxr> I think changed, didn't the qt3 package installed it all, too ?
09:14 < capey> dont really remember how qt3 package worked
09:20 -!- Xor [n=oscar@11.Red-80-38-62.staticIP.rima-tde.net] has joined #t2
09:20 < Xor> hi all
09:20 -!- Xor is now known as Guest77211
09:21 -!- Guest77211 is now known as Xor_
09:24 < capey> hi Xor
09:52 < rxr> i all
09:52 < rxr> hi all, even - moin moin
10:05 < CIA-28> rene * r34805 /trunk/package/database/sqlite/sqlite.desc: * updated sqlite (3.6.19 -> 3.6.20)
10:09 < rxr> http://code.google.com/p/stressapptest/
10:35 < CIA-28> rene * r34806 /trunk/package/base/stressapptest/ (. stressapptest.cache stressapptest.desc): * added (Google's) stressapptest (1.0.1)
11:12 < rxr> hm, interesting:
11:12 < rxr> selected item doesn't fit into memory.
11:12 < rxr> ^- while loading memtest86 on the nokia booklet, this helped:
11:12 < rxr> http://bugs.centos.org/view.php?id=3104
11:16 < rxr> http://www.t2-project.org/packages/stressapptest.html
11:20 -!- Baldzius [n=aldas@87-198-133-94.ptr.magnet.ie] has joined #t2
11:20 < Baldzius> moin
11:20 < rxr> moin moin Baldzius
11:20 < Baldzius> hey rxr
11:24 < CIA-28> rene * r34807 /trunk/package/develop/ginac/ginac.desc: * updated ginac (1.5.3 -> 1.5.5)
11:24 < CIA-28> rene * r34808 /trunk/package/multimedia/libnice/libnice.desc: * updated libnice (0.0.9 -> 0.0.10)
11:24 < CIA-28> rene * r34809 /trunk/package/network/dante/dante.desc: * updated dante (1.1.19 -> 1.2.0)
11:24 < CIA-28> rene * r34811 /trunk/package/network/duplicity/duplicity.desc: * updated duplicity (0.6.05 -> 0.6.06)
11:24 < CIA-28> rene * r34810 /trunk/package/security/strongswan/strongswan.desc: * updated strongswan (4.3.4 -> 4.3.5)
11:32 < rxr> new linux kernels are getting more and more unstable
11:32 < rxr> unbelievable
11:32 < rxr> since 2.6.28 resume (from sleep) is broken on my PPC G4 Cube (worked for over 5 years!!!)
11:33 < rxr> then 2.6.31 has typos in the PPP code, oopsed our gateway on DSL (PPPOE) load quite some times until I updated to 2.6.32
11:33 < rxr> (make the above 2.6.30 vs 2.,6.31)
11:33 < rxr> then, 2.6.31 has some more typos in the RFKill stuff, oopses on many machines, including the Nokia Booklet, already fixed in git
11:33 < rxr> then, if this is all not enough, I get random memory corruption on some machines
11:34 < rxr> (guess why I just added stressapptest, ...
11:34 < rxr> and now run memtest86 (which did not load on the machines, due to the posted years old loader issue, damnit, now modded, loaded as netbsd kernel, heck)
11:34 < rxr> memtest86 does not show memory corruption on the machines
11:35 < rxr> guess I can not test old kernels, and where that broke
11:35 < rxr> or finally move on to some micro kernel thing, because
11:35 < rxr> frankly this instability gets on my nerves
11:35 < rxr> neither some rfkill fluffery, nor pppoe or other network rounting code should memory clobber a modern system, IN ANY CASE
11:36 < rxr> I want' any f*cking low speed, low priority code run as unpriveledged process, crash like it wants, but not the damn base schedule and memory manager (of my envisioned micro kernel)
11:36 * rxr headshake
11:36 < capey> :)
11:36 < rxr> should just boot windows 7 on the nokia booklet, try to use that for a week, month
11:37 < rxr> too bad OSs don't sell very well, otherwise I might start my own, sell it, ...
11:37 < capey> havent tryed windows 7 yet, but guess i'm force to try it some time
11:38 < rxr> well, I even purchased a copy, because the windows 7 started edition that came with my Booklet had no background wallpaper change option, no Aero, ...
11:38 < rxr> so now I own a genuine "Starter edition" (that sucks) and some "Home Premium" version :-)
11:39 < Baldzius> how fast Win7 Starter was booting on your booklet?
11:39 < rxr> the later allows to change the wallpaper, comes with Aero ...
11:39 < rxr> faster than OS X :-)
11:39 < capey> :)
11:40 < rxr> I can meassure the regular windows 7 when the memtest run some more passes
11:40 < Baldzius> there are rumours going around that it should boot in 10sec
11:40 < capey> i guess the starter edition is pretty useless, well maybe on low power stuff
11:40 < rxr> the start edition only lives on a recovery backup in the shelf
11:40 < rxr> I think the starter and regular did not make a big difference
11:40 < rxr> maybe none at all
11:40 < Baldzius> heh
11:40 < rxr> it's just a normal windows with the background changer, areo #ifdef'ed out, ..
11:41 < rxr> I doubt it booted in 10s to full operation on this 1.6GHz Atom Z
11:41 < rxr> has a slow 1.8" HD anyway, maybe with SSD, some slightly faster processor
11:41 < rxr> but windows 7 boots faster than former versions of windows, in any case
11:42 < Baldzius> this might not be the truth
11:42 < rxr> no?
11:42 < Baldzius> boot time is when you have fully functioning system
11:42 < rxr> well, after all I avoided windows the last decade, only saw it sporadically :_)
11:42 < Baldzius> not your desktop background
11:43 < rxr> yeah, that's why I wrote above: " I doubt it booted in 10s to full operation ..."
11:44 < Baldzius> win7 in *real* test is slower than Vista
11:44 < Baldzius> looking for youtube video which shows that
11:45 < rxr> yeah - I also keep telling people that win7 is a 'sed s/Vista/7/ PR thing, and that performance is about equal :-)~
11:45 < rxr> well, one youtube video is one thing, some massive test is more of a proof
11:45 < rxr> however, I read in some PC centric sites, that the Win 7 Beta's felt faster than the final gold master
11:46 < rxr> they speculated the beta and RC builds had some checking, DRM stuff #ifdef'ed out to receive better PR for the lunch
11:46 < capey> well yeah, it's basically vista, with some minor changes in system, not asking admin privleges on every action and so forth
11:46 < rxr> and the final master of 7 is more like Vista was
11:47 < rxr> but I'm not a Windows expert, dd not run tests, did not dissect :-)
11:47 < capey> sadly i used it daily
11:47 < capey> used=use
11:47 < rxr> yeah - that's life
11:47 < rxr> will probably also use it more in the next decade than in the last
11:48 < rxr> switch between Linux, OS X, WIndows
11:48 < rxr> after all stability wise it'snow better than 95, 3.11
11:48 < rxr> :-)
11:49 < capey> heh yeah :)
11:53 < Baldzius> http://www.youtube.com/watch?v=88aal60AqBs&feature=player_embedded
11:53 < Baldzius> it's a real test with FF starting up automatically and loading the page
11:54 < Baldzius> win 7 loses 5 sec to wista
11:56 < rxr> :-)
11:58 < capey> well those speeds really depend on hardware and installed system...
11:59 < capey> software,etc... but overall i agree on the result
11:59 < rxr> maybe they used the identical hardware, fresh, clean, install for the test?
11:59 < capey> maybe
11:59 < capey> i hope so
11:59 * rxr under T2 x86_64 without flash at the moment, so no video :-)
12:00 < capey> :)
12:06 < rxr> yeah - gnash never really worked for me, yet
12:08 < Baldzius> http://labs.adobe.com/downloads/flashplayer10.html
12:08 < rxr> was that 64bit?
12:08 < Baldzius> it is
12:08 < rxr> oh really, now, ?
12:08 < Baldzius> http://download.macromedia.com/pub/labs/flashplayer10/libflashplayer-10.0.32.18.linux-x86_64.so.tar.gz
12:08 < Baldzius> not sure how stable it is
12:08 < rxr> :-)
12:10 < rxr> ok - lunch - cu
12:51 < rxr> http://t2-project.org/packages/stressapptest.html
13:05 < rxr> Hardware Error: miscompare on CPU 1(0x2) at 0x5bd98480(0x0:DIMM Unknown): read:0x0000000004000000, reread:0x0000000004000000 expected:0x0400000004000000
13:06 < rxr> Hardware Error: miscompare on CPU 1(0x2) at 0x49363240(0x0:DIMM Unknown): read:0x0000000004000000, reread:0x0000000004000000 expected:0x0400000004000000
13:07 < rxr> this google stressapptest thing
13:07 < rxr> I accidently might have found a culprit: acpi-cpufreq
13:10 < rxr> now without:
13:10 < rxr> Stats: Starting SAT, 848M, 300 seconds
13:10 < rxr> Log: Seconds remaining: 290
13:10 < rxr> Log: Seconds remaining: 280
13:10 < rxr> Log: Seconds remaining: 270
13:10 < rxr> Log: Seconds remaining: 260
13:10 < rxr> Log: Seconds remaining: 250
13:10 < rxr> Log: Seconds remaining: 240
13:10 < rxr> Log: Seconds remaining: 230
13:10 < rxr> Log: Seconds remaining: 220
13:10 < rxr> Log: Seconds remaining: 210
13:10 < rxr> Log: Seconds remaining: 200
13:10 < rxr> Log: Seconds remaining: 190
13:10 < rxr> Log: Seconds remaining: 180
13:10 < rxr> Log: Seconds remaining: 170
13:10 < rxr> no error, yet
13:13 < rxr> Stats: Completed: 166746.00M in 300.03s 555.76MB/s, with 0 hardware incidents, 0 errors
13:13 < rxr> Stats: Memory Copy: 166746.00M at 555.80MB/s
13:15 < rxr> but suspend resume still corrupts
13:17 -!- mqueiros_ [n=mqueiros@217.70.75.212] has quit [Client Quit]
13:24 < rxr> strange
13:24 -!- mqueiros [n=mqueiros@217.70.75.212] has joined #t2
13:24 < rxr> does not help reproducable
13:24 < rxr> fresh reboot, without the acpi module and issues again
13:29 < rxr> acpi=off appears to help more reliable :-(
13:30 < rxr> no, doesn't Hardware Error: miscompare on CPU 0(0x1) at 0x621dc4c0(0x0:DIMM Unknown): read:0xfffffffffff7ffff, reread:0xfffffffffff7ffff expected:0xfff7fffffff7ffff
14:01 -!- capey [n=mp@telakka.wapice.com] has quit ["leaving"]
15:53 < rxr> another server looks healthier:
15:53 < rxr> Stats: Found 0 hardware incidents
15:53 < rxr> Stats: Completed: 432122.00M in 120.02s 3600.36MB/s, with 0 hardware incidents, 0 errors
15:53 < rxr> Stats: Memory Copy: 432122.00M at 3601.01MB/s
15:53 < rxr> Stats: File Copy: 0.00M at 0.00MB/s
16:15 < koan> what's the problem rxr?
16:15 < rxr> hi :-)
16:31 -!- aloof [n=kjj@pool-74-107-128-126.ptldor.fios.verizon.net] has joined #t2
16:31 -!- aloof [n=kjj@pool-74-107-128-126.ptldor.fios.verizon.net] has quit [Client Quit]
16:31 -!- aloof [n=kjj@pool-74-107-128-126.ptldor.fios.verizon.net] has joined #t2
16:33 -!- Baldzius [n=aldas@87-198-133-94.ptr.magnet.ie] has quit ["Leaving"]
17:09 < rxr> koan: problem is linux stability on the Nokia Booklet 3G :-)
17:10 < rxr> (beside other regressions in latest linux kernels)
17:15 < rxr> I even run the windows 7 install dvd memory test since some time, now ...
17:20 < CIA-28> rene * r34812 /trunk/package/network/liferea/liferea.desc: * updated liferea (1.4.28 -> 1.6.0)
17:42 -!- koan [n=koan@unaffiliated/koan] has quit ["Lost terminal"]
17:42 -!- koan [n=koan@unaffiliated/koan] has joined #t2
18:00 < rxr> wow, I'm impressed
18:01 < rxr> not only does DragonflyBSD boot on the Nokia Booklet (and some other hardware here around me in the office), it even is able to suspend, resume (ACPI S3)
18:01 < rxr> this motivates me to port this google memory check thing, to BSD
18:01 < rxr> well, at least try, see if dragonfly does show simillar memory corruption as Linux does on the Nokia Booklet
18:08 -!- mckoan is now known as mckoan|away
18:58 < koan> nice
19:06 < rxr> zeah
19:06 < rxr> yeah
19:07 < rxr> just setup a dragonfly virtual machine on our main server
19:07 < rxr> this thing needs some more evaluation (was on the TOP of my TODO for some time, now ;-)
19:41 < rxr> ok - gotta go home, cu
19:41 < rxr> http://rene.rebe.name/2009-11-05/firts-steps-with-an-bsd-dragonfly/
20:04 < mqueiros> rxr: had some fun some years ago, playing with OpenBSD. Never tried DragonFly or even NetBSD or FreeBSD...
20:11 -!- mckoan|away is now known as mckoan
21:14 < rxr> mqueiros: :-)
21:14 * rxr never neither
21:23 -!- xperia2 [n=chatzill@zux182-249.adsl.green.ch] has joined #t2
21:24 < xperia2> hello to all. have a question abaout using t2 for building a own arm based distro
21:24 < xperia2> would like to have a automated script or builder that fetch all 24 hours the sources and build the changed packages
21:25 < xperia2> does anybody know if something like this is possible ?
21:28 < koan> couldn't you just run a cron job that does a svn up and then start a build?
21:28 < rxr> sure
21:28 < rxr> svn up and then:
21:29 < rxr> scripts/Create-ErrList -cfg arm-dist -newremove
21:29 < rxr> or -newdelete
21:30 < rxr> one just wipes the logs and thus re-schedules the new package, the other uninstalls them before (delets all files of the package)
21:30 < rxr> and then the usuaul scripts/Build-Target -cfg arm-dist # etc.
21:32 < xperia2> hmmm i have thinked more in the direction of tinderbox http://www-archive.mozilla.org/projects/tinderbox/
21:33 < xperia2> a very good example http://tinderbox.openembedded.net/search/
21:33 < rxr> xperia2: whatever, koan mean the "how to invoke", I the technical T2 "re-schedule changed" hint
21:33 < xperia2> okay
21:34 < rxr> damnit, the fun of BSD vs. GNU/Linux strerorr_r, one returns an int/status, the other the string pointer, fun
21:35 < rxr> The XSI-compliant strerror_r() is preferred for portable applications.
21:35 < rxr> It returns the error string in the user-supplied buffer buf of length
21:35 < rxr> buflen.
21:35 < rxr> The XSI-compliant strerror_r() function returns 0 on success; on error,
21:35 < rxr> -1 is returned and errno is set to indicate the error.
21:36 < xperia2> well okay in this case it would be not very hard to svn up the sources and rebuild the distro. i understand. sorry for this question but i am totally new to t2.
21:37 < xperia2> exist somewhere a crashcourse for creating a small gui less distro for the start.
21:37 < xperia2> with t2
21:37 < rxr> sure, welcome!
21:38 < rxr> you can find the some guides as "quick starts" on the t2 side, as well as a "partical intro" in the article "zine" section and then the full detail handbook
21:39 < rxr> ok - cu again
21:40 < rxr> good n8
21:40 < xperia2> good night rxr. and thanks !
21:42 < rxr> I
21:42 < rxr> 'm back in some 8h+
21:43 < xperia2> okay we see us sure tomorow. only one quick question
21:43 < xperia2> i am planning to do the compile stuff on a ubuntu server
21:43 < xperia2> would this makes problems or do i need a t2os for doing all the build proccess
21:44 < xperia2> just studing atm http://www.t2-project.org/documentation/buildintro.html
21:50 < rxr> it should work to build on other linux', e.g. ubuntu, though sometimes tiny issues occur
21:51 < rxr> e.g. on ubuntu it might be neccessary to change the system shell to /bin/bash (from the default dash)
21:51 < rxr> that is the only ubuntu standard issue I can think of right now
21:51 < rxr> n8
22:29 < CIA-28> aldas * r34813 /trunk/package/kde/k3b/k3b.desc: * updated k3b (1.0.95-r1015056 -> 1.68.0alpha3)
23:12 -!- mckoan is now known as mckoan|away
23:49 -!- cb88 [n=chase@209.charlotte-04rh15-16rt.nc.dial-access.att.net] has joined #t2
23:52 -!- cb88 [n=chase@209.charlotte-04rh15-16rt.nc.dial-access.att.net] has quit [Read error: 54 (Connection reset by peer)]
23:52 -!- cb88 [n=chase@209.charlotte-04rh15-16rt.nc.dial-access.att.net] has joined #t2
23:54 < cb88> I have mailed lists@t2-project.org with the subject subscribe t2 and haven't received any response... minimalist@t2-project.org also isn't responding except for an email I received due to a spelling error in my 'suscribe t2' line
23:55 < cb88> is it possible that yahoo mail is attaching something extra to my email causing it to fail or somthing?
23:55 -!- cb88 [n=chase@209.charlotte-04rh15-16rt.nc.dial-access.att.net] has quit [Read error: 131 (Connection reset by peer)]
23:56 -!- cb88 [n=chase@209.charlotte-04rh15-16rt.nc.dial-access.att.net] has joined #t2
--- Log closed Fri Nov 06 00:00:17 2009
04:27 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: mjungwirth__, koan
04:28 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: @ChanServ, mtr, mckoan|away
04:28 -!- Netsplit over, joins: @ChanServ
04:28 -!- Netsplit over, joins: mckoan|away, mtr
04:29 -!- Netsplit over, joins: koan, mjungwirth__
08:10 -!- capey [n=mp@telakka.wapice.com] has joined #t2
08:22 < capey> hmm... should there be another qt package that also builds qtdesigner et al or should the current qt4 package to be changed to use the "all of qt" package? seems a bit overkill to change the current package since mostly only need the qt libs not the dev stuff
08:40 -!- mckoan|away is now known as mckoan
09:13 < rxr> I think changed, didn't the qt3 package installed it all, too ?
09:14 < capey> dont really remember how qt3 package worked
09:20 -!- Xor [n=oscar@11.Red-80-38-62.staticIP.rima-tde.net] has joined #t2
09:20 < Xor> hi all
09:20 -!- Xor is now known as Guest77211
09:21 -!- Guest77211 is now known as Xor_
09:24 < capey> hi Xor
09:52 < rxr> i all
09:52 < rxr> hi all, even - moin moin
10:05 < CIA-28> rene * r34805 /trunk/package/database/sqlite/sqlite.desc: * updated sqlite (3.6.19 -> 3.6.20)
10:09 < rxr> http://code.google.com/p/stressapptest/
10:35 < CIA-28> rene * r34806 /trunk/package/base/stressapptest/ (. stressapptest.cache stressapptest.desc): * added (Google's) stressapptest (1.0.1)
11:12 < rxr> hm, interesting:
11:12 < rxr> selected item doesn't fit into memory.
11:12 < rxr> ^- while loading memtest86 on the nokia booklet, this helped:
11:12 < rxr> http://bugs.centos.org/view.php?id=3104
11:16 < rxr> http://www.t2-project.org/packages/stressapptest.html
11:20 -!- Baldzius [n=aldas@87-198-133-94.ptr.magnet.ie] has joined #t2
11:20 < Baldzius> moin
11:20 < rxr> moin moin Baldzius
11:20 < Baldzius> hey rxr
11:24 < CIA-28> rene * r34807 /trunk/package/develop/ginac/ginac.desc: * updated ginac (1.5.3 -> 1.5.5)
11:24 < CIA-28> rene * r34808 /trunk/package/multimedia/libnice/libnice.desc: * updated libnice (0.0.9 -> 0.0.10)
11:24 < CIA-28> rene * r34809 /trunk/package/network/dante/dante.desc: * updated dante (1.1.19 -> 1.2.0)
11:24 < CIA-28> rene * r34811 /trunk/package/network/duplicity/duplicity.desc: * updated duplicity (0.6.05 -> 0.6.06)
11:24 < CIA-28> rene * r34810 /trunk/package/security/strongswan/strongswan.desc: * updated strongswan (4.3.4 -> 4.3.5)
11:32 < rxr> new linux kernels are getting more and more unstable
11:32 < rxr> unbelievable
11:32 < rxr> since 2.6.28 resume (from sleep) is broken on my PPC G4 Cube (worked for over 5 years!!!)
11:33 < rxr> then 2.6.31 has typos in the PPP code, oopsed our gateway on DSL (PPPOE) load quite some times until I updated to 2.6.32
11:33 < rxr> (make the above 2.6.30 vs 2.,6.31)
11:33 < rxr> then, 2.6.31 has some more typos in the RFKill stuff, oopses on many machines, including the Nokia Booklet, already fixed in git
11:33 < rxr> then, if this is all not enough, I get random memory corruption on some machines
11:34 < rxr> (guess why I just added stressapptest, ...
11:34 < rxr> and now run memtest86 (which did not load on the machines, due to the posted years old loader issue, damnit, now modded, loaded as netbsd kernel, heck)
11:34 < rxr> memtest86 does not show memory corruption on the machines
11:35 < rxr> guess I can not test old kernels, and where that broke
11:35 < rxr> or finally move on to some micro kernel thing, because
11:35 < rxr> frankly this instability gets on my nerves
11:35 < rxr> neither some rfkill fluffery, nor pppoe or other network rounting code should memory clobber a modern system, IN ANY CASE
11:36 < rxr> I want' any f*cking low speed, low priority code run as unpriveledged process, crash like it wants, but not the damn base schedule and memory manager (of my envisioned micro kernel)
11:36 * rxr headshake
11:36 < capey> :)
11:36 < rxr> should just boot windows 7 on the nokia booklet, try to use that for a week, month
11:37 < rxr> too bad OSs don't sell very well, otherwise I might start my own, sell it, ...
11:37 < capey> havent tryed windows 7 yet, but guess i'm force to try it some time
11:38 < rxr> well, I even purchased a copy, because the windows 7 started edition that came with my Booklet had no background wallpaper change option, no Aero, ...
11:38 < rxr> so now I own a genuine "Starter edition" (that sucks) and some "Home Premium" version :-)
11:39 < Baldzius> how fast Win7 Starter was booting on your booklet?
11:39 < rxr> the later allows to change the wallpaper, comes with Aero ...
11:39 < rxr> faster than OS X :-)
11:39 < capey> :)
11:40 < rxr> I can meassure the regular windows 7 when the memtest run some more passes
11:40 < Baldzius> there are rumours going around that it should boot in 10sec
11:40 < capey> i guess the starter edition is pretty useless, well maybe on low power stuff
11:40 < rxr> the start edition only lives on a recovery backup in the shelf
11:40 < rxr> I think the starter and regular did not make a big difference
11:40 < rxr> maybe none at all
11:40 < Baldzius> heh
11:40 < rxr> it's just a normal windows with the background changer, areo #ifdef'ed out, ..
11:41 < rxr> I doubt it booted in 10s to full operation on this 1.6GHz Atom Z
11:41 < rxr> has a slow 1.8" HD anyway, maybe with SSD, some slightly faster processor
11:41 < rxr> but windows 7 boots faster than former versions of windows, in any case
11:42 < Baldzius> this might not be the truth
11:42 < rxr> no?
11:42 < Baldzius> boot time is when you have fully functioning system
11:42 < rxr> well, after all I avoided windows the last decade, only saw it sporadically :_)
11:42 < Baldzius> not your desktop background
11:43 < rxr> yeah, that's why I wrote above: " I doubt it booted in 10s to full operation ..."
11:44 < Baldzius> win7 in *real* test is slower than Vista
11:44 < Baldzius> looking for youtube video which shows that
11:45 < rxr> yeah - I also keep telling people that win7 is a 'sed s/Vista/7/ PR thing, and that performance is about equal :-)~
11:45 < rxr> well, one youtube video is one thing, some massive test is more of a proof
11:45 < rxr> however, I read in some PC centric sites, that the Win 7 Beta's felt faster than the final gold master
11:46 < rxr> they speculated the beta and RC builds had some checking, DRM stuff #ifdef'ed out to receive better PR for the lunch
11:46 < capey> well yeah, it's basically vista, with some minor changes in system, not asking admin privleges on every action and so forth
11:46 < rxr> and the final master of 7 is more like Vista was
11:47 < rxr> but I'm not a Windows expert, dd not run tests, did not dissect :-)
11:47 < capey> sadly i used it daily
11:47 < capey> used=use
11:47 < rxr> yeah - that's life
11:47 < rxr> will probably also use it more in the next decade than in the last
11:48 < rxr> switch between Linux, OS X, WIndows
11:48 < rxr> after all stability wise it'snow better than 95, 3.11
11:48 < rxr> :-)
11:49 < capey> heh yeah :)
11:53 < Baldzius> http://www.youtube.com/watch?v=88aal60AqBs&feature=player_embedded
11:53 < Baldzius> it's a real test with FF starting up automatically and loading the page
11:54 < Baldzius> win 7 loses 5 sec to wista
11:56 < rxr> :-)
11:58 < capey> well those speeds really depend on hardware and installed system...
11:59 < capey> software,etc... but overall i agree on the result
11:59 < rxr> maybe they used the identical hardware, fresh, clean, install for the test?
11:59 < capey> maybe
11:59 < capey> i hope so
11:59 * rxr under T2 x86_64 without flash at the moment, so no video :-)
12:00 < capey> :)
12:06 < rxr> yeah - gnash never really worked for me, yet
12:08 < Baldzius> http://labs.adobe.com/downloads/flashplayer10.html
12:08 < rxr> was that 64bit?
12:08 < Baldzius> it is
12:08 < rxr> oh really, now, ?
12:08 < Baldzius> http://download.macromedia.com/pub/labs/flashplayer10/libflashplayer-10.0.32.18.linux-x86_64.so.tar.gz
12:08 < Baldzius> not sure how stable it is
12:08 < rxr> :-)
12:10 < rxr> ok - lunch - cu
12:51 < rxr> http://t2-project.org/packages/stressapptest.html
13:05 < rxr> Hardware Error: miscompare on CPU 1(0x2) at 0x5bd98480(0x0:DIMM Unknown): read:0x0000000004000000, reread:0x0000000004000000 expected:0x0400000004000000
13:06 < rxr> Hardware Error: miscompare on CPU 1(0x2) at 0x49363240(0x0:DIMM Unknown): read:0x0000000004000000, reread:0x0000000004000000 expected:0x0400000004000000
13:07 < rxr> this google stressapptest thing
13:07 < rxr> I accidently might have found a culprit: acpi-cpufreq
13:10 < rxr> now without:
13:10 < rxr> Stats: Starting SAT, 848M, 300 seconds
13:10 < rxr> Log: Seconds remaining: 290
13:10 < rxr> Log: Seconds remaining: 280
13:10 < rxr> Log: Seconds remaining: 270
13:10 < rxr> Log: Seconds remaining: 260
13:10 < rxr> Log: Seconds remaining: 250
13:10 < rxr> Log: Seconds remaining: 240
13:10 < rxr> Log: Seconds remaining: 230
13:10 < rxr> Log: Seconds remaining: 220
13:10 < rxr> Log: Seconds remaining: 210
13:10 < rxr> Log: Seconds remaining: 200
13:10 < rxr> Log: Seconds remaining: 190
13:10 < rxr> Log: Seconds remaining: 180
13:10 < rxr> Log: Seconds remaining: 170
13:10 < rxr> no error, yet
13:13 < rxr> Stats: Completed: 166746.00M in 300.03s 555.76MB/s, with 0 hardware incidents, 0 errors
13:13 < rxr> Stats: Memory Copy: 166746.00M at 555.80MB/s
13:15 < rxr> but suspend resume still corrupts
13:17 -!- mqueiros_ [n=mqueiros@217.70.75.212] has quit [Client Quit]
13:24 < rxr> strange
13:24 -!- mqueiros [n=mqueiros@217.70.75.212] has joined #t2
13:24 < rxr> does not help reproducable
13:24 < rxr> fresh reboot, without the acpi module and issues again
13:29 < rxr> acpi=off appears to help more reliable :-(
13:30 < rxr> no, doesn't Hardware Error: miscompare on CPU 0(0x1) at 0x621dc4c0(0x0:DIMM Unknown): read:0xfffffffffff7ffff, reread:0xfffffffffff7ffff expected:0xfff7fffffff7ffff
14:01 -!- capey [n=mp@telakka.wapice.com] has quit ["leaving"]
15:53 < rxr> another server looks healthier:
15:53 < rxr> Stats: Found 0 hardware incidents
15:53 < rxr> Stats: Completed: 432122.00M in 120.02s 3600.36MB/s, with 0 hardware incidents, 0 errors
15:53 < rxr> Stats: Memory Copy: 432122.00M at 3601.01MB/s
15:53 < rxr> Stats: File Copy: 0.00M at 0.00MB/s
16:15 < koan> what's the problem rxr?
16:15 < rxr> hi :-)
16:31 -!- aloof [n=kjj@pool-74-107-128-126.ptldor.fios.verizon.net] has joined #t2
16:31 -!- aloof [n=kjj@pool-74-107-128-126.ptldor.fios.verizon.net] has quit [Client Quit]
16:31 -!- aloof [n=kjj@pool-74-107-128-126.ptldor.fios.verizon.net] has joined #t2
16:33 -!- Baldzius [n=aldas@87-198-133-94.ptr.magnet.ie] has quit ["Leaving"]
17:09 < rxr> koan: problem is linux stability on the Nokia Booklet 3G :-)
17:10 < rxr> (beside other regressions in latest linux kernels)
17:15 < rxr> I even run the windows 7 install dvd memory test since some time, now ...
17:20 < CIA-28> rene * r34812 /trunk/package/network/liferea/liferea.desc: * updated liferea (1.4.28 -> 1.6.0)
17:42 -!- koan [n=koan@unaffiliated/koan] has quit ["Lost terminal"]
17:42 -!- koan [n=koan@unaffiliated/koan] has joined #t2
18:00 < rxr> wow, I'm impressed
18:01 < rxr> not only does DragonflyBSD boot on the Nokia Booklet (and some other hardware here around me in the office), it even is able to suspend, resume (ACPI S3)
18:01 < rxr> this motivates me to port this google memory check thing, to BSD
18:01 < rxr> well, at least try, see if dragonfly does show simillar memory corruption as Linux does on the Nokia Booklet
18:08 -!- mckoan is now known as mckoan|away
18:58 < koan> nice
19:06 < rxr> zeah
19:06 < rxr> yeah
19:07 < rxr> just setup a dragonfly virtual machine on our main server
19:07 < rxr> this thing needs some more evaluation (was on the TOP of my TODO for some time, now ;-)
19:41 < rxr> ok - gotta go home, cu
19:41 < rxr> http://rene.rebe.name/2009-11-05/firts-steps-with-an-bsd-dragonfly/
20:04 < mqueiros> rxr: had some fun some years ago, playing with OpenBSD. Never tried DragonFly or even NetBSD or FreeBSD...
20:11 -!- mckoan|away is now known as mckoan
21:14 < rxr> mqueiros: :-)
21:14 * rxr never neither
21:23 -!- xperia2 [n=chatzill@zux182-249.adsl.green.ch] has joined #t2
21:24 < xperia2> hello to all. have a question abaout using t2 for building a own arm based distro
21:24 < xperia2> would like to have a automated script or builder that fetch all 24 hours the sources and build the changed packages
21:25 < xperia2> does anybody know if something like this is possible ?
21:28 < koan> couldn't you just run a cron job that does a svn up and then start a build?
21:28 < rxr> sure
21:28 < rxr> svn up and then:
21:29 < rxr> scripts/Create-ErrList -cfg arm-dist -newremove
21:29 < rxr> or -newdelete
21:30 < rxr> one just wipes the logs and thus re-schedules the new package, the other uninstalls them before (delets all files of the package)
21:30 < rxr> and then the usuaul scripts/Build-Target -cfg arm-dist # etc.
21:32 < xperia2> hmmm i have thinked more in the direction of tinderbox http://www-archive.mozilla.org/projects/tinderbox/
21:33 < xperia2> a very good example http://tinderbox.openembedded.net/search/
21:33 < rxr> xperia2: whatever, koan mean the "how to invoke", I the technical T2 "re-schedule changed" hint
21:33 < xperia2> okay
21:34 < rxr> damnit, the fun of BSD vs. GNU/Linux strerorr_r, one returns an int/status, the other the string pointer, fun
21:35 < rxr> The XSI-compliant strerror_r() is preferred for portable applications.
21:35 < rxr> It returns the error string in the user-supplied buffer buf of length
21:35 < rxr> buflen.
21:35 < rxr> The XSI-compliant strerror_r() function returns 0 on success; on error,
21:35 < rxr> -1 is returned and errno is set to indicate the error.
21:36 < xperia2> well okay in this case it would be not very hard to svn up the sources and rebuild the distro. i understand. sorry for this question but i am totally new to t2.
21:37 < xperia2> exist somewhere a crashcourse for creating a small gui less distro for the start.
21:37 < xperia2> with t2
21:37 < rxr> sure, welcome!
21:38 < rxr> you can find the some guides as "quick starts" on the t2 side, as well as a "partical intro" in the article "zine" section and then the full detail handbook
21:39 < rxr> ok - cu again
21:40 < rxr> good n8
21:40 < xperia2> good night rxr. and thanks !
21:42 < rxr> I
21:42 < rxr> 'm back in some 8h+
21:43 < xperia2> okay we see us sure tomorow. only one quick question
21:43 < xperia2> i am planning to do the compile stuff on a ubuntu server
21:43 < xperia2> would this makes problems or do i need a t2os for doing all the build proccess
21:44 < xperia2> just studing atm http://www.t2-project.org/documentation/buildintro.html
21:50 < rxr> it should work to build on other linux', e.g. ubuntu, though sometimes tiny issues occur
21:51 < rxr> e.g. on ubuntu it might be neccessary to change the system shell to /bin/bash (from the default dash)
21:51 < rxr> that is the only ubuntu standard issue I can think of right now
21:51 < rxr> n8
22:29 < CIA-28> aldas * r34813 /trunk/package/kde/k3b/k3b.desc: * updated k3b (1.0.95-r1015056 -> 1.68.0alpha3)
23:12 -!- mckoan is now known as mckoan|away
23:49 -!- cb88 [n=chase@209.charlotte-04rh15-16rt.nc.dial-access.att.net] has joined #t2
23:52 -!- cb88 [n=chase@209.charlotte-04rh15-16rt.nc.dial-access.att.net] has quit [Read error: 54 (Connection reset by peer)]
23:52 -!- cb88 [n=chase@209.charlotte-04rh15-16rt.nc.dial-access.att.net] has joined #t2
23:54 < cb88> I have mailed lists@t2-project.org with the subject subscribe t2 and haven't received any response... minimalist@t2-project.org also isn't responding except for an email I received due to a spelling error in my 'suscribe t2' line
23:55 < cb88> is it possible that yahoo mail is attaching something extra to my email causing it to fail or somthing?
23:55 -!- cb88 [n=chase@209.charlotte-04rh15-16rt.nc.dial-access.att.net] has quit [Read error: 131 (Connection reset by peer)]
23:56 -!- cb88 [n=chase@209.charlotte-04rh15-16rt.nc.dial-access.att.net] has joined #t2
--- Log closed Fri Nov 06 00:00:17 2009