To use your own IRC client, join channel #boycottnovell in FreeNode.
_Hicham_ | but the choice of WTL is what is making things hard to port to Linux | Mar 06 11:06 |
---|---|---|
_Hicham_ | the first problem of WTL is that it is undocumented | Mar 06 11:06 |
oiaohm | You are building something as a testing ground. | Mar 06 11:07 |
_Hicham_ | it is a Microsoft's Open Source Library | Mar 06 11:07 |
oiaohm | You don't think that is going to last long term. | Mar 06 11:07 |
_Hicham_ | it is | Mar 06 11:07 |
oiaohm | Yep you use what ever your programmers in your group know. | Mar 06 11:07 |
oiaohm | Now where it decides to last long term you have major head aches. | Mar 06 11:07 |
_Hicham_ | no one knows WTL | Mar 06 11:08 |
_Hicham_ | that is the question | Mar 06 11:08 |
_Hicham_ | they must have worked with the guy who wrote WTL | Mar 06 11:08 |
oiaohm | Some university gradueds do. | Mar 06 11:08 |
_Hicham_ | not very much | Mar 06 11:08 |
_Hicham_ | since it is undocumented | Mar 06 11:08 |
oiaohm | http://wtl.sourceforge.net/ << What do you mean undocumented? | Mar 06 11:09 |
_Hicham_ | yes | Mar 06 11:09 |
_Hicham_ | look for the doc there | Mar 06 11:09 |
_Hicham_ | you wont find anything | Mar 06 11:09 |
oiaohm | Guess what some Uni do a subject on. | Mar 06 11:10 |
_Hicham_ | possible | Mar 06 11:10 |
oiaohm | Documenting code | Mar 06 11:10 |
_Hicham_ | there have been some documentation projects for it | Mar 06 11:10 |
oiaohm | Becides WTL does not port. | Mar 06 11:10 |
oiaohm | It depends on ATL and MS interals. | Mar 06 11:10 |
_Hicham_ | yes | Mar 06 11:10 |
_Hicham_ | that is my point | Mar 06 11:11 |
_Hicham_ | you are finally getting it | Mar 06 11:11 |
oiaohm | Google was building a testing platform. | Mar 06 11:11 |
oiaohm | Not something ever with the idea of being a used product. | Mar 06 11:11 |
_Hicham_ | it has an official release now | Mar 06 11:11 |
oiaohm | That happens big ass headache. | Mar 06 11:11 |
oiaohm | To correct. | Mar 06 11:11 |
_Hicham_ | it is not beta anymore | Mar 06 11:12 |
oiaohm | It would be almost simpler to recode the complete thing. | Mar 06 11:12 |
schestowitz | _Hicham_: I'm still here | Mar 06 11:12 |
_Hicham_ | recode the complete thing? | Mar 06 11:12 |
oiaohm | WTL is not the only problem. | Mar 06 11:12 |
_Hicham_ | do u know how many libraries they used? | Mar 06 11:12 |
green999 | so does chrome then have nasty dependencies? | Mar 06 11:12 |
_Hicham_ | yes | Mar 06 11:12 |
oiaohm | The secuirty framework google developer created does not port either. | Mar 06 11:12 |
oiaohm | WTL is not even a major problem. | Mar 06 11:13 |
_Hicham_ | so they are not intending to port it to Linux | Mar 06 11:13 |
_Hicham_ | oiaohm : tell me about the real problem? | Mar 06 11:13 |
oiaohm | They are intending to port to Linux _Hicham_ | Mar 06 11:13 |
_Hicham_ | how? | Mar 06 11:13 |
_Hicham_ | by rewriting the code completely? | Mar 06 11:13 |
_Hicham_ | If you take for example | Mar 06 11:14 |
_Hicham_ | Google Earth | Mar 06 11:14 |
oiaohm | Yep by rewriting large blocks of code completely. | Mar 06 11:14 |
oiaohm | Because there is no other option for some of them. | Mar 06 11:14 |
_Hicham_ | it has been developed from the beginning to be cross-platform | Mar 06 11:14 |
_Hicham_ | and for that, they used Qt | Mar 06 11:14 |
oiaohm | Because Google Earth was not a experment. | Mar 06 11:14 |
oiaohm | Google developers learnt a really hard leason with chrome | Mar 06 11:15 |
oiaohm | Never ever thing an expermient will not end up suck in existance. | Mar 06 11:15 |
oiaohm | thing/think | Mar 06 11:15 |
_Hicham_ | ok, if you believe so | Mar 06 11:16 |
_Hicham_ | i never red that Chrome is an experiment | Mar 06 11:16 |
_Hicham_ | all the official annoucements said it is the brand new browser from Google | Mar 06 11:16 |
oiaohm | Did you see all the leaks and talk about it before the offical release. | Mar 06 11:17 |
_Hicham_ | not all of them | Mar 06 11:17 |
oiaohm | Google developers were already in it up to there neck before the release. | Mar 06 11:17 |
schestowitz | Google should have thought about their code prior to hastily throwing this browser out there. | Mar 06 11:18 |
schestowitz | Maybe Microsofters on staff | Mar 06 11:18 |
green999 | xiti stopped making maps: | Mar 06 11:18 |
green999 | http://www.xitimonitor.com/en-us/browse... | Mar 06 11:18 |
_Hicham_ | Roy understands me | Mar 06 11:18 |
green999 | is there a 2008 or 2009 map that is public? | Mar 06 11:18 |
oiaohm | http://code.google.com/p/chromium/... chromium to Linux still a on going process | Mar 06 11:18 |
schestowitz | green999: no cache? | Mar 06 11:18 |
schestowitz | oiaohm: many people use Firefox if it comes with Linux | Mar 06 11:19 |
schestowitz | Google faces an uphill battle here. | Mar 06 11:19 |
oiaohm | I know | Mar 06 11:20 |
_Hicham_ | Google can't catch up with Firefox | Mar 06 11:20 |
schestowitz | http://www.youtube.com/watch?v... | Mar 06 11:20 |
oiaohm | Really some of google secuirty ideas are great. | Mar 06 11:20 |
_Hicham_ | Firefox is already very popular | Mar 06 11:20 |
oiaohm | I would like to see them in Firefox | Mar 06 11:21 |
_Hicham_ | like private browsing? | Mar 06 11:21 |
oiaohm | Crash protection. | Mar 06 11:21 |
oiaohm | One tab crashes that is all that goes. | Mar 06 11:21 |
_Hicham_ | crash protection? | Mar 06 11:21 |
green999 | nothing to cache | Mar 06 11:21 |
green999 | they seem to have stopped making the maps | Mar 06 11:21 |
oiaohm | Each tab in chromium is a independant process. | Mar 06 11:21 |
green999 | the old ones are still there and the data shows that MSIE is on average on the way out as far as usage goes | Mar 06 11:22 |
oiaohm | So yes can go complete splat without effecting the main interface. | Mar 06 11:22 |
green999 | yes ff is very popular, but how popular in which countries | Mar 06 11:22 |
green999 | MSIE doesn't go away until you erase windows | Mar 06 11:22 |
_Hicham_ | MSIE is the base of Windows | Mar 06 11:22 |
oiaohm | Really you get people trying to run MSIE in Wine. | Mar 06 11:23 |
oiaohm | That gets really interesting to say the least. | Mar 06 11:23 |
_Hicham_ | I run MSIE on Wine from time to time | Mar 06 11:23 |
_Hicham_ | for MSIE only websites | Mar 06 11:23 |
green999 | h: it's there and can be triggered by activex exploits in other programs | Mar 06 11:25 |
green999 | it's there until the machine no longer has windows | Mar 06 11:25 |
oiaohm | They are there in wine version as well. | Mar 06 11:25 |
oiaohm | Yes eep. | Mar 06 11:25 |
oiaohm | Record so far is 200 viruses running in wine. | Mar 06 11:26 |
_Hicham_ | but ActiveX controls can't do much to Wine | Mar 06 11:26 |
_Hicham_ | oiaohm : are they harmful? | Mar 06 11:26 |
oiaohm | Depends how dumb the user is. | Mar 06 11:26 |
oiaohm | sudo wine | Mar 06 11:26 |
oiaohm | Pray. | Mar 06 11:26 |
oiaohm | Wine + root + viruses = 1 dead Linux box. | Mar 06 11:27 |
_Hicham_ | sudo wine? | Mar 06 11:27 |
_Hicham_ | never heard of that | Mar 06 11:27 |
_Hicham_ | who is dumb enough to do that? | Mar 06 11:27 |
oiaohm | Reason lot of viruses cannot tell the difference between a Linux elf file and a PE file when running inside wine. | Mar 06 11:27 |
oiaohm | So they infect everything. | Mar 06 11:27 |
schestowitz | sudo make sandwich | Mar 06 11:27 |
green999 | speaking of wine, i notice that google's picasa is not running on linux | Mar 06 11:27 |
green999 | it runs on wine | Mar 06 11:27 |
oiaohm | Including kernel image. | Mar 06 11:27 |
oiaohm | So yep 100 percent dead Linux if someone is that dumb. | Mar 06 11:28 |
green999 | s: http://xkcd.com/149/ | Mar 06 11:28 |
oiaohm | Read darwin awards there are humans out there that dumb _Hicham_ | Mar 06 11:28 |
green999 | what goal does google have in using picasa to undermine opengl? | Mar 06 11:28 |
schestowitz | GOOG: "We've also noticed a more disturbing trend: in many of these cases, the patents being asserted against us are owned by — and in a surprising number of cases, are even “invented” by — patent lawyers themselves." http://googlepublicpolicy.blogspot.com/200... | Mar 06 11:29 |
_Hicham_ | Roy : a lot of developers include a special "love" target in their Makefiles, so when you type "make love", you are answered properly :D | Mar 06 11:29 |
schestowitz | find target | Mar 06 11:29 |
oiaohm | http://picasa.google.com/linux/ So not using the Linux releation green999? | Mar 06 11:30 |
schestowitz | _Hicham_: at Microsoft corporation they make "War" a target | Mar 06 11:31 |
schestowitz | That pleases their devs when they use gcc | Mar 06 11:32 |
green999 | o: http://picasa.google.com/linux/faq.html#24 | Mar 06 11:33 |
green999 | o: picasa is not for linux, it's for wine. it's a trojan for introducing directx | Mar 06 11:33 |
green999 | as in all-your-photos-are-belong-to-bill | Mar 06 11:34 |
oiaohm | Not really green999 | Mar 06 11:34 |
_Hicham_ | DirectX is very bad | Mar 06 11:34 |
oiaohm | Wined3d is not MS direct x | Mar 06 11:34 |
green999 | o: install it without wine... | Mar 06 11:34 |
_Hicham_ | it is one of the things that kept linux away from games | Mar 06 11:34 |
_Hicham_ | it is also bad that developers rely on it | Mar 06 11:35 |
oiaohm | Really in time why will it matter. | Mar 06 11:35 |
_Hicham_ | Roy, I think that GNU should provide an IDE which is as good as MSVC | Mar 06 11:35 |
oiaohm | Galuim3d and DRI2 interface allow multiable graphical rendering systems like opengl and others to be used side by side. | Mar 06 11:36 |
_Hicham_ | that way, devs won't go to Billy | Mar 06 11:36 |
_Hicham_ | are they used in games? | Mar 06 11:36 |
_Hicham_ | that is the question | Mar 06 11:36 |
oiaohm | wined3d is used by emulators | Mar 06 11:36 |
oiaohm | on windows. | Mar 06 11:36 |
oiaohm | To be correct virtual machines. | Mar 06 11:37 |
oiaohm | Wine is a wrapper to opengl. | Mar 06 11:37 |
_Hicham_ | but emulation is not a correct solution | Mar 06 11:37 |
_Hicham_ | opengl is supported natively on linux | Mar 06 11:37 |
oiaohm | Galuim3d and DRI2 do allow for direct x to be done native as well. | Mar 06 11:38 |
green999 | opengl is supported natively on most systems AFAIK | Mar 06 11:38 |
green999 | No linux version of picasa: http://picasa.google.com/support/bi... | Mar 06 11:38 |
oiaohm | Wine as forced to go threw opengl due to the design of the video stack in X11. | Mar 06 11:38 |
*MinceR_ is now known as MinceR | Mar 06 11:38 | |
*jose (n=jose@adsl-233-56-48.mia.bellsouth.net) has joined #boycottnovell | Mar 06 11:38 | |
_Hicham_ | especially for games, you have to do a lot of optimizations during build process | Mar 06 11:38 |
oiaohm | Also opengl has not worked perfectly in X11 either. | Mar 06 11:38 |
oiaohm | X11 was only designed for one application. | Mar 06 11:38 |
_Hicham_ | oiaohm : did you try it? | Mar 06 11:39 |
oiaohm | to use opengl at a time. | Mar 06 11:39 |
green999 | Specifically, no linux version of picasa: http://picasa.google.com/support/bin... | Mar 06 11:39 |
*schestowitz accumulates links for bn | Mar 06 11:39 | |
oiaohm | Green999 offical released version no. | Mar 06 11:39 |
oiaohm | There is a beta if you know it. | Mar 06 11:39 |
green999 | Vector for introducing directx: http://picasa.google.com/support/bin/a... | Mar 06 11:40 |
jose | oiaohm, sorry if this is a dumb question (just got on) but what did you mean by "X11 was only designed for one application?" | Mar 06 11:40 |
_Hicham_ | oiaohm : why compiz is such a success in linux then? | Mar 06 11:40 |
green999 | AFAIK all systems except ms vista run opengl natively | Mar 06 11:40 |
oiaohm | You call compiz a success | Mar 06 11:40 |
_Hicham_ | yes | Mar 06 11:40 |
green999 | even a locked-in system like os x uses opengl natively, | Mar 06 11:40 |
jose | schestowitz, do you have a suggestion for me getting in touch with Red Hat legal? | Mar 06 11:41 |
oiaohm | On everything bar NVIDIA ie everything that follows X11 design. It had a race condition. | Mar 06 11:41 |
green999 | os x has but doesn't use x11 | Mar 06 11:41 |
_Hicham_ | you are probably talking about Xorg 1.6 | Mar 06 11:41 |
jose | i haven't yet attempted to get emails and addresses. It might be no problem, but i thought i would ask here first since you have contacts | Mar 06 11:41 |
oiaohm | 1.6 fixes up one section of the problem _Hicham_ | Mar 06 11:42 |
oiaohm | After 1.6 multiable graphical frameworks are supported as well as more than 1 application. | Mar 06 11:42 |
oiaohm | Before that only 1 was techincally supported. | Mar 06 11:42 |
oiaohm | jose one opengl or equal application at a time. | Mar 06 11:43 |
oiaohm | PS os x is using wined3d | Mar 06 11:44 |
oiaohm | in picasa | Mar 06 11:44 |
jose | forgive me because i don't stay up on this but which extension would we be talking about, oiaohm | Mar 06 11:44 |
oiaohm | So the opengl support is not special as such. | Mar 06 11:44 |
jose | so we are talking implementation and not protocol? | Mar 06 11:45 |
oiaohm | Both jose | Mar 06 11:45 |
oiaohm | DRI 1 implementation has very major limits. DRI 2 fixes most of those limits. | Mar 06 11:45 |
oiaohm | compiz was operating by a hack on DRI1 so it failed badly from time to time. | Mar 06 11:46 |
oiaohm | race condition faults are some of the most random you can have. | Mar 06 11:47 |
oiaohm | It was annoying doing support in winehq telling people that compiz would fail them and they would say it has never faild me yet then have them come bad a few days latter with opengl not working | Mar 06 11:48 |
oiaohm | because compiz had failed and stuffed up opengl access. | Mar 06 11:48 |
jose | dri is part of an architecture (design), but that would be implementation unless the dri related protocols had this limitation. | Mar 06 11:48 |
jose | i'm just a bit curious. | Mar 06 11:48 |
jose | i'll be gone soon, but might as well learn some details about dri if i can | Mar 06 11:49 |
oiaohm | dri had protocals for drivers to use. | Mar 06 11:49 |
oiaohm | Yep they were stuffed. | Mar 06 11:49 |
jose | let me ask this way, dri2 changes the interfaces in order to fix some problems or simply for more functionality etc? | Mar 06 11:50 |
oiaohm | Fix problems | Mar 06 11:50 |
oiaohm | Fix major problems that would cause X11 to fail completely. | Mar 06 11:50 |
jose | are docs for dri2 available (short of reading tons of emails or source code)? | Mar 06 11:50 |
oiaohm | Dri 2 was only created when dri1 was found to be impossable to repair because if its design. | Mar 06 11:50 |
oiaohm | I would have to dig back into x.org mailing list. | Mar 06 11:51 |
*_Hicham_ has quit (Read error: 60 (Operation timed out)) | Mar 06 11:51 | |
oiaohm | Or into x.org documentation. | Mar 06 11:51 |
oiaohm | Neither is exaclty fun. | Mar 06 11:51 |
jose | is this relevant http://lists.freedesktop.org/archives/xorg/ ? | Mar 06 11:52 |
jose | x.org documentation.. do you mean an extension? | Mar 06 11:52 |
jose | a formal extension doc? | Mar 06 11:52 |
oiaohm | DRI is the default driver design for xorg | Mar 06 11:53 |
oiaohm | Not the place were you want bad design. | Mar 06 11:53 |
oiaohm | It is covered in the main x.org docs it is nasty reading some of the ones about it in the faq are simpler reading. | Mar 06 11:55 |
oiaohm | Opps mailing list not faq. | Mar 06 11:55 |
jose | I've read this http://en.wikipedia.org/wiki/Direct_R... and a write-up Keith and various other pages.. but it's difficult to remember details if you haven't worked on the code | Mar 06 11:56 |
oiaohm | X Developers' Summit for 2007 work on DRI2 started. The new rendering infrastructure improves several shortcomings of the old design. Among its significant improvements: the lack of internal locks and proper support for offscreen rendering, so that compositing and XVideo/OpenGL applications are properly managed. << that line | Mar 06 11:57 |
oiaohm | Call DRI 1 wild west of Driver design. | Mar 06 11:58 |
oiaohm | It was a lot of good luck that it even worked. | Mar 06 11:58 |
oiaohm | Lack of internal locks meant nothing stoped two applications from writing into the same data struct at the same time. | Mar 06 11:59 |
oiaohm | Generating nice random result. | Mar 06 11:59 |
jose | locks are most likely an implementation detail | Mar 06 11:59 |
jose | or internal api.. that sort of thing | Mar 06 11:59 |
oiaohm | Yes and no. | Mar 06 12:00 |
jose | but it could be exposed in a high level proto | Mar 06 12:00 |
green999 | trolls attacking OOo's independence from MS:http://news.zdnet.co.uk/software/0... | Mar 06 12:00 |
jose | this is something dri2 would likely fix | Mar 06 12:00 |
oiaohm | It is really handy at high level to know at times if a lock is held. | Mar 06 12:00 |
oiaohm | So you can do something else instead. | Mar 06 12:00 |
jose | yes, at times, and especially for performance reasons | Mar 06 12:01 |
jose | as you said | Mar 06 12:01 |
oiaohm | Native support of off screen rendering withotu applications knowing. | Mar 06 12:02 |
oiaohm | Makes doing compiz 1000 times simpler. | Mar 06 12:02 |
jose | dri involves drivers and low level stuff.. at that level yes... but for applications.. maybe or not | Mar 06 12:02 |
oiaohm | No hooking of opengl and other things. | Mar 06 12:02 |
oiaohm | Ie dri2 exposes extra features so hacks are not required any more. | Mar 06 12:02 |
jose | do you know of design docs for compiz? | Mar 06 12:03 |
jose | where does it hook up to the system? | Mar 06 12:03 |
oiaohm | Compiz is being completely recoded for dri2 | Mar 06 12:03 |
oiaohm | It was hooking a stack of opengl functions. | Mar 06 12:03 |
oiaohm | To prevent them overwriting compiz memory. | Mar 06 12:03 |
oiaohm | in the video card. | Mar 06 12:04 |
oiaohm | Yes another problem with DRI 1 | Mar 06 12:04 |
oiaohm | No memory management in video card so you can overwrite another programs memory in there. | Mar 06 12:04 |
oiaohm | As I say that DRI 1 worked at all was more good luck. | Mar 06 12:04 |
jose | what do you mean by hooking (stack of) opengl ? | Mar 06 12:05 |
oiaohm | opengl functions jose | Mar 06 12:05 |
jose | "hooking"? | Mar 06 12:06 |
oiaohm | Intercepting jose | Mar 06 12:06 |
jose | and modifying so that different apps did not step on each other? | Mar 06 12:06 |
oiaohm | Ie instead of call going straight from application to opengl it would go to compiz first for corrections to prevent house of cards failing in. | Mar 06 12:06 |
oiaohm | Problem was when new opengl functions where released compiz would be out of sync so house of cards would come down. | Mar 06 12:07 |
oiaohm | WIne uses lots of only recently released opengl functions. | Mar 06 12:07 |
oiaohm | Basically only true cure was to fix the driver stack. | Mar 06 12:08 |
oiaohm | Its the same as was pulseaudio is doing to Linux audio now. | Mar 06 12:09 |
oiaohm | Instead of fixing the driver stack stick somethign else on top and hope it keeps on working. | Mar 06 12:09 |
jose | let me ask (I'm not sure about extension details): opengl calls (maybe they don't all work the same) are sent to X? where would compiz intercept? | Mar 06 12:09 |
*_Hicham_ (n=hicham@41.249.17.92) has joined #boycottnovell | Mar 06 12:09 | |
jose | oiaohm, i can ask questions as long as you are willing to answer.. i won't get upset if you want to cool off.. but let me know. | Mar 06 12:10 |
jose | pulse audio is a solution where you add another interface on top? | Mar 06 12:11 |
_Hicham_ | pulse audio is a universal sound server | Mar 06 12:11 |
_Hicham_ | it aim to fill the gap between distro and desktop environments | Mar 06 12:11 |
_Hicham_ | it supports nearly all sound architectures | Mar 06 12:12 |
jose | _Hicham_ let me quote what oiaohm said before you got on | Mar 06 12:12 |
_Hicham_ | alsa | Mar 06 12:12 |
_Hicham_ | oss | Mar 06 12:12 |
_Hicham_ | arts | Mar 06 12:12 |
_Hicham_ | ok | Mar 06 12:12 |
_Hicham_ | go ahead | Mar 06 12:12 |
jose | <oiaohm> Basically only true cure was to fix the driver stack. | Mar 06 12:12 |
jose | <oiaohm> Its the same as was pulseaudio is doing to Linux audio now. | Mar 06 12:12 |
jose | <oiaohm> Instead of fixing the driver stack stick somethign else on top and hope it keeps on working. | Mar 06 12:12 |
_Hicham_ | ? | Mar 06 12:12 |
_Hicham_ | it has nothing to do with driver stach | Mar 06 12:13 |
jose | the driver stack conversation is about dri1 vs dri1 | Mar 06 12:13 |
_Hicham_ | stack | Mar 06 12:13 |
jose | dri2 | Mar 06 12:13 |
jose | he (?) then threw in the analogy to pulse audio | Mar 06 12:13 |
_Hicham_ | pulse audio is an abstraction of sound architectures | Mar 06 12:14 |
jose | i think the idea is that pulse audio would leverage something else underneath rather than implement something new | Mar 06 12:14 |
_Hicham_ | Linux needs that | Mar 06 12:14 |
jose | right | Mar 06 12:14 |
_Hicham_ | yes | Mar 06 12:14 |
jose | i'm following then.. just wasnt' sure | Mar 06 12:14 |
_Hicham_ | that is the purpose | Mar 06 12:14 |
_Hicham_ | what is beneath is fixed at the kernel level | Mar 06 12:14 |
jose | yes | Mar 06 12:14 |
jose | you can do both.. fix and redesign below.. to an extent | Mar 06 12:15 |
_Hicham_ | one thing : you can't fix everything but creating new architecture | Mar 06 12:15 |
jose | right | Mar 06 12:15 |
_Hicham_ | coz there is a major issue : compatibility | Mar 06 12:15 |
jose | "you can't fix everything but creating new architecture" | Mar 06 12:16 |
_Hicham_ | you have to take into account the various apps that rely on some specific architecture | Mar 06 12:16 |
jose | you meant to say "by creating" right? | Mar 06 12:16 |
_Hicham_ | yes | Mar 06 12:16 |
oiaohm | Abstraction always introduces a sync problem. | Mar 06 12:16 |
_Hicham_ | it was just a typing mistake | Mar 06 12:16 |
oiaohm | Alsa adds a new feature. | Mar 06 12:16 |
oiaohm | Pulse audio does it wrong darm alsa application breaking all over the place. | Mar 06 12:16 |
_Hicham_ | what do alsa introduce? | Mar 06 12:16 |
_Hicham_ | which alsa application is that? | Mar 06 12:17 |
oiaohm | From time time alsa does add new functions. | Mar 06 12:17 |
oiaohm | For realtime and other features. | Mar 06 12:17 |
jose | you can't always map cleanly | Mar 06 12:17 |
_Hicham_ | PulseAudio will get notified | Mar 06 12:17 |
jose | look at different sorts of threading semantics for example | Mar 06 12:17 |
oiaohm | Coded in two places instead of 1. | Mar 06 12:17 |
oiaohm | Always risks sync errors. | Mar 06 12:17 |
_Hicham_ | there is always a risk of sync error | Mar 06 12:18 |
oiaohm | Inside 1 development tree it simpler. | Mar 06 12:18 |
_Hicham_ | but that is the price to pay for compatibility | Mar 06 12:18 |
_Hicham_ | here we go again | Mar 06 12:18 |
_Hicham_ | you are against linux spirit | Mar 06 12:18 |
_Hicham_ | you can restrict devs | Mar 06 12:19 |
oiaohm | Linux kernel has branches centrally managed. | Mar 06 12:19 |
_Hicham_ | instead, you can coordinate between them | Mar 06 12:19 |
_Hicham_ | as freedesktop.org do | Mar 06 12:19 |
oiaohm | Yes pulseaudio independant tree. | Mar 06 12:19 |
oiaohm | To alsa | Mar 06 12:19 |
_Hicham_ | yes it is | Mar 06 12:19 |
oiaohm | Really there needs to be a tree for audio with everything in it. | Mar 06 12:20 |
schestowitz | jose: mail me, I'll give you E-mails | Mar 06 12:20 |
jose | thanx schestowitz | Mar 06 12:20 |
_Hicham_ | not necessarily | Mar 06 12:20 |
oiaohm | So sync errors are less likely. | Mar 06 12:20 |
jose | i'll do later | Mar 06 12:20 |
oiaohm | _Hicham_: history of Linux kernel tells you that independant trees just bring trouble. | Mar 06 12:20 |
_Hicham_ | so u propose that all sound apps must move to alsa? | Mar 06 12:21 |
_Hicham_ | linux kernel is another story | Mar 06 12:21 |
oiaohm | Not the sound apps the thing that predend to be drivers. | Mar 06 12:21 |
oiaohm | Ie all the sound servers and driver structs. | Mar 06 12:21 |
_Hicham_ | drivers are independant | Mar 06 12:21 |
jose | foss apps can be patched by anyone so as to leverage a better incompat arch.. but if this is not done, you can still include multiple interfaces/libraries at once.. eg that some things like sound perhaps may not work unless coordinated | Mar 06 12:21 |
oiaohm | pulseaudio pretends to be an Alsa driver to applications. | Mar 06 12:21 |
_Hicham_ | drivers are independants from sounds servers | Mar 06 12:21 |
_Hicham_ | pulseaudio is not a driver | Mar 06 12:22 |
oiaohm | once you start pretending to be a driver you are one. | Mar 06 12:22 |
_Hicham_ | it intercepts the streams | Mar 06 12:22 |
oiaohm | You have to provide the same interfaces or things will break. | Mar 06 12:22 |
_Hicham_ | and direct them accordingly | Mar 06 12:22 |
_Hicham_ | sounds interface at the low level is the same | Mar 06 12:23 |
_Hicham_ | but you don't work with that is sound apps | Mar 06 12:23 |
_Hicham_ | in sound app | Mar 06 12:23 |
_Hicham_ | in sound apps, you generally work with an abstraction layer | Mar 06 12:23 |
_Hicham_ | like alsa | Mar 06 12:23 |
_Hicham_ | arts | Mar 06 12:24 |
_Hicham_ | esd | Mar 06 12:24 |
_Hicham_ | oss | Mar 06 12:24 |
oiaohm | Issue is pulseaudio sound interfaces provided as alsa | Mar 06 12:24 |
oiaohm | arts is dead | Mar 06 12:24 |
oiaohm | It is offically marked for end of life. | Mar 06 12:24 |
_Hicham_ | sound interfaces are not marked as alsa | Mar 06 12:24 |
_Hicham_ | ALSA exists as a plugin for PulseAudio | Mar 06 12:24 |
_Hicham_ | that is very different | Mar 06 12:25 |
oiaohm | And it don't work. | Mar 06 12:25 |
_Hicham_ | it does for me | Mar 06 12:25 |
oiaohm | ALSA and OSS have bridges between them. | Mar 06 12:25 |
oiaohm | That have been kept in sync. | Mar 06 12:25 |
_Hicham_ | ye | Mar 06 12:25 |
_Hicham_ | they have both plugins in PA | Mar 06 12:25 |
oiaohm | PA's don't work correctly. | Mar 06 12:26 |
_Hicham_ | wow | Mar 06 12:26 |
_Hicham_ | just tell which distro do u use? | Mar 06 12:26 |
_Hicham_ | PulseAudio is shipped with Ubuntu and Fedora | Mar 06 12:26 |
oiaohm | Own. | Mar 06 12:26 |
_Hicham_ | and is known to work very well on these | Mar 06 12:26 |
_Hicham_ | currently im running ubuntu | Mar 06 12:26 |
oiaohm | There are sets of applications where it don't work well _Hicham_ | Mar 06 12:26 |
_Hicham_ | and it works flawlessly | Mar 06 12:27 |
_Hicham_ | give me an example | Mar 06 12:27 |
_Hicham_ | and I will try | Mar 06 12:27 |
_Hicham_ | give an example of an app that dont work in PA | Mar 06 12:27 |
oiaohm | Do you know what PA answer to applications that don't work correctly as been. | Mar 06 12:27 |
_Hicham_ | silence again | Mar 06 12:27 |
oiaohm | Ask that project to make a PA driver. | Mar 06 12:28 |
oiaohm | Then they don't have to fix the issue. | Mar 06 12:28 |
schestowitz | green999: ZDNET freaks: "Sun disputes claims OpenOffice is a fading project " | Mar 06 12:28 |
schestowitz | the headline suggests (again) that OOo is dying or something | Mar 06 12:28 |
oiaohm | To be correct they are listed on the PA site if you know where to look _Hicham_ | Mar 06 12:28 |
schestowitz | They spread doubt like this on a regular basis | Mar 06 12:28 |
schestowitz | Something about "dead horse" too | Mar 06 12:28 |
oiaohm | Currently I am stuck on dial up speed. | Mar 06 12:28 |
_Hicham_ | oiaohm : I didn't see but It's possible | Mar 06 12:28 |
_Hicham_ | if they don't rely on an abstration layer | Mar 06 12:29 |
_Hicham_ | at all | Mar 06 12:29 |
_Hicham_ | Roy : ZDNet is a M$ puppy | Mar 06 12:29 |
oiaohm | http://www.pulseaudio.org/wiki/PerfectSetup << Start reading | Mar 06 12:30 |
_Hicham_ | OpenOffice is going far beyond the Expectations | Mar 06 12:30 |
oiaohm | Disobey them as see how long pulseaudio lasts _Hicham_ | Mar 06 12:30 |
_Hicham_ | oiaohm : i will read that | Mar 06 12:30 |
_Hicham_ | Roy : why OOo would die? | Mar 06 12:30 |
_Hicham_ | it is well sponsored | Mar 06 12:30 |
_Hicham_ | by Sun itself | Mar 06 12:30 |
_Hicham_ | and SPI | Mar 06 12:31 |
_Hicham_ | why would it die? | Mar 06 12:31 |
oiaohm | Novell wants it dead | Mar 06 12:31 |
oiaohm | So Go-oo can take over. | Mar 06 12:31 |
jose | is part of pulse audio a reimplementation of the various sound interfaces? | Mar 06 12:32 |
jose | that would make sense | Mar 06 12:32 |
oiaohm | If pulseaudio alsa and oss interfaces worked there should be zero need to redirect pulseaudio. | Mar 06 12:32 |
_Hicham_ | I think that the problem is that how to create an abstraction layer above all the sound architectures | Mar 06 12:34 |
_Hicham_ | it is not an easy task | Mar 06 12:34 |
oiaohm | Really there only needs to be 2 | Mar 06 12:34 |
oiaohm | at max 3 | Mar 06 12:34 |
oiaohm | because that is how many kernel level interfaces there are. | Mar 06 12:35 |
oiaohm | Some how I don't think pulseaudio blocks fireware interfaced sound. | Mar 06 12:35 |
oiaohm | fireware/firewire | Mar 06 12:36 |
oiaohm | pulseaudio follows in the long history of don't fix lower down stick something on top. | Mar 06 12:37 |
oiaohm | X11 toolkits started that off. | Mar 06 12:37 |
jose | http://rudd-o.com/en/linux-and-free-... | Mar 06 12:37 |
*_Hicham_ has quit (Remote closed the connection) | Mar 06 12:37 | |
MinceR | sounds like what someone who doesn't understand X would say :> | Mar 06 12:38 |
oiaohm | Might sound that way at first MinceR | Mar 06 12:38 |
oiaohm | Ok xlib old version not multithread | Mar 06 12:38 |
*_Hicham_ (n=hicham@41.249.17.92) has joined #boycottnovell | Mar 06 12:38 | |
oiaohm | So toolkits wrap hacks on top to support multithread | Mar 06 12:38 |
schestowitz | _Hicham_: yes, ZDNet is a puppet | Mar 06 12:38 |
jose | oiaohm, i think x extensions can bypass all the older stuff as much as they want for the most part.. it's a new interface | Mar 06 12:38 |
oiaohm | So you end up with slower code. | Mar 06 12:39 |
schestowitz | I wrote about this a lot, with proof | Mar 06 12:39 |
MinceR | ic | Mar 06 12:39 |
schestowitz | I could repeat :-) | Mar 06 12:39 |
schestowitz | ZDNet is to tech what Fox News is to politics | Mar 06 12:39 |
oiaohm | Then one day someday someone wake up and says lets fix this and creates xcb that is multithreaded. | Mar 06 12:39 |
oiaohm | then shock horrer the core of the X11 server is single threaded. | Mar 06 12:39 |
oiaohm | no one had found out because xlib was single threaded and toolkits where hiding it. | Mar 06 12:40 |
oiaohm | Yes that is X11 history for you. | Mar 06 12:40 |
jose | oiaohm, for threadedness are we talking implementation detail? | Mar 06 12:40 |
oiaohm | X11 is old | Mar 06 12:40 |
oiaohm | So it was using internal round robing threading. | Mar 06 12:40 |
_Hicham_ | there is a proposition for a reimplementation of Xlib | Mar 06 12:41 |
oiaohm | Yep only ever using a max of 1 kernel thead even if there was more. | Mar 06 12:41 |
_Hicham_ | have you heard of that | Mar 06 12:41 |
oiaohm | Xlib has been reimplemented since then. | Mar 06 12:41 |
oiaohm | It called xcb | Mar 06 12:41 |
_Hicham_ | yes, i heard about that | Mar 06 12:41 |
oiaohm | Yes I know X11 extreamally well and painfully well. | Mar 06 12:41 |
oiaohm | History of stupidity in it is classic. | Mar 06 12:42 |
_Hicham_ | did you participate to XCB? | Mar 06 12:42 |
oiaohm | Take a great design and do everything to stuff it up. | Mar 06 12:42 |
oiaohm | Some testing of xcb _Hicham_ | Mar 06 12:42 |
_Hicham_ | and how does it perform? | Mar 06 12:43 |
oiaohm | xlib these days is a emultation on top of xcb | Mar 06 12:43 |
oiaohm | There is no xlib really left | Mar 06 12:43 |
MinceR | doesn't sound all that stupid now :> | Mar 06 12:43 |
_Hicham_ | so you see, the major problem is compatibility | Mar 06 12:43 |
oiaohm | xcb is more raw protocal. | Mar 06 12:43 |
oiaohm | xlib cut of a lot of raw protocal access. | Mar 06 12:44 |
oiaohm | So applications have less crud with xcb between them and the X11 server. | Mar 06 12:44 |
oiaohm | Than with xlib. | Mar 06 12:44 |
jose | it seems to me you now have xlib interface and xcb interface | Mar 06 12:45 |
oiaohm | Yep. | Mar 06 12:45 |
oiaohm | so you have backwards compadiblity. | Mar 06 12:45 |
jose | ie, an extension though perhaps not an official "x extension" | Mar 06 12:45 |
oiaohm | Nothing about the x11 protocal says you have to use xlib | Mar 06 12:45 |
jose | yeah, but if xlib is worse, then it deserves worse performance | Mar 06 12:45 |
oiaohm | It never did. | Mar 06 12:46 |
jose | i did think xlib was very close to the protocol by definition | Mar 06 12:46 |
MinceR | do toolkits use xcb already? | Mar 06 12:46 |
oiaohm | Some do. | Mar 06 12:46 |
jose | i don't do much x programming, btw | Mar 06 12:46 |
oiaohm | Lot still use xlib | Mar 06 12:46 |
oiaohm | It takes a long time to change a code base. | Mar 06 12:46 |
MinceR | do gtk and qt do? | Mar 06 12:46 |
oiaohm | gtk still xlib qt xcb in places. | Mar 06 12:47 |
jose | qt just got a speed upgrade.. they likely reworked a number of things at the lower level interfaces i would guess | Mar 06 12:47 |
oiaohm | Yep right on the money jose | Mar 06 12:47 |
oiaohm | Pulseaudio breaks the basic rule of fast applications. | Mar 06 12:48 |
oiaohm | You want the least ammout of code between you and the output. | Mar 06 12:48 |
oiaohm | GEM will even allow applications to bypass X11 completely and still render in a X11 window. | Mar 06 12:48 |
oiaohm | Graphical stack is on the right path. | Mar 06 12:49 |
oiaohm | 1 central interface management is required for audio. Going out side kernel provided is not going to ensure that. | Mar 06 12:50 |
MinceR | we have alsa for that | Mar 06 12:50 |
MinceR | but how would you implement network transparent audio? | Mar 06 12:51 |
jose | MinceR, have an init call | Mar 06 12:51 |
jose | here you switch to native alsa or pulse audio masquerading as alsa | Mar 06 12:52 |
jose | right? | Mar 06 12:52 |
MinceR | rught | Mar 06 12:52 |
MinceR | s/u/i/ | Mar 06 12:52 |
jose | assuming alsa is good enough for all local purposes | Mar 06 12:52 |
jose | if not then we need to fix alsa | Mar 06 12:52 |
MinceR | well, i could choose the alsa device as needed | Mar 06 12:52 |
oiaohm | Alsa devices are not restricted on what they can be interfacing with. | Mar 06 12:52 |
jose | i'm not an alsa expert by any means, btw | Mar 06 12:53 |
jose | so i may get lost here and there in the conversation | Mar 06 12:53 |
oiaohm | Also pulseaudio is going to become more of a pain as container tech gets more broader used. | Mar 06 12:54 |
oiaohm | How is pulsaudio sitting on pulseaudio then sitting on pulseaudio again going to perform. | Mar 06 12:54 |
jose | vms and such? "container tech" | Mar 06 12:54 |
oiaohm | cgroups in Linux. | Mar 06 12:54 |
oiaohm | Or zones in solarias. | Mar 06 12:55 |
_Hicham_ | oiaohm : that is not the final solution | Mar 06 12:55 |
oiaohm | That is the problem _Hicham_ | Mar 06 12:55 |
_Hicham_ | PA provides a migration ABI through its plugin | Mar 06 12:55 |
_Hicham_ | all applications are moving to it | Mar 06 12:55 |
_Hicham_ | the final goal is that all applications provide interfaces for it | Mar 06 12:55 |
oiaohm | Not going to happen. | Mar 06 12:56 |
_Hicham_ | since it is the product of a long collaboration | Mar 06 12:56 |
oiaohm | Wine project has flatly refused to include a pulseaudio driver. | Mar 06 12:56 |
_Hicham_ | oiaohm : why not? | Mar 06 12:56 |
oiaohm | It has a alsa driver and a oss driver that are both in kernel. | Mar 06 12:56 |
oiaohm | They are are a big enough problem to maintain. | Mar 06 12:57 |
oiaohm | Sorry Wine is not alone there. | Mar 06 12:57 |
_Hicham_ | maybe because of lack of resources | Mar 06 12:57 |
_Hicham_ | whose else? | Mar 06 12:57 |
oiaohm | Its because you can disable pulse audio | Mar 06 12:57 |
oiaohm | You still need the other interfaces. | Mar 06 12:57 |
oiaohm | Pulseaudio is simplely extra work that can be better spent else where. | Mar 06 12:57 |
oiaohm | Also Pulseaudio is not a suitable long term solution. | Mar 06 12:58 |
oiaohm | cgroup tech will come to Linux. | Mar 06 12:58 |
oiaohm | There is no reason why audio interface providing the same per application volume could not be done threw it. | Mar 06 12:58 |
_Hicham_ | what is the solution then oiaohm? | Mar 06 12:58 |
oiaohm | Information in a cgroup travels with the process it self. | Mar 06 12:59 |
oiaohm | Rememember cgroup also provide application suspend to disk. | Mar 06 12:59 |
oiaohm | With pulseaudio you can guess what kinds of fun that causes. | Mar 06 13:00 |
oiaohm | Basically its future tech incompadible. | Mar 06 13:00 |
jose | http://www.mjmwired.net/ke... | Mar 06 13:00 |
MinceR | the wine project refusing to do it doesn't mean it won't happen | Mar 06 13:00 |
oiaohm | It wont happen MinceR | Mar 06 13:01 |
oiaohm | Its final. | Mar 06 13:01 |
MinceR | ever heard of forks and distro patches? :> | Mar 06 13:01 |
oiaohm | Wine releases every 2 weeks | Mar 06 13:01 |
oiaohm | You try keeping a patch in sync with that. | Mar 06 13:01 |
oiaohm | Heck most distrobutions cannot keep up building wine let alone custom patching it. | Mar 06 13:02 |
MinceR | the audio-related code doesn't necessarily change | Mar 06 13:02 |
MinceR | also, distros don't offer every release | Mar 06 13:02 |
oiaohm | Every distribution who has done custom patches has given up on it with wine. | Mar 06 13:02 |
MinceR | they usually take one, test it and take it to stable, then provide only minor updates | Mar 06 13:03 |
oiaohm | Its so simple to stuff it up completely. | Mar 06 13:03 |
MinceR | doesn't mean they won't try. | Mar 06 13:03 |
oiaohm | Wine stable is basically a paper weight for application support. | Mar 06 13:03 |
oiaohm | Wine development great support alterations all the time. | Mar 06 13:03 |
MinceR | i mean distro stable, not wine :> | Mar 06 13:03 |
jose | the community will grow in size over time.. fwiw | Mar 06 13:03 |
MinceR | i've seen "stable" packages made from a particular daily snapshot :> | Mar 06 13:04 |
oiaohm | Remember wine is trademarked | Mar 06 13:04 |
oiaohm | So yep you will not be able to call it wine either. | Mar 06 13:04 |
MinceR | so it won't be wine? | Mar 06 13:04 |
oiaohm | It is not happening in wine ever. | Mar 06 13:05 |
_Hicham_ | wine is offered pre packaged for most distro | Mar 06 13:05 |
oiaohm | If you brance wine off you will have to support it yourself. | Mar 06 13:05 |
oiaohm | branch wine off you will have to support it yourself. | Mar 06 13:05 |
oiaohm | Not something you want to be left with. | Mar 06 13:05 |
oiaohm | Pulseaudio need to work out how they are going to support cgroups. | Mar 06 13:06 |
*PetoKraus has quit (Read error: 110 (Connection timed out)) | Mar 06 13:06 | |
MinceR | which might simply mean reapplying a patch until it breaks, and then forward porting it. :> | Mar 06 13:06 |
oiaohm | You know the problem with that. | Mar 06 13:07 |
oiaohm | Wine is a extreamally complex program. | Mar 06 13:07 |
oiaohm | a patch may apply and take about 5 days before you find out you have stuffed it. | Mar 06 13:07 |
_Hicham_ | very very very very complex in fact | Mar 06 13:07 |
oiaohm | By then you have already released it to your users. | Mar 06 13:08 |
oiaohm | Its also the reason why wine does not want any unrequired parts. | Mar 06 13:08 |
_Hicham_ | no disto bothers with wine as it is prepackaged | Mar 06 13:08 |
oiaohm | Its insanely complex to start off with. | Mar 06 13:08 |
oiaohm | Pulseaudio really need to work out how they are going to support the features cgroup is providing. | Mar 06 13:09 |
oiaohm | Currently Pulseaudio is forcing it and all applications using it to be suspended at once. | Mar 06 13:09 |
oiaohm | Until thing Pulseaudio is nothing more than a short term thing that will have to be replaced in time. | Mar 06 13:10 |
jose | let me ask, cgroup info can be discovered at the time a process uses (pulse)audio, yes? | Mar 06 13:11 |
oiaohm | cgroup is a kernel level feature. | Mar 06 13:11 |
jose | is there a specific problem with cgroup pulseaudio (an example)? | Mar 06 13:12 |
oiaohm | When it suspends to disk problem here is that pulseaudio will not find out what exacly happend to the application. | Mar 06 13:12 |
oiaohm | So only way you can prevent application and pulseaudio from being in different state is to suspend both with each other. | Mar 06 13:12 |
oiaohm | Because pulseaudio is not kernel and is not cgroup aware enough. | Mar 06 13:13 |
jose | but making it aware | Mar 06 13:13 |
jose | presents what problem? | Mar 06 13:13 |
oiaohm | Yep most likely equals recoding it. | Mar 06 13:13 |
oiaohm | And adding kernel features to support it. | Mar 06 13:13 |
oiaohm | GEM for the graphical stack is such a example. | Mar 06 13:14 |
jose | but interfaces is what counts.. what would be the issue here | Mar 06 13:14 |
oiaohm | Everything that is needed at the restore point the cgroup suppend better save. | Mar 06 13:14 |
oiaohm | That is kinda not possiable if its not tracked for cgroup. | Mar 06 13:15 |
oiaohm | GEM provides a way to do it for graphical memory blocks. X11 protocal has a resend command for basic 2d graphics. So graphics can work with cgroup. | Mar 06 13:16 |
jose | is anything more needed besides simply knowing the suspend flag (timestamp?) | Mar 06 13:16 |
oiaohm | Notice something here pulseaudio does not allow it half done state to be saved or for cgroup to know that it need to leave it longer. | Mar 06 13:16 |
oiaohm | Volume information for application is stored in Pulseaudio. | Mar 06 13:17 |
oiaohm | So cgroup restored application will restore at wrong volume. | Mar 06 13:17 |
oiaohm | The list of problems go on and on. | Mar 06 13:17 |
jose | but can a flag lead to pulseaudio setting its proper state? | Mar 06 13:18 |
oiaohm | There need to be a audio cgroup really to store some of th information | Mar 06 13:18 |
oiaohm | the | Mar 06 13:18 |
jose | audio cgroup | Mar 06 13:18 |
jose | ok | Mar 06 13:18 |
jose | but isn't that necessary | Mar 06 13:18 |
oiaohm | Ie kernel mode alteration. | Mar 06 13:18 |
MinceR | i think alsa should do per-application volume control for local apps | Mar 06 13:18 |
jose | no matter what audio arch you use | Mar 06 13:18 |
MinceR | perhaps in dmix | Mar 06 13:19 |
oiaohm | By putting it in a audio cgroup the information can be connected to the process of the application it self. | Mar 06 13:19 |
oiaohm | Instead of hidden in a sound server. | Mar 06 13:20 |
oiaohm | Also allows differnet sound servers to use the same data. | Mar 06 13:20 |
jose | the libraries are part of the process right (they get an image right)? | Mar 06 13:20 |
oiaohm | Memory management is also part of it. | Mar 06 13:20 |
oiaohm | Current pulse audio application dies. | Mar 06 13:20 |
oiaohm | How does pulse find out. How does kernel make sure nothing of an application is left in memory. | Mar 06 13:21 |
oiaohm | There does need to be a memory manager for audio as well. | Mar 06 13:21 |
oiaohm | Linked to the applications process. | Mar 06 13:21 |
oiaohm | Setting up cgroups means everything need to be trackable. | Mar 06 13:22 |
oiaohm | No more that might have come from X application. | Mar 06 13:22 |
oiaohm | Basically there is a lot of work to make pulseaudio shipshape for the future. | Mar 06 13:23 |
oiaohm | It might be just simpler to fix ALSA up to do pulseaudio features and forget that pulseaudio ever existed. | Mar 06 13:23 |
jose | let me ask as i can't follow the details here... | Mar 06 13:24 |
jose | what would be difference in disk suspend with or without cgroups (for pulseaudio) | Mar 06 13:24 |
jose | ? | Mar 06 13:25 |
oiaohm | cgroups allows per application suspend to disk. | Mar 06 13:25 |
jose | ok .let me think | Mar 06 13:25 |
oiaohm | Also cgroups these days do the system wide suspend to disk too. | Mar 06 13:26 |
jose | fine.. now wouldn't the suspend effect give the app time to prepare? | Mar 06 13:26 |
jose | the preparation can alert the daemon | Mar 06 13:26 |
oiaohm | cgroup is only checking kernel level data. | Mar 06 13:26 |
jose | ditto when you come back online | Mar 06 13:26 |
oiaohm | Application does not know it has been suspended. | Mar 06 13:26 |
oiaohm | Only on restore could you do something. | Mar 06 13:27 |
oiaohm | You have a real data location problem. | Mar 06 13:27 |
jose | but the suspending of the processes.. shouldn't these be notified somehow or some queryable flag set (that pulse audio could keep track of or hook into)? | Mar 06 13:27 |
oiaohm | There is no flag for good reasons. | Mar 06 13:27 |
MinceR | hm, what if apps connected to the sound server over sockets? | Mar 06 13:27 |
jose | MinceR, I don't follow.. how does cgroups relate to the socket question? | Mar 06 13:29 |
jose | [I have to ask questions to keep up, sorry.] | Mar 06 13:29 |
MinceR | afaik the discussion is about how we keep track of audio clients | Mar 06 13:29 |
oiaohm | Application connected via socket will be suspended along with target application if cgroup cannot id split. Zoned off cgroup you would not be able to talk to another application outside it with a socket anyhow. | Mar 06 13:29 |
jose | [even if they might be dumb in some cases] | Mar 06 13:29 |
MinceR | how common is it to zone off typical audio apps? | Mar 06 13:29 |
oiaohm | Remember cgroups allow running multiable incompadible distributions side by side. | Mar 06 13:30 |
oiaohm | On the same kernel. | Mar 06 13:30 |
MinceR | i don't know a lot about cgroups | Mar 06 13:30 |
oiaohm | There is one mother of a headache. | Mar 06 13:30 |
MinceR | i took a look at that txt file | Mar 06 13:30 |
MinceR | (didn't read much of it) | Mar 06 13:30 |
oiaohm | Ever used openvz MinceR | Mar 06 13:31 |
MinceR | nope | Mar 06 13:31 |
MinceR | only vmware, virtualbox and KVM | Mar 06 13:31 |
MinceR | (i prefer KVM) | Mar 06 13:31 |
oiaohm | Ok KVM without needing extra copy of kernel MinceR | Mar 06 13:31 |
oiaohm | To run another distribution. | Mar 06 13:31 |
MinceR | ic | Mar 06 13:32 |
MinceR | so it's like UML? | Mar 06 13:32 |
oiaohm | Not really. | Mar 06 13:32 |
oiaohm | 1 Linux kernel. | Mar 06 13:32 |
oiaohm | Multibale process tables | Mar 06 13:32 |
oiaohm | Memory management keeping everything apart that should be appart. | Mar 06 13:32 |
oiaohm | Even some disk level management. | Mar 06 13:33 |
jose | oiaohm, do you mean "multiple" when you say multibale and multiable? | Mar 06 13:33 |
oiaohm | Yep oiaohm | Mar 06 13:33 |
oiaohm | darm | Mar 06 13:33 |
jose | i could not figure that out | Mar 06 13:33 |
jose | you meant jose, right? | Mar 06 13:33 |
oiaohm | I am having dyslexia on a few words at moment. | Mar 06 13:33 |
jose | :-) | Mar 06 13:33 |
jose | it doesn't bother me, but i want to try and keep up so i have to get inside your head | Mar 06 13:34 |
MinceR | :) | Mar 06 13:34 |
jose | :-) | Mar 06 13:34 |
oiaohm | cgroups add a complete new level of complexity. | Mar 06 13:35 |
MinceR | what's the difference between UML, vserver and openvz? | Mar 06 13:35 |
oiaohm | vserver and openvz tech are both being merged to form cgroup. | Mar 06 13:35 |
MinceR | oh. | Mar 06 13:35 |
oiaohm | UML is a lInux kernel running inside Linux. | Mar 06 13:35 |
oiaohm | Solarias Zones take it to a completely different level. | Mar 06 13:36 |
oiaohm | Inside a Solarias Zone it can be a completely different OS emulated by the kernel. | Mar 06 13:36 |
oiaohm | The levels of trouble you are talking about long term with cgroup for pulseaudio if it don't support it will be kinda insane. | Mar 06 13:37 |
oiaohm | Really how much of pulseaudio information about applications volume controls and the like could not be attached to the application process data. | Mar 06 13:38 |
oiaohm | It solve the problem why the pulseaudio developer said dmix could never do it. | Mar 06 13:38 |
oiaohm | Pulseaudio has basically been developed in the wrong level with the wrong tech. | Mar 06 13:39 |
jose | let me ask this because it's not clear to me.. what is wrong with an app stopping contact with pulse audio server for T time and then continuing? (suspend to disk) | Mar 06 13:43 |
jose | --- oiaohm | Mar 06 13:44 |
oiaohm | If its like halfway threw sending a audio block to pulse audio. | Mar 06 13:44 |
oiaohm | You could get a really nasty noise on the other end. | Mar 06 13:44 |
jose | ok | Mar 06 13:44 |
oiaohm | Lets just say it better if we avoid that event. | Mar 06 13:44 |
oiaohm | Also volume control information for the application. | Mar 06 13:46 |
jose | you'd want the full transmission unit to take place plus a cue that suspend is happening | Mar 06 13:46 |
jose | why isn't this up to the kernel to manage | Mar 06 13:46 |
oiaohm | that is stored inside the pulseaudio | Mar 06 13:46 |
jose | if you add support to the kernel, then pulseaudio could later tap into it | Mar 06 13:46 |
oiaohm | So restored applicaiton could come back with either no volume or full voulume. | Mar 06 13:46 |
oiaohm | exactly jose | Mar 06 13:46 |
oiaohm | Pulseaudio need to add some kernel support to work long term. | Mar 06 13:47 |
oiaohm | But them you kill the major problem the developer said Pulseaudio had to exist. | Mar 06 13:47 |
oiaohm | Dmix could not know the application volume of a process. | Mar 06 13:47 |
oiaohm | That disappears completely does it not with kernel storage. | Mar 06 13:48 |
jose | but you care when you come back.. isn't the data brought back? | Mar 06 13:48 |
oiaohm | Other evil of cgroups you many not be restoring on the same machine you supended on. | Mar 06 13:48 |
oiaohm | So yes you cannot depend on anything not traveling with the application knowning anything. | Mar 06 13:49 |
jose | it still seems like something for the low level people to work out first and then pulseaudio could add support | Mar 06 13:49 |
oiaohm | first a protocal/standard of what they want has to be designed. | Mar 06 13:50 |
jose | you are right that if we want smooth audio, the kernel will have to have special support for that case | Mar 06 13:50 |
oiaohm | Kernel mode developers could add a struct but if not all the needed information is store its worthless. | Mar 06 13:50 |
jose | adding something new like cgroups seems will require coordination among various kernel level items.. like alsa or whatever | Mar 06 13:50 |
oiaohm | Basicaly the audio guys need to put there head together. | Mar 06 13:51 |
jose | i still imagine that pulseaudio would be fixable.. (from what i can tell in this high level discussion) | Mar 06 13:51 |
oiaohm | Memory management of audio streams in a cgroup trackable way. | Mar 06 13:51 |
jose | cgroup is quite new right.. so you are right about the audio coordination still needed perhaps | Mar 06 13:52 |
oiaohm | Would add some nice headaches. | Mar 06 13:52 |
oiaohm | cgroup new by kernel standard basics have been in there for over 9 months. | Mar 06 13:53 |
jose | i think i noticed that fedora 11 will include that (i think that is where i saw a reference to cgroup) | Mar 06 13:53 |
oiaohm | cgroup developers are not working on the audio section because they don't have agreement between the audio framework. | Mar 06 13:53 |
oiaohm | cgroups developers are also waiting on DRI2 particular GEM. | Mar 06 13:54 |
oiaohm | To make X11 work well. | Mar 06 13:54 |
jose | the scenario mentioned .. of having a buffer that gets emptied automatically by the kernel as the app goes to sleep would be required in most cases for sound asthetics | Mar 06 13:54 |
jose | also, you want a flag of some sort. but that can just tag on to the sound data being sent off | Mar 06 13:54 |
jose | i looked up gem and got this: http://www.reddit.com/r/linux/comments/... | Mar 06 13:55 |
jose | http://sourceforge.net/projects/pd-gem/ | Mar 06 13:56 |
oiaohm | Only intel drivers support it at this time jose | Mar 06 13:56 |
oiaohm | pd-gem is a differnet thing | Mar 06 13:57 |
jose | oh | Mar 06 13:57 |
oiaohm | http://en.wikipedia.org/wiki... | Mar 06 13:58 |
oiaohm | Its basically a memory manager tracking graphical resoruces. | Mar 06 13:58 |
oiaohm | Applications windows and so on can be connected to the application and suspend because of it. | Mar 06 13:59 |
oiaohm | There is a equal framework for network stuff. | Mar 06 13:59 |
oiaohm | There is nothing for audio that provides a tracking framework like that. | Mar 06 14:00 |
oiaohm | In Linux | Mar 06 14:00 |
oiaohm | Windows Vista and up does have a tracking framework. | Mar 06 14:00 |
oiaohm | Its really stupid if for some reason something has laged you have killed the application and a bell from an application still plays. | Mar 06 14:01 |
oiaohm | Yes pulseaudio can do that. | Mar 06 14:01 |
jose | supposedly gem got introduced in this email thread: http://sourceforge.net/mailarchive/forum.php?... | Mar 06 14:02 |
oiaohm | Yep | Mar 06 14:03 |
oiaohm | Issue now is driver support jose | Mar 06 14:03 |
oiaohm | There is lag between new framework and enough drivers so its useful. | Mar 06 14:03 |
*oiaohm has quit (Remote closed the connection) | Mar 06 14:04 | |
schestowitz | Wiki just restored: http://boycottnovell.com/wiki/in... | Mar 06 14:13 |
schestowitz | The host didn't even tell me why it went corrupt (the DB) | Mar 06 14:13 |
schestowitz | http://www.microturfs.org/ | Mar 06 14:15 |
*zer0c00l (n=zer0c00l@210.212.255.131) has joined #boycottnovell | Mar 06 14:17 | |
schestowitz | Further exposition of Richard Steel/Newham and the presidency of the Czech Republic (funded by Microsoft): http://boycottnovell.com/2009/03/0... | Mar 06 14:33 |
*jose has quit ("Leaving") | Mar 06 14:45 | |
_Hicham_ | Roy, M$ and Novell are probably targetting the website | Mar 06 14:53 |
schestowitz | Which one? | Mar 06 15:02 |
_Hicham_ | boycottnovell.com | Mar 06 15:03 |
schestowitz | Yes, they leave comments there. | Mar 06 15:03 |
_Hicham_ | Miguel de Icaza must be very angry | Mar 06 15:07 |
_Hicham_ | I dont understand this man | Mar 06 15:07 |
_Hicham_ | he was one the great gnome founders | Mar 06 15:07 |
_Hicham_ | then he went to Microsoft for some pennies? | Mar 06 15:08 |
_Hicham_ | tainting the whole linux community | Mar 06 15:08 |
_Hicham_ | as the famous the saying is telling : In a world without walls and fences, who needs windows and gates? | Mar 06 15:09 |
_Hicham_ | who needs M$ in Linux? | Mar 06 15:09 |
*macabe (n=macabe@cpe-67-240-216-105.rochester.res.rr.com) has joined #boycottnovell | Mar 06 15:13 | |
*macabe__ has quit (Read error: 54 (Connection reset by peer)) | Mar 06 15:13 | |
schestowitz | _Hicham_: Miguel has been with Microsoft since the beginning | Mar 06 15:14 |
schestowitz | It's not as though he 'turned' to them. | Mar 06 15:14 |
schestowitz | He started with Linux after he had attempted (without success) to get a job at Microsoft | Mar 06 15:15 |
_Hicham_ | very deceiving | Mar 06 15:16 |
_Hicham_ | I don't know why gnome project still uses Novell Apps | Mar 06 15:17 |
_Hicham_ | remarkably, evolution | Mar 06 15:17 |
_Hicham_ | anything from Novell is a risk | Mar 06 15:18 |
_Hicham_ | what if they port evolution to Mono? | Mar 06 15:18 |
_Hicham_ | should gnome incorporate mono by default? | Mar 06 15:18 |
MinceR | then they should keep to the last non-mono version | Mar 06 15:19 |
MinceR | and gnome shouldn't incorporate mono | Mar 06 15:19 |
_Hicham_ | it is a very embarrassing issue | Mar 06 15:19 |
MinceR | also, its hig should be rewritten without the apple crap | Mar 06 15:19 |
_Hicham_ | how can u tell? | Mar 06 15:19 |
_Hicham_ | the evil must be removed from the roots | Mar 06 15:20 |
MinceR | it's easy to tell whether it uses mono or not | Mar 06 15:20 |
_Hicham_ | I am not talking about that | Mar 06 15:20 |
MinceR | try running it on a system without mono; inspect the source code | Mar 06 15:20 |
_Hicham_ | i am talking about the fact that gnome will incorporate it as an essential package always, but with mono | Mar 06 15:21 |
_Hicham_ | that would a real pb | Mar 06 15:21 |
_Hicham_ | who is the real sponsor of mono? | Mar 06 15:21 |
_Hicham_ | of gnome i mean | Mar 06 15:21 |
_Hicham_ | sorry | Mar 06 15:21 |
_Hicham_ | isnt it RedHat? | Mar 06 15:21 |
_Hicham_ | what I know is that RedHat is the official hoster of Gnome | Mar 06 15:21 |
*mib_0w8gi6 (i=44e625da@gateway/web/ajax/mibbit.com/x-09beec59b2484a47) has joined #boycottnovell | Mar 06 15:22 | |
*mib_0w8gi6 has quit (Client Quit) | Mar 06 15:22 | |
*paramahamsa (n=phamsa@59.92.60.31) has joined #boycottnovell | Mar 06 15:40 | |
schestowitz | Hi, paramahamsa | Mar 06 15:44 |
schestowitz | _Hicham_: GNOME has many sponsors | Mar 06 15:45 |
paramahamsa | schestowitz: hi. | Mar 06 15:51 |
paramahamsa | I don't quite understand. Why is Novell bad? | Mar 06 15:51 |
schestowitz | It helps Microsoft harm Free software. | Mar 06 15:55 |
schestowitz | For Novell's benefit, which isn't. | Mar 06 15:55 |
paramahamsa | How? Is there a page documenting the details? | Mar 06 15:56 |
_Hicham_ | Roy : Gnome hoster is RedHat, no? | Mar 06 16:01 |
schestowitz | paramahamsa: http://boycottnovell.com/faq | Mar 06 16:05 |
schestowitz | http://news.zdnet.com/2100-9595_22-276211.html "Make no mistake, this is intended to force Tom Tom to violate the GPL, or change to Microsoft embedded software." | Mar 06 16:08 |
schestowitz | _Hicham_: technically, yes, I guess. | Mar 06 16:08 |
*PetoKraus (n=pk@fsf/member/petokraus) has joined #boycottnovell | Mar 06 16:11 | |
paramahamsa | Wow. I didn't even know all this was happening. Have to look it up. What are people doing about it? | Mar 06 16:11 |
schestowitz | Stating the obvious here: | Mar 06 16:12 |
schestowitz | Bad Economy = Easier Military Recruiting < | Mar 06 16:12 |
schestowitz | http://www.prwatch.org/node/8257 > | Mar 06 16:12 |
schestowitz | OT BTW | Mar 06 16:12 |
schestowitz | paramahamsa: What people should do is pressure Novell.. and pressure others not to be infected by Novell's poison (Moonlight. Mono/.NET, etc). | Mar 06 16:13 |
schestowitz | All that corruption could lead to another war. :-( | Mar 06 16:14 |
paramahamsa | What exactly is Novell trying to do? Get money? | Mar 06 16:14 |
schestowitz | Join forces with Microsoft | Mar 06 16:15 |
schestowitz | Microsoft pays them for it | Mar 06 16:15 |
schestowitz | Sweden's EU presidency logo sets sights on climate < http://www.euractiv.com/en/opinion/sweden-e... > | Mar 06 16:18 |
schestowitz | Global Warming May Get Its Very Own Top Level Domain < http://www.techcrunch.com/2009/03/04/global-... > | Mar 06 16:18 |
paramahamsa | Do you have any active groups that actually meet and discuss things? like LUGS? | Mar 06 16:19 |
schestowitz | Which things specifically? | Mar 06 16:20 |
*schestowitz sees China's already rising to superpower status... China says worried about arrest warrant on Sudan's Bashir < http://uk.reuters.com/article/oilRpt... > | Mar 06 16:21 | |
paramahamsa | Anything interesting. Novell, Global Warming. I'm passionate about both. | Mar 06 16:21 |
_Hicham_ | Roy's group is the most active one | Mar 06 16:25 |
_Hicham_ | you can't find such a group on the net :-D | Mar 06 16:25 |
schestowitz | :-) | Mar 06 16:25 |
paramahamsa | The thing is, I'm in India. Nothing here? | Mar 06 16:27 |
paramahamsa | Where is Roy's group? | Mar 06 16:27 |
schestowitz | http://techdirt.com/articles/200... Moody: "No Doubt: Buy A Concert Ticket, Download All Our Songs - make money on the scarce goods, give away the abundant ones" | Mar 06 16:27 |
schestowitz | What do you mean by "Group"? | Mar 06 16:27 |
_Hicham_ | paramahamsa : you are now in Roy's Group | Mar 06 16:28 |
paramahamsa | I mean a bunch of people in a particular geographic location who physically meet, | Mar 06 16:29 |
_Hicham_ | ah | Mar 06 16:29 |
_Hicham_ | Maybe in the US | Mar 06 16:29 |
_Hicham_ | do you have any meetings in the US Roy? | Mar 06 16:30 |
schestowitz | The 'disease' (Linux) is spreading in American: Cuba Gets an (Open) Hand from Brazil < http://opendotdotdot.blogspot.com/2009/03/c... > | Mar 06 16:30 |
schestowitz | *American nations... | Mar 06 16:30 |
schestowitz | We have ManLUG here. | Mar 06 16:30 |
paramahamsa | The US is a little far for me to be travelling every now and then... | Mar 06 16:30 |
schestowitz | First one in the UK.... but I hardly attend it anymore. Linux grew past the LUGs | Mar 06 16:31 |
paramahamsa | Who is Roy? | Mar 06 16:31 |
schestowitz | Do LUGs still matter? < http://www.linux.com/feature/50878 > | Mar 06 16:31 |
schestowitz | paramahamsa: I guess the Internet changed a lot too, not just at the coding level. | Mar 06 16:32 |
schestowitz | http://www.worldchanging.com/arc... "The Latest on Climate Science, Solutions, and Politics Jack Bauer becomes first-ever carbon-neutral torturer as Rupert Murdoch says “Climate change poses clear, catastrophic threats”" | Mar 06 16:32 |
schestowitz | Even Murdoch! | Mar 06 16:32 |
schestowitz | Riding with the first cowboys – in 3500 BC < http://www.newscientist.com/article/dn167... > | Mar 06 16:34 |
*_Hicham_ has quit (leguin.freenode.net irc.freenode.net) | Mar 06 16:36 | |
schestowitz | The Microsoft press is whitewashing the crimes of a family that for decades has defrauded the United States government: http://www.techflash.com/microsoft/... | Mar 06 16:38 |
paramahamsa | schestowitz: do you hang around on this channel all day? | Mar 06 16:39 |
schestowitz | When I'm home | Mar 06 16:40 |
schestowitz | What is this eWeek Europe? | Mar 06 16:42 |
schestowitz | http://www.eweekeurope.co.uk/comment... | Mar 06 16:42 |
paramahamsa | What's your profession? | Mar 06 16:42 |
schestowitz | And Pete Judge is from ZDNet UK | Mar 06 16:42 |
schestowitz | eWeek is on bankruptcy protection of something (Ziff Davis) with very few writers left... | Mar 06 16:42 |
schestowitz | paramahamsa: research | Mar 06 16:42 |
schestowitz | Protest against MS: http://www.eweekeurope.co.uk/comment/mi... | Mar 06 16:42 |
paramahamsa | schestowitz: what research? | Mar 06 16:43 |
schestowitz | Mostly FOSS stuff (not paid for) | Mar 06 16:44 |
schestowitz | Conference on the European patent system, Brussels, Belgium < http://cordis.europa.eu/fetch?CALLER=... > | Mar 06 16:45 |
schestowitz | ping zoobab01 | Mar 06 16:46 |
paramahamsa | schestowitz: Do you have a random-link-generator or something? | Mar 06 16:47 |
schestowitz | I know someone who's part of the WG7 (Open Source) and we need info.. | Mar 06 16:47 |
schestowitz | schestowitz: random? No, why? | Mar 06 16:47 |
schestowitz | zoobab01: They need to deal with SAP and CompTIA -- those that want to call Open Source some criminal business. One document has been leaked to Wikileaks and therefore one can request the entirety of the documents of this strategy panel (perhaps the one of all the workgroups). Maybe we can request the access to documents, via the European Commission... | Mar 06 16:48 |
schestowitz | Which person/address should I contact? I can see this link, which explains some of it, but it's in French: http://eur-lex.europa.eu/Lex... | Mar 06 16:49 |
paramahamsa | schestowitz: not quite random, but you seem to be throwing links out faster than I can read the pages! | Mar 06 16:49 |
zoobab01 | DG enterprise & DG Infso are the one organising the ESS | Mar 06 16:50 |
schestowitz | What's their E-mail addy? | Mar 06 16:50 |
schestowitz | Maybe we can pull copies of the documents | Mar 06 16:50 |
schestowitz | It makes it hard for Zuck et al to toss rubbish at em | Mar 06 16:50 |
schestowitz | Transparency for open source panels, no? :-) | Mar 06 16:50 |
schestowitz | I spoke to Zuck (and trapped him) too, but I doubt he wants to talk further, for obvious reasons. | Mar 06 16:51 |
zoobab01 | mmh | Mar 06 16:53 |
zoobab01 | European Software Strategy is a closed forum | Mar 06 16:54 |
zoobab01 | that's the answer you will get from the COmmission | Mar 06 16:54 |
schestowitz | haha :-) | Mar 06 16:54 |
zoobab01 | but it is good to show the light on the darkness | Mar 06 16:54 |
schestowitz | Close forum on FOSS. | Mar 06 16:54 |
schestowitz | *Closed | Mar 06 16:54 |
zoobab01 | and get an official answer | Mar 06 16:54 |
schestowitz | Who would you contact? | Mar 06 16:54 |
schestowitz | To have it written down that they refuse access would be valuable | Mar 06 16:55 |
schestowitz | Like you guys did re: ACTA | Mar 06 16:55 |
zoobab01 | wait | Mar 06 16:55 |
schestowitz | This gets people more angry | Mar 06 16:55 |
schestowitz | Then it gets those behind closed doors uncomfortable/nervous | Mar 06 16:55 |
zoobab01 | you can request only the ACT submission | Mar 06 16:55 |
zoobab01 | because it was leaked | Mar 06 16:55 |
schestowitz | Like the BRM... | Mar 06 16:55 |
zoobab01 | so if they refuse, it does not matter so much | Mar 06 16:56 |
zoobab01 | if they release the contrib, you can ask for more | Mar 06 16:56 |
schestowitz | Who can this be requested from? | Mar 06 16:56 |
schestowitz | I see names in the docs. | Mar 06 16:56 |
schestowitz | Female ones too | Mar 06 16:56 |
schestowitz | Are those like panel mods? | Mar 06 16:56 |
schestowitz | The forces of evil make a move: http://209.85.229.132/search?q=cache:... | Mar 06 17:13 |
schestowitz | DS Lite to be killed off < http://www.theinquirer.net/inquire... > for the better :-) | Mar 06 17:15 |
schestowitz | Let's hope that Big Blue has no intent of associating itself with corruptions: < http://www.theinquirer.net/inquirer/new... > (IBM set to buy Satyam) | Mar 06 17:17 |
schestowitz | Earthquake felt across Melbourne < http://www.itwire.com/content... > | Mar 06 17:20 |
schestowitz | No wonder Mahalo has been accused of spamming: http://www.webpronews.com/topnews/20... | Mar 06 17:22 |
schestowitz | I notice links there to Net Applications, which is not statistics; ti's junk science and it's sponsored by Apple and Microsoft. | Mar 06 17:26 |
MinceR | it's the Intelligent Design of market share :) | Mar 06 17:26 |
schestowitz | On the Eighth Day, God gave us Windows Vista. Hallelujah. | Mar 06 17:27 |
schestowitz | http://news.bbc.co.uk/1/hi/technolo... | Mar 06 17:28 |
schestowitz | MSBBC goes on a Wikipedia rant: http://news.bbc.co.uk/1/hi/uk_pol... | Mar 06 17:29 |
schestowitz | There are some more obscure companies here < http://au.sys-con.com/node/863852/... > which claim to deliver Web surveys | Mar 06 17:31 |
PetoKraus | hehe, want to see a bit of statistics? | Mar 06 17:31 |
PetoKraus | http://pk.gjhak.sk/old/.s... | Mar 06 17:31 |
schestowitz | Hehe. http://www.statcounter.com/microsoft_office_live/ (StatCounter provides free customisable hit counters, visitor tracking and website stats for Microsoft Office Live). Keep it where it's warm.. | Mar 06 17:32 |
schestowitz | PetoKraus: whose class? | Mar 06 17:33 |
PetoKraus | my | Mar 06 17:33 |
PetoKraus | chemistry | Mar 06 17:33 |
PetoKraus | it's just one 20 marks class test | Mar 06 17:33 |
PetoKraus | i though i was rubbish when I had only 14.5 out of twenty | Mar 06 17:34 |
PetoKraus | but a wee bit of statistics | Mar 06 17:34 |
PetoKraus | and i'm feeling better :D | Mar 06 17:34 |
schestowitz | Beware stats in general.. | Mar 06 17:34 |
schestowitz | Especially when the dataset is too diverse and large | Mar 06 17:34 |
PetoKraus | not if you are the one making them :D | Mar 06 17:34 |
schestowitz | These stats 'companies' have a business model | Mar 06 17:34 |
schestowitz | Giving truth is rarely their bsuiness model | Mar 06 17:34 |
schestowitz | They admit it's biased | Mar 06 17:35 |
schestowitz | Net Applications for example | Mar 06 17:35 |
schestowitz | They also change them after publication | Mar 06 17:35 |
schestowitz | So buy consulting from then... they might do some 'magic' | Mar 06 17:35 |
schestowitz | Or give them your biased (population wise) logs | Mar 06 17:35 |
schestowitz | Rome meeting snubs intelligent design, creationism < http://hosted.ap.org/dynamic/stories/E/E... > | Mar 06 17:37 |
PetoKraus | schestowitz: sure | Mar 06 17:37 |
PetoKraus | but again | Mar 06 17:37 |
schestowitz | 'The foundation has criticized intelligent design in the past and says on its Web site that it doesn't support any research or programs that "deny large areas of well-documented scientific knowledge."' | Mar 06 17:38 |
PetoKraus | if you are the one making the statistics | Mar 06 17:38 |
PetoKraus | you don't have to fear :D | Mar 06 17:38 |
PetoKraus | you are manipulating these numbers any way you want | Mar 06 17:38 |
PetoKraus | so... ;) | Mar 06 17:38 |
schestowitz | Happy with ours :-) | Mar 06 17:38 |
schestowitz | BN: half the world uses Linux | Mar 06 17:38 |
schestowitz | Ms: "Just the Facts" (/WHICH/ facts?) | Mar 06 17:39 |
schestowitz | Here's why Digg is broken: USocial CEO: 'We're gaming Digg' < http://latimesblogs.latimes.com/techn... >. It would be better if Digg just died and emitted people to sites that are not gamed. I hardly contribute to Digg anymore. | Mar 06 17:45 |
*_Hicham_ (n=hicham@41.249.17.92) has joined #boycottnovell | Mar 06 17:59 | |
schestowitz | http://www.linuxtoday.com/news_story.php3?... "A recent article showed a presentation pie chart by MS CEO Balmer, who stated Linux was a bigger threat than the MAC, depicted Linux with a slightly bigger market share than Apple's 10%." | Mar 06 18:02 |
MinceR | if apple's is 10%, that is | Mar 06 18:03 |
schestowitz | Yeah :-D | Mar 06 18:03 |
schestowitz | In the US maybe. | Mar 06 18:03 |
schestowitz | US=world | Mar 06 18:04 |
*MinceR is offworld | Mar 06 18:04 | |
_Hicham_ | it has a larger market share | Mar 06 18:04 |
_Hicham_ | otherwise, they wouldn't have done the "Get the Facts" Campaign | Mar 06 18:05 |
MinceR | Get the FUD | Mar 06 18:07 |
schestowitz | http://www.linuxtoday.com/news_story.ph... "Seti@home (the most popular project) has Linux-client numbers above 10%. " | Mar 06 18:08 |
schestowitz | > MinceR is offworld | Mar 06 18:10 |
schestowitz | Slipped off the wedge of the disc? | Mar 06 18:11 |
MinceR | went through the Gate | Mar 06 18:11 |
MinceR | "door to heaven" | Mar 06 18:11 |
PetoKraus | how's the stairway? | Mar 06 18:12 |
schestowitz | Money over freedom at Sun: http://weblog.infoworld.com/openre... | Mar 06 18:12 |
*zer0c00l has quit (Remote closed the connection) | Mar 06 18:13 | |
_Hicham_ | freedom first | Mar 06 18:13 |
schestowitz | SUSE is slowing down: http://arstechnica.com/open-source/... | Mar 06 18:13 |
PetoKraus | oh god | Mar 06 18:15 |
PetoKraus | (18:19:36) Scherff, Aline: yea im absolutely gutted...my supervisor wanted the data finished so he can look at it over the weekend and the lab will prolly not be available on mon...and tues im never in...and wed is the deadline | Mar 06 18:15 |
PetoKraus | (18:19:46) Peter Kraus: well | Mar 06 18:15 |
PetoKraus | (18:19:57) Peter Kraus: do you HAVE to do it in the lab? | Mar 06 18:15 |
PetoKraus | (18:19:58) Peter Kraus: i mean | Mar 06 18:15 |
PetoKraus | (18:20:09) Peter Kraus: from what i know, it's a computer program | Mar 06 18:16 |
PetoKraus | (18:20:42) Scherff, Aline: and its licenced and costs like a fortune plus it only runs on hi end pcs | Mar 06 18:16 |
MinceR | can the lab be accessed over the net? | Mar 06 18:16 |
PetoKraus | it's psychology lab | Mar 06 18:17 |
PetoKraus | running on windows machines | Mar 06 18:17 |
PetoKraus | with novell zenworks | Mar 06 18:17 |
PetoKraus | the short answer is: unless you hack it | Mar 06 18:17 |
PetoKraus | which shouldn't be that hard | Mar 06 18:17 |
PetoKraus | but even 1st year computing science students don't get SSH access to the uni | Mar 06 18:17 |
schestowitz | No wonder it's a psychology lab.. | Mar 06 18:17 |
Balrog | hello everyone | Mar 06 18:17 |
schestowitz | :-) | Mar 06 18:17 |
Balrog | here every cis student gets SSH access | Mar 06 18:17 |
Balrog | :) | Mar 06 18:18 |
PetoKraus | I KNOW | Mar 06 18:18 |
schestowitz | SSHISH! | Mar 06 18:18 |
Balrog | not windows remote login, though :) | Mar 06 18:18 |
PetoKraus | i'm really disappointed at the state of software in my uni | Mar 06 18:20 |
Balrog | explain how it is? | Mar 06 18:20 |
*zer0c00l (n=zer0c00l@210.212.255.131) has joined #boycottnovell | Mar 06 18:21 | |
PetoKraus | well | Mar 06 18:21 |
PetoKraus | the wireless for instance. I couldn't connect to the unsecured wireless today with wicd | Mar 06 18:21 |
Balrog | here wireless sucks | Mar 06 18:21 |
PetoKraus | but that might be just some misconfig on my side | Mar 06 18:21 |
PetoKraus | but anyway | Mar 06 18:21 |
Balrog | but it's the buildings | Mar 06 18:21 |
PetoKraus | it's unsecured, but! you have to use vpn | Mar 06 18:22 |
PetoKraus | which wouldn't be a problem if they used normal vpn | Mar 06 18:22 |
Balrog | not here ... you use captive-portal | Mar 06 18:22 |
Balrog | (around here) | Mar 06 18:22 |
trmanco | schestowitz, posting on identi.ca eh! :-P | Mar 06 18:22 |
Balrog | just a login page | Mar 06 18:22 |