00:07.05 | *** join/#openmoko-cdevel pabs3 (~pabs@2001:470:5:b71:a83f:ee9b:5b8b:d244) |
02:39.05 | *** join/#openmoko-cdevel pabs3 (~pabs@2001:470:5:b71:a83f:ee9b:5b8b:d244) |
04:06.26 | *** join/#openmoko-cdevel jackh (~quassel@27.115.44.230) |
04:47.23 | *** join/#openmoko-cdevel GNUtoo-netbook (~gnutoo@host254-162-dynamic.1-79-r.retail.telecomitalia.it) |
05:13.04 | *** join/#openmoko-cdevel Martix (~martix@ip-62-245-123-103.net.upcbroadband.cz) |
07:12.10 | *** join/#openmoko-cdevel GNUtoo-hplaptop (~GNUtoo@host254-162-dynamic.1-79-r.retail.telecomitalia.it) |
07:24.13 | *** join/#openmoko-cdevel ao2 (~ao2@2001:1418:117::1) |
07:38.17 | *** join/#openmoko-cdevel plotr (~name@unaffiliated/plotr) |
08:09.29 | *** join/#openmoko-cdevel ThibG (~ThibG@spike.sitedethib.com) |
08:14.53 | *** join/#openmoko-cdevel plotr1 (~name@178.185.46.106) |
08:32.03 | *** join/#openmoko-cdevel [Rui] (~rms@a85-138-36-247.cpe.netcabo.pt) |
08:34.31 | [Rui] | heya |
08:35.36 | [Rui] | I'm having a few problems trying to boot the current shr image, I added the following to append-GTA02 in hopes it would help but... not really... glamo_mci.sd_max_clk=500000 glamo_mci.sd_drive=3 |
08:36.01 | [Rui] | it always hangs with mmcblk0 error -16 requesting status |
08:36.19 | [Rui] | this sd card had a previous image from january I think |
08:38.49 | [Rui] | I also tried the aurora image, but all I had was a white screen |
08:40.33 | [Rui] | with glamo_mci.sd_max_clk=1000000 it looks even worse |
09:18.04 | *** join/#openmoko-cdevel mirage335 (~mirage335@mirage335.dyndns.org) |
09:24.16 | *** join/#openmoko-cdevel tg (~tg@irc.tgbit.net) |
09:41.31 | *** join/#openmoko-cdevel Martix (~martix@21.74.broadband2.iol.cz) |
09:46.56 | *** join/#openmoko-cdevel GarthPS (~quassel@2a01:e35:2f5c:e670:2400:fb59:280:ea21) |
10:55.25 | *** join/#openmoko-cdevel NIN101 (~NIN@p5DD291C0.dip0.t-ipconnect.de) |
12:58.51 | *** join/#openmoko-cdevel ttyS3 (~o@gateway/tor-sasl/omlib) |
13:28.59 | *** join/#openmoko-cdevel paulk-desktop (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
13:35.46 | *** join/#openmoko-cdevel GNUtoo-desktop (~GNUtoo@host254-162-dynamic.1-79-r.retail.telecomitalia.it) |
13:37.00 | *** join/#openmoko-cdevel morphis (~morphis@dslb-092-076-169-021.pools.arcor-ip.net) |
13:39.24 | morphis | GNUtoo-desktop: ping |
13:39.30 | GNUtoo-desktop | pong |
13:41.42 | morphis | just read your mail |
13:41.50 | morphis | you got any further with the camera |
13:42.11 | GNUtoo-desktop | look in #replicant.... |
13:42.31 | morphis | ah |
13:51.40 | *** join/#openmoko-cdevel DocScrutinizer (~halley@openmoko/engineers/joerg) |
13:54.25 | *** join/#openmoko-cdevel DocScrutinizer05 (~HaleBopp@openmoko/engineers/joerg) |
13:54.32 | CIA-69 | freesmartphone.org: 03morphis 07utilities * ra8dd68fe306b 10/android/brcm_patchram_plus/ (11 files in 2 dirs): android: add brcm_patchram_plus utilitiy need to initialize bluetooth |
14:00.57 | CIA-69 | freesmartphone.org: 03morphis 07utilities * r69371f6ecc82 10/android/brcm_patchram_plus/configure.ac: android: brcm_patchram_plus: bump version to 4.1.1 as it's from this android version |
14:18.40 | *** join/#openmoko-cdevel jonwil (~jonwil@27-33-137-199.static.tpgi.com.au) |
15:50.43 | *** join/#openmoko-cdevel ttyS3 (~o@gateway/tor-sasl/omlib) |
15:51.09 | *** join/#openmoko-cdevel DocScrutinizer05 (~HaleBopp@openmoko/engineers/joerg) |
15:55.23 | *** join/#openmoko-cdevel DocScrutinizer05 (~HaleBopp@openmoko/engineers/joerg) |
16:23.28 | *** join/#openmoko-cdevel pabs3 (~pabs@2001:470:5:b71:a83f:ee9b:5b8b:d244) |
16:55.38 | *** join/#openmoko-cdevel pabs3 (~pabs@2001:470:5:b71:a83f:ee9b:5b8b:d244) |
17:59.07 | *** join/#openmoko-cdevel pabs3 (~pabs@2001:470:5:b71:a83f:ee9b:5b8b:d244) |
18:01.03 | *** join/#openmoko-cdevel chomwitt (~chomwitt@ppp-94-64-45-176.home.otenet.gr) |
19:32.22 | nully | Could i please have a trac account? this is getting very old... fast. |
19:35.36 | DocScrutinizer05 | nully: hm? |
19:36.21 | nully | shr-project.ort.. accounts are by request only... ive been asking for almost a week now. |
19:36.40 | DocScrutinizer05 | aaah shr |
19:37.09 | DocScrutinizer05 | no idea who's managing that |
19:37.15 | nully | no one does |
19:37.37 | PaulFertser | mrmoku: should know for sure |
19:37.39 | nully | ... and this is why the registration should not have been disabled. IMHO. |
19:38.14 | PaulFertser | Bad spammers were too bad... |
19:38.16 | DocScrutinizer05 | o/ PaulFertser |
19:38.21 | PaulFertser | hey :) |
19:39.00 | DocScrutinizer05 | wonders if yawning can be fatal in itself |
19:40.14 | nully | anyways: account plz&thx. |
19:41.06 | *** join/#openmoko-cdevel pabs3 (~pabs@2001:470:5:b71:a83f:ee9b:5b8b:d244) |
19:41.29 | PaulFertser | nully: are you always here, how to contact you? |
19:41.32 | DocScrutinizer05 | anyways nfc/bye |
19:41.45 | PaulFertser | DocScrutinizer05: good night! |
19:45.48 | *** join/#openmoko-cdevel ThibG (~ThibG@spike.sitedethib.com) |
19:52.00 | *** join/#openmoko-cdevel NIN101 (~NIN@p5DD28A04.dip0.t-ipconnect.de) |
20:08.38 | jake42 | nully: JaMa and mrmoko (and others?) have trac admins access |
20:11.16 | nully | PaulFertser: Always here, and on the mailing list. Thanks for caring :) |
20:11.51 | nully | jake42: Yeah... But no response sooooooooo I complain in public spaces. |
20:12.15 | nully | Maybe we need more people with trac admin access? |
20:12.33 | jake42 | nully: yes, that would be a good start |
20:14.02 | nully | A trac plugin is them answer... Im willing to look at it, but im on the road this weekend... But it seems like it must be written... as the alternitive is not working. |
20:14.21 | nully | s/them/the\ |
20:15.24 | jake42 | a few days sooner or later don't matter |
20:15.56 | nully | Never know :) there could be confused users who dont irc that no one knows about :) |
20:16.14 | jake42 | shr is really slow going atm |
20:16.30 | jake42 | what devices are you using with shr? |
20:16.32 | jake42 | btw |
20:17.22 | nully | 2 gta01s a n900 and any other phone that walks into our freegeek that might work. |
20:18.14 | jake42 | is a recent shr version running on gta01? |
20:22.49 | nully | One of them, yes :) But i didnt do much testing past placing a call... this was about 8 months ago... the other still runs an older version. |
20:24.21 | jake42 | I'm already amazed it's running at all :) |
20:25.21 | nully | SHR is good stuff :) |
20:26.12 | jake42 | we need more devs though |
20:26.49 | *** join/#openmoko-cdevel GNUtoo-desktop (~GNUtoo@host254-162-dynamic.1-79-r.retail.telecomitalia.it) |
20:27.09 | nully | Build it, and they will come. (so they tell me) |
20:28.49 | jake42 | they're probably right |
20:29.14 | jake42 | first shr needs to get more stable, then we'll see |
20:30.35 | nully | One step at a time. |
20:30.45 | jake42 | morphis is doing good work on fso |
20:30.59 | nully | Im on the get a trac wiki account sep =P |
20:31.32 | nully | s/sep/step/ |
20:32.18 | jake42 | not until recently I tried to stop spam by getting some bearstech admin to enable some antispam trac plugins |
20:33.05 | GNUtoo-desktop | hmmm I don't know who's admin.... |
20:33.08 | jake42 | they didn't help much and it was decided to turn the registration off |
20:33.35 | nully | ya... thoes confused me when i was doing research... they seem to not offer any traditional methods for anti spam :/ |
20:34.40 | jake42 | which is really confusing – yes |
20:35.00 | nully | is still confuzzled about trac not supporting such basic ideas. |
20:35.26 | nully | Iz okay: thats why we use free software. |
20:35.39 | jake42 | we should switch to redmine or something ;) |
20:36.02 | nully | Trac is okay... I love the doxygen plugin.OM |
20:37.20 | nully | Anyways i gotta run... Ill be here, awaiting an account :) |
20:37.42 | nully | Thanks for your help jake42 i appreciate it. |
20:37.45 | jake42 | yw |
20:37.51 | jake42 | just be patient :) |
20:38.15 | GNUtoo-desktop | I'm not admin I think |
20:38.21 | GNUtoo-desktop | so I cannot add accounts |
20:38.36 | GNUtoo-desktop | mrmoku, is gone for good it seems |
20:38.43 | GNUtoo-desktop | maybe JaMa has admin |
20:38.49 | GNUtoo-desktop | but he's not on IRC |
20:42.12 | nully | patients is a virtue :) this is my first response to my request :) |
21:03.03 | *** join/#openmoko-cdevel nschle85 (~kvirc@178-27-189-81-dynip.superkabel.de) |
21:04.16 | nschle85 | GNUtoo-desktop: hello, i have a short question: is the mali 400 GPU fully supported by open source drivers ? (including 2d/3d accell) ? |
21:04.57 | nschle85 | GNUtoo-desktop: or in other words, do i need properiatry drivers for mali 400 ? |
21:07.04 | GNUtoo-desktop | nschle85, I don't think it's fully supported but there is a project for reverse engineering it, you should look at the status |
21:07.21 | GNUtoo-desktop | last time I looked they were still in the reversing phase |
21:07.36 | GNUtoo-desktop | that is to say no free driver was written but only test-cases |
21:08.05 | GNUtoo-desktop | https://gitorious.org/lima |
21:09.06 | nschle85 | GNUtoo-desktop: ok thank you |
21:09.17 | GNUtoo-desktop | nschle85, it's unclear if I should continue to spend time on the n900 or not.... |
21:10.41 | nschle85 | GNUtoo-desktop: are your changes available upstream ? |
21:11.18 | GNUtoo-desktop | yes of course |
21:11.24 | GNUtoo-desktop | they are fully pushed in branches |
21:11.35 | GNUtoo-desktop | but since JaMa doesn't care about telephony |
21:11.45 | GNUtoo-desktop | and that you don't seem to care anymore about n900.... |
21:12.36 | GNUtoo-desktop | btw here's a summ-up of all reverse enginering project for embedded graphics drivers: |
21:12.39 | nschle85 | GNUtoo-desktop: makes it sense to merge your changes upstream ? |
21:12.41 | GNUtoo-desktop | http://www.phoronix.com/scan.php?page=news_item&px=MTE0MDA |
21:12.47 | GNUtoo-desktop | no |
21:12.50 | GNUtoo-desktop | it's not ready yet |
21:13.04 | GNUtoo-desktop | there is no mic handling |
21:13.13 | GNUtoo-desktop | that is to say the remote can't ear you |
21:14.54 | nschle85 | GNUtoo-desktop: is it possible to fix mic handling and merge that result and fix audio forwarder based on your merged results ? |
21:15.50 | GNUtoo-desktop | yes of course it's possible |
21:15.55 | GNUtoo-desktop | someone just have to do it |
21:16.10 | nschle85 | GNUtoo-desktop: or do you mean mic handling is missing in audio forwarder ? or scenarios ? |
21:16.24 | GNUtoo-desktop | I don't record any sound |
21:16.27 | GNUtoo-desktop | so it can't work |
21:16.37 | GNUtoo-desktop | if I add code for recording sound |
21:16.41 | GNUtoo-desktop | it would probably work |
21:17.01 | *** part/#openmoko-cdevel GNUtoo-desktop (~GNUtoo@host254-162-dynamic.1-79-r.retail.telecomitalia.it) |
21:17.15 | *** join/#openmoko-cdevel GNUtoo-desktop (~GNUtoo@host254-162-dynamic.1-79-r.retail.telecomitalia.it) |
21:17.22 | GNUtoo-desktop | oops sorry |
21:17.24 | GNUtoo-desktop | wrong window |
21:19.32 | nschle85 | GNUtoo-desktop: you are talking about audio forwarder ? |
21:19.42 | GNUtoo-desktop | yes |
21:19.54 | GNUtoo-desktop | n900 audio forwarder |
21:21.03 | nschle85 | but why not merge and let implement recording audio implement later ? |
21:21.40 | GNUtoo-desktop | first it can't even work the way you want |
21:21.45 | GNUtoo-desktop | it depend on beeing realtime |
21:21.56 | GNUtoo-desktop | and that must be added in the fsoaudiod service |
21:22.04 | GNUtoo-desktop | *service file from systemd |
21:22.43 | nschle85 | GNUtoo-desktop: ?? pardon ? |
21:23.16 | GNUtoo-desktop | the code needs to be able to run realtime |
21:23.22 | GNUtoo-desktop | systemd prevent that |
21:23.31 | GNUtoo-desktop | unless you disable in a config file |
21:23.45 | GNUtoo-desktop | or permit it in the .service file for fsoaudiod |
21:24.38 | nschle85 | GNUtoo-desktop: i dont understand ? why realtime ?? is it polling something ? |
21:24.44 | GNUtoo-desktop | yes |
21:24.57 | GNUtoo-desktop | else it's too late |
21:25.01 | GNUtoo-desktop | and it loose buffers |
21:25.37 | nschle85 | GNUtoo-desktop: is there no possibility of async io ? |
21:26.43 | nschle85 | GNUtoo-desktop: thats ther reason why i wanted to get warm with huawei 177 and async io programming in vala |
21:27.55 | GNUtoo-desktop | I tried async vala stuff first |
21:28.02 | GNUtoo-desktop | it didn't work |
21:28.21 | GNUtoo-desktop | the only way was to use the same way than the pulse plugin |
21:28.35 | nschle85 | GNUtoo-desktop: how is the ofono stuff implmented ? its also polling ? |
21:29.00 | GNUtoo-desktop | I just copied what the pulseaudio plugin did |
21:29.20 | GNUtoo-desktop | maybe look at the code |
21:30.08 | GNUtoo-desktop | http://git.freesmartphone.org/?p=cornucopia.git;a=blob;f=fsoaudiod/src/plugins/gsmvoice_alsa_cmtspeechdata/cmthandler.vala;h=943fdaae88b65d726a22837f126d753805dfdbfc;hb=870f94138f97223ec06eca67b06ad58a4a87d066#l133 |
21:37.33 | nschle85 | GNUtoo-desktop: wow its really polling |
21:39.10 | nschle85 | GNUtoo-desktop: polling may have some advantages in general: the polling device cannot be overdriven by async events... |
21:39.35 | GNUtoo-desktop | it's the only way that works |
21:39.42 | GNUtoo-desktop | and note that epoll is rather efficent |
21:39.51 | GNUtoo-desktop | despite of its name |
21:40.20 | nschle85 | what kind of file is the fds ? |
21:42.17 | CIA-69 | freesmartphone.org: 03GNUtoo 07cornucopia * r058d37d54095 10/fsodeviced/conf/nokia_n900/alsa-default/ (headset stereoout): (log message trimmed) |
21:42.17 | CIA-69 | freesmartphone.org: fsodeviced: fix the nokia N900 stereoout and headset scenarios. |
21:42.17 | CIA-69 | freesmartphone.org: Without that fix fsodeviced will refuses to start and fails with that message: |
21:42.17 | CIA-69 | freesmartphone.org: fsodeviced: control.c:1680: snd_ctl_elem_list_get_id: Assertion `idx < obj->used' failed. |
21:42.17 | CIA-69 | freesmartphone.org: That is because the 'Media Volume' control is not a hardware control but a software one |
21:42.17 | CIA-69 | freesmartphone.org: (created by /etc/asound.conf): it appears only after playing some sound. |
21:42.18 | CIA-69 | freesmartphone.org: And so when fsodeviced starts no sound has been played yet, so it's not there, |
21:42.39 | GNUtoo-desktop | nschle85, file descriptors |
21:42.57 | nschle85 | GNUtoo-desktop: i know but what is behind it ? |
21:43.20 | GNUtoo-desktop | the kernel |
21:43.41 | nschle85 | GNUtoo-desktop: its a serial device ? |
21:43.46 | GNUtoo-desktop | https://en.wikipedia.org/wiki/File_descriptor |
21:44.04 | GNUtoo-desktop | it's a number |
21:44.09 | GNUtoo-desktop | like 0,1,2,3,4,... |
21:47.25 | nschle85 | GNUtoo-desktop: a select on this file descriptor does not work ? |
21:48.15 | GNUtoo-desktop | I don't remember |
21:48.35 | GNUtoo-desktop | anyway pulseaudio does poll |
21:48.37 | GNUtoo-desktop | so we do poll |
21:48.44 | GNUtoo-desktop | unless you want something not working well |
21:48.55 | GNUtoo-desktop | and in fact it's rt_poll |
21:49.12 | nschle85 | GNUtoo-desktop: no if select works we should use select |
21:49.20 | GNUtoo-desktop | why? |
21:49.42 | nschle85 | GNUtoo-desktop: you cannot miss any buffer |
21:49.56 | GNUtoo-desktop | yes |
21:50.13 | GNUtoo-desktop | and why can I miss buffers with rtpoll? |
21:51.15 | nschle85 | GNUtoo-desktop: hmm... you are polling |
21:52.40 | GNUtoo-desktop | it's epoll |
21:52.41 | nschle85 | GNUtoo-desktop: if you use select you have to implement some logic not do overdrive the phone with audio stuff but |
21:53.18 | GNUtoo-desktop | wikipedia says: |
21:53.19 | GNUtoo-desktop | epoll is a scalable I/O event notification mechanism for Linux |
21:53.31 | GNUtoo-desktop | https://en.wikipedia.org/wiki/Epoll |
21:53.47 | GNUtoo-desktop | forget about its name |
21:53.52 | GNUtoo-desktop | just look at the article |
21:54.06 | DocScrutinizer05 | POLL(2) Linux Programmer's Manual POLL(2) NAME |
21:54.08 | DocScrutinizer05 | <PROTECTED> |
21:54.19 | DocScrutinizer05 | poll is not "polling" |
21:54.22 | GNUtoo-desktop | indeed |
21:54.29 | GNUtoo-desktop | that's what I am trying to explain to him |
21:54.36 | DocScrutinizer05 | poll() is a wait on multiple events |
21:54.38 | GNUtoo-desktop | you get notified by the kernel |
21:55.18 | DocScrutinizer05 | SEE ALSO |
21:55.20 | DocScrutinizer05 | <PROTECTED> |
21:55.52 | nschle85 | GNUtoo-desktop: is the ofono code using epoll ? |
21:56.27 | GNUtoo-desktop | nschle85, it's not ofono, the *pulseaudio plugin* uses rt_poll from pulseaudio wich put the poll in realtime |
21:56.55 | DocScrutinizer05 | DESCRIPTION |
21:56.56 | DocScrutinizer05 | <PROTECTED> |
21:57.24 | GNUtoo-desktop | ah I used poll instead of epoll it seems |
21:57.29 | GNUtoo-desktop | anyway.... |
21:57.47 | GNUtoo-desktop | I guess it's ok the same |
22:01.06 | nschle85 | GNUtoo-desktop: sure ? |
22:01.21 | GNUtoo-desktop | epoll has better performance |
22:01.26 | GNUtoo-desktop | but I don't monitor that much fds |
22:02.24 | GNUtoo-desktop | ah I've buffer underruns because of the high CPU usage.... |
22:04.03 | nschle85 | GNUtoo-desktop: DocScrutinizer05: hmm i am confused. what is the solution using select(epoll) ? |
22:04.43 | DocScrutinizer05 | solution for what? |
22:05.04 | nschle85 | DocScrutinizer05: missing buffers |
22:05.30 | DocScrutinizer05 | both select() and poll() wait for one of several async system calls "returning" |
22:05.51 | DocScrutinizer05 | I don't know of any missing buffers |
22:06.49 | DocScrutinizer05 | nschle85: the point is: for mic->modem you need to use modem as clock-source |
22:07.14 | *** join/#openmoko-cdevel Wuzzi (5cc43224@gateway/web/freenode/ip.92.196.50.36) |
22:07.16 | DocScrutinizer05 | and this clock-source is via special timing adjust messages the modem sends |
22:07.33 | GNUtoo-desktop | I'm not there yet |
22:07.37 | GNUtoo-desktop | I'm modem->alsa |
22:07.42 | GNUtoo-desktop | *I'm at |
22:08.01 | DocScrutinizer05 | this should be pretty straightforward task |
22:08.07 | GNUtoo-desktop | yes |
22:08.17 | GNUtoo-desktop | if done correctly tough |
22:08.43 | DocScrutinizer05 | well, probably vala doesn't help to do it correctly |
22:08.51 | Wuzzi | GNUtoo-desktop: not sure it helps, but there is a epoll patch "Prevent creating circular epoll structures" that isn't in n900 kernel. |
22:09.19 | GNUtoo-desktop | ok |
22:09.29 | GNUtoo-desktop | DocScrutinizer05, indeed |
22:10.06 | GNUtoo-desktop | DocScrutinizer05, in what use cases epoll is more performant? |
22:11.36 | Wuzzi | GNUtoo-desktop: did you see "tlv320aic3x: Fix codec pll configure bug" patch? |
22:11.44 | GNUtoo-desktop | no |
22:12.42 | Wuzzi | GNUtoo-desktop: http://davinci-linux-open-source.1494791.n2.nabble.com/PATCH-V3-1-1-ASoC-tlv320aic3x-Fix-codec-pll-configure-bug-td7580116.html (look for patch v1-v2, too) |
22:13.00 | Wuzzi | might be relatet to audio issues. |
22:13.40 | GNUtoo-desktop | ok |
22:13.44 | GNUtoo-desktop | but audio wokrs fine |
22:14.06 | GNUtoo-desktop | the problem is the modem<->alsa forwarder |
22:14.11 | GNUtoo-desktop | and the modem is cmt_speech |
22:14.28 | Wuzzi | and that epoll thing http://lkml.org/lkml/2011/2/12/108 |
22:14.37 | Wuzzi | ok |
22:15.11 | Wuzzi | was all i had to say. good night. |
22:16.05 | DocScrutinizer05 | GNUtoo-desktop: I heard of epoll first time right now, no idea if and how much better it is over poll() |
22:16.19 | GNUtoo-desktop | ok |
22:16.24 | DocScrutinizer05 | to me it seems it's more convenient to use maybe |
22:16.35 | DocScrutinizer05 | but possibly with more overhead |
22:17.12 | nschle85 | DocScrutinizer05: so poll() and epoll() has same implementation ? |
22:17.31 | DocScrutinizer05 | errr, I wouldn't think so |
22:17.54 | DocScrutinizer05 | probably epoll using poll |
22:19.10 | DocScrutinizer05 | also: >>The epoll API is Linux-specific. Some other systems provide similar mechanisms, for example, FreeBSD has kqueue, and Solaris has /dev/poll.<< |
22:19.23 | DocScrutinizer05 | while poll probably is even POSIX |
22:20.27 | DocScrutinizer05 | DESCRIPTION |
22:20.28 | DocScrutinizer05 | <PROTECTED> |
22:20.30 | DocScrutinizer05 | <PROTECTED> |
22:20.39 | GNUtoo-desktop | isn't it the other way arround? poll() using epoll() for performances reasons? |
22:21.15 | DocScrutinizer05 | nope, evidently epoll is more complex and feature-rich than poll |
22:22.53 | nschle85 | GNUtoo-desktop: DocScrutinizer05: iam a java programmer and what i would like to prefer is something like select(). why does it not work with the modem ? |
22:24.01 | DocScrutinizer05 | nschle85: poll can wait for other events than just file-io aiui |
22:24.11 | GNUtoo-desktop | http://www.xmailserver.org/linux-patches/nio-improve.html |
22:24.21 | GNUtoo-desktop | what about the graphs here ^^^ |
22:25.51 | DocScrutinizer05 | nschle85: >>poll() performs a similar task to select(2): it waits for one of a set of file descriptors to become ready to perform I/O.<< so I have no idea what are the delicate differences |
22:29.28 | DocScrutinizer05 | nschle85: http://www.unixguide.net/network/socketfaq/2.14.shtml |
22:30.43 | nschle85 | DocScrutinizer05: and so i dont understand why poll is used instead of select |
22:30.59 | DocScrutinizer05 | yup |
22:31.14 | DocScrutinizer05 | probably select is the even better option here |
22:34.30 | nschle85 | DocScrutinizer05: ok so ill try to port to use a select statement |
22:38.32 | nschle85 | GNUtoo-desktop: in case of performance, i would like to prefer select. but someone needs to implements protection against overdriving the whole system |
22:44.21 | nschle85 | GNUtoo-desktop: DocScrutinizer05: ill leave now... good night |
22:44.34 | DocScrutinizer05 | sorry been afk |
22:44.47 | GNUtoo-desktop | see you later |
22:44.59 | DocScrutinizer05 | nshfine (try use select). And good night :-) |
22:45.33 | DocScrutinizer05 | yay, he must have had a need for bathroom or sth |
22:46.39 | nschle85 | DocScrutinizer05: GNUtoo-desktop: bye :-) ??? what are you talking about bathroom ? |
22:47.04 | *** join/#openmoko-cdevel ttyS3 (~o@gateway/tor-sasl/omlib) |
22:47.07 | DocScrutinizer05 | heh |
22:47.16 | DocScrutinizer05 | nschle85: I'm a fool |
22:47.29 | DocScrutinizer05 | tried to tab-expand nsh |
22:47.55 | DocScrutinizer05 | so I thought you already left chan |
22:48.12 | DocScrutinizer05 | ;-) |
22:48.53 | *** join/#openmoko-cdevel pabs3 (~pabs@2001:470:5:b71:a83f:ee9b:5b8b:d244) |
22:49.38 | DocScrutinizer05 | nschle85: (using select) sure thing. If you got alternatives, then the simplest one is usually the best one. at least performance wise |
22:49.39 | nschle85 | DocScrutinizer05: sorry i cannot follow so bye |
22:50.15 | DocScrutinizer05 | [2012-07-15 00:44:59] <DocScrutinizer05> nsh<tab>fine (try use select). And good night :-) |
22:50.39 | DocScrutinizer05 | thinks "oh, nshle must have left the channel already |
22:50.50 | DocScrutinizer05 | nschle85: now all clear? :-) |
22:51.22 | DocScrutinizer05 | never mind, good night pal :-) |
22:53.45 | nschle85 | DocScrutinizer05: no , you are confusing me much more ... but its ok... bye |
22:54.23 | *** part/#openmoko-cdevel nschle85 (~kvirc@178-27-189-81-dynip.superkabel.de) |
22:58.26 | GNUtoo-desktop | I'll try to int err = sleep(me,timeout); if it fails I'll be back |
22:58.32 | GNUtoo-desktop | bye |