Back to Home / #mythtv / 2003 / 07 / Prev Day | Next Day
#mythtv IRC Logs for 2003-07-20

00:27<term>now I need a way to be able to drop a cdrom into the drive, have it automounted, and have mythvideo scan it
00:27<term>same for dvds..
01:10-!-Ripp [~chris@] has joined #mythtv
01:24<Captain_Murdoch>Chutt: did the bug you fixed with slave backends have to do with the guy's problem on the -dev list?
01:25<Chutt>i haven't gotten that message yet, so i can't reply and say it's fixed
01:25<Chutt>speakeasy's having intermittant problems getting email from the lists
01:25<Captain_Murdoch>ok. wanted to make sure it wasn't something I broke. :)
01:26<Chutt>it was broken by the changed livetv->recording stuff
01:29<Captain_Murdoch>I'm going to commit that playbackbox status images stuff in a little while but am leaving them commented out in the blue theme so they're there for reference until we can get some better images. :) also going to commit something to let you unset a bookmark by hitting the bookmark key a 2nd time.
01:30<Chutt>can you just make it clear the bookmark once it's been used?
01:30<Chutt>that was the old behavior
01:32<Captain_Murdoch>ok. that's easy since I already had the code to clear it. do you mind me keeping the code in so if you hit it twice it clears it as well?
01:32<Captain_Murdoch>that way it will clear automatically but if you want to force clear you can.
01:37-!-mecraw [] has quit []
01:37-!-mikegrb [] has quit []
01:37-!-bigguy [] has quit []
01:37-!-Scrye [] has quit []
01:37-!-mdz [] has quit []
01:38-!-mecraw [] has joined #mythtv
01:38-!-mikegrb [] has joined #mythtv
01:38-!-bigguy [] has joined #mythtv
01:38-!-mdz [] has joined #mythtv
01:38-!-Scrye [] has joined #mythtv
01:49-!-_shad [] has quit [Read error: 60 (Operation timed out)]
02:20-!-lmatter [] has quit ["Client exiting"]
03:38-!-_shad [] has joined #mythtv
03:53-!-choenig [] has joined #mythtv
04:13-!-Drikus_ [] has joined #mythtv
04:27-!-Viddy [] has quit [Read error: 110 (Connection timed out)]
04:28-!-Viddy [] has joined #mythtv
04:52-!-bline [] has joined #mythtv
05:02<bline>Chutt: you around?
05:40-!-Viddy [] has quit []
05:41-!-mikegrb [] has quit []
05:41-!-Scrye [] has quit []
05:41-!-mecraw [] has quit []
05:41-!-mdz [] has quit []
05:41-!-bigguy [] has quit []
05:41-!-yebyen [] has quit []
05:41-!-radsaq [] has quit []
05:41-!-ahbritto [] has quit []
05:41-!-__keturn [] has quit []
05:41-!-hfb [] has quit []
05:41-!-poptix [] has quit []
05:41-!-Drikus_ [] has quit []
05:41<-- Snow-Manhas quit (
05:41-!-just1nux [] has quit []
05:41-!-jfm [] has quit []
05:41-!-Captain_Murdoch [] has quit []
05:41-!-fulbert [] has quit []
05:41-!-tdb30_ [] has quit []
05:41-!-Chutt [] has quit []
05:41-!-bline [] has quit []
05:41-!-_shad [] has quit []
05:41-!-sc00p [] has quit []
05:41<-- paulproteus|lapthas quit (
05:41-!-thor [] has quit []
05:41-!-Ripp [] has quit []
05:41-!-rkulagow_ [] has quit []
05:41-!-rkulagow___ [] has quit []
05:41-!-GreyFoxx [] has quit []
05:41-!-ca1vin [] has quit []
05:41-!-Justin_ [] has quit [Nick collision]
05:41-!-Justin_ [] has joined #mythtv
05:41-!-bline [] has joined #mythtv
05:41-!-Viddy [] has joined #mythtv
05:41-!-Drikus_ [] has joined #mythtv
05:41-!-_shad [] has joined #mythtv
05:41-!-Scrye [] has joined #mythtv
05:41-!-mdz [] has joined #mythtv
05:41-!-bigguy [] has joined #mythtv
05:41-!-mikegrb [] has joined #mythtv
05:41-!-mecraw [] has joined #mythtv
05:41-!-Ripp [~chris@] has joined #mythtv
05:41-!-hfb [] has joined #mythtv
05:41-!-sc00p [] has joined #mythtv
05:41-!-ahbritto [] has joined #mythtv
05:41-!-rkulagow_ [] has joined #mythtv
05:41--> Snow-Man( has joined #mythtv
05:41-!-yebyen [] has joined #mythtv
05:41--> paulproteus|lapt( has joined #mythtv
05:41-!-just1nux [] has joined #mythtv
05:41-!-jfm [] has joined #mythtv
05:41-!-fulbert [] has joined #mythtv
05:41-!-tdb30_ [] has joined #mythtv
05:41-!-poptix [] has joined #mythtv
05:41-!-Chutt [] has joined #mythtv
05:41-!-Captain_Murdoch [] has joined #mythtv
05:41-!-__keturn [] has joined #mythtv
05:41-!-radsaq [] has joined #mythtv
05:41-!-ca1vin [] has joined #mythtv
05:41-!-thor [~thor@] has joined #mythtv
05:41-!-GreyFoxx [] has joined #mythtv
05:41-!-rkulagow___ [~mythtv@] has joined #mythtv
05:49-!-sfr [] has joined #mythtv
06:50-!-ahbritto [] has quit []
07:04-!-ahbritto [] has joined #mythtv
07:21-!-sfr [] has quit []
07:21-!-mikegrb [] has quit []
07:21-!-Scrye [] has quit []
07:21-!-mecraw [] has quit []
07:21-!-mdz [] has quit []
07:21-!-Viddy [] has quit []
07:21-!-bigguy [] has quit []
07:22-!-sfr [] has joined #mythtv
07:22-!-Viddy [] has joined #mythtv
07:22-!-mecraw [] has joined #mythtv
07:22-!-mikegrb [] has joined #mythtv
07:22-!-bigguy [] has joined #mythtv
07:22-!-mdz [] has joined #mythtv
07:22-!-Scrye [] has joined #mythtv
08:18-!-mikegrb [] has quit []
08:18-!-Viddy [] has quit []
08:18-!-sfr [] has quit []
08:18-!-Scrye [] has quit []
08:18-!-mecraw [] has quit []
08:18-!-mdz [] has quit []
08:18-!-bigguy [] has quit []
08:18-!-sfr [] has joined #mythtv
08:18-!-Viddy [] has joined #mythtv
08:18-!-mecraw [] has joined #mythtv
08:18-!-mikegrb [] has joined #mythtv
08:18-!-bigguy [] has joined #mythtv
08:18-!-mdz [] has joined #mythtv
08:18-!-Scrye [] has joined #mythtv
08:36-!-PeteCool [] has joined #mythtv
09:07-!-PeteCool [] has quit ["Client exiting"]
09:46-!-bline [] has quit ["Client exiting"]
09:47-!-bigguy [] has quit [Read error: 60 (Operation timed out)]
10:04-!-bigguy [] has joined #mythtv
10:11-!-mikegrb [] has quit []
10:11-!-Viddy [] has quit []
10:11-!-sfr [] has quit []
10:11-!-Scrye [] has quit []
10:11-!-mecraw [] has quit []
10:11-!-mdz [] has quit []
10:11-!-sfr [] has joined #mythtv
10:11-!-Viddy [] has joined #mythtv
10:11-!-mecraw [] has joined #mythtv
10:11-!-mikegrb [] has joined #mythtv
10:11-!-mdz [] has joined #mythtv
10:11-!-Scrye [] has joined #mythtv
11:02-!-paulproteus [] has joined #mythtv
11:11-!-Drikus_ [] has quit ["toedeledoki"]
11:55-!-Captain_Murdoch [] has quit ["X-Chat!"]
11:58-!-Drikus_ [] has joined #mythtv
12:13-!-Captain_Murdoch [] has joined #mythtv
12:15<paulproteus|lapt>rizzi: I notice you're also in Rochester! Cool!
13:02<paulproteus|lapt>MythMusic doesn't seem to support reading the ID3 data of VBR MP3 files. The files in question have ID3V1 and ID3V2; they're purchased from eMusic. Anyone else experience this?
13:05<sfr>hm, i have some vbr mp3 files and only use id3v2 tags. i haven't seen what you describe, sorry.
13:05<paulproteus|lapt>Hmm, okay. I'll remove the V1 tags and see what happens.
13:17-!-dopez [] has joined #mythtv
13:22<paulproteus|lapt>Hmm. I dropped the V1 tag, and Myth still considers it "Unknown".
13:30-!-jkolb [] has joined #mythtv
13:37-!-Edgan [] has joined #mythtv
13:59--> Peit|Home( has joined #mythtv
14:00<Peit|Home>images for the channels in the program guide, they live on the backend don't they?
14:07<Peit|Home>cool, is there a size limit (pixel) on the images, or are they dynamically resized?
14:08<Chutt>i think they're all resized to be the same
14:08<Chutt>on load
14:08<Chutt>don't hold me to that, though
14:10<Peit|Home>it will/should load both .jpg and .gif yeah?
14:10<Chutt>should load anything qt supports
14:10<Peit|Home>cool, wonder why it's not loading these then
14:10<Peit|Home>heh, watch live tv, seg fault
14:11<Peit|Home>i think it's the ivtv driver.. or at least i hope it is :)
14:11-!-brunes [] has joined #mythtv
14:12<brunes>does mythtv have a bugzilla or somewhere where I can report a major bug?
14:12<Chutt>the mailing list.
14:12<Chutt>or here
14:12<Chutt>peit|home, no, it's likely something else.
14:12<brunes>its one a mentioned before... mythmusic does not wokr properly on any non 44,100 khz sound cards... which includes all NForce boards
14:13<Chutt>my nforce board works rather perfectly at 44.1khz
14:13<Chutt>or any other sampling rate.
14:13<brunes>which is a real bummer since nforce based systems are perfect for myth
14:13<Chutt>those boards are _not_ locked to 48khz
14:13<brunes>well you must not be using the NVidia audio drivers
14:13<Chutt>funnily enough, i am
14:13<brunes>the nvidia audio drivers are locked at 48,000
14:13<Chutt>works fine with alsa, too
14:13<brunes>at least the rev. 256 ones
14:13<brunes>which is the latest
14:14<Chutt>the fact that mythmusic doesn't resample audio has been reported before
14:14<Chutt>but, as usual, no one has submitted a patch
14:14<Peit|Home>Chutt: mind if i just paste you the kernel panic dump before i run the backtrace?
14:14<brunes>i dont know enough to patch it or I would have already
14:14<Chutt>so i really don't care.
14:14<brunes>I dont know anything about resampling audo or codecs..
14:14<Chutt>peit|home, oh, you got a kernel panic?
14:15<Chutt>then it's probably the driver =)
14:15<Peit|Home>Chutt: yeah, devide my zeno on what i think is saa7115[0]:
14:15<Peit|Home>it was working before i went away as well :)
14:15-!-Justin_ [] has quit [Read error: 60 (Operation timed out)]
14:16<brunes>See, you need the latest NForce audio to enable SPDIF out.. it doesnt work with the previous versions, which may be non locked as you say.. so its liek either I cant use mythmusic or I can't play DVDs with digital sound
14:17<Chutt>or you could learn how to code and fix it up
14:17<Chutt>and let everyone else benefit from it
14:17<brunes>I do know how to code.. quite well..
14:17<Chutt>then what's the problem?
14:17<brunes>but I dont know dick all about libavcodec, and from looking at it its pretty complex
14:18-!-kolb [] has joined #mythtv
14:18<brunes>so I'd rather just use freevo then spend liek a week figuring out this dumbo resamplig problem
14:18<Chutt>ReSampleContext *blah = audio_resample_init(2, 2, 48000, 44100);
14:18-!-jkolb [] has quit [Read error: 104 (Connection reset by peer)]
14:18<Chutt>audio_resample(blah, output_buffer, input_buffer, number_of_input_samples);
14:19<Chutt>write the output_buffer
14:19<Chutt>and you're done.
14:19-!-kolb is now known as jkolb
14:19<brunes>if you already know what to do why isn't it patched?
14:19* brunesboggles
14:19<Chutt>add about 10 lines of code on top of that to get the actual values, make sure the buffers are large enough
14:19<Chutt>because i don't give a damn about fixing other people's problems if they have no intention of trying to fix them themselves, first
14:20<paulproteus|lapt>Chutt: Any reason my MythMusic install isn't picking up ID3 tags in VBR MP3s but is picking them up in FLAC files abnd CBR MP3s?
14:20<Chutt>especially for something as easy as this
14:20<Chutt>paulproteus, send me one of the files
14:20<paulproteus|lapt>You'll delete it when you're done?
14:20<Chutt>you said you bought it from emusic?
14:20* paulproteus|laptcringes about violating his contract
14:20<brunes>man I sure am glad I don't react that way when people post bugs to projects I work on...
14:20<Chutt>i've got a lifetime subscription, used to work for them :p
14:21<paulproteus|lapt>Oooh, r0x0r :)
14:21<Chutt>but, yeah, i'll delete it
14:21<Chutt>brunes, if _you_ can't be bothered to write 20 lines of code, why should i?
14:22<brunes>what if I did not know C++?
14:22* paulproteus|laptpoints out brunes said he'd rather ditch Myth than fix it
14:22<Chutt>that's the lamest excuse
14:22<Chutt>if you know c, you know enough c++ to fix minor bugs like this
14:22<Chutt>since it's the same language.
14:22<brunes>who said I know C?
14:26-!-jkolb [] has quit []
14:26-!-Edgan [] has quit []
14:26-!-mikegrb [] has quit []
14:26-!-Viddy [] has quit []
14:26-!-sfr [] has quit []
14:26-!-Scrye [] has quit []
14:26-!-mecraw [] has quit []
14:26-!-mdz [] has quit []
14:26-!-Captain_Murdoch [] has quit []
14:26-!-brunes [] has quit []
14:26-!-bigguy [] has quit []
14:26-!-yebyen [] has quit []
14:26-!-radsaq [] has quit []
14:26-!-__keturn [] has quit []
14:26-!-ahbritto [] has quit []
14:26-!-hfb [] has quit []
14:26-!-poptix [] has quit []
14:26-!-Drikus_ [] has quit []
14:26-!-paulproteus [] has quit []
14:26-!-tdb30_ [] has quit []
14:26-!-fulbert [] has quit []
14:26-!-jfm [] has quit []
14:26-!-just1nux [] has quit []
14:26-!-Chutt [] has quit []
14:26<-- Snow-Manhas quit (
14:26<-- Peit|Homehas quit (
14:26-!-thor [] has quit []
14:26<-- paulproteus|lapthas quit (
14:26-!-sc00p [] has quit []
14:26-!-_shad [] has quit []
14:26-!-dopez [] has quit []
14:26-!-rkulagow___ [] has quit []
14:26-!-GreyFoxx [] has quit []
14:26-!-ca1vin [] has quit []
14:26-!-rkulagow_ [] has quit []
14:26-!-Ripp [] has quit []
14:27-!-brunes [] has joined #mythtv
14:27-!-Captain_Murdoch [] has joined #mythtv
14:27-!-bigguy [] has joined #mythtv
14:27-!-yebyen [] has joined #mythtv
14:27-!-radsaq [] has joined #mythtv
14:27--> Peit|Home( has joined #mythtv
14:27-!-dopez [] has joined #mythtv
14:27-!-Drikus_ [] has joined #mythtv
14:27-!-paulproteus [] has joined #mythtv
14:27-!-_shad [] has joined #mythtv
14:27-!-Ripp [~chris@] has joined #mythtv
14:27-!-sc00p [] has joined #mythtv
14:27-!-rkulagow_ [] has joined #mythtv
14:27--> Snow-Man( has joined #mythtv
14:27--> paulproteus|lapt( has joined #mythtv
14:27-!-just1nux [] has joined #mythtv
14:27-!-jfm [] has joined #mythtv
14:27-!-fulbert [] has joined #mythtv
14:27-!-tdb30_ [] has joined #mythtv
14:27-!-poptix [] has joined #mythtv
14:27-!-Chutt [] has joined #mythtv
14:27-!-ca1vin [] has joined #mythtv
14:27-!-thor [~thor@] has joined #mythtv
14:27-!-rkulagow___ [~mythtv@] has joined #mythtv
14:27-!-GreyFoxx [] has joined #mythtv
14:27-!-jkolb [] has joined #mythtv
14:27-!-ahbritto [] has joined #mythtv
14:27-!-hfb [] has joined #mythtv
14:27-!-__keturn [] has joined #mythtv
14:27-!-Edgan [] has joined #mythtv
14:27-!-sfr [] has joined #mythtv
14:27-!-Viddy [] has joined #mythtv
14:27-!-mecraw [] has joined #mythtv
14:27-!-mikegrb [] has joined #mythtv
14:27-!-mdz [] has joined #mythtv
14:27-!-Scrye [] has joined #mythtv
14:32-!-__keturn [] has quit []
14:32-!-ahbritto [] has quit []
14:32-!-hfb [] has quit []
14:32-!-poptix [] has quit []
14:32-!-jfm [] has quit []
14:32-!-just1nux [] has quit []
14:32-!-Chutt [] has quit []
14:32<-- Snow-Manhas quit (
14:32-!-tdb30_ [] has quit []
14:32-!-Drikus_ [] has quit []
14:32-!-paulproteus [] has quit []
14:32-!-fulbert [] has quit []
14:32-!-thor [] has quit []
14:32<-- paulproteus|lapthas quit (
14:32-!-sc00p [] has quit []
14:32-!-_shad [] has quit []
14:32<-- Peit|Homehas quit (
14:32-!-rkulagow___ [] has quit []
14:32-!-rkulagow_ [] has quit []
14:32-!-Ripp [] has quit []
14:32-!-GreyFoxx [] has quit []
14:32-!-ca1vin [] has quit []
14:32-!-dopez [] has quit []
14:32-!-Captain_Murdoch [] has quit []
14:32-!-brunes [] has quit []
14:32-!-yebyen [] has quit []
14:32-!-bigguy [] has quit []
14:32-!-radsaq [] has quit []
14:34-!-GreyFoxx [] has joined #mythtv
14:34-!-rkulagow___ [~mythtv@] has joined #mythtv
14:34-!-thor [~thor@] has joined #mythtv
14:34-!-ca1vin [] has joined #mythtv
14:34-!-Chutt [] has joined #mythtv
14:34-!-poptix [] has joined #mythtv
14:34-!-tdb30_ [] has joined #mythtv
14:34-!-fulbert [] has joined #mythtv
14:34-!-jfm [] has joined #mythtv
14:34-!-just1nux [] has joined #mythtv
14:34--> paulproteus|lapt( has joined #mythtv
14:34--> Snow-Man( has joined #mythtv
14:34-!-rkulagow_ [] has joined #mythtv
14:34-!-sc00p [] has joined #mythtv
14:34-!-Ripp [~chris@] has joined #mythtv
14:34-!-_shad [] has joined #mythtv
14:34-!-paulproteus [] has joined #mythtv
14:34-!-Drikus_ [] has joined #mythtv
14:34-!-dopez [] has joined #mythtv
14:34--> Peit|Home( has joined #mythtv
14:34-!-radsaq [] has joined #mythtv
14:34-!-yebyen [] has joined #mythtv
14:34-!-bigguy [] has joined #mythtv
14:34-!-Captain_Murdoch [] has joined #mythtv
14:34-!-brunes [] has joined #mythtv
14:34-!-ahbritto [] has joined #mythtv
14:34-!-hfb [] has joined #mythtv
14:34-!-__keturn [] has joined #mythtv
15:12<paulproteus|lapt>Chutt: Was that helpful at all?
15:12<Chutt>working on it
15:12<Chutt>was busy with something else
15:13<paulproteus|lapt>No worries, thanks for everything so far.
15:15<Chutt>simple little bug
15:15-!-Justin_ [] has joined #mythtv
15:16<Chutt>it'll be fixed in cvs in a little bit
15:17<paulproteus|lapt>Excellent! Thank you tremendously.
15:19<paulproteus|lapt>Now, if only I could get this libcdaudio glitch fixed.
15:20<paulproteus|lapt>(And if only SF would actually send me a subscribe message!)
15:20<Chutt>that i can't fix =)
15:20<Chutt>i dunno, i really wouldn't bother contacting the libcdaudio guy
15:20<Chutt>i doubt he can help
15:20<Chutt>considering that the project's essentially dead
15:20* paulproteus|laptblinks
15:21<paulproteus|lapt>Could you put a "special, crazy, duct-tape mode" that uses cdparanoia or cdrecord to check for CD?
15:21<Chutt>it still uses libcdaudio to get freedb info
15:21<Chutt>so it really would need to be fixed there
15:24* paulproteus|laptblinks
15:41<paulproteus|lapt>But it won't be fixed.
15:41* paulproteus|laptblinks
15:47<paulproteus|lapt> ?
15:47<paulproteus|lapt>You're probably not going to want to rip out all your dependencies to libcdaudio.
15:48<Chutt>i don't want to, no
15:48<Chutt>but someone else could
15:48* paulproteus|laptducks
15:48<Chutt>i wouldn't be opposed to that
15:49<paulproteus|lapt>I've lived in Java and LISP for the past two years, but I did some AP computer science in C++ back in high school.
15:49<paulproteus|lapt>Do you think it would be difficult?
15:49<paulproteus|lapt>I've never dealt with linking, and am mildly afraid.
15:50<Chutt>honestly, no idea
15:50<paulproteus|lapt>I suppose this is a better question: How closely married are MythMusic and libcdaudio?
15:50<Chutt>they're not
15:51<paulproteus|lapt>That's convenient.
15:52<paulproteus|lapt>If I weren't appallingly over-committed already, I'd do it.
15:52<paulproteus|lapt>For now, I'll just wait and see if I get time, or if happens to happen behind my back.
15:53* paulproteus|laptsighs and needs to buy some time-management skills ;)
15:55<jkolb>Chutt: Something that was discussed on -dev a while back, and I've been noticing it lately: The usleep(50) in weathercomms.cpp(line 154) makes my weather data update timeout way early (within 3s w/agg 15). Changing the 50 to a 200000 as was suggested makes it work.
15:56<jkolb>I think it may have something to do with the fact that this box doesn't often have anything else running.
15:56<jkolb>The explanation that was given was something to the effect of, "Well, it won't usually return for 20000us even though I only say 50, so it works most of the time."
15:56<jkolb>That doesn't happen on an unloaded box.
15:57<Chutt>usleep's guaranteed to sleep for one timeslice
15:58<jkolb>That doesn't seem to be long enough for me.
15:58<Chutt>unless you've changed things, that's 10ms
15:58<jkolb>I'm not sure if 200000 is the right number to put in there, but it doesn't timeout anymore. I'm not sure how long it would wait if it really couldn't reach the server.
15:59<brunes>there fixed the resampling :P
15:59<Chutt>brunes, patch?
15:59<Chutt>or did you 'fix' it
15:59<brunes>i 'fixed' it
16:00<brunes>as in I just removed all the libmad crap and spwan an external player
16:01<Chutt>yeah, that'll work _real_ well.
16:01<brunes>works great :P
16:01<Chutt>no vis
16:02<jkolb>Dropping that machine's packets, it times out in twenty-two-one-thousand at aggressiveness 15.
16:02<brunes>xmms has better vis
16:02<Chutt>which plugins?
16:02<brunes>there are tons of xmms vis plugins..
16:02<Chutt>converting them takes a whole half hour if they're not crap
16:02<Chutt>most of the xmms vis plugins are horrible
16:03<brunes>well seeing how I have no interest in the libmad player while it is broken, I also have no interest in the visualizations associated with it
16:04<Chutt>it's not broken.
16:04<Chutt>you want a feature added, and you're not willing to write the feature.
16:04<Chutt>if you weren't so whiney about it, i'd probably do it for you
16:05<sfr>hey, Chutt you're not feeling well? ;-)
16:05<brunes>I dont know how to write the feature
16:05<brunes>and I have bette rthings to do then fix bugs you're too lazy to fix. I'll just bypass it
16:06<Chutt>it's not a bug.
16:06<brunes>yes it is
16:06<Chutt>it's a feature that isn't implemented
16:06<brunes>its not a feature, it is a bug
16:06<Chutt>it works fine on non-crappy hardware
16:06<Chutt>with actual working drivers
16:06<brunes>so 44,100 is better sound quality then than 48,000 now?
16:06<Chutt>um, if that's the original source, yes.
16:06<brunes>funny i did not get that nocice of change sin the laws of physics
16:07<Chutt>resampling in software doesn't magically add quality
16:07<brunes>I have mnay file swhose original source is 48,000
16:07<Chutt>it'll play 48khz files in 48khz
16:07<Chutt>amazingly enough
16:07<brunes>seemingly not, since it does nto work
16:07<Chutt>you're playing 44.1khz files back as 48khz
16:08<Chutt>without resampling
16:08<brunes>no, I am tryng to play 48 khz files at 48khz
16:08<Chutt>considering that that works perfectly fine
16:08<brunes>but I would prefer it to not crap out when it plays a 44,1000 file
16:08<Chutt>you're not.
16:13<brunes>Regardless of your opinion of if it is a bug if mythmusic does not wrk on a large portion of hardware with up to date drivers, you still have a very poor attitude. Someone comes in and repots a bug and all you have to say is "fix it yourself". Who is to say the person even knows how to write C or C++ or anything at all ? Nnot only that but you're acknowledgement that you know how to fix it and that it would take all of
16:13<brunes>'t because "I should fix it because I reported it" is assinine Attitudes like that just discourage people from trying OS software altogether. A simple "Yes it is a known problem and is on the TODO list" would have sufficed instead of the piss poor attitude.
16:13<brunes> </Rant>
16:13<Chutt>whine whine whine
16:13<Chutt>bitch bitch bitch
16:13<Chutt>moan moan moan
16:13<-- brunes( has left #mythtv ("Client Exiting")
16:13<Chutt>i should start charging people for coding time
16:14<tdb30_>put up a site with a paypal link... I'll pay your some cash to work on features ;)
16:15<vektor>he wasn't playing a 48khz file was he.
16:15<Chutt>no, he wasn't
16:15<Chutt>he wants software resampling of the audio
16:15<vektor>he was playing an mp3 right
16:16<Chutt>and isn't willing to write the 20 lines of code to do so
16:16<Chutt>i've got better things to do
16:16<Chutt>and besides, my hardware works properly =)
16:16<vektor>yeah :)
16:16<vektor>software resampling is stupid.
16:29<Chutt>wow, 5 minutes of work, and it mostly works
16:29<jkolb>Chutt: Are there any no-no CFLAGS for mythtv?
16:29<Chutt>not that i know of
16:30<Chutt>oh, -fno-exceptions and -fno-rtti don't work, of course
16:30<jkolb>Not using those.
16:30<jkolb>-march=pentium3 -O2 -pipe -fomit-frame-pointer -fforce-addr -falign-functions=4 -ffast-math -mmmx -msse makes mythfrontend die on playback.
16:31<Chutt>die how?
16:31<vektor>that looks terribly bad.
16:31<Chutt>generally, you don't need to futz with the options
16:31<vektor>jkolb: why are you doing that?
16:31<jkolb>It just goes away.
16:31<jkolb>Gentoo ebuild just sets those.
16:31<jkolb>I suppose I could comment out the part that replaces the CFLAGS.
16:31<vektor>i really don't understand why gentoo people always think that they should use all these aggressive-looking CFLAGS
16:32<vektor>if there's some piece of code that can be optimized using some compiler options, then make a patch based on its asm output, don't compile for a specific arch.
16:32<vektor>-march=pentium3 is just silly
16:32<jkolb>I've not had any problems with other packages.
16:33<Chutt>should i commit this audio resampling crap or not
16:34<vektor>how do you resample?
16:34<vektor>not linear interpolation, i hope :)
16:34<Chutt>just using a function in libavcodec
16:34<vektor>oh that
16:34<vektor>sounds phearsome
16:34<Chutt>let's see what it does, i hadn't looked at it at all
16:35<Chutt>it's essentially doing linear interpolation
16:35<vektor>so, not high quality then
16:35<vektor>that will have audible artifacts
16:36<Chutt>that took all of 10 minutes
16:36<Chutt>had to work around a linking issue was all
16:40<jkolb>So, with using_alsa, is the audio setup screen supposed to look any different, or do I just put in the alsa devices for output and mixer?
16:40<Chutt>it's the same
16:41<Chutt>the guy that submitted the alsa patch hasn't finished it yet
16:41<Chutt>you have to edit the db a bit for the native alsa output to work
16:41<Chutt>paulproteus, your files should be fixed. you just need to clear out the metadata entries in the db for those files
16:41<Chutt>and update to current cvs, of course
16:43<Chutt>see, no problem at all fixing actual bugs =)
16:44<Chutt>actually, let's fix that in a better manner
16:47--> daddy-o( has joined #mythtv
16:47-!-daddy-o is now known as mattfelsen
16:48<jkolb>So, uh, what would be the setting I have to mess with in the db to get native alsa output?
16:48<Chutt>that i couldn't tell you
16:48<Chutt>try searching the mailing list archives for 'foom'
16:49<Chutt>i think he may mention it in the email about the patch
16:52<jkolb>Yay. AudioOutputDevice had to be set to ALSA:default
17:08<Chutt>you can probably even make it do all the special stuff with rmix and that
17:09<Chutt>isn't that the software mixing code?
17:09<Chutt>or dmix or i dunno
17:09<Chutt>you can do all sorts of stuff with an .asoundrc
17:09<Chutt>sample rate conversion, software mixing
17:09<Chutt>add filters
17:09<jkolb>I had no idea.
17:10<jkolb>I just wanted it so I didn't have to load snd-pcm-oss and snd-mixer-oss.
17:11-!-creepy [~lightning@] has joined #mythtv
17:15<Chutt>you'll still need it for mythmusic, though
17:15<Chutt>that wasn't ever converted over to it
17:16<jkolb>So it seems.
17:25<sfr>creepy, got mythtv up and running?
17:27<Chutt>people whose mta bounces aren't parsed by mailman suck.
17:27<Chutt>especially when i get a good hundred or so bounces in my inbox
17:35-!-Drikus_ [] has quit ["toedeledoki"]
17:38<-- mattfelsenhas quit ()
17:40<thor>Chutt, where did that resamoke stuff come from?
17:41<Captain_Murdoch>thor: libavcodec I think he said
17:42<thor>just checking the scrollback
17:42<thor>people with attitude
17:42<creepy>sfr nope, having exams
17:43<creepy>and myth does the same thing
17:43<creepy>but we may know why now :)
17:43<creepy>tv_grab_de suxass
17:43<creepy>well Chutt said at least
17:44<creepy>and the channel listing db may not suffice
17:44<creepy>maybe I need a whole DB, maybe I have no chance ;) but you could set it up somehow so I dunno
17:45<thor>Chutt, never mind ... I'm up to speed
17:47-!-creepy [] has quit ["gone.."]
17:48<sfr>creepy, why works fine, i'm not using it directly, but let it save into a file.
18:06<Chutt>thor, i was curious to see how much code it'd actually take
18:07<thor>not much
18:07<thor>(depending on ability)
18:07<thor>I had just gotten in and saw the commits
18:08-!-moegreen [] has joined #mythtv
18:08<Chutt>i didn't feel like working out the linking issues
18:09<Chutt>so i just included the resample.c file from libavcodec in there
18:09<Chutt>moegreen, hi
18:09<thor>exercise is left to the reader?
18:09<Chutt>i suppose
18:10<Chutt>isn't really important either way
18:10<thor>figured out conditional compiling
18:10<thor>(not that its very complicated :-)
18:10<thor>things are coming
18:11<thor>slightly hard to debug plugins sometimes
18:11<thor>if it's an unresolved
18:14<thor>if you have an unresolved symbol
18:14<thor>it's hard to know
18:14<thor>link doesn't know yet
18:14<thor>and then mythfrontend mythdvd just tells you it doesn't work
18:14<thor>... this is a very minor point ...
18:16<Chutt>you may be able to change the link type on load
18:17<Chutt>so it'd report missing syms there
18:17<thor>I wouldn't bother ... just happened a couple of times last night
18:18<thor>I was futzing around in main() to change what gets checked and passed around on startup
18:18<thor>and ended up scratching my head a few times
18:18<thor>... "well, it compiles .... so what's the problem ... "
18:20<thor>things get confusing, when you're (stupid and) simultaneously writing a GUI frontend (as a plugin), a non-GUI backend, a communications protocol (on the fly), a theme.xml file, library widgets, graphics, and linking it all to a new library you don't really understand.
18:22-!-just1nux [] has quit [Read error: 60 (Operation timed out)]
18:23<jkolb>To quote Chutt: whine, whine, whine
18:23* jkolbgrins
18:23<Chutt>jkolb, i just upped the usleep in mythweather, btw
18:24<thor>I think moegreen (moegreen?) thought a usec was 1/1000th of a second ...
18:24<Chutt>it is
18:24<jkolb>Uh, 1e-6
18:24<Chutt>it's a timeslice
18:25<thor>1/1000000 no?
18:25<Chutt>unless you've got a modified kernel that changes what HZ is
18:25<jkolb>Timeslices are e-3?
18:25<Chutt>even :p
18:25<thor>all I know is that 500,000 usecs seems an awful lot like half a second (at least around here)
18:26<Chutt>but usleep doesn't sleep for a usec
18:26<jkolb>Then what's the point of giving usec granularity in the call?
18:26<jkolb>Eventual hardware support?
18:26<Chutt>well, it's more granular than sleep
18:26<Chutt>i basically use usleep to say 'get me off the scheduler'
18:27<Chutt>though, that's probably going to change at some point in the kernel
18:27<Chutt>so i probably shouldn't, but it's a habit
18:27<thor>funny, I use it to say, "wait 1/5 of a second to see if the socket we want to talk to appears"
18:28<jkolb>Does Xv suffer if I don't have dri working? As in, is that why X has been using 30-someodd % of my cpu?
18:28<Chutt>what video card?
18:28<jkolb>i815 on board
18:28<vektor>jkolb: there are lots of things that can hurt XVIDEO performance.
18:28<Chutt>it's probably doing a bunch of stuff in software
18:28<vektor>Chutt: no it isn't.
18:28<vektor>the i815 is pretty good.
18:29<Chutt>ah, ok
18:29<vektor>and the AGP chipset rocks.
18:29<Chutt>well, that's what high X performance generally means
18:29<vektor>jkolb: Run x11perf -shmput500
18:29<vektor>jkolb: and show me the results in /msg
18:30<jkolb>Um, no results to show. "X Error of failed request: BadMatch (invalid parameter attributes)"
18:31<vektor>I beg your pardon?
18:31<vektor>You're running X locally, right?
18:31<vektor>like, no crazy remote setup or anything?
18:31<jkolb>Well, I ran that command from a ssh session on that box with DISPLAY=:0
18:32<jkolb>A white box appeared on the screen, and then went away.
18:32<vektor>can you please post or email me your X log file
18:35<vektor>What's up with the agpgart errors?
18:35<jkolb>The warnings?
18:35<vektor>you're using some square pixel 16:9 modes, interesting. i hope you set your displaywidth and displayheight correctly
18:36<vektor>um, yeah, the "warnings" :)
18:36<jkolb>I have a 16:9 tv
18:36<vektor>does it take VGA input?
18:36<jkolb>I have a scan converter between them.
18:36<jkolb>Yeah, well, it was the only way.
18:36<Chutt>there, that mythweb profile patch is in now too
18:36<vektor>sounds terrible
18:37<vektor>jkolb: anyway, is something busted with your agpgart?
18:37<jkolb>You mean the 'allocation of blah blah' warning? That's what I'm trying to fix right now.
18:37<vektor>still, why didn't x11perf run
18:37<jkolb>glxinfo ran. Said I didn't have dri.
18:38<vektor>nothing to do with GL, dude
18:38<vektor>try 'x11perf -putimage500'
18:38<vektor>i'm trying to test the speed of your video memory writes
18:38<jkolb>Rebooting right now. Ran across a page telling me to disable the kernel DRM and install a DRM package for 4.3.0
18:38<vektor>if -shmput500 fails, that means somsething is just _broken_ with your X
18:39<jkolb>It'll be up here in a moment.
18:39<vektor>oh, yeah, you do need the right DRM module for XVIDEO.
18:39<vektor>That's sort of a 'thing' with the i815.
18:40<vektor>I think it has to do with how the AGP transfers get set up.
18:40<Chutt>vektor, have you looked at XvMC at all?
18:40<vektor>It's not typical.
18:40<vektor>Chutt: sure, well, i 'reviewed' the API when they posted it however long ago.
18:40<vektor>we had some discussions about interlaced images and stuff.
18:40<Chutt>i've been trying to figure it out, and, well, can't =)
18:40<vektor>ok, what's the issue?
18:40<Chutt>nothing, really
18:41<Chutt>someone posted some patches to libavcodec to start getting support for that there
18:41<vektor>it's very low level
18:41<Chutt>and apparently the matching stuff is already in mplayer
18:41<vektor>like, you need to do it as part of a partial mpeg2 decoder
18:41<Chutt>but i was just having problems figuring out how it all worked together
18:41<vektor>like, you have to parse the stream a bunch
18:41<Chutt>and massage the data a little
18:41<vektor>and then you have images which get rendered after you set them up
18:41<Chutt>then send it off to it
18:41<vektor>and it renders to an off-screen image
18:41<vektor>and then you have it display later
18:42<vektor>so it's exclusively for display
18:42<vektor>you can't get back the pixels
18:42<Chutt>do you know if the calling program has access to the data?
18:42<Chutt>ah, ok
18:42<vektor>it's not interesting to me since it makes it more difficult to do an OSD
18:42<Chutt>so without an additional overlay, i can't really use it
18:42<vektor>they have an overlay
18:42<vektor>usually for subtitles and stuff
18:42<vektor>and you can draw there
18:42<vektor>i don't know how good it is though.
18:43<vektor>i just don't like the architecture ;)
18:43<Chutt>it seemed rather obscure to me
18:43<vektor>but for an epia or something :/
18:43<vektor>like, i just find it sad that boards like the epia suck so bad
18:43<Chutt>even though it'd just be for mpeg2
18:43<vektor>that we have to go back to hardware APIs and stuff
18:43<Chutt>well, we'll have to see what via's mpeg stuff looks like
18:44<vektor>drawing pixels is so much easier for writing software _once_ and not having to redesign your whole bloody architecture.
18:44<jkolb>vektor: It must just be broken. Same error msg.
18:44<vektor>jkolb: but do you have the right DRM module loaded now?
18:44<vektor>let's at least start there
18:44<vektor>or what is broken
18:44<jkolb>glxinfo still reports no dri.
18:44<vektor>x11perf ?
18:44<jkolb>x11perf -putimage500
18:44<vektor>did you enable DRI in your X config file?
18:45<jkolb>Yeah, Load "dri" is there.
18:45<vektor>and load "glx" ?
18:45<vektor>and you have Mode "0666" in section "dri" ?
18:45<vektor>let me see your log file again
18:46<jkolb>I just added the Mode 0666 line. Not sure if I restarted the XServer after that.
18:46<jkolb>Lemme try that.
18:47<jkolb>That didn't fix it.
18:48<vektor>post your new log
18:48<jkolb>Wait, didn't have i810 loaded.
18:48<vektor>you didn't?
18:48<vektor>well there you go.. :)
18:48<jkolb>Nope. Trying again.
18:48<jkolb>No joy.
18:48<vektor>post your new log
18:49<vektor>is agpgart loaded? :)
18:49<vektor>it better be
18:49* jkolbwhistles innocently
18:49<jkolb>Wait. No, I built that into the kernel.
18:49<vektor>eww, don't do that :)
18:49<vektor>post your log
18:49<vektor>post your log
18:49<vektor>post your log
18:50<vektor>hey waitasec
18:50<vektor>what depth are you at?
18:50<vektor>are you at 24bpp ?
18:50<vektor>then DRI won't work
18:50<vektor>you must be at 16bpp
18:50<vektor>hardware limitation
18:51<vektor>sorry, i didn't think of that.
18:51<vektor>and you're still getting that gart error.
18:51<jkolb>I sewar I didn't see anything about that in the docs.
18:51<jkolb>er, swear, even.
18:51<vektor> The driver supports hardware accelerated 3D via the Direct Rendering Infrastructure (DRI), but only in
18:51<vektor> depth 16.
18:51<vektor>that's in 'man i810' here.
18:52<jkolb>Well, sure if you read the *right* docs, it's there.
18:52<vektor>write a better doc system and send it to
18:52<vektor>i'm sure they'll take it.
18:52<jkolb>Yay! direct rendering: Yes
18:52<vektor>nobody is maintaining these docs.
18:52<vektor>now, about your xvideo performance.
18:53<vektor>get that and run it
18:53<vektor>the compile line is at the top of the .c file
18:53<vektor>we'll compare with that since x11perf is busted (why? i don't get that)
18:54<Chutt>i don wanna write an osd menu
18:54<jkolb>It's dying in X_GetImage, if that helps.
18:56<vektor>dying ??
18:56<jkolb>How do I link the X libs?
18:56<vektor>dude i told you to look at the top of the file
18:56<jkolb>Well, not dying. That's the request that failed in x11perf
18:56<vektor>to see the compile line
18:56<vektor>Chutt: what do you mean?
18:56<Chutt>for mythtv
18:56<vektor>Chutt: use my list widget
18:56<vektor>it's cool
18:56<Chutt>i've got all the code
18:56<Chutt>i just need to write a little wrapper
18:57<Chutt>and some control bits
18:57<vektor>for xvmc?
18:57<vektor>or for in general
18:57<Chutt>for the osd menu
18:57<Chutt>just being lazy
18:57<Chutt>2400 FPS
18:57<Chutt>with xvtest
18:57<jkolb>Hey, look. A pretty picture. In tye-dyed colors.
18:57<vektor>i get 1200 FPS
18:57<vektor>jkolb: how many fps ?
18:57<Chutt>that's because your computer sucks.
18:57<vektor>nah, it's because my video card driver sucks
18:57<Chutt>not that mine's good anymore
18:57<vektor>jkolb: what?
18:57<jkolb>[18:57] <Chutt> 2400 FPS
18:58<jkolb>60.171781 FPS
18:58<vektor>but no, it's because i'm limited by p3 copy speed
18:58<vektor>jkolb: that means something is broken.
18:58<jkolb>I hate copy-on-select sometimes.
18:58<vektor>jkolb: you don't have mtrrs set up correctly
18:58<jkolb>Yay b0rked config.
18:58<vektor>borked kernel
18:58<vektor>did you compile that kernel?
18:59<vektor>did you forget to enable MTRRs ?
18:59<vektor>type 'cat /proc/mtrr'
18:59<jkolb>I compiled it, and if I didn't enable MTRRs, I'll be the first to hit myself with a large, solid object.
18:59<jkolb>Woo. I don't have to hit myself.
19:00<Chutt>I prefer APUE for that sorta violence
19:00<Chutt>big and hefty
19:00<paulproteus|lapt>Yeah, I think we'd all rather that just have been the issue.
19:00<jkolb>You're more than welcome to come over here and beat the everloving crap out of me if you fix this.
19:00* paulproteus|laptducks when other people realize he just got here
19:00<Chutt>paulproteus, your tags are fixed, btw
19:00<paulproteus|lapt>Thanks, Chutt!
19:00<paulproteus|lapt>(grumble, grumble, libcdaudio, grumble, grumble)
19:01<Chutt>you'll have to empty the existing data out of the music metadata table in the database, though
19:01<vektor>jkolb: ok so if you get that low FPS from xvtest, i get the impression that system->video memory copies are extremely slow for some reason.
19:01<vektor>jkolb: you're not using X 4.3 from CVS are you?
19:01<paulproteus|lapt>Okay, when I get the updated version, I'll come back asking how to do that.
19:01<jkolb>No, 4.3.0
19:01<vektor>jkolb: where did you get X 4.3 from?
19:01<jkolb>Uh, Gentoo?
19:01<vektor>that blows
19:01<paulproteus|lapt>Gentoo sucks.
19:02<vektor>ok guys, don't turn this into a gentoo bash
19:02<vektor>although we want it to be:)
19:02<Chutt>distros are distros
19:02<vektor>jkolb: ok so, did you compile it yourself?
19:02<jkolb>It didn't seem like such a bad idea at the time.
19:02<jkolb>Yes, I did.
19:02<vektor>like, did you compile X 4.3?
19:02<vektor>do you have the source tree there?
19:02<jkolb>With those same godawful flags I sent earlier.
19:02<vektor>oh you fool :)
19:02<vektor>bad call
19:02<jkolb>I have the compressed source.
19:02<vektor>just the compressed source?
19:02<Chutt>it should be fine
19:02<vektor>not the source you built?
19:03<vektor>well, uncompress it i need you to run a grep on it.
19:03<vektor>i appreciate you doing this, there have been a lot of people complaining about very similar performance problems.
19:03<vektor>it's nice to have another machine to 'check'.
19:03<vektor>in a sense. :/
19:04<vektor>i'm hoping that this is related to this bug:
19:04<jkolb>Trust me, I have no problem being a guinea pig if it helps anyone out. Especially when it helps me out.
19:04<vektor>i have already scheduled myself to watch a swashbuckling movie tonight though so hopefully we get this fixed asap :)
19:05<vektor>pirates of the caribbean was so good i've decided to seek out more pirate movies
19:05<jkolb>The PotC movie is actually fun to see.
19:05<jkolb>You have to find ones that don't take themselves seriously.
19:05<jkolb>Like Goonies!
19:06<vektor>tonight we're watching 'the sea hawk'
19:06<vektor>1940, errol flynn, classic swashbuckler ;)
19:08<vektor>is it uncompressed yet?
19:08<vektor>run 'grep --recursive O_SYNC .'
19:08<jkolb>I uncompressed the wrong one first.
19:08<vektor>in the tree
19:09<Chutt>rkulagow, hi =)
19:10<rkulagow_>hey. busy week. did i miss anything?
19:10<Chutt>not really
19:10<-- rkulagow_( has left #mythtv
19:10<jkolb>Getting some matches.
19:10-!-rkulagow [] has joined #mythtv
19:10<rkulagow>that was an accident.
19:10<rkulagow>i'll bbl; working on basement.
19:11<Chutt>i want to finish my basement
19:11<jkolb>vektor: Done grepping
19:11<vektor>jkolb: /msg me the results
19:12<rkulagow>it's easy. just get fired, and have no excuse for sitting in front of the computer all day.
19:12<rkulagow>you know, "working".
19:12<Chutt>i can't come up with a floor plan that allows for enough room for speakers
19:12<Chutt>is my main problem
19:12<jkolb>Uh, it may be easier for me to post them. Lemme throw them in a file.
19:13<jkolb>It will be when I'm done.
19:13<vektor>will you be around in, um, two hours?
19:14<jkolb>Yeah, I can be around.
19:14<vektor>i want you to post these two .c files:
19:14<vektor>you have 5 minutes :)
19:15<vektor>and then i'll be back in 2 hrs
19:15<-- sfr( has left #mythtv ("Client exiting")
19:15<jkolb>Hah! Done!
19:17<jkolb>Whoa. glxgears doesn't draw properly.
19:17<vektor>i'll be back later
19:17* vektorgone
19:17<jkolb>At *all*
19:17<paulproteus|lapt>Oh, Chutt, I hope you deleted that file ;)
19:17<paulproteus|lapt>jkolb: Are you that Gentoo freak?
19:17* jkolbsighs
19:17<jkolb>Yes, I suppose I am.
19:18* paulproteus|laptshakes his head
19:20* ahbrittois amazed at all the gentoo envy
19:24<jkolb>paulproteus|lapt: Why do you hate Gentoo so?
19:37<paulproteus|lapt>I hate Gentoo because of this story:
19:37<paulproteus|lapt>I finally prepared my Xbox to run Linux.
19:38<paulproteus|lapt>I installed Gentoo; never used it, may as well give it a fair chance.
19:38<paulproteus|lapt>I installed mythtv, after much trouble.
19:38<paulproteus|lapt>(As in, I learned how to use "unstable". Okay, fine, not much trouble.)
19:39<paulproteus|lapt>Then I figured I'd run it.
19:39<paulproteus|lapt>It died.
19:39<paulproteus|lapt>You know why?
19:39<jkolb>No, why?
19:39<paulproteus|lapt>Because my QT wasn't compiled with MySQL support.
19:39<paulproteus|lapt>Either Gentoo should know that, and recompile QT for MySQL support, or THE LIBS SHOULD JUST FREAKING WORK.
19:39<paulproteus|lapt>They're libraries. Come freaking on.
19:40<jkolb>Yeah, that should default to having mysql support, instead of default to no mysql support.
19:40<paulproteus|lapt>Just because the guy who built my Qt didn't have MySQL installed doesn't mean my installing Myth should b0rk.
19:40<paulproteus|lapt>That's braindamaged.
19:41<paulproteus|lapt>I shouldn't need to install MySQL just to run Myth.
19:41<paulproteus|lapt>What if I'm using it on some machine where I don't *want* MySQL?
19:41<jkolb>Uh, what? How are you running myth without mysql?
19:41<paulproteus|lapt>The server, I mean.
19:41<jkolb>You don't need mysql installed on a gentoo box to have mysql support in the qt libs.
19:42<paulproteus|lapt>And Gentoo should not have made the stupid mistakes it did. It should know about my Qt lacking MySQL.
19:42<paulproteus|lapt>That's the biggie.
19:42<paulproteus|lapt>And then I just installed Debian.
19:42<jkolb>Yeah, it really should keep a record of what CFLAGS, CXXFLAGS, USE, etc were used to compile a package.
19:43<jkolb>Then if you tried to emerge a package that required a flag, and one of the supporting packages was built without that flag, at least give you the option of rebuilding the supporting packages.
19:43<jkolb>Or something like that.
19:44<ahbritto>submit a patch
19:44* ahbrittoducks
19:54-!-bline [] has joined #mythtv
19:54<jkolb>Well, X is down to ~4% CPU on playback now.
19:56<Viddy>you could make a perl script to mimic mysql
19:57<jkolb>Wouldn't get quite the same performance out of it, though.
19:57<Viddy>so you wouldnt have to have mysql at all ;)
19:58<paulproteus|lapt>That's be hilarious. :)
19:59<jkolb>Well, goom still takes up all of my cpu, but at least it's displaying in realtime now.
20:08<Chutt>have you set the options to cut the resolution in half for goom?
20:08<jkolb>No. It looks so much nicer without that.
20:08<Chutt>should help a little
20:08<jkolb>Yeah, I know.
20:08<Chutt>yeah, but =)
20:09<Chutt>my nehemiah can't run the thing at all, even at quarter res
20:09<jkolb>It was having trouble at quarter res without DRI.
20:10-!-choenig [] has quit ["wuuuusch..."]
20:11<paulproteus|lapt>So, from my Gentoo experience, I realied it was a debugging nightmare.
20:12<jkolb>It's possible to install qt debs without mysql support.
20:12<jkolb>The default ones may even be lacking that.
20:12<paulproteus|lapt>That's fine.
20:12<paulproteus|lapt>apt-get install libqt3-mysql
20:12<paulproteus|lapt>But I can't do that, see.
20:13<jkolb>USE="mysql" emerge qt
20:13<paulproteus|lapt>And no one will complain if I install myth without mysql support.
20:13<paulproteus|lapt>*That*'s the kicker.
20:13<paulproteus|lapt>And Gentoo has no way of knowing if I did, as I understand.
20:13<paulproteus|lapt>Which makes it a debugging nightmare.
20:14<jkolb>Are you saying that Gentoo won't complain if you install myth without USE=mysql?
20:15<paulproteus|lapt>Yes. And it has no way of knowing if that's the case.
20:16<jkolb>Right. Which is why I would like it if their tbz2 files had package buildinfo in them. They could go into /etc/portage/<package> or something, so you could easily check to see what all the settings were for any given package currently installed.
20:17<bline>it keeps a list of what USE flags you used in /var/db/pkg
20:17<jkolb>Does it?
20:17<bline>qpkg has ways to display them, part of gentoolkit
20:17<paulproteus|lapt>I didn't USE=mysql when I recompiled QT the next time.
20:17<paulproteus|lapt>I just emerged it.
20:17<paulproteus|lapt>After emerging mysql.
20:17<paulproteus|lapt>And, gee, it worked.
20:17<paulproteus|lapt>I then installed Debian.
20:18<bline>just put mysql in USE inside /etc/make.conf
20:18<paulproteus|lapt>But I didn't do that.
20:18<jkolb>Sweet. This is exactly what I was just saying I wanted.
20:19<paulproteus|lapt>And thus, checking for that USE variable is meaningless to my installation.
20:19<paulproteus|lapt>Qt "happened" to pick up MySQL.
20:19<paulproteus|lapt>And Gentoo didn't note that down.
20:19<bline>sounds like a bug, go report it
20:19<paulproteus|lapt>I don't know the terminology well enough to report it.
20:20<paulproteus|lapt>I guess if someone tells me more detailedly what to say, I will.
20:20<paulproteus|lapt>I'm much happier in Debian land, though, let me tell you. And one can install Debian packages from source, which seems to be Gentoo's only big advantage.
20:20<paulproteus|lapt>(And Gentoo's strange init setup; that might be cool, I don't now.)
20:20<bline>fine just sit here an wine, that helps
20:21<paulproteus|lapt>Well, I try to avoid whining, by switching to Debian. ;)
20:21<bline>gentoo requires half a brain, you should go back to debian
20:21<paulproteus|lapt>I'd say Debian also requires a brain.
20:22<paulproteus|lapt>But, anyway, I guess it's only fair for you to flame Debian since I spent all this time flaming your system :).
20:22<bline>I didn't like debian, and I never flamed it
20:23<paulproteus|lapt>Well, I'm really not just flaming. My complaints are serious.
20:23<paulproteus|lapt>You do have a reason to complain about my avoidance of filing the bug, though.
20:23<paulproteus|lapt>Yeah, good point. I'll report it sometime tomorrow, from work, when I need a break.
20:23<paulproteus|lapt>And then I'll at least make life easier for those "still in Gentoo".
20:23* paulproteus|laptducks
20:24<paulproteus|lapt>Honestly, though, can Gentoo just "notice" that Qt "happened to pick up" MySQL?
20:24<paulproteus|lapt>If it can, great. If not, that's silly. Either way, I'll file.
20:25<jkolb>You mean can Gentoo check to see that qt was built with mysql support?
20:25<bline>you have no idea how it works I can see. gentoo ebuilds are shell scripts that use the configure that comes with the tar file to set things up
20:26<bline>in this case, the package maintainer probably did not notice that Qt automatically picked up Qt
20:26<bline>or he didn't care, the latter I can see being reasonable
20:26<bline>if you already have mysql installed, there is not reason not to build that part of Qt
20:27<bline>but it is also possible that it is a bug in the Qt configure script, though I doubt it
20:27<jkolb>Qt doesn't build in mysql support unless you have the mysql flag, IIRC.
20:29<bline>the ebuild adds "-plugin-sql-mysql -I/usr/include/mysql -L/usr/lib/mysql" if USE has mysql in it, it does nothing more I just looked
20:30<bline>so either it is a Qt bug or you are mistaken
20:30<bline>not refering to you jkolb
20:30* paulproteus|laptblinks and reads through these again
20:30<jkolb>I figured.
20:31<paulproteus|lapt>I'm pretty sure I just recompiled and it worked. Maybe I just recompiled and I installed Debian too soon to tell.
20:31<paulproteus|lapt>I suppose I should get a pet Gentoo install to see if that actually happened before reporting the bug.
20:33<jkolb>You hadn't changed your USE flags since compiling qt the first time?
20:33<paulproteus|lapt>I seem to recall it picked up the Qt.
20:33<paulproteus|lapt>Er, that Qt picked up MySQL.
20:34<paulproteus|lapt>But I may have neglected to test it.
20:34<paulproteus|lapt>I do seem to recall seeing something, maybe compile screens, that implied that.
20:40<jkolb>Ah. I think I know what x11perf is failing. My screen is 960x540, and I found a page that claims I must have y >= 600;
20:44-!-kc8apf [] has joined #mythtv
20:45<kc8apf>anyone alive?
20:45<jkolb>You could say that.
20:45<kc8apf>is anyone working on getting mythtv to work on PPC?
20:46<jkolb>I have no idea.
20:49-!-ahbritto [] has quit ["Client exiting"]
20:53<kc8apf>i managed to get it to compile, but I think i messed up when I fixed the functions using IO
20:56<jkolb>I can't help you, don't know the codebase really. Someone else here may be able to eventually, though.
21:03<jkolb>Well, I got x11perf -shmput500 to run finally. 273.0/sec
21:08<Chutt>captain_murdoch, thanks for fixing that mythweb stuff
21:24<Captain_Murdoch>I've been emailing the guy back and forth some. he sent me another patch that I'm getting ready to apply.
21:33<vektor>jkolb: discover anything?
21:33<jkolb>Yeah. x11perf requires vres > 603. I'm running 960x540.
21:33<vektor>well, that explains that one.
21:33<jkolb>So I bumped my virtual to 1920x1080, and got results.
21:34<vektor>what did you get for -shmput500 ?
21:34<jkolb>How was the movie, btw?
21:34<vektor>at what bitdepth did you get 273 ?
21:34<vektor>the movie rocked
21:34<vektor>much swashing and buckling
21:35<vektor>ok, 273 is a bit low, but not _terrible_
21:35<jkolb>Oh, hey. Lemme run xvtest again
21:35<jkolb>Lookee there. 60.17
21:35<jkolb>Well, yeah.
21:35<vektor>ok, so wtf.
21:36<jkolb>Did anything strike your fancy in those O_SYNC files?
21:36<vektor>i had suspected that the gentoo X 4.3 might have the patch that opened /dev/mem using O_SYNC that was screwing up SuSE.
21:36<vektor>but by your files it isn't (unless you're an ia64 machine) ;-)
21:36<jkolb>Um, no. Celeron 850
21:36<vektor>a celeron?
21:36<jkolb>Uh, yeah.
21:37<vektor>so maybe your x11perf numbers _are_ normal ;-)
21:37<jkolb>xvtest also?
21:37<jkolb>Oh, good.
21:37<vektor>xvtest is just terrible
21:37<jkolb>Well, not good, but good as in I can do better.
21:37<jkolb>Or something.
21:38<jkolb>Lemme try installing the latest xfree package again.
21:38<vektor>i want more specs on your system
21:39<jkolb>Ok. MX3S-T mobo, 256M PC100. It's a cheap system. I built it to be low(ish) power, and quiet.
21:41<vektor>still, you're only getting 78MB/sec upload speed to the card
21:41<vektor>through XVIDEO
21:42<vektor>and that's about half of what it should be for a P3 ... i can't see that being normal for your celeron
21:42<vektor>especially since you're getting about 130MB/sec with RGB32 uploads
21:42<vektor>so something is just busted
21:42<jkolb>glxtest also looks freaky when I run it. Like it's not converting CR -> CRLF.
21:42<jkolb>er, LF->CRLF
21:45<jkolb>Hm. Their newest xfree package fixes gl, but breaks xv.
21:45<jkolb>150fps in glxgears
21:45<vektor>breaks it how?
21:45<vektor>150fps is terrible
21:46<jkolb>The overlay doesn't get blitted to screen. Just a blue box.
21:46<jkolb>150fps is better than it stepping across the screen.
21:46<jkolb>Reinstalling r2
21:53<jkolb>Well sonofabitch. glxgears runs properly now with 4.3.0-r2
21:53<jkolb>Also ~150fps
21:54<jkolb>In xvtest, all of the CPU is being used by X, if that helps any.
21:54<jkolb>Not by xvtest.
21:54<vektor>of course
21:55<vektor>i mean, it's X that's writing to video memory
21:55<vektor>and that's what is slow
21:55<jkolb>Oh, yeah.
21:55<vektor>copies in system memory are cheap
21:56<jkolb>Now, a lot of the pages I've been bringing up have talked about setting Option "AgpMode". If I put that in my XF86Config, though, the log claims that it's not used.
22:00<vektor>that's only for the NVIDIA binary drivers ..
22:00<vektor>no it isn't
22:00<vektor>i810 uses it too
22:00<vektor>i think
22:00<jkolb>That's what all these pages claim.
22:00<vektor>anyway, that won't help you
22:01<vektor>that's only for DMA'ed stuff.
22:01<vektor>and the i810 driver uses a memcpy
22:02<vektor>so wtf
22:02<vektor>wtf wtf wtf.
22:03<jkolb>Um, I had managed to disable drm there. glxgears is now displaying all freaky again, and giving me ~375fps
22:03<jkolb>xvtest, otoh, has dropped to 56.6fps now. wtf?
22:04<jkolb>Oh, mythbackend is doing something.
22:09<jkolb>Could it be a colorspace thing? Maybe the 815 doesn't support YUY2? What colorspace does Myth use?
22:09<vektor>myth uses YV12
22:09<vektor>a 4:2:0 colourspace
22:09<vektor>and the 815 supports YUY2 just fine
22:09<vektor>get that
22:09<vektor>same deal
22:09<vektor>and see how many fps you get
22:10<jkolb>~57fps, same as xvtest
22:11<jkolb>This time the CPU is split between the two, though.
22:11<vektor>that's not great though
22:13-!-mecraw [] has quit [Read error: 54 (Connection reset by peer)]
22:13<jkolb>You were expecting lower?
22:13<vektor>no you should see better performance
22:14<bline>is 256/sec slow?
22:14<vektor>it's not as bad as i would have thought though
22:14<vektor>bline: hmm?
22:14<jkolb>bline: I hate you. <g>
22:14<vektor>well it's a bit unfair since this app does too much work
22:14<bline>from -shmput500
22:14<vektor>the xvtest app is much better for a test
22:14-!-Edgan [] has quit ["Client exiting"]
22:14<vektor>bline: oh, at what bitdepth?
22:15<bline>depth 24, is what what you mean?
22:16<bline>xvinfo | grep depth:
22:16<bline> depth: 24
22:16<bline> depth: 24
22:16<vektor>ok, then that's pretty good
22:16<vektor>NVIDIA binary drivers?
22:16<jkolb>Wow. -putimage500 gives me 60-61/sec
22:16<vektor>yeah, they rock.
22:16<vektor>jkolb: don't use putimage500
22:17<vektor>jkolb: use shmput500
22:17<vektor>Chutt: from shmput, yeah?
22:17<jkolb>shmput500 was 273.
22:17<vektor>jkolb: at 16bpp though
22:17<vektor>jkolb: divide by 2 if you're at 16bpp to get your MB/sec
22:17<Chutt>i've been pondering upgrading this box
22:17<vektor>i need to upgrade my box
22:17<vektor>it's way too dated
22:18<bline>Chutt: what card?
22:18<Chutt>gf4 ti4200
22:19<bline>I would have gotten a gf4, but at the time I didn't think nvtv supported it
22:19<Chutt>it didn't when i got the card
22:19-!-Edgan [] has joined #mythtv
22:26<vektor>jkolb: do you have some like APM stuff set up?
22:27<jkolb>I think I included APM support, yeah.
22:28<jkolb>Yeah, APM, Enable at boot, Make CPU Idle calls, RTC in GMT, and Allow Interrupts
22:28<jkolb>No ACPI
22:29<vektor>can you shut that off?
22:29<jkolb>By rebuilding the kernel, yeah
22:30<vektor>no i mean like in the bios ?
22:30<jkolb>Oh. I dunno.
22:30<jkolb>Lemme try
22:34<jkolb>No, I can only disable ACPI support.
22:34<vektor>ok, can you try that?
22:34<vektor>what good would that do
22:34<kc8apf>anyone familiar with vsync.c?
22:34<jkolb>Already was
22:34<vektor>kc8apf: sure
22:34<vektor>kc8apf: i wrote it
22:34<kc8apf>good, could you change it to support PPC?
22:34<kc8apf>PPC uses memory mapped IO
22:35<kc8apf>no inb or outb
22:35<kc8apf>or ioperm for that matter
22:35<vektor>those inb/outb statements never get hit
22:35<vektor>they won't work anyway
22:35<kc8apf>didn't figure they would
22:35<kc8apf>but it prevents the source from compiling
22:35<vektor>they only work on luckily-configured older VGA cards in PCs.
22:35<vektor>then comment it out
22:35<kc8apf>I have
22:35<vektor>those functions aren't called
22:36<-- dopezhas quit ()
22:36<kc8apf>but a simple ifdef __powerpc__ would fix it permently
22:36<vektor>even better would be to remove the functions
22:36<kc8apf>or that
22:36<vektor>i'll do it in my copy and mention it to chutt.
22:37<kc8apf>now i need to look at why I get segfaults.....
22:57<Captain_Murdoch>quote from the -users mailing list concerning changing the hostname fields in the database if you change your hostname... "This seems like an item that should only be entered one time on a config page." my thought... "it is, when you install your OS". :)
22:57<thor>I'm waiting for Chutt to get back to him on that ...
22:58<jkolb>Ah, I do miss the humor of the -users list.
22:58<paulproteus|lapt>I seem to get segfaults while changing my playlist in MythMusic. I should recompile, get unstripped binaries, and submit the core dumps?
22:58<thor>recompile for debugging
22:58<thor>top line
22:58* paulproteus|laptnods
22:59<thor>may need to do the same thing to mythlibs
22:59<-- rkulagow___(~mythtv@ has left #mythtv
22:59<thor>depending on where the problem is
22:59<thor>what do you mean by "changing" a playlist ?
22:59<paulproteus|lapt>I mean I do these things:
22:59<paulproteus|lapt>I'm playing music
23:00<paulproteus|lapt>I choose Edit Playlist (3, iirc)
23:00<paulproteus|lapt>I choose new songs, and uncheck some in the current playlist
23:00<paulproteus|lapt>I accept that (Esc? I can't qutie recall what key)
23:00<paulproteus|lapt>And boom, MythMusic dies.
23:00<paulproteus|lapt>It only happens from time to time.
23:00<thor>well, yeah, a backtrace from gdb with a copy built for debugging would help
23:01<jkolb>vektor: New kernel installed, still 60.17fps xvtest
23:01<paulproteus|lapt>I figured it might be related to the specific things I selected, but I selected the same songs on restarting the program and it just works.
23:02-!-rkulagow___ [] has joined #mythtv
23:02<thor>it's probably the visual
23:02<thor>see if it happens on goom but not on ... say ... Gears
23:06<Captain_Murdoch>jkolb: just for reference sake, I'm getting 60.3 FPS on xvtest on my Celeron 466 w/ builtin i810 video under XFree v4.2.0
23:07<jkolb>Celeron 850 i815 4.3.0
23:07<jkolb>Should be faster.
23:08<Captain_Murdoch>that explains why in some fast-moving scenes I see the screen redraw. :) need to get a better card in there.
23:08<jkolb>?? The video is only 30fps.
23:09<Captain_Murdoch>yeah but the card is slow. :)
23:09<jkolb>But being able to do > 60fps isn't going to make the video look any different.
23:10<Captain_Murdoch>maybe doesn't explain it, but the card is slow I can tell going from there to other machines that are the same speed with different video cards.
23:10<Captain_Murdoch>keep in mind that's only 256x256 also in xvtest right?
23:10<jkolb>Looks to be, yeah.
23:10<Captain_Murdoch>so higher res would be slower fps
23:11<Captain_Murdoch>assumedly. :)
23:11<jkolb>But playing back the video from myth only takes ~3% of the CPU for X.
23:11<jkolb>I have lots of idle CPU when playing myth's videos.
23:11<jkolb>Which is creepy.
23:12<Captain_Murdoch>ditto here, that's why I have only a Celeron 466 for my main frontend system. it's more than powerful enough to watch mythtv and divx files.
23:13<kc8apf>hmm, out of curiosity, has anyone tried myth on an alpha?
23:14<kc8apf>or better yet, is there anyway that the backend can be compiled without QT?
23:14<Captain_Murdoch>on linux you mean?
23:15<Captain_Murdoch>I think the backend needs qt but not necessarily X
23:15<kc8apf>I have a headless alpha with lots of harddrive space that I would love to use for capture
23:15<kc8apf>I guess I could build X without any servers
23:16<kc8apf>just the libs for forwarding
23:16<Captain_Murdoch>backend doesn't need a display, I think it just uses non-display qt stuff.
23:16<thor>kc8apf, you can compile both qt and X on headless machines
23:16<kc8apf>can you do the setup without a display as well?
23:16<kc8apf>thor: i know
23:17<Captain_Murdoch>setup can be run remote over ssh
23:17<Captain_Murdoch>or you could even run it somewhere else and hand-edit the DB afterwards. :)
23:19--> jkolb-lurk( has joined #mythtv
23:19<jkolb>vektor: I'm going to sleep now. I'l putting up my XF86Config at (I moved all the other files in there as well). If you see anything in there that you think may help, please let my alter ego here know.
23:19<jkolb>The XFree86.0.log is from the most recent startup, too.
23:19<-- jkolbhas quit ()
23:35<thor>Chutt, well that was disappointing
23:42-!-just1nux [] has joined #mythtv
23:46<just1nux>anyone reported a problem where PVR-250 streams loose audio after switching inputs? Its odd, works fine audio-wise when changing channel but as soon as i switch inputs its gone, current myth and ivtv.
23:50-!-Timon [] has joined #mythtv
23:55<-- hfb( has left #mythtv ("Client exiting")