Back to Home / #mythtv / 2003 / 05 / Prev Day | Next Day
#mythtv IRC Logs for 2003-05-26

00:24-!-Timon [] has quit ["using sirc version 2.211+KSIRC/1.2.4"]
00:29-!-Timon [] has joined #mythtv
00:30-!-Timon [] has quit [Client Quit]
00:31-!-Timon [] has joined #mythtv
01:01-!-Timon [] has quit ["using sirc version 2.211+KSIRC/1.2.4"]
01:01-!-jrh [] has quit [Read error: 104 (Connection reset by peer)]
01:01-!-jrh [] has joined #mythtv
03:12-!-FryGuy [] has quit ["peace and protection - if you don't try it, you can't tell me it sucks. -"]
03:12-!-yebo [] has quit [Read error: 104 (Connection reset by peer)]
03:58-!-FryGuy [] has joined #mythtv
04:06-!-kiwi_uk [] has joined #mythtv
04:12-!-choenig [] has joined #mythtv
04:44-!-choenig [] has quit [Remote closed the connection]
05:29-!-kiwi_uk [] has quit ["Peace and Protection 4.22"]
05:33-!-FryGuy [] has quit [Read error: 104 (Connection reset by peer)]
05:34-!-FryGuy [] has joined #mythtv
05:44-!-FryGuy [] has quit [Read error: 104 (Connection reset by peer)]
05:44-!-froggy19 [] has joined #mythtv
05:44-!-froggy19 is now known as FryGuy
06:01-!-FryGuy [] has quit [Read error: 104 (Connection reset by peer)]
06:01-!-froggy19 [] has joined #mythtv
06:01-!-froggy19 is now known as FryGuy
06:12-!-FryGuy [] has quit [Read error: 104 (Connection reset by peer)]
06:54-!-choenig [] has joined #mythtv
07:51-!-bline [] has quit [Read error: 104 (Connection reset by peer)]
08:47-!-bline [] has joined #mythtv
09:26-!-bline [] has quit ["Client exiting"]
09:52-!-bline [] has joined #mythtv
11:40-!-bline [] has quit [Read error: 113 (No route to host)]
11:44-!-bline [] has joined #mythtv
11:57-!-rkulagow [] has quit [Read error: 110 (Connection timed out)]
12:04-!-Drikus_ [] has joined #mythtv
12:35-!-kiwi_uk [] has joined #mythtv
12:45-!-kiwi_uk [] has quit [Read error: 104 (Connection reset by peer)]
15:33-!-justin [] has joined #mythtv
16:01-!-itsame [] has joined #mythtv
16:07<-- itsamehas quit ()
16:10-!-TheAsp [] has joined #mythtv
16:10<TheAsp>funny, the patch i wrote like 4 months ago was reimplemented this afternoon...
16:12<TheAsp>or is it my patch?
16:15<TheAsp>close but not the same
16:25-!-rkulagow [] has joined #mythtv
16:46<mdz_>is there an osd.xml for oldosd? or is oldosd obsolete?
16:50<Captain_Murdoch>TheAsp: you mean what I put in?
16:51<Captain_Murdoch>I didn't remember a patch. I thought you were working on some form of editor that would let you change things like profile, "record only on weekdays", etc..
16:52<TheAsp>no, the press i to get the infordialog in the scheduled recordings screen patch
16:52<TheAsp>i'm working on the autodeletion stuff, but need ui stuff to config it
16:53<TheAsp>so i was going to write something like that menu, but gave up
16:53<TheAsp>the ui.xml stuff was just new at that point
16:54<Captain_Murdoch>that's what I committed, the 'I' key thing. just creates an InfoDialog and then refreshes afterwards. I didn't add much code, just called infodialog like the epg does.
16:54<TheAsp>neither did i
17:15-!-FryGuy [] has joined #mythtv
17:37-!-bline [] has quit ["Client exiting"]
17:43-!-TheAsp [] has quit [Remote closed the connection]
17:43-!-Drikus_ [] has quit ["toedeledoki"]
18:21-!-choenig [] has quit [Remote closed the connection]
18:22<mdz_>I'm getting quite a bit of this:
18:22<mdz_>uitypes.o: Going to be hard to request an update without a parent
18:22<mdz_>whose code is that?
18:23<mdz_>is that normal?
18:23<mdz_>(for now anyway)
18:23<Chutt>i dunno
18:23<Chutt>it's not finished yet
18:23<mdz_>Chutt: the gcc guys yelled at us for mythtv's default CFLAGS
18:23<Chutt>why's that?
18:24<mdz_>-O6 because it's no different from -O3, -fexpensive-optimizations because it's on at -O2 and higher
18:24<mdz_>and -funroll-loops because it probably shouldn't be on for every file
18:24<Chutt>yeah, whatever
18:24<mdz_>I'm using just -O3 -march=pentiumpro -fomit-frame-pointer now
18:25<mdz_>current CVS is looking pretty good so far
18:25<Chutt>-fmmx or -fsse would be appropriate
18:25<Chutt>as well
18:25<Chutt>or is it -mmmx and -msse
18:26<mdz_>would you object to changing the defaults to remove -fexpensive-optimizations and use -O3 instead of -O6 at least?
18:27<mdz_>should be no functional change, makes the command line a bit shorter
18:28<mdz_>I think I'd like to remove -funroll-loops as well, but whatever
18:33<Chutt>whatever you want
18:33<Chutt>i don't use those defaults anyway
18:49<mdz_>just got some weird behaviour while watching an in-progress recording
18:49<mdz_>I tried to skip back, and it seemed to skip forward
18:49<mdz_>it was fast, though, definitely using the seek table
18:50<mdz_>it looks like it's being written in batches, every 15 seconds or so
18:50<mdz_>to the db
18:59<Morph>hrm. still no live tv for me.
18:59* Morphrebuilds
19:03-!-Sg4 [] has joined #mythtv
19:10<Morph>hmm no luck.
19:19<Morph>Chutt: has something changed recently in regards to ivtv and mythtv? that would cause this to not work?
19:20<Chutt>everything's working for me with all that
19:20<Morph>hmm hmm
19:20<Morph>you're using latest ivtv cvs?
19:20<Chutt>i did go in and drop the bitrate of the video so that the crappy nehemiah had an easier time decoding, though =)
19:21-!-hfb [] has quit [Read error: 110 (Connection timed out)]
19:21<Morph>I suspected ivtv but mplayer grabs video from the device..mythtv doesnt seem to attempt anything.
19:22<Chutt>best thing to do would be to compile in debug mode
19:22<Chutt>get it to hang
19:22<Morph>its in debug mode :)
19:22<Chutt>switch to the gdb term, hit ctl-c
19:23<Chutt>to break it, then generate the backtrace like it says how to in the mythtvdocs
19:23<Chutt>err, insert a space there =)
19:23<Morph>did that. didnt see anything out of the ordinary. looking around now.
19:23<Chutt>well, mail it to me
19:24<Chutt>i'll check it out
19:24<Morph>k. one sec.
19:24<Chutt>later, though, i'll bbiab =)
19:30<Viddy>la vista
19:45<mdz_>Chutt: is oldosd obsolete now, or does it just need an osd.xml?
19:46<thor_>mdz, where are you getting "uitypes.o: Going to be hard to request an update without a parent"
19:46<mdz_>thor_: I'm not sure; I see it on the terminal where I started mythfrontend when I come back after using it
19:46<mdz_>when I'm navigating the menus I assume
19:46<mdz_>11 of them
19:47<mdz_>in a row
19:47<thor_>probably on theme parsing ... I'll fix it
19:48* Morphfires off the backtrace
19:49<thor_>mdz, it's just me walking over moegreen's code ... setting UI text objects ... will be gone when I commit tonight. Sorry about that
19:49<mdz_>thor_: no problem, didn't break anything for me
19:55<Morph>hm mythweather works :)
19:57<mdz_>hmm, it's not that it's going forward, it's that it doesn't skip at all
19:57<mdz_>doesn't happen on another frontend I started up to test
19:57<mdz_>ah, unknown position errors
19:57<mdz_>unknown position: 130980
19:58<mdz_>meaning it wasn't in the positionmap
20:45-!-billytwowilly [] has joined #mythtv
20:48<billytwowilly>Since the last piece of news (that is almost a month old now) Hints at a release soon and I can't find any mention of a release or pending release on the mailing lists archives, anyone care to speculate as to when a release will be coming?
20:53-!-billytwowilly [] has quit ["Client exiting"]
20:53-!-billytwowilly [] has joined #mythtv
20:56<mdz_>where is the font set for the new EPG?
20:59-!-billytwowilly [] has quit ["Client exiting"]
21:34<Chutt>in the ui
21:34<Chutt>.xml file
21:35<Chutt>mdz, it just needs a new theme xml written for it
21:35<Chutt>should be easy to translate over
21:36<thor_>Chutt, have you tried that tarball I sent last night
21:36<Chutt>no, sorry :(
21:36<Chutt>been really busy with other stuff
21:37<Chutt>morph, if you see this, could you repeat that step on the backend process?
21:37-!-hfb [] has joined #mythtv
21:38<Chutt>morph, and get a trace of all threads? (shows how to in the mythtv howto)
21:43<Chutt>thor, lemme compile it up right now
21:44<mdz_>Chutt: is the positionmap updated at all while watching an in-progress recording?
21:44<mdz_>or will it only have what was there when it starts watching?
21:44<Chutt>it should be grabbing it from the backend
21:44<Chutt>but i think captain_murdoch might've broken that
21:45<mdz_>it didn't seem to be, for me, earlier
21:45<mdz_>looking at it, I was thinking, is there any reason not to add entries as we go along if they don't exist for whatever reason?
21:45<mdz_>right now it does if (!haspositionmap)
21:45<Chutt>ya know, that's probably the issue
21:46<Chutt>before, you'd not have any position map
21:46<Chutt>now, you get a partial one when you start playing the in-progress recording
21:46<Chutt>so that fucks it up
21:46<Captain_Murdoch>haspositionmap doesn't get set true unless it loads the positionmap/seektable when it first starts, so it should always try to get the latest if it needs it.
21:46<mdz_>but even if it did, it wouldn't have any entries ahead of where it's watching
21:46<mdz_>so no seek table for skipping forward
21:46<Chutt>no seek table for seeking backwards
21:47<Chutt>it'll grab the stuff from the backend for forward
21:47<thor_>Chutt, don't run on it a box with code that isn't saved somewhere else
21:47<Chutt>but not for stuff that it (the frontend) has seen itself
21:47<thor_>it can lock up pretty hard ... I'm solving that now
21:47<Chutt>thor, heh
21:47<Chutt>i'm just recompiling mythtv to try out bruce's stuff
21:47<Chutt>then i'll do mythmusic
21:47<thor_>no hurry
21:48<Morph>Chutt: just saw your messages.
21:48<Chutt>morph, cool, mind doing that?
21:48<Morph>not at all. doing it now. noticed something on the backend that im looking at now..something with the mysql query.
21:48<Chutt>captain_murdoch, main issue is that seeking forward will grab from the backend, rewinding won't
21:49<Captain_Murdoch>so if there's no entry in the positionmap for where you're trying to seek to it should either fetch from the backend (preferably) or the db.
21:49<Chutt>and if it has a seektable (like an inprogress recording does now), it won't append data to that seektable as it plays back
21:49<Chutt>best thing to do would be to get rid of the if !positionmap tests for adding to the map
21:49<Chutt>and to try and load it from the db if it's not in the local map
21:50<mdz_>Chutt: do you think it should be if (it's not in the map already) or just unconditional?
21:50<mdz_>I can't think of any time it would be different
21:50<Captain_Murdoch>haspostitionmap is only set true when it loads a positionmap at startup. so that !haspositionmap will execute whenever there was no positionmap at startup.
21:50<Chutt>that used to catch in-progress recordings
21:51<Chutt>since they wouldn't have a seektable
21:51<Chutt>now they do, though
21:51<mdz_>in-progress recordings have a positionmap at start
21:51<mdz_>but an incomplete one
21:51<Morph>it works.
21:51<Morph>database issues.
21:51<Captain_Murdoch>this goes back to what we were talking about the other day then.
21:51<mdz_>so I start watching, I get a partial seek table, I watch past where the seek table ends, and then try to skip
21:51<mdz_>and those entries are not there
21:51<Captain_Murdoch>if livtv or watching recording then don't set haspositionmap to true.
21:51<Chutt>morph, what was it?
21:51<mdz_>(skip backwards, that is)
21:52<Morph>Chutt: it was missing some columns so it wasnt reading in the info for the tuner..
21:52<mdz_>Captain_Murdoch: so, don't set haspositionmap when loading from the db?
21:52<mdz_>or try to determine if it's complete or not?
21:52<Captain_Murdoch>no, just if loading from the DB and it's livetv or watching a recording.
21:52<Captain_Murdoch>sorry, inprogress recording
21:52<Chutt>morph, wasn't it complaining about that?
21:53<Captain_Murdoch>watchingrecording, not watching-recorded
21:53<Morph>Chutt: not before, no. I only noticed it when it was running from gdb.
21:53<Morph>Chutt: this was the backend running in gdb when I noticed the DB error.
21:53<mdz_>does it know at that point if it's incomplete or not?
21:54<Captain_Murdoch>mdz, if it's livetv or watching an in-progress recording then by definitiion it's incomplete
21:54<mdz_>I am not sure that it does
21:54<mdz_>but does nuppeldecoder know that?
21:54<mdz_>isn't it transparent from the ringbuffer up?
21:55<Captain_Murdoch>yes. there are 2 variables, livetv and watchingrecording.
21:55<Captain_Murdoch>the player knows about the livetv because it can get the positionmap from the encoder
21:56<Captain_Murdoch>so I can just make the 2 decoders check those vars before setting haspotitionmap = true.
21:56<Captain_Murdoch>if either is set then don't set haspositionmap
21:56<Chutt>that'll work
21:56<mdz_>and rename haspositionmap to hasfullpositionmap :-P
21:58<Chutt>compiled mythtv in debug mode
21:58<Chutt>no cpu =)
21:59<Captain_Murdoch>is that a good meh or a bad meh? so I'll add checks for !livetv and !watchingrecording and then rename all haspositionmap to hasfullpositionmap
21:59<Chutt>no no, you're fine
21:59<Captain_Murdoch>ok. it'll be in cvs in a few minutes.
22:00<Chutt>how 'bout the setlength calls?
22:00<Chutt>those should also not be used if it's 'live'
22:03-!-Sg4 [] has quit [Read error: 60 (Operation timed out)]
22:04<Captain_Murdoch>ok. so wrap the hasfullpositionmap = true and the setfilelength() call.
22:04<mdz_>totalLength, but not filelength, right?
22:06<Captain_Murdoch>so totalframes = what?
22:07-!-PeteCool [] has joined #mythtv
22:07<PeteCool>any of you guys able to connect to tarek's website? Doesn't work here :(
22:07<Captain_Murdoch>I don't see it setting either of those after it calls nvr_enc->FillPositionMap in dofastforward
22:08* Captain_Murdochdoesn't remember the url off the top of his head.
22:11<Captain_Murdoch>no connection here either.
22:14<hachi>what's the program I'm looking for on the lists to convert the nupple video files to something else?
22:14<PeteCool>hachi: you need to use a patched mencoder (mencoder comes with mplayer)
22:16<hachi>ahh, I'll patch it then, I asumed it was a standalone product
22:17<hachi>project, even
22:17<PeteCool>hachi: search the archives for the patch, I think Matt Zimmermann posted it
22:18<hachi>he has it on his site
22:19<PeteCool>hachi: can you paste the URL please? I didn't know he had a site :)
22:20<hachi>somewhere in that tree
22:22<hachi>oh wonderful... I know I saw it on there somewhere
22:22<Captain_Murdoch>Chutt, mdz_: should m_parent->SetFileLength() be called at all if it's livetv or recording?
22:23<Chutt>it wasn't set before
22:23<PeteCool>I'm getting a 403 going there
22:23<Captain_Murdoch>ok, that's what I thought. but avformatdecoder.cpp has code that predicts the length and frames based upon the bitrate if haspositonmap is false.
22:23<hachi>PeteCool, mythtv/ belongs on the end of that again
22:24<hachi>eep, /mythtv/
22:24<Chutt>but that gets overwritten later
22:25<hachi>PeteCool: documented here
22:25<Chutt>thor, hey, it kinda works now =)
22:26<thor_>turn off accelerators
22:26<Captain_Murdoch>chutt, ok. thanks.
22:26<Chutt>they're off
22:27<thor_>I've just greaty sped up the redrawing, and figured where it was going hard (that's fixed). Lot faster now
22:28<thor_>First time people hit left arrow, it will kinda throw them
22:29<Chutt>might it make sense to combine the pause/play button?
22:29<Chutt>or do you want to keep them separate?
22:29<thor_>makes no differene to me
22:30<thor_>(there the same on the QKeyEvent intercepting level)
22:31<Chutt>bruce's stuff is worse :(
22:38<Captain_Murdoch>hasFullPositionMap fix is in cvs.
23:04-!-paperclip [] has quit []
23:05-!-paperclip [] has joined #mythtv
23:06-!-hfb [] has quit ["Client exiting"]
23:09<hachi>I just searched on google for mythtranscode, only one match that doesn't give me anything, does anyone have a pointer?
23:20<thor_>Chutt, you sure you want a "no tree" option even with acclerators on?
23:20-!-rkulagow [] has quit [Read error: 104 (Connection reset by peer)]
23:22<Morph>hm openbox makes a decent WM to run with myth.
23:23<thor_>anyone know anything about: ?? Nice dedicated MythMusic box at $250.00
23:25<thor_>or even a frontend ... "integrated MPEG decoder" ??
23:30-!-rkulagow [] has joined #mythtv
23:32<mdz_>Captain_Murdoch: thanks
23:42<Captain_Murdoch>if I broke it, I should have to fix it. :)