Back to Home / #mythtv / 2004 / 10 / Prev Day | Next Day
#mythtv IRC Logs for 2004-10-24

---Logopened Sun Oct 24 00:00:17 2004
00:05--- ---> DanMeister [] has joined #MythTV
00:05DanMeister hey all
00:06--- <--- DanMeister [] has left #MythTV ()
00:08dopester "solve my problem NOW" :)
00:08Beirdo heh
00:08Beirdo how's life in dopester-land today?
00:09dopester decent
00:09Majestik Hmm.. can I suggest a solution?
00:09* Majestik puts on the bofh hat
00:09dopester fixing stuff, but also finding more broken stuff in the process
00:09Beirdo yeah, that's par for the course, dopester :)
00:13Majestik There's no way to switch between fullscreen and windowed with a single key, is there?
00:22Majestik Ok.. and the followup question to that one, since I think I know the answer, is is that something that can be added?
00:25Beirdo you volunteering?
00:25Beirdo :)
00:26dopester sounds like it to me :)
00:26Majestik Not a chance.. noone wants my code to enter into mythtv
00:27Majestik manpages are bad enough, and I'm backlogged on that
00:28Beirdo heh
00:29Majestik I remember what happened last time we had this sort of conversation..
00:52hadees did anyone ever post a patch for supporting screen savers durning certain times like paused tv or on the menu
00:52hadees i saw something on it in the mailing list when i was searching through for something else
00:53dopester i think i remember someone complaining baout it but thats about it
00:57hadees dopester, complaining about it? this is the thread i saw about it;#45676
00:58dopester that seems different that what i remember
00:59dopester i also use the word complaining loosely..
00:59hadees dopester, you don't happen to know who this guy is do you?
01:00dopester nope sorry man
01:00hadees or anyone for that matter
01:02Majestik Ummm.. just email him?
01:06hadees Majestik, = ) what do you think i was just doing
01:08Majestik Ok, fine. I'm too used to talking in -users
01:09hadees i tried asking in users first but i figured the dev channel would know if some one added a patch for it
01:10dopester its not like it was a completely off topic question.. it wasn't "whats the best card to use with myth?" :)
01:11hadees i am guessing he never added it though but how much could have changed in the code since then... actully how much could have changed i haven't looked in the code in a while
01:11hadees he wrote it for like .12
01:14hadees mythtv uses a screensaver disable command right? and an enable?
01:15--- <<-- cmorgan [] has quit (Remote closed the connection)
01:25--- ---> hays [] has joined #mythtv
01:26hays ok, so I'm trying to put together a mythTV box... i want it to be able to play movies and output the stuff in high def and dolby 5.1... Will an SB Audigy work with the standard VGA cable output to a projector?
01:26hays shoot wrong channel
01:51--- ---> mocker [] has joined #mythtv
02:33--- ---> hamsta [] has joined #mythtv
02:33--- <--- hamsta [] has left #mythtv ("Ich guck mal woanders ;)")
03:47--- ---> hamsta [] has joined #mythtv
04:45--- ---> Narkov [] has joined #mythtv
04:46Narkov i have two tuner cards working under analog and a DVB do i specify which one to use for livetv? i have correctly setup priorities for recording but it always seems to go for the analog tuner for livtv
04:48mocker Narkov: Try #mythtv-users
04:48Narkov yeah sorry there now
04:48mocker :)
05:09--- ---> Omnic_ [] has joined #mythtv
05:10--- <<-- Omnic [] has quit (Read error: 104 (Connection reset by peer))
05:14--- Netsplit <-> quits: DonLKSAB, kRutOn, jwhite, hays, adante, dopester, GreyFoxx, Octane, Morph, bline, (+43 more, use /NETSPLIT to show all of them)
05:27--- <<-- racer [] has quit (Read error: 110 (Connection timed out))
05:27--- <<-- m0j0_ [] has quit (Read error: 104 (Connection reset by peer))
05:30--- ---> Omnic_ [] has joined #mythtv
05:30--- ---> Narkov [] has joined #mythtv
05:30--- ---> hamsta [] has joined #mythtv
05:30--- ---> mocker [] has joined #mythtv
05:30--- ---> hays [] has joined #mythtv
05:30--- ---> Majestik [] has joined #mythtv
05:30--- ---> Octane [] has joined #mythtv
05:30--- ---> stoffel [] has joined #mythtv
05:30--- Netsplit over, joins: o_cee
05:30--- ---> yebyen_ [] has joined #mythtv
05:30--- ---> meshuga [] has joined #mythtv
05:30--- ---> flatronf700B [] has joined #mythtv
05:30--- ---> Viddy [] has joined #mythtv
05:30--- ---> nulltank_ [] has joined #mythtv
05:30--- ---> jeffpc_ [] has joined #mythtv
05:30--- Netsplit over, joins: Captain_Murdoch
05:30--- ---> poptix [] has joined #mythtv
05:30--- ---> kados [] has joined #mythtv
05:30--- ---> pigeon [] has joined #mythtv
05:30--- Netsplit over, joins: kvandivo_
05:30--- ---> nullman [] has joined #mythtv
05:30--- ---> BleedingE [] has joined #mythtv
05:30--- ---> riksta [] has joined #mythtv
05:30--- Netsplit over, joins: vagrant
05:30--- ---> kRutOn [] has joined #mythtv
05:30--- ---> Chutt [] has joined #mythtv
05:30--- ---> orangebits [] has joined #mythtv
05:30--- ---> onyxsoft [] has joined #mythtv
05:30--- ---> bbeattie [~bbeattie@] has joined #mythtv
05:30--- ---> Rince [] has joined #mythtv
05:30--- ---> jams [] has joined #mythtv
05:30--- ---> netman [] has joined #mythtv
05:30--- ---> dhofstra [] has joined #mythtv
05:30--- ---> Morph [] has joined #mythtv
05:30--- ---> dopester [] has joined #mythtv
05:30--- ---> Jayakumar__ [] has joined #mythtv
05:30--- ---> jfm [] has joined #mythtv
05:30--- ---> CoaxD [] has joined #mythtv
05:30--- ---> splat1 [] has joined #mythtv
05:30--- ---> Snow-Man [] has joined #mythtv
05:30--- ---> truls [] has joined #mythtv
05:30--- ---> DonLKSAB [] has joined #mythtv
05:30--- ---> stevenh [~xev@] has joined #mythtv
05:30--- Netsplit over, joins: bline
05:30--- ---> Wildgoose [] has joined #mythtv
05:30--- Netsplit over, joins: GreyFoxx
05:30--- ---> Matrix9 [] has joined #mythtv
05:30--- Netsplit over, joins: [M-M]
05:30--- ---> adante [] has joined #mythtv
05:30--- ---> mdz [] has joined #mythtv
05:30--- ---> Beirdo [~gjhurlbu@beirdo.user] has joined #mythtv
05:30--- Netsplit over, joins: taco
05:30--- ---> jwhite [] has joined #mythtv
05:31--- ---> m0j0 [] has joined #mythtv
05:31--- <--- Narkov [] has left #mythtv ("Kopete 0.8.2 :")
05:40--- Netsplit <-> quits: DonLKSAB, kRutOn, jwhite, hays, adante, dopester, GreyFoxx, Octane, Morph, bline, (+43 more, use /NETSPLIT to show all of them)
05:45--- Netsplit over, joins: o_cee, Captain_Murdoch, kvandivo_, vagrant, [M-M], bline, GreyFoxx, taco, m0j0, Omnic_ (+43 more)
06:20--- <<-- hamsta [] has quit (Read error: 60 (Operation timed out))
06:21--- <<-- mocker [] has quit (Remote closed the connection)
06:21--- ---> mocker [] has joined #mythtv
06:25--- ---> hamsta [] has joined #mythtv
06:48--- ---> cello [] has joined #mythtv
06:52--- <<-- cello [] has quit (Client Quit)
07:07--- ---> gfiend [] has joined #mythtv
07:10--- <<-- GeckoFiend [] has quit (Read error: 110 (Connection timed out))
07:38--- <<-- jwhite [] has quit (Read error: 113 (No route to host))
07:39--- <<-- hays [] has quit ("Leaving")
07:55--- <<-- netman [] has quit (Read error: 60 (Operation timed out))
07:57--- <<-- mocker [] has quit ("brb.")
07:57--- ---> mocker [] has joined #mythtv
07:57--- <--- mocker [] has left #mythtv ()
07:58--- ---> mocker [] has joined #mythtv
08:07--- ---> netman [] has joined #mythtv
08:48--- <<-- Octane [] has quit (Read error: 104 (Connection reset by peer))
08:52--- ---> Octane [] has joined #mythtv
09:08--- ---> jwhite [] has joined #mythtv
09:13--- <<-- hamsta [] has quit ("Bin jetzt weeeeech")
09:16--- ---> anduin [] has joined #mythtv
09:47--- User: *** gfiend is now known as GeckoFiend
09:47--- ---> linuxwhor1 [] has joined #mythtv
10:01--- <<-- linuxwhore [] has quit (Read error: 110 (Connection timed out))
10:11anduin the lcddevice.h include in mythdvd's main.cpp is string wrong (need <mythtv/lcddevice.h> instead of "lcddevice.h"
10:11anduin )
10:11anduin er still
10:24GeckoFiend anduin it compiled just fine here...
10:28anduin I don't see how, mtd is bad as well (changes MythContext ctor)
10:31GeckoFiend anduin I just fixed the header location for mythdvd
10:31GeckoFiend I probably missed the mtd check in
10:31anduin thanks!
10:33GeckoFiend constructor fixed as well, I forgot to check it in
10:49--- ---> dc_ [] has joined #mythtv
10:54stoffel GeckoFiend: good morning
10:58* mocker fails miserably at creating a splash.xpm.gz for his boot screen.
10:59GeckoFiend stoffel morning
11:01stoffel GeckoFiend: as you're working your way through the patch backlog, will you also apply my translations patch?
11:03GeckoFiend stoffel is that the one you put in in the "call for patches" thread?
11:03stoffel yupp
11:04GeckoFiend yeah i'll grab it in just a few
11:07GeckoFiend /var/lib/mythcvs/mythtv/i18n/mythfrontend_de.ts,v <-- mythfrontend_de.ts
11:07GeckoFiend new revision: 1.57; previous revision: 1.56
11:07GeckoFiend it's in
11:08stoffel hey, that was fast, thx
11:08GeckoFiend heh I figured I better do it now before I got caught up in this new test box I'm building and fvorgot about it
11:18jams hmm your new dev box is just nothing but trouble
11:18* GeckoFiend is starting to hate Suse
11:18GeckoFiend I'm sooo close to being able to take this out for a spin but stupid shit keeps cropping up.
11:19jams i know that feeling with suse
11:20GeckoFiend hrm anyone know what package might include I seem to only have the .a on this machine for some stupid reason
11:20jams start yast and do a search
11:21GeckoFiend my dev box conversion to Suse went fairly well, it's just the test box I'm building that's being a pain.
11:21GeckoFiend yast turned up nothing
11:21jams do you have the xfree86 compat libs installed
11:24GeckoFiend I'm not seeing any compat type package in yast either under compat or xfree86
11:24jams alright which suse is this ?
11:24GeckoFiend 9.1
11:25GeckoFiend ROFL the wife is playing sims 2 and one of her MEN gave birth to alien twins
11:30GeckoFiend trying to get apt to resolve this
11:33--- ---> jwhite_ [] has joined #mythtv
11:34jams in suse 9 /usr/X11R6/lib/ is in XFree86-compat-libs-
11:35jams my suse 9.1 is only a live edition
11:36GeckoFiend I just noted that I don't have the shared version on my dev box either :/
11:37--- ---> hays [] has joined #mythtv
11:40hays cool, my gigabyte nforce2 motherboard may have spdif out (gotta see if its supported in linux i suppose)
11:44hays any distro recommendation for mythtv? I'm thinking Gentoo
11:44jams hays these questions are a better fit for #mythtv-users
11:45hays damnit, I did it again. sorry
11:45riksta hays, if its nforce sound then itll prolly be the intel8x0 which has spdif support
11:46hays riksta, awesome. This means I can consider upgrading my box as an excuse. :)
11:46riksta go over to -users can talk there
11:48--- <<-- jwhite [] has quit (Read error: 113 (No route to host))
12:38dopester what is programid used for in the program table? is it safe to shove the dvb-eventID into this field or should I add a seperate field?
12:39riksta hey dopester mate
12:39--- ---> piksel [~piksel@] has joined #mythtv
12:40--- <--- piksel [~piksel@] has left #mythtv ()
12:40GeckoFiend dopester it's used to identify the program for the scheduler. As long as the dvb-eventID is a unique ID for a given show it should be ok
13:02--- ---> KillerBunny [] has joined #mythtv
13:04--- ---> JeffH [] has joined #mythtv
13:05JeffH is it easy to connect an existing dvd player through Myth-TV or would it be better to just buy a dvd player for my myth box?
13:08--- <--- JeffH [] has left #mythtv ("Kopete 0.9.0 :")
13:14--- ---> FreqMod [] has joined #mythtv
13:25riksta man, got some crazy ass thunder storm going on here
13:25--- ---> FreeZ [] has joined #mythtv
13:29--- <<-- FreqMod [] has quit (Remote closed the connection)
13:35GeckoFiend JeffH /topic
13:36--- <<-- dc_ [] has quit (Remote closed the connection)
13:53--- ---> cmorgan [] has joined #mythtv
14:10--- ---> gfiend [] has joined #mythtv
14:13--- <<-- FreeZ [] has quit (Remote closed the connection)
14:13dopester geckofiend: the eventID is unique for each event per channel
14:14dopester so that should be ok.. its not perfect saving a uint16_t in a varchar() field but seems silly to add 1 more column to the table
14:16--- ---> Baylink [] has joined #mythtv
14:17--- <<-- GeckoFiend [] has quit (Nick collision from services.)
14:17--- User: *** gfiend is now known as GeckoFiend
14:31dopester gfiend: does filldata.cpp generate this programId if the xmltv provider doesn't provide one?
14:33dopester i can only see if referenced by the dd code
14:36GeckoFiend not sure to be honest
14:53--- <<-- hadees [~hadees@hadees.sustaining.supporter.pdpc] has quit (Read error: 110 (Connection timed out))
15:16onyxsoft Wildgoose: just posted a link to the latest audio update with volume on the dev list.
15:26--- ---> hadees [~hadees@hadees.sustaining.supporter.pdpc] has joined #mythtv
15:31--- <<-- riksta [] has quit (Read error: 232 (Connection reset by peer))
15:38--- ---> krash314 [] has joined #mythtv
15:38--- ---> o_c [] has joined #mythtv
15:41--- <<-- o_cee [] has quit (Client Quit)
15:48--- <<-- mike2 [] has quit ("Leaving")
15:49Chutt geckofiend, please don't apply the audio patches - i've got them mostly done in my tree
15:49--- ---> dc_ [] has joined #mythtv
15:57Beirdo Heya Chutt, how's the weekend treating you?
15:57GeckoFiend Chutt I was waiting for the dust to settle on those. I thought they were a work in progress still anyway
16:02Chutt yeah, mainly
16:02Chutt but it shouldn't break oss, which is all i care about
16:03Chutt weekend's fine, though =)
16:03Beirdo That's good to hear. :) A good weekend's a good thing. Gotta enjoy the time off, etc.
16:16--- <<-- Majestik [] has quit ("Leaving")
16:19--- <<-- adante [] has quit (Remote closed the connection)
16:19--- ---> adante [] has joined #mythtv
16:22stoffel hmpf, translating all these messages feels like fighting a hydra. you finish 5 messages and PLOP, 15 new messages appear after a cvs update :(
16:23Wildgoose Chutt: That's great to hear. The new stuff is only additions to the current stuff, so it would be great if we could get the bulk of it in, and then I will just submit incremental patches from there.
16:24Wildgoose onyxsoft: I'm not convinced about how you are doing the volume control stuff?
16:25Wildgoose Why bother with the seperate volume control object?
16:27Wildgoose I think a better method would be to move the volume control stuff (unique to each layer) into the AudioOutput class. You then need a class still to handle general volume control stuff
16:28Wildgoose However, I think this can't go in AudioOutputBase, because it's not REQUIRED that this ever be implemented (see the DirectX output for example)
16:28Wildgoose So we need VolumeBase, which is inherited by AudioOutput and then job's done.
16:28Wildgoose Why not go this way?
16:29--- ---> riksta [] has joined #mythtv
16:37onyxsoft Wildgoose: It was the easiest/quickest way I could get it implemented without breaking other stuff.
16:38onyxsoft At least the OSS and ALSA volume have been refactored so it should be easier to do the type of volume control you wanted now ;-) Maybe?
16:40--- <<-- jeffpc_ [] has quit ("Client exiting")
16:46--- <<-- m0j0 [] has quit ("Leaving")
16:46Wildgoose Yeah, I think it should be very easy. In fact I think it only needs a minor refactoring of the stuff you have done. Which is why I am unsure why you didn't do it?
16:47Wildgoose Just merge the volumealsa code into AudioOutputAlsa. Hooking the init stuff into OpenDevice, and the close stuff into CloseDevice
16:47Wildgoose Then make AudioOutput inherit from VolumeControl (or whatever you want to call it), and put all generic code there.
16:49Wildgoose I would also suggest taking the time to reconcile a few of those functions. It's not really a clean interface right now. For example do we really need an "increment" volume function?
16:49Wildgoose Do we really need a "toggleMute" function?
16:50Wildgoose Why not change this to a MonoToStereo function..?
16:50Wildgoose (somewhere in the player, or perhaps even in the AudioOutput class?)
16:50cmorgan Wildgoose: hey man
16:51--- <<-- riksta [] has quit (Remote closed the connection)
16:51jams Wildgoose, did you go by the name mrgoose a few months ago ?
17:04--- <<-- dopester [] has quit (Read error: 60 (Operation timed out))
17:08Wildgoose nope not me
17:08Wildgoose wildgoose is my surname (really)
17:09Wildgoose hey Chris
17:09cmorgan Wildgoose: i just put in that fix to jack_write() that i for some time couldn't figure out what you meant by it ;-)
17:09cmorgan let me check things in so you can update and see whats there now
17:10Wildgoose Cool
17:11Wildgoose You decided what to do about the free space calcs?
17:12cmorgan Wildgoose: yeah, i just put the old ones back and renamed the new ones to super long names like JACK_GetBytesStoredRealTimeApproximation() ;-)
17:13cmorgan i'm not usually a huge fan of long function names but if they make it easy to see wtf is going on its worth it
17:15cmorgan Wildgoose: alright things are checked in, you see them?
17:21--- ---> LLyric [] has joined #mythtv
17:23Wildgoose I still REALLY don't like having both functions though!
17:23cmorgan heh
17:23cmorgan internally JACK_GetBytesStored() is used to compute the amount of free space
17:23Wildgoose I also can't see any need for a getfreespace function which doesn't tell me how much free space there is? (ie the approximate version)
17:24Wildgoose Honestly, the only right answer here is the approximate one for BytesBuffered
17:24cmorgan Wildgoose: its mostly the other side of JACK_GetPosition(), if people really want it then they can use it ;-)
17:24Wildgoose Think what the alsa driver or the oss driver would tell you
17:24Wildgoose I'm not sure what you mean by that?
17:24cmorgan Wildgoose: thats my point. if you want the realtime approximation of the bytes stored then use that function
17:24cmorgan Wildgoose: but internally the non-adjusted version is used to compute the free space
17:25Wildgoose Yeah, but don't make the interface cluttered
17:26Wildgoose I like what you have got, but I think it's almost at the point where we have to rationalise what we have got before we end up supporting silly stuff because it's used by the end clients
17:27Wildgoose I think you are looking at it a lot from the point of view of the API, and perhaps not from the intended use?
17:27cmorgan Wildgoose: there is a use for JACK_GetBytesStored() and also for the real-time equivalent of that, so both are available
17:27Wildgoose For example, we could potentially chuck out all the Bytes stored stuff and just have a "latency" function...?
17:28cmorgan Wildgoose: latency in terms of time? bytes?
17:28Wildgoose OK, so I don't see the point for JACK_GetBytesStored? Educate me a little
17:28Wildgoose It wouldn't matter really
17:28Wildgoose "frames" for preference
17:28cmorgan Wildgoose: well i use it internally to determine the amount of free space ;-)
17:29cmorgan Wildgoose: frames is a good way to express latency...
17:30cmorgan Wildgoose: i guess i'm not understanding your concern, that there are functions that no one should need to use?
17:30Wildgoose Yeah, basically I think you are in danger of function overload and supporting a load of unnecessary stuff going forward
17:30Wildgoose A nice clean (concise) interface is always my goal
17:31Wildgoose always trying to *remove* stuff...
17:31Wildgoose Perhaps what would be useful would be to break the public and internal interfaces a bit more
17:31Wildgoose Otherwise people will use your internal functions and then you will find it hard to change them later
17:31cmorgan Wildgoose: i mean i can see the use for knowing how many bytes are stored in bio2jack AND getting a real-time approximation of that if necessary
17:31Wildgoose Please don't see this as criticism. It's more just trying to make things better for the future really
17:32Wildgoose I think you miss the point
17:32truls future is f...ed anyway ;)
17:32truls live for the now
17:33Wildgoose For example, I can't think of any use off the top of my head to know those two different bytes numbers
17:33Wildgoose Dunno if you can?
17:33Wildgoose Externally that is...
17:33Wildgoose Taking a step back and designing a new interface for example: Blocking I/O basically needs:
17:33cmorgan Wildgoose: xmms-jack uses JACK_GetBytesStored() when it is winding down its playback
17:33--- ---> choenig [] has joined #mythtv
17:34Wildgoose a) Write some bytes, b) Get Output latency, c) How many bytes can we write without blocking, d) ...can't think of another one...
17:34cmorgan Wildgoose: i suppose the real-time value would work, the issue is that we might buffer enough such that xmms-jack would reset things before the buffer completely flushed out
17:34Wildgoose Yeah, but XMMS-Jack *should* be useing GetBytesStoredApproximate + Latency + JackLatency
17:35cmorgan Wildgoose: xmms-jack also uses JACK_GetBytesFreeSpace(), i suppose it could compute that from JACK_GetBytesStored() and the max buffer size
17:35Wildgoose FreeSpace == c)
17:35Wildgoose So you have that
17:35Wildgoose You also know when to reset things, that's b)
17:36cmorgan Wildgoose: so you'd suggest removing either the free space or the bytes stored since you can compute one from the other, assuming you know the amount buffered
17:36Wildgoose I think you must be resetting things too early right now to be honest. You have forgotten to count the bytes in the jack engine
17:36stoffel is Kenneth around here?
17:36Wildgoose Well, although it's nice to know total bytes buffered, this is barely a d).
17:37Wildgoose Have a look at, say, the mplayer audio API. It's as I describe above. a), b) and c)
17:37cmorgan Wildgoose: its alright to reset the client as long as the bytes are into the engine
17:38Wildgoose Agreed, but now your starting to "know" something about the output layer. In general purpose clients this is just too specific. They shouldn't know this stuff
17:38Wildgoose Agreed, "high-performance" stuff is going to need to know this kind of thing, but I think that this kind of app will need a full Jack layer anyway
17:38cmorgan Wildgoose: well thats my point, if bio2jack says i've got no bytes stored then its fine to reset things, the bytes are played into jack
17:38Wildgoose For example, I have an app which measures impulse response of rooms, and needs sub sample accurate playback and recording...
17:39Wildgoose *that* is a high performance app for example
17:39cmorgan alright, so i'd like to drop either JACK_GetBytesFreeSpace() or JACK_GetBytesStored()
17:40cmorgan and as you mentioned, free space is a pretty common thing apps want
17:40Wildgoose OK, well, I could buy your need, but I would still ask you whether it matters that you reset the output layer while there are still two unplayed buffers in jack, or whether you let it underflow a buffer or two before filling it with new data. Can humans hear the difference? Or do you then get a click between songs?
17:41Wildgoose In fact, why are you resetting it at all?
17:41cmorgan Wildgoose: internally i don't reset, i'm just saying hypothetically
17:41Wildgoose Yeah, but hypothesise a real scenario?
17:42cmorgan Wildgoose: the number should be accurate. if all bytes have left bio2jack then the api should reflect that
17:42Wildgoose If you were going to reset then you will get a click/silence, so who cares if it's one sample long, or two samples long...?
17:42cmorgan Wildgoose: so do you need the real-time approxmation of the bytes stored in bio2jack?
17:42Wildgoose Yes
17:42Wildgoose Otherwise you can't get AV sync
17:42cmorgan but you don't think anyone needs the actual amount of bytes in jack?
17:42cmorgan in bio2jack rather?
17:43Wildgoose Well, not off the top of my head
17:43Wildgoose Suggest remove it and add it back if someone can think of a reason...
17:43Wildgoose (hope it's not me...)
17:44cmorgan and i can fix xmms-jack to instead compute JACK_GetBytesStored() from the max buffered bytes and the current free space
17:44cmorgan any other function calls you think are extranious?
17:44Wildgoose Ohhh, well if you need it, then don't remove it
17:44Wildgoose However, lets go back and examine your use for a second
17:44cmorgan Wildgoose: oh i really don't. i can compute it from two other values so its not a big deal
17:44Wildgoose Why do you need the exact amount and not the approx amount
17:44cmorgan Wildgoose: like you say, it should simplify things
17:44Wildgoose No, lets not compute silly things
17:44Wildgoose It will only break later
17:44cmorgan Wildgoose: yes, thats my point
17:45Wildgoose OK, so why do you need buffered bytes, not estimated
17:45Wildgoose Trying to get my head around that
17:45cmorgan Wildgoose: i don't, the estimated is probably fine
17:45cmorgan Wildgoose: if i really need the actual buffered bytes i can compute that.
17:45Wildgoose So you won't do TotalBuffered - FreeSpace?
17:46cmorgan Wildgoose: nahh, i don't think its necesary, the approximate value works just as well
17:46Wildgoose Yeah, but if you DO need it, then by all means add the function. I just don't think that you will find a need for it, and someone will instead use it by mistake
17:46cmorgan Wildgoose: if anything it would mean xmms would start to play the next song at a more correct time
17:46Wildgoose Agreed
17:46cmorgan Wildgoose: nahh, the user level app can compute that value
17:46Wildgoose Now, lets do some more rationalisation
17:47cmorgan Wildgoose: alright, so i'll can JACK_GetBytesStored() now that we have the real-time approximation
17:47Wildgoose Lets remove the latency functions and roll them into the BytesStored...
17:47Wildgoose hang on
17:47Wildgoose OK, go back to your last point
17:47Wildgoose Yes, lets keep BytesStored == estimated values
17:47cmorgan i'm still keeping the new name btw
17:48cmorgan which latency functions are you speaking of?
17:48Wildgoose And FreeSpace = Amount of bytes I can write *NOW*, so this probably means not estimated
17:48cmorgan yep, free space will not be an estimate
17:48Wildgoose Hang on, lets just finish off these two functions
17:48Wildgoose OK, so we agree on this then?
17:48Wildgoose In fact of course you don't need to change anything because internally we probably need all these functions
17:49cmorgan yep, i'll make the free space function not rely on jack_getbytesbuffered
17:49Wildgoose Really all we are saying is that the external interface is getting trimmed
17:49cmorgan i think i can trim the internal interface as well
17:50Wildgoose Sure, but really it's the external interface that I would like to see simplified
17:51Wildgoose For example just looking at the mplayer driver you could see the kinds of error that people will make
17:51Wildgoose OK, so lets keep in mind, that my ideal (simple) interface is a) write, b) get output latency, c) free space
17:51cmorgan yep
17:51cmorgan your next point? ;-)
17:52Wildgoose So I would like to trim back the GetLatency, and BufferSize functions and roll them into the BufferedBytes function...
17:52Wildgoose WHat do you think?
17:52cmorgan JACK_GetJackLatency()?
17:54Wildgoose Yeah, consider that the mplayer/myth "get_delay" functions both look like this:(JACK_GetBytesStored(driver) + approx_bytes_in_jackd + latency)
17:54cmorgan hmm
17:54Wildgoose Where latency==GetJackLatency and approx_bytes == your new function
17:54cmorgan yeah
17:54Wildgoose to get the Jack callback size
17:55cmorgan i could go for making that change
17:55Wildgoose In fact I need to double that number assuming that there are two buffers...
17:55cmorgan since it really gives the end app more of an idea of how many bytes of latency there are between them and physical output
17:55Wildgoose So really, I would like the bio2jack to estimate all those other latencies (monitoring them if they change!) and then tell me the answer
17:55cmorgan well i'll ask jack each time
17:56Wildgoose Indeed, and it's essential for AV sync otherwise the voices are ahead of the video
17:56cmorgan you call JACK_GetBytesStored() and i'll do the real-time approximation I do now and then add on the latency through jack to th eoutput device
17:56Wildgoose Excellent
17:56Wildgoose So you need to add jack "latency" (ie stuff in the graph) and then also the number of buffers * buffer size
17:57cmorgan i don't know that amount
17:57Wildgoose Basically if we later improve this algorith, then everyone gets the benefits downstream
17:57cmorgan or rather
17:57Wildgoose Whereas now they need to recode to use the new functinos
17:57cmorgan thats the amount of the bytes that jack wants during each callback
17:57Wildgoose I think you should have a function to get the number of periods?
17:57cmorgan and then subtract away the time between the now and last callback to compute the bytes that have been played out
17:57cmorgan hmm
17:58cmorgan i'd have to look at the jack api more carefully to see if thats the case, you want to take a look? ;-)
17:58Wildgoose Exactly. What we really want to know is the estimated number of frames between the top of the buffer, and the frame leaving the soundcard right now
17:58cmorgan yep
18:00cmorgan so i think its pretty simple, bio2jack buffer + jack port latency + bytes buffered in jack - playback_rate * timesincelastcallback
18:00cmorgan or rather bio2jack buffered bytes is what i meant
18:02Wildgoose Hmm, I see what you mean - looked at the api and I can't see a function to get number of buffers...
18:02Wildgoose Better stick to estimating it to be two...
18:02Wildgoose hang on, gotta go for a sec
18:03--- ---> Matt [matt@matt.staff.freenode] has joined #mythtv
18:03--- <<-- KillerBunny [] has quit (Read error: 104 (Connection reset by peer))
18:07cmorgan Wildgoose: i've got to head out for dinner man
18:08cmorgan Wildgoose: i'd like to hear which other functions you want to roll together though
18:08cmorgan Wildgoose: i'll try to get the latest round of changes checked in tonight as well
18:18--- ---> cfreeze [] has joined #mythtv
18:19cfreeze anyone else getting segfaults from the latest cvs on MythContextGetLCDDeviceEv being an undefined symbol?
18:19cfreeze happens loading all the plugins, but not the tv
18:20--- <<-- choenig [] has quit (Remote closed the connection)
18:21cfreeze LCD_DEVICE wasn't defined either
18:21cfreeze I didn't remove the comment marker rather
18:21cfreeze looks like it has to be defined for the current CVS to build/work
18:22cfreeze builds ok, but the segs I mean
18:24LLyric Are y'all interested in end-user myth segfaults?
18:24cfreeze looks like some ifdefs are missing that's all
18:25Wildgoose cmorgan: Great. I think we achieved some stuff here. Gotta go now. Having a kitchen and bath room fitted next week... So may be slow to respond.
18:26Wildgoose However, basically, I think we should pull lots of stuff out of the public api and then only put it back if someone needs it. Less future support that way, and it will be easier for new users to make use of it
18:26Wildgoose Bye for now
18:27--- <<-- hays [] has quit ("Leaving")
18:27--- <<-- Baylink [] has quit ("User disconnected")
18:27LLyric I take it you didn't want end-user myth segfaults, then?
18:33--- ---> choenig [] has joined #mythtv
18:34--- User: *** Omnic_ is now known as Omnic
18:36--- <<-- choenig [] has quit (Remote closed the connection)
18:37Beirdo LLyric: the docs state how to submit such coredump information, I believe
18:37* LLyric looks
18:41--- <<-- linuxwhor1 [] has quit (Client Quit)
18:41--- ---> linuxwhore [] has joined #mythtv
18:45stoffel GeckoFiend: the mythmusic patch from Kenneth hasn't been applied completely, databasebox.cpp hasn't been updated
18:46stoffel see;#87616 ; the mythmusic.patch
18:46stoffel in the last post of the thread
18:48GeckoFiend commited
19:04--- <<-- linuxwhore [] has quit ("Download Gaim:")
19:10--- ---> dopester [] has joined #mythtv
19:12* LLyric instals gdb so he can chase a segfault
19:12LLyric Probably bad program information, but shouldn't cuse a segfault
19:20--- <<-- dc_ [] has quit (Read error: 104 (Connection reset by peer))
19:20--- ---> dctanner [] has joined #mythtv
19:21--- ---> Majestik [] has joined #mythtv
19:21--- User: *** dctanner is now known as dc_
19:28LLyric Where's the bugzilla about mythtv?
19:29LLyric I've got a repeatable segmentation fault, looks like QT
19:29GeckoFiend send a backtrace to the dev list
19:30GeckoFiend what distro are you using?
19:30LLyric FC2 with atrpms
19:30LLyric Not subscribed, just trying to be helpful
19:30GeckoFiend post it to the users list then
19:31LLyric I've got a guess what it is - testing
19:35LLyric Nope.
19:35LLyric Thought it didn't like my channel icon
19:38--- <<-- hadees [~hadees@hadees.sustaining.supporter.pdpc] has quit (Success)
19:39--- ---> hadees [] has joined #mythtv
19:52--- <<-- cfreeze [] has quit ("BitchX: often imitated, never duplicated!")
19:58--- <<-- dc_ [] has quit (Remote closed the connection)
20:00--- <<-- DonLKSAB [] has quit (Read error: 110 (Connection timed out))
20:02* LLyric idles, waiting for the source to download, so he can produce a patch for his segfault
20:03LLyric Can I post three lines of trace here, in case it looks faimilar to anyone?
20:03Snow-Man Sure
20:03jams GeckoFiend, you have mail
20:03Snow-Man Well, provided you have debugging symbols
20:04LLyric Well, there are in those three lines
20:04LLyric #0 0x05a07b37 in operator== () from /usr/lib/qt-3.3/lib/
20:04LLyric #1 0x05a07bc6 in operator!= () from /usr/lib/qt-3.3/lib/
20:04LLyric #2 0x06b9e939 in GuideGrid::paintChannels () from /usr/lib/
20:04LLyric Don't have symbols in libqt, but can probabyl spot it in GuideGrid
20:04Snow-Man erm.
20:04Snow-Man That looks mighty useless.
20:05LLyric Yeah
20:05Snow-Man Guess I should have said "unless it's useless"
20:06LLyric I've got more, but it just comes out of the message pump
20:06LLyric WHich isn't any more useful
20:06Snow-Man Well, recompile w/ debugging symbols and stuff.
20:06LLyric paintChannels <-- paintEvent <-- QWidget:Event <-- QApplication:internalNotify
20:07GeckoFiend jams got it thanks
20:18--- ---> mechou [] has joined #mythtv
20:31--- <--- mechou [] has left #mythtv ("Kopete 0.8.3 :")
20:44Chutt LLyric, for a backtrace to be useful, you must compile in debug mode. What you pasted hasn't been.
20:45LLyric "debug mode" is just -g?
20:46Chutt no
20:46Chutt debug mode is turning on debug mode like the docs tell you to.
20:59--- ---> CyberKnet [] has joined #mythtv
21:08--- ---> m0j0 [] has joined #mythtv
21:08--- <<-- m0j0 [] has quit (Client Quit)
21:09--- ---> m0j0 [] has joined #mythtv
21:19--- ---> stoffel_ [] has joined #mythtv
21:31--- <<-- stoffel [] has quit (Read error: 110 (Connection timed out))
21:33--- ---> ta[i]nted [] has joined #mythtv
21:58--- ---> linuxwhore [] has joined #mythtv
22:06--- <--- ta[i]nted [] has left #mythtv ()
22:19--- <<-- anduin [] has quit ()
22:35--- <<-- Matrix9 [] has quit (Read error: 104 (Connection reset by peer))
23:14--- <<-- linuxwhore [] has quit (Connection timed out)
23:21--- ---> linuxwhore [] has joined #mythtv
23:40--- ---> mdz_ [] has joined #mythtv
23:47--- <<-- mdz [] has quit (Read error: 110 (Connection timed out))
23:50--- <<-- CyberKnet [] has quit ("Dedicated to all I ever was, and what I never will be...")
23:55--- User: *** mdz_ is now known as mdz
23:57--- <--- mdz [] has left #mythtv ("Client exiting")
23:58--- <<-- linuxwhore [] has quit ("Download Gaim:")
---Logclosed Mon Oct 25 00:00:20 2004