Back to Home / #mythtv / 2004 / 08 / Prev Day
#mythtv IRC Logs for 2004-08-21

---Logopened Sat Aug 21 00:00:45 2004
00:03--- ---> snowblind [] has joined #mythtv
00:17--- User: *** Rince is now known as RinceMasterOfDNS
00:23--- ---> billytwowilly [] has joined #mythtv
00:26--- User: *** RinceMasterOfDNS is now known as Rince
00:27Chutt captain_murdoch, tricky.
00:31--- ---> Wilbur [] has joined #mythtv
00:31--- <<-- duckeo [] has quit (Read error: 54 (Connection reset by peer))
00:45dopester hey chutt
00:46Chutt good morning
00:47dopester wow it is morning
00:50--- ---> KarlosII [] has joined #mythtv
00:51Chutt i wonder if there's any patches i missed
00:54Captain_Murdo| almost think I have it. just have the various status routines fill in a map of qstrings instead of calling SetItemText themselves, then have updateContent() display the lines starting at the proper index, seems to be working ok for my job queue status info. now to figure out what I did wrong when using a qvaluevector that caused it not work when it works fine with a qmap. :(
00:54Chutt heh
00:55Chutt cool.
00:55Chutt i wonder if anyone would mind if i just up and break the ringbuffer
00:55Captain_Murdo| lets me scroll properly though. can see all queued jobs. need to adapt the other status boxes as well.
00:56Captain_Murdo| sure, you could do that and people wouldn't even notice if I broke something by committing that big decoder patch that consolidates the common functions like dorewind/dofastforward into decoderbase.cpp
00:56Chutt hehe
00:57Chutt that came up at work today, actually =)
00:57Chutt well, not that, but that situation
00:57Chutt i broke the build, checked in a fix a few minutes later
00:57Chutt but someone else broke it another way between the two, so i still got fingered for the breakage after i fixed my problem
00:58Chutt go ahead and commit your patch, though
00:58Chutt i'm sure people'll bitch if you broke something =)
00:58Captain_Murdo| :) I notice that someone committed a change to uitypes.cpp to get it compiling on older gcc's. I went to fix that yesterday and saw my diff conflicted with current CVS.
00:59Chutt yeah, i did that
00:59Captain_Murdo| works fine for nuppel and avfformatdecoder, can't test ivtvdecoder. and need to test that micro-jump stuff that was added for transcoded files that have that extra frame map.
00:59Chutt the guy submitted an updated patch, so i fixed the at's while i was applying it
01:00Captain_Murdo| there's one more in there I found this afternoon. line 1570.
01:00Chutt you going to fix it?
01:00Chutt i wouldn't worry too much about breaking ivtvdecoder
01:00Chutt it should be similar enough
01:00Chutt that it's easy to fix if needed
01:01Chutt and that's a small minority of users
01:01Captain_Murdo| yeah, I'll commit the at() fix when I do my next commit;
01:02Captain_Murdo| ok. do you think it's ok for the content box on the status screen to update automatically when you scroll on the left hand side?
01:02Chutt doesn't matter to me
01:03Captain_Murdo| I'm gonna enable that then when I get around to committing this status page for the job queue.
01:11Chutt rkulagow, are you still around?
01:14--- <<-- cmorgan [] has quit (Remote closed the connection)
01:18Chutt rkulagow, if you see this, what version of qt are you using?
01:49Captain_Murdo| cool, even have the up/down arrows working. scrolling on that listbox was easier than the playbackbox because there's no need for a highlighted bar to sit in the middle, you just scroll the whole block of text.
01:51--- User: *** david_ is now known as Viddy
01:57--- <<-- billytwowilly [] has quit ("Leaving")
02:10--- ---> ads [] has joined #mythtv
02:12--- User: *** ads is now known as adante
02:16--- <<-- Wilbur [] has quit (" - empower your web presence")
02:24--- <<-- KarlosII [] has quit (Read error: 104 (Connection reset by peer))
02:24--- ---> KarlosII [] has joined #mythtv
02:37--- <<-- KarlosII [] has quit (Read error: 104 (Connection reset by peer))
02:38--- ---> KarlosII [] has joined #mythtv
02:39snowblind chutt: do you use the cvs version of myth on your main setup
02:39--- <<-- Netslayer [] has quit (Remote closed the connection)
02:56--- <<-- KarlosII [] has quit (Read error: 104 (Connection reset by peer))
02:56--- ---> KarlosII [] has joined #mythtv
02:59--- ---> KarlosII_ [] has joined #mythtv
02:59--- <<-- KarlosII [] has quit (Read error: 104 (Connection reset by peer))
03:11--- <<-- KarlosII_ [] has quit (Read error: 104 (Connection reset by peer))
03:12--- ---> KarlosII_ [] has joined #mythtv
03:16--- ---> KarlosII__ [] has joined #mythtv
03:16--- <<-- KarlosII_ [] has quit (Read error: 104 (Connection reset by peer))
03:18--- <<-- KarlosII__ [] has quit (Read error: 104 (Connection reset by peer))
03:18--- ---> KarlosII__ [] has joined #mythtv
03:50--- <<-- KarlosII__ [] has quit ("Leaving")
04:01--- ---> gfiend [] has joined #mythtv
04:11--- ---> choenig [] has joined #mythtv
04:19--- <<-- GeckoFiend [] has quit (Read error: 110 (Connection timed out))
04:46--- ---> mithro [] has joined #mythtv
05:04--- <<-- snowblind [] has quit ()
05:53--- ---> vagrant [chip@] has joined #mythtv
06:11--- <<-- Agrajag- [] has quit (Read error: 113 (No route to host))
07:04--- ---> Killerbun [] has joined #mythtv
08:04--- User: *** gfiend is now known as GeckoFiend
08:06--- ---> lucas3333 [] has joined #mythtv
08:22--- <<-- nrh1703 [] has quit (".<UPP>.")
08:40--- <<-- Killerbun [] has quit (Read error: 104 (Connection reset by peer))
08:40--- ---> billytwowilly [] has joined #mythtv
08:44--- User: *** GeckoFiend is now known as Gecko_GoneForDay
09:06--- <<-- vagrant [chip@] has quit (Read error: 60 (Operation timed out))
09:28--- <<-- choenig [] has quit (Remote closed the connection)
09:50--- <--- prophecy__ [] has left #mythtv ()
10:08--- ---> choenig [] has joined #mythtv
10:24--- ---> cmorgan [] has joined #mythtv
10:53--- ---> snowblind [] has joined #mythtv
10:53--- <<-- adante [] has quit ("Lost terminal")
10:53--- ---> adante [] has joined #mythtv
11:20--- <<-- pigeon [] has quit (Connection timed out)
11:38--- ---> lisadaveh [] has joined #mythtv
11:56* Beirdo is away: blah
11:56* Beirdo is back (gone 00:00:03)
12:09--- ---> sfr_ [] has joined #mythtv
12:09Chutt hmm
12:09Chutt code up new tv stuff, or play gamecube
12:10--- <<-- stoffel [] has quit (Read error: 110 (Connection timed out))
12:10thor_ new tv, then gamecube
12:10Chutt either one's going to take all day :p
12:10thor_ ah
12:10cmorgan heh
12:10thor_ well then gamecube of course
12:11Beirdo it's the weekend.. playing games takes priority
12:11thor_ all work and no play makes Chutt a pedantic bastard
12:11Chutt _but_
12:11Chutt i also have to sand the walls in the basement
12:11Chutt so, hmmm
12:11Chutt no coding today =)
12:11thor_ heh
12:11Beirdo real life gets in the way far too often, doesn't it?
12:18thor_ bah ...
12:19thor_ my little brain cannot figure out how to avoid iterating over a content to tree and (pre-checking) metadata items to reflect the state of a given playlist
12:23--- User: *** sfr_ is now known as stoffel
12:31Chutt thor, unless the db is redone
12:31Chutt that's probably the only way
12:32thor_ hmmm
12:32thor_ db redone ?
12:32thor_ ah, right
12:33Chutt playlist format changed
12:33Chutt ok
12:33Chutt so, if live tv becomes a series of files
12:34Chutt what's the behavior if the user is in the earliest file, and it wants to delete that
12:34Chutt to emulate a ring buffer, essentially
12:34thor_ bah, go sand mud
12:34--- <<-- choenig [] has quit (Remote closed the connection)
12:35thor_ in memory, what I really want is a many to many mapping so I can just iterate over the playlist
12:36thor_ hmmm .... many to many with pointers
12:36thor_ err, no ... just one to many
12:37thor_ fast lookups in Dict's don't work .... that's one to one
12:37Chutt there's always a multimap
12:38thor_ sounds promising
12:38Chutt map with multiple entries, standard stl
12:38thor_ and some kind of find() that returns a list ?
12:39bobnvic Chutt: what's the behavior now if the user is at the beginning of the ringbuffer and reaches the set size limit?
12:39* thor_ really needs to stop relying on Qt for wrapping STL in well documented classes
12:39--- ---> choenig [] has joined #mythtv
12:40thor_ mmmm ... looks value based
12:40thor_ well, pointes as values
12:40thor_ would work
12:41--- <--- ipso [] has left #mythtv ("Client exiting")
12:42bobnvic Chutt: could have one of two options - 1) stop recording tv, keeping the existing files so the user finsih watching what they paused on.
12:42bobnvic 2) bump them to the next file and delete the one they were on and keep recording
12:44bobnvic I'd probably prefer 1) since it would make more sense to me. If I paused tv and went away for a while, I'd probably want to pick up where I left off, otherwise, why pause?
12:44--- ---> o_cee [] has joined #mythtv
12:44Chutt bobnvic, first question, it unpauses you
12:44Chutt i don't like stopping recording, though
12:44Chutt hmm
12:46bobnvic maybe you could stop recording until the user was past that file, then delete it and continue recording - I can see lots of problems with that though
12:46Chutt yup
12:47Chutt maybe i'll just delete the next recording
12:47bobnvic any idea how much time each file would include, 5 minutes, 30 minutes, etc? I assume it would depend on data rate, but in general
12:48bobnvic I would think that the data that is least important would be the data that is eithe behind or the farthest in front of the pause point.
12:49o_cee Chutt: will the themes need to be re-done (more than to take advantage of new stuff) to work with mythui?
12:50Chutt yes.
12:50Chutt bobnvic, a single program, up to 2 hours
12:50o_cee allright, good to know
12:50Chutt was what i was thinking
12:50Chutt longer things would get chunked
12:50rkulagow_ chutt: according to dpkg -l qt*, it's 3.2.3-4 (this is on debian testing)
12:50bobnvic maybe delete the last recorded file. would need some sort of message to let the user know that there is a gap of x minutes in recording
12:50Chutt rkulagow, hmm
12:50Chutt ok
12:54o_cee Chutt: know i aksed you this before, but would it be possible to have a hidden dir in cvs where i could put my original files for creating the titles and stuff like that for gant?
12:54Chutt a hidden dir?
12:54Chutt no.
12:54o_cee yeah, or at least something thats ignored, doesn't need to be downloaded to everyone checking out cvs
12:55bobnvic with the file scheme couldn't you rewind past channel changes? that would seem like a definite plus.
12:56Chutt bobnvic, yup
12:58bobnvic wouldn't this also allow you to record on two tuners when using PIP so that you don't lose your recording when switching between them?
13:00thor_ .lower_bound() .upper_bound() <--- that's the ticket
13:00bobnvic I suppose if you wanted to be really smart about the file usage you could look at the program guide and see when the program that they are paused on will end or has ended.
13:02--- ---> splat1 [] has joined #mythtv
13:02bobnvic i take it you're going to have each show start a new file, so that's why you said delete the next recording. am i following you now?
13:03Chutt yup
13:03Chutt gotta go, back later
13:03bobnvic excellent, sounds great
13:05bobnvic as usual, you all are way ahead of me
13:26dopester yeah! im able to grab half of the dvb si tables now
14:03dopester autoscan is inching closer now
14:08--- ---> vagrant [] has joined #mythtv
14:52dopester got myth scanning at least dvb-s now..
14:52dopester gotta make it verify against the db now
15:20--- ---> Killerbun [] has joined #mythtv
15:20--- <<-- Killerbun [] has quit (Client Quit)
15:31--- ---> Killerbun [] has joined #mythtv
15:50--- ---> gfiend [] has joined #mythtv
15:53--- ---> woodwizzle [] has joined #mythtv
15:56--- <<-- Captain_Murdoch [] has quit (Read error: 110 (Connection timed out))
15:57--- <<-- Gecko_GoneForDay [] has quit (Read error: 104 (Connection reset by peer))
15:58thor_ Chutt, as it seems impossible to build a lib and an app in the same dir with qmake, ok if I create mythui/lib/ and mythui/test ?
16:02thor_ or would we rather put just the test app in a subdir that does not get built by default ?
16:02--- <<-- vagrant [] has quit (No route to host)
16:03--- ---> Netslayer [] has joined #mythtv
16:04--- ---> vagrant [chip@] has joined #mythtv
16:26--- ---> AridWork [] has joined #mythtv
16:32thor_ I officially think that UIListBtnTypeItem::CalcDimensions() is crap
16:36--- <<-- Killerbun [] has quit (Read error: 110 (Connection timed out))
16:38dopester Ive noticed the use of pthread and qmutex is there one that should be used over the other?
16:39thor_ I think (?) almost everything is QThread at this point
16:40dopester since im messing with dvbsections.cpp should i change them to qthreads?
16:40dopester i kinda figured out was deprecated, but didnt want to touch that stuff if there was a reason to have both
16:41thor_ hmmmm
16:42dopester there is a note in the cvs about it being detached threads "until someone can clean this crap up"
16:42thor_ if you are fundamentally redoing things, then it's probably not a bad idea .... how long has it been like that .... checking cvs ....
16:43dopester ages
16:44dopester i probbably will create a base class for it anyway since a good deal of that code could be re-used by ATSC or OPENCABLE table parsers i bet
16:44thor_ right ... well, you probably want to check with Chutt on this in any case
16:45thor_ but if I were really doing a fundamental rewrite, I would redo it with QThreads
16:47dopester aight
16:48thor_ then again, the likelyhood of me even understanding dvbsections is vanishingly small
16:49dopester heh
16:49dopester its taken me a while to fully understand it
16:52--- ---> Captain_Murdoch [] has joined #mythtv
17:33--- ---> Killerbun [] has joined #mythtv
17:33--- <<-- Killerbun [] has quit (Client Quit)
17:33--- ---> Killerbun [] has joined #mythtv
17:46thor_ this model/view stuff in listbttn is absolutely killing me
17:49--- <<-- jams [] has quit (Read error: 104 (Connection reset by peer))
---Logclosed Sat Aug 21 18:01:44 2004