Back to Home / #mythtv / 2003 / 04 / Prev Day | Next Day
#mythtv IRC Logs for 2003-04-17

09:36--> Youare now talking on #mythtv
09:36-!-Topic for #mythtv is http://www.mythtv.org/
09:36-!-Topic for #mythtv set by Viddy at Sun Mar 9 06:04:45
09:42-!-Ndiin [ruler@ip68-3-77-245.ph.ph.cox.net] has joined #mythtv
11:05-!-StarHeart [edgan@64-42-21-228.atgi.net] has joined #mythtv
11:24-!-ply [~ply@12-246-43-38.client.attbi.com] has joined #mythtv
11:25-!-rkulagow_ [] has quit [Read error: 54 (Connection reset by peer)]
11:29-!-hfb [~hbarrett@lsanca1-ar2-4-60-014-253.lsanca1.dsl-verizon.net] has joined #mythtv
11:49-!-rkulagow_ [~rkulagow@12-207-131-29.client.attbi.com] has joined #mythtv
11:49<rkulagow_>simply parrots spam?
11:49<Chutt>heh
11:49<Chutt>someone didn't change the To address
11:59<moegreen>Chutt: conflict resolutions is about done. I'm going to change around the UIListType class to allow any number of fonts to be used (instead of the UIFontPairType)
12:16<Chutt>cool =)
12:17<Chutt>for different colors, etc?
13:02<-- rkulagow_(~rkulagow@12-207-131-29.client.attbi.com) has left #mythtv
13:02-!-rkulagow_ [~rkulagow@12-207-131-29.client.attbi.com] has joined #mythtv
13:24<thor_>Chutt, should "options" go in the theme, or in settings?
13:24<thor_>Example --> widgets that have keyboard accelerators
13:31-!-Rizzi [~mark@roc-24-93-16-37.rochester.rr.com] has joined #mythtv
13:31<moegreen>Chutt: yeah, mainly for the different colors in the conflict screen, but I think it will be useful elsewhere
13:33<thor_>moegreen, to get my music-ui.xml file installed, do I need to anything other than add the lines to mythmusic.pro ?
13:33<thor_>(any other qmake things I need)
13:36<moegreen>thor_: that's it, I'd just put it in the mythtv/themes/default directory
13:36<moegreen>then if a themedesigner wants to change it, the xmlparse will look the theme's directory first, then fallback to the default dir
13:37<thor_>yup ... good
13:37<rkulagow_>anyone here running mandrake cooker?
13:37<rkulagow_>(other than me)
13:39<moegreen>thor_: what'd you think? pretty easy to code? yes/no?
13:39<thor_>yeah .. not bad ... taking me longer than expected, but that's only because I keep going off to do other things and then have to remind myself how it's all supposed to work.
13:42<thor_>moegreen, draworder is just for things to get painted on top of things, yes? Or is there something else to it?
13:44<moegreen>thor_: it's for painting, it's probably just best to draw layers 0-8 after you set the ui object's data, that gives the theme designer the most freedom
13:45<thor_>yup ... ok
13:51<hurdel>offtopic here: anybody know where to get the 4images software? their site appears to be down
13:51-!-rkulagow_ [] has quit [Read error: 104 (Connection reset by peer)]
13:51-!-rkulagow_ [~rkulagow@12-207-131-29.client.attbi.com] has joined #mythtv
14:36<thor_>moegreen: can a container have buttons in it?
14:49<moegreen>thor_: I didn't code a button widget yet. The container can only have UIType objects
14:50<thor_>right ... ok ... I guess I'll do that
14:56* yebyenblums
15:31<rkulagow_>chutt, that error message (referring to section 17 of the HOWTO) about the seg fault that the user is talking about on -users was updated on 2003-03-27, so the guy is way out of date with CVS if that is indeed what he's running.
15:32<Chutt>it's probably 0.8
15:32<Chutt>but there's the first error, too =)
15:33<rkulagow_>yep, and i specifically added text to the HOWTO after people complained of my terseness in the ML.
15:33<Chutt>if he was running current cvs, it'd die after the first error
15:33<Chutt>as i added an exit() there
15:34<rkulagow_>ah, that's good. i also like how various modules are checking if current lib != compiled against lib.
15:38<Captain_Murdoch>would people see benefit in a patch that allows you to see the title/description/etc info from the recorded table while watching a recording? like when you hit 'I' while watching TV, this would popup the same info while watching a recording.
15:40<Chutt>captain_murdoch, sure, i suppose
15:40<Captain_Murdoch>I was thinking of chaning the 'I' key in playback to popup the same box that it does in LiveTV mode, but then have to move the DoPosition() popup to another key. do you mind that or is changing keys non-kosher?
15:42<Captain_Murdoch>I in LiveTV does program info, in playback it does position so this would standardize what the key does. always program info whether live or playback
15:46<Chutt>i is information
15:46<Chutt>or whatnot
15:46<Chutt>position info during playback is information, and more useful than the program info
15:46<Chutt>especially since the program info was already on the playback screen
15:47<Captain_Murdoch>ok, so use another key. fine by me. :) I'll keep it handy (not in cvs) in case anyone requests something like that and can assign to a new key if I add it.
15:48<Captain_Murdoch>I guess once we have actor/year/etc. info in the database it might be more handy.
15:48<rkulagow_>possibly dumb question, but here goes. i've installed a bare-bones mandrake 9.1 distro; didn't include KDE or GDK. the fonts in mythtv (including the main screen with all the buttons) are tiny and practically unreadable. any thoughts as to what i may have forgotten to include that would cause this?
15:49<rkulagow_>fonts in eterm seems reasonable.
15:49-!-Rizzi [] has quit [Read error: 110 (Connection timed out)]
16:16<Chutt>hrm
16:17<Chutt>did it install the ttf fonts?
16:19<Chutt>ok
16:19<Chutt>if it can't get the audio device for playback
16:19<Chutt>should it a) print out an error message that people will ignore or b) exit
16:21<Timon>Both
16:21<Chutt>well, exiting implies printing out an error message first
16:22<Timon>Ok, then b
16:22<thor_>c) decide what to do based on user setting in db
16:22<Timon>Unless you make a) use a popup window of some sort instead of spewing to stdout
16:23<Timon>or stderr
16:23<Chutt>i'm just going to make it exit
16:23<thor_>(I was joking)
16:23<Timon>Works for me!
16:24<Chutt>i think the only value i should get is ebusy
16:32-!-Rizzi [~mark@roc-24-93-16-37.rochester.rr.com] has joined #mythtv
16:34<yebyen>whoa
16:34<yebyen>I see someone's working on mythtv right now, heh
16:34<Chutt>hm?
16:34<yebyen>heh, I just updated like 10 minutes ago, now there's something changed in NuppelVideoPlayer.cpp
16:36<Chutt>oh, that
16:36<yebyen>what did you do to it? ^_~
16:36<yebyen>HEH
16:37<Chutt>like 10 lines of code to make it not block on trying to open the audio dsp
16:37<Chutt>and print an error and exit if it can't
16:42<Chutt>heh
16:42<Chutt>Issac,
16:42<Chutt><someone's name> from the PVR Discussion Group told me you needed some more hardware.
16:43<Chutt>I like what you are doing.
16:43<Chutt>How can we help?
16:43<yebyen>w00t
16:43<Chutt><guy>
16:43<Chutt>Hauppauge
16:43<yebyen>haha, sweet
16:43<thor_>digg 'em
16:43<yebyen>"Guy at Hauppauge,
16:43<yebyen>Dude!
16:43<yebyen>--Isaac"
16:44<Chutt>names removed to protect the innocent
16:44<Chutt>:p
16:44<Chutt>and i think the VIA guy is actually sending me hardware
16:44<yebyen>monger :)
16:44<Chutt>he somehow got the impression that'd i'd be willing to add support for their hardware decoder
16:44<yebyen>HEH
16:44<Chutt>i think
16:44<yebyen>well,
16:44<yebyen>free hardware, either way
16:45<Chutt>i'm not writing code for cheap hardware :p
16:45-!-greeble [dunne@homeworld.rutgers.edu] has joined #mythtv
16:45<yebyen>no, free hardware means you don't have to do anything.
16:45<thor_>on the hauppauge stuff, could you ask them to give developper resources to ivtv?
16:45<yebyen>ivtv?
16:45<Chutt>thor, i _believe_ they are
16:45<Chutt>but not officially
16:45<thor_>Ahh
16:46<thor_>get rid of whine!
16:46<yebyen>heh
16:47<yebyen>anything significant gone into mythtv cvs in the last 6 days or so? :)
16:47<yebyen>it's compiling right now
16:47<yebyen>(I saw that something changed in the commercial detection code, so I have to try...)
16:48<thor_>Chutt, on a completely unrelated topic, what do you thing of a themed music playback screen that is a nock 'off of tv playback
16:48<thor_>including playlists on the left (with Active as the default)
16:48<Chutt>that might be interesting
16:49<thor_>it's suprisingly useful
16:49<thor_>can I can buttons?
16:49<Chutt>i like them as status indicators
16:49<Chutt>and they're useful for more limited remotes
16:49<thor_>yeah ... it's just that they don't exist in UITypes yet
16:50<Chutt>heh
16:50<Chutt>buttons aren't difficult to add, though
16:50<thor_>I'll leave them to the end
16:50<yebyen>oh, shit
16:50<Chutt>though, the uitypes stuff doesn't really have any idea of focus
16:50<yebyen>transgaming is slashdotted
16:51<thor_>and can a container handle focus across button groups?
16:51<Chutt>i dunno
16:51<thor_>I thought this was going to be easier than it's turning out
16:51<Chutt>heh
16:51<thor_>anyway ... back to work
16:52<thor_>(reminding me of playlists)
16:52<thor_>oh ... one more thing ... have you tried the hostname'd playlists cvs.sql >
16:52<thor_>s/>/?
16:53<Chutt>not yet
16:53<Chutt>timon, you're going to send an updated patch to the list sometime?
17:02-!-hootie [] has quit [Read error: 110 (Connection timed out)]
17:05<Timon>Chutt: Yeah. I want to clean it up a bit, and do a final bit of testing. I just got some weirdness that I want to figure out if its a fluke, or something wrong
17:09<Chutt>ok
17:31<rkulagow_>chutt: how'd you like that email i forwarded from sweden? another guy who wants to make a commercial project out of myth.
17:31-!-PhracturedBlue [] has quit [Read error: 54 (Connection reset by peer)]
17:31-!-PBlue_away [~tcpaiqgvn@cs6668127-96.austin.rr.com] has joined #mythtv
17:33<Chutt>so's the 'It half works!' guy who can't read your docs
17:34<rkulagow_>the more bangs, the less likely there's good info in the message
17:34<Chutt>from his page, he says he's using a p3-500
17:35<Chutt>and is curious as to what the possible bottlenecks may be
17:35<rkulagow_>never heard back from any of the people who were jumping at the chance to make debian install docs for inclusion into the HOWTO. and to think, they thought that MDZs approval was going to be the bottleneck. it's more like time and motivation.
17:35<Chutt>heh
17:44<Timon>inman: You here?
17:49-!-Rizzi [] has quit [Read error: 110 (Connection timed out)]
17:52<rkulagow_>this font thing is pissing me off.
17:52-!-nziarek [nziarek@pcp010412pcs.unplugged.mu.edu] has joined #MythTV
17:53<rkulagow_>nziarek: thanks for the pdf
17:53<nziarek>rkulagow: yes sir
18:01-!-rcaskey [~rcaskey@adsl-154-7-85.asm.bellsouth.net] has joined #mythtv
18:04<thor_>I don't like themes very much
18:04<nziarek>like the ones available, or making the program themable?
18:04<thor_>the latter
18:05<nziarek>that I can understand (I can understand not liking themes, too, but that hurts :)
18:06<thor_>I just seem to be spending a lot of time calculating intersections on QPainter events ...
18:06<thor_>... anyway ... the stuff is great ... I'm just being impatient
18:06<nziarek>i can imagine it is pretty time consuming, especially to do it "right"
18:07<nziarek>feel free to come here and vent as you need to, and then get back to work!
18:07<nziarek>=)
18:07<thor_>yup
18:07<Timon>Your a slave driver nziarek :-)
18:08<rcaskey>what's being done?
18:08<nziarek>that is my business minor talking :)
18:08<nziarek>the above statement was very motivational :)
18:08<Timon>hahah
18:09<Timon>Which theme did you make nziarek?
18:09<nziarek>iulius and color bars
18:09<nziarek>iulius is ongoing (http://www.ziarek.com/mythtv)
18:09<Timon>ok. I like the iulius osd with mythblue for the theme
18:09<Timon>But I also like corners by ben@handcoder for the osd
18:10<rcaskey>Screenshots around?
18:10<nziarek>i like that OSD...as I clean up iulius, i hope to get to something a little closer to that
18:10<Timon>Man, what is it with people using squintesimal fonts on websites?
18:11<nziarek>does that mean small?
18:11<Timon>That means uber small
18:11<nziarek>sorry :) i use some small fonts for my menu, it is just a look :)
18:12<Timon>At least you don't FORCE the small fonts.
18:12<Timon>I hate going to sites where I change the text zoom in mozilla and the fonts don't change size
18:12<rcaskey>nziarek: can you put up some screenshots?
18:13<nziarek>yeah, for now if you look at "testing" those are mockups that are pretty close
18:13<nziarek>i will put some up tonight, though
18:13<rcaskey>oh
18:13<rcaskey>just make sure it looks good at 640x480
18:13<nziarek>deal
18:13<rcaskey>even with my TV some text is kinda hard to make out
18:22<yebyen>this commercial skip is really kicking ass
18:30<moegreen>thor_: you shouldn't need to be calculating intersections in qpainter, just have your program update the button and lists and respond to keystrokes - then let the uitypes take care of the rest
18:50<Timon>inman: You here?
18:54<Captain_Murdoch>Timon: I coded up that browse feature I mentioned. :) still needs some tweeking and some limits set on how far in time you can browse but it works for seeing program descriptions and changing the channel. need to track down an occasional lockup also.
18:54<Timon>What browse feature (Refresh my memory)
18:55<Captain_Murdoch>the satellite/cable-box type feature where you can browse through program descriptions while still watching the current show.
18:55<Captain_Murdoch>weren't we talking about that yesterday?
18:55<Timon>Ohh, I mis-read your statement. I read it as "Timon: You coded up that browse. . . ."
18:55<Timon>Yeah, it was me
18:55<moegreen>i think timon was hammered at the time
18:56<Timon>I wasn't hammered. I was pretty happy, but not hammered
18:56<moegreen>heh
18:56<Timon>Captain_Murdoch: Good deal!
18:57<Timon>moegreen: Captain_Murdoch either of you want to help me in doing a last bit of debugging with the mod?
18:57<Captain_Murdoch>might need to figure out a better way to do the channel stuff. right now it has to figure out the channel ordering on it's own so it's a bigger mod than I'd like.
18:57<Timon>I keep getting a segfault that I can't figure out why its happening. The numbers are all within acceptable ranges
18:57<Captain_Murdoch>can you reproduce every time?
18:57<Timon>what about if you just query the db and sort it in the db
18:58<Timon>Yup
18:58<Timon>Its some how releated to this: int vector = channame_vector.size() - (times_pressed % channame_vector.size());
18:58<Captain_Murdoch>I use sql to sort, but since there are 4 sort methods, and I allow going up/down in the list, then I have to have quite a few different sql statements to do the sort (1 per case)
18:59<Captain_Murdoch>can channame_vector.size() ever be equal to zero?
18:59<Timon>Yes, but thats caught up above
19:00<Timon>if ( channame_vector.size() == 0 )
19:00<Timon> {
19:00<Timon> AddPreviousChannel();
19:00<Timon> return;
19:00<Timon> }
19:00<Captain_Murdoch>where's the segfault occurring, in TV::SetPreviousChannel?
19:01<thor_>moegreen, but everything I've looked at intercepts QPainter events, does some rect()'s, and then uses logic to call subelements that need to repaint themselves (mythweather, progfind, playbackbox, etc.)
19:01<Timon>Heres an example of how its supposed to work (And what my debug output verifies) Say your on chan 2, then change to 3, then 4, then 5. If you hit "H" once it goes to 4. If your on 5 and hit "H" 4 times, it should wrap back to 5. And its confirmed with a cout. But it crashes.
19:01<moegreen>thor_: right, but the size of those is nothing you have to worry about
19:02<Timon>I dunno, I didn't bother to run it in gdb :-)
19:02<Timon>But based on the cout's that run/don't get run it seems to happen in there
19:02<thor_>moegreen, ok .... as I said above, I'm just being impatient
19:02<moegreen>thor_: let the theme designer determine the size, just setup a rect with variables like is done in the playbackbox, then have the parseContainer assign those variables.
19:03<thor_>yup yup
19:05<thor_>at some point, it would be great to have UITypes do signals and slots ... I think that would make wiring things up easier for total novices like me
19:05<Captain_Murdoch>maybe try calling AddPreviousChannel() inside TV::LiveTV() right after it sets activerecorder = recorder = testrec;
19:05<Chutt>thor, yup
19:05<Chutt>more object like =)
19:06<Chutt>hrm
19:06<Chutt>the QThread stuff spawns a new thread each time
19:06<thor_>you say that like it's a bad thing
19:07<Chutt>i'm making a small thread pool for the backend
19:07<Chutt>want to keep threads living, so there's no creation/deletion time
19:07<thor_>... hang on ... every start() on a single instance creates a new thread?
19:07<Timon>thor_: I think I figured it out. . . I didn't init times_pressed when it gets created
19:08<Timon>I know thats contributed to a lot of the weirdness I've noticed.
19:09<Captain_Murdoch>Timon: should there be a -1 on the end of that mod line? if you have 4 entries and press h 4 times then 4 - 4 % 4 = 4 when the vector has 0-3 right?
19:09<thor_>Timon, this that usleep --> threads --> timers --> mutex --> no, not mutx problem?
19:09<Timon>not a mutex problem. Captain_Murdoch Your right on the -1. I had it on there under a different implimentation, just forgot to put it back
19:11<Timon>thor_: Thanks for the LiveTV suggestion. I didn't even think to look there to have it save the channel initially
19:14<Chutt>thor, it waits for a running thread to exit first, but, yeah, creates a new thread
19:15<thor_>Timon, did I suggest that? I amy be loosing my memory faster than I thought?
19:16<Timon>thor_: Sorry, that was Captain_Murdoch not you. I'm the one loosing the marbles
19:16<thor_>Chutt, that seems odd ... could you fake it with an event loop in start than sleeps?
19:16<thor_>s/than/that
19:18<Chutt>thor, that's what i did
19:18<Chutt>but, used a waitcondition instead of sleeping
19:19<thor_>wow, I feel almost clever
19:19<Chutt>pretty easy stuff
19:19<Chutt>=)
19:19<Chutt>well, not really _that_ easy
19:19<Chutt>but
19:19<Chutt>should improve responsiveness of the backend
19:20<thor_>just the instantiation/destruction overhead ...
19:20<Chutt>right, and now it's not doing that except at startup
19:20<Chutt>since the thread's just sitting in a loop
19:20<thor_>... sounds to me like you're desperate to avoid cracking the nut on per-recording settings .... =)
19:20<Chutt>naw
19:21<Chutt>this is for rkulagow
19:21<Chutt>since he's been having problems with the backend locking =)
19:23<rcaskey>mdz_: you around?
19:23<thor_>well, he deserves everything he asks for. I've never seen docs so up to date ... although you wouldn't know it from the -user list
19:24<Chutt>also
19:24<Chutt>this should help multiple frontends
19:24<rcaskey>win win :)
19:25<Chutt>if a backend query were to take a long time before (say, lots of work to do for conflict resolution)
19:25<rcaskey>are there any plans to include a setup routine for lirc sa well?
19:25<Chutt>that would make anything else that was doing stuff (playback/livetv) pause
19:29<Captain_Murdoch>Chutt: I've coded up a "browse" feature for LiveTV but have a question. This is available on most digital cable and sat boxes, it lets you browse the descriptions from other shows/channels while still watching LiveTV fullscreen. User hits ESC to exit browse mode or enter/space to change channels. I've got it working pretty good (few minor bugs still working out), but in order to work I have to perform sql queries to find out the
19:29<Captain_Murdoch>p/down using thei user's desired channel ordering. I'm wondering if it's ok to do this inside tv_play.cpp. I looked at channel.cpp but didn't think I could reuse any of that.
19:29<Timon>Chutt: I submitted the patch for previous channel functionality
19:29<Chutt>captain_murdoch, i'd almost prefer if that went through the backend
19:30-!-Rizzi [~mark@roc-24-93-16-37.rochester.rr.com] has joined #mythtv
19:30<Chutt>all that stuff already lives in there
19:30<Captain_Murdoch>ok, so the frontend could ask the backend what's the next channel up from X?
19:30<Chutt>timon, thanks
19:30<Chutt>captain_murdoch, yup
19:30<Timon>np
19:30<Chutt>or, 'just give me the next channel data up from X'
19:30<Chutt>since i assume you're still asking for the data from the backend?
19:31<Captain_Murdoch>I'm using an sql query to figure out the next channel/show up/down (you can browse all 4 directions) and then using ProgramInfo to get the program on that channel at the right time.
19:32<Captain_Murdoch>actualy sql is only channel, programinfo is show.
19:32<Chutt>ah
19:32<Chutt>hmm
19:32<Chutt>yeah, i'd ask the backend for the next channel up
19:32<Chutt>since it has all that logic already
19:32<Captain_Murdoch>so it uses sql query to get next channel up or down and uses sql query for next startime before or after then once it's got a chanid and starttime it calls programinfo to get show info.
19:32<Captain_Murdoch>ok, I'll look into it further then, thanks.
19:33<Chutt>it could even just have the backend do all the work
19:33<Chutt>and just return a programinfo struct
19:34<Captain_Murdoch>what would you send to the backend to request? channel & starttime & direction-to-move?
19:34<Chutt>yeah
19:35<yebyen>hm
19:37<Captain_Murdoch>ok, sounds good.
19:38<-- Rizzi(~mark@roc-24-93-16-37.rochester.rr.com) has left #mythtv
19:38<Timon>Is there one OSD method you could call to hide all visible sets?
19:39<Chutt>nope
19:39<Chutt>but why would you want to hide everything?
19:40<Timon>Chutt: Makig a mod to UpdateOSD(). Such as that if you hit "I" when the OSD is visible, it will hide it. Like my TV does
19:40-!-Rizzi_ [~blah@roc-24-93-16-37.rochester.rr.com] has joined #mythtv
19:40<Chutt>right
19:40<Chutt>but you shouldn't hide _everything_
19:40<Timon>Which ones should be hidden?
19:41<Chutt>what if there's something important up, like a dialog that's expecting an answer?
19:41<Timon>Ohh, well then I shouldn't hide that :-)
19:41<Timon>if I hide "program_info" and "channel_number" should that be pretty incompusing?
19:42<Chutt>yup
19:42<Timon>Or are there any other containers to include/
19:42<Chutt>that should be it
19:42<Timon>Ok, thanks
19:43<thor_>Chutt, moegreen ... is it just me, or is it odd that I have to create things to store rectangle values ... so that I can know what to do during a painter event.
19:43<thor_>Why ... in fact, am I doing Painter events?
19:43<Chutt>might be wise to move that into the top level containers
19:44<Chutt>then in the paintEvent() you can just iterate over the top level containers
19:44<Chutt>and ask them if they need redrawn
19:44<Chutt>thor, that's what triggers drawing =)
19:44<thor_>the containers know depths and dimensions
19:44<Chutt>they might as well know where they are, too
19:45<thor_>and there are widgets in there
19:45<thor_>and Qt knows how to call painter events on its widgets
19:45<thor_>so why am I in the middle?
19:45<Chutt>there aren't widgets in there
19:45<Chutt>is the thing
19:45<Chutt>everything's getting drawn on the background
19:46<thor_>it's visualizations all over again
19:46<thor_>=)
19:46<-- rcaskey(~rcaskey@adsl-154-7-85.asm.bellsouth.net) has left #mythtv ("Client exiting")
19:46<Chutt>heh
19:46-!-rcaskey [~rcaskey@adsl-154-7-85.asm.bellsouth.net] has joined #mythtv
19:46<Chutt>conceivably, qt could eventually not be used to draw anything, even
19:46<Chutt>well
19:46<Chutt>not really
19:46<Chutt>since i wouldn't want to do the settings stuff without it
19:47<rcaskey>anyone have an OsX machine here?
19:48<thor_>"there are no widgets in there" ... my god ... you are correct .... I'm "making" all the widgets.
19:48<thor_> .... oh ... that makes more sense then.
19:48<Chutt>in normal qt stuff, each widget has its own window
19:49<Chutt>making it difficult, to say the least, to do the fun transparency effects
19:49<Chutt>and overlap things
19:49<thor_>yup ... yup ...
19:56<thor_>.... no widgets ... so ... uhm ... buttons are going to have to be text, pixmaps and methods (signals and slots) that have nothing to do with QButton's
19:57<Chutt>yup
19:57<Chutt>and it'll have to know where focus is, etc
19:57<thor_>holly crap
19:58<Chutt>kinda like the themedmenu stuff, probably
19:58<Chutt>and the themedmenu stuff could be made to use the uitypebutton stuff whenever that's done
19:58<Chutt>hehe
19:58<thor_>well ... better get back to it then ...
19:58<Chutt>don't you wish you picked easy things =)
19:59<thor_>no TV for me tonight
19:59<moegreen>bah...a little legwork and you can sit back while others (nziarek et al) do their magic
20:00<thor_>yup ... efficient division of labour ...
20:01<inman>Timon: i am here.
20:01<rcaskey>is there a .9 feature list in the works?
20:01<Chutt>finish up the UI makeover and the pvr-250 support
20:01<Timon>inman: Maybe you can help me with your suggestion for times_pressed % vector.size()
20:02<inman>okay; is your patch in cvs?
20:02<inman>i didn't see it there..
20:02<Chutt>inman, not yet
20:02<Timon>No, but I sent it to the mailing list
20:02<rcaskey>is Myth on Knoppix?
20:02<Chutt>it'll be there later tonight
20:02<inman>cool, i'll apply it from the ml.
20:02<rcaskey>it had xawtv a few months ago
20:02<Timon>inman: I had to rip out that line of code so I'll recode it for you to paste into tv_play.cpp
20:03<inman>whah?
20:04<Timon>replace the two lines of int vector = blah blah with this: int vector = channname_vector.size() - (times_pressed % (channame_vector.size());
20:05<inman>right, line 1544?
20:05<Timon>And 1564
20:05<Timon>You can also get rid of the if ( vector < 0 ) lines
20:07<inman>uh huh
20:07<Timon>Whats taht?
20:07<inman>i was reading email, nm.
20:07<Timon>k
20:08<Timon>You reading the message I posted along with the patch?
20:08<inman>i'm getting unresolved symbols on 'tr'.
20:08<rcaskey>so have plans for after the fact reencoding been scrapped?
20:09<Timon>tr?
20:09<Timon>what line? There is no var tv
20:09<inman>which version of qt are you using?
20:09<Timon>err tr
20:09<Timon>that I'm aware of
20:09<Timon>qt 3.0
20:09<inman>line 921
20:10<inman>i think i had better `make distclean`.
20:10<Timon>Dunno about that stuff. I didn't touch it
20:10<Timon>hehe, ok :-)
20:10<inman>i put in my patch; gimme a couple minutes to rebuild and test it.
20:10<inman>or, i can send it to you.
20:11<Timon>what's it do?
20:11<inman>i think it's funny that i made you go through so much effort to implement my multiple-history idea.
20:11<inman>hah
20:12<inman>it's good though, i think.
20:12<inman>you know, having several previous channels.
20:13<Timon>yup
20:13-!-StarHeart [] has quit ["Client exiting"]
20:15<mdz_>rcaskey: yes?
20:15<rcaskey>btw, is mythtv pretty smart about turning the screensaver on and off when there are on screen dialogs and such to rpevent burn in?
20:15<mdz_>I don't think myth touches the screensaver settings at all
20:16<rcaskey>do you maintain the deb repository?
20:16<Timon>inman: You get it compiled and running yet?
20:16<mdz_>yes
20:16<inman>not yet, got a couple other things building.
20:16<Timon>k
20:16<rcaskey>what do you think about including packages for lirc and such?
20:17<mdz_>there are already lirc packages in Debian proper
20:17<rcaskey>well binaries
20:17<mdz_>not sure what you mean, there are already binary packages for lirc in Debian proper
20:17<mdz_>or are you talking about the kernel modules?
20:17<rcaskey>yeah
20:17<mdz_>hmm
20:19<mdz_>ah, I think I know why there aren't any such packages in Debian
20:19<mdz_>some of the lirc options need to be set at compile time
20:19<mdz_>inexplicably
20:19<mdz_>for instance, to use the actisys 200l
20:20<rcaskey>I think most work fine though
20:20<mdz_>what about the serial port settings?
20:20<rcaskey>mdz_: I thought we had kinda standardized on those pretty much
20:20<mdz_>are there module parameters for that?
20:21<mdz_>or do you have to compile in which serial port it's using and that sort of thing?
20:21<rcaskey>not sure, definately don't know much about that sort of thing
20:21<mdz_>I have no problem putting the packages in my repository if someone wants to build them
20:21<rcaskey>I just always accepted the default
20:22<_shad>I couldn't get the debian lirc stuff working, but compiling from source worked.
20:22<mdz_>I have an actisys 200L, so I have to build it differently
20:22<rcaskey>ill build ones for x86 as soon as I figure out how to issue packages for 2.4.20-1, it seems to insist on 2.4.20
20:22<rcaskey>mdz_: i just compiled em all in
20:24<moegreen>mdz__: I'll probably make a deb for MythWeather for the next release ...
20:25<moegreen>though I've never made a deb before :)
20:26<rcaskey>moegreen: everyone I talk to about myth asks about mythweather
20:26<rcaskey>It's useful, but its not the kind of thing I expected to get any attention, you know?
20:34<Timon>inman: You get it compiled and running yet?
20:34<inman>yeah, on the phone, gimme a few.
20:34<inman>(this is trivial).
20:34<Timon>k
20:35<rcaskey>I'd really like to get mythmusic in gear but the way it handles MP3's makes it kinda unmanagable because my mp3s are on a remote share
20:36<rcaskey>and crawling 60 gigs of mp3s at 30k/sec is not so good
20:36<Timon>60 gigs?
20:36<Timon>Thats insane
20:38<inman>try using flac.
20:40<inman>i'm ready for self-employment, aiiieee.
20:41<Timon>why's that?
20:42<inman>i'm a recovering sysadmin.
20:42<inman>okay, what do you want, the line to replace your line?
20:42<inman> int vector = (channame_vector.size() - times_pressed - 1) % channame_vector.size();
20:43<Timon>Dude, I KNEW it had to be something that simple
20:43<inman>well, you were right.
20:44<inman>er, take out the vector < 0 check, too.
20:44<Timon>Yeah, did that already
20:48<Timon>inman: I found something a bug with it, but I don't think its nessicarlly a bad thing. The channel your currently on will show up twice in the cycle
20:48<inman>i know
20:49<inman>i didn't know if you desired that behavior or not.
20:49<inman>you mean with my method or yours or both?
20:49<inman>are you talking about the occurence of channels in the vector multiple times?
20:50<inman>or that it starts off showing the current channel and then moves to the next in the queue?
20:50<Timon>Not the occurence of the channel within the vector multiple times. Thats desired.
20:51<Timon>Ok, I guess it wasn't a prob with your code. Just with how the channels got into my vector
20:51<inman>yes; the current channel should be added to the vector only when switching channels.
20:51<Timon>Yeah
20:55-!-rcaskey [] has quit [Read error: 104 (Connection reset by peer)]
20:57-!-rcaskey [~rcaskey@adsl-154-30-48.asm.bellsouth.net] has joined #mythtv
21:02<Timon>inman: Do you think previously entered channels should be filtered?
21:02<inman>probably not, it'll confuse people.
21:03<Timon>k
21:09<mdz_>rcaskey: back now
21:09<mdz_>rcaskey: what I was saying is that some of the options are mutually exclusive
21:10<mdz_>rcaskey: you can build lirc_sir either for the ACT200L, or for the homebrew receiver
21:10<mdz_>rcaskey: but you can't support both with the same binary module
21:10<mdz_>moegreen: I guess you haven't looked in my myth package repository :-)
21:13<moegreen>heh, no I didn't until now. :)
21:16<rkulagow_>chutt: thanks for that threaded stuff. i'll let you know how it works out.
21:17<rkulagow_>thor: thanks for the kudos. yes, it's a thankless task to see the same questions over and over, especially when someone says "i read the HOWTO" and it's obvious that they didn't. ah, well. if you and chutt wrote better code we wouldn't need docs. =)
21:19<Timon>hah
21:19<Timon>inman: I'm going to submit the updated patch
21:19<Timon>Sound good?
21:19<inman>congrats. :-)
21:19<Timon>heh
21:19<inman>did you fix the channel-addition-to-vector issue?
21:21<Timon>What chan addition to vector issue?
21:23<inman>you know, with the current channel existing in the vector prior to it ever having changed.
21:23<Timon>Oh, well that issue :-)
21:24<inman>if you start watching tv, immediately hitting 'h' should do nothing.
21:24<inman>it could maybe print 'no previous channel, dimwit' in the osd...
21:24<Timon>Yeah
21:24<Timon>I just fixed that. The other thing I want is to make it configurable in setup
21:25<Timon>Such as that the user could configure it to act like the prev channel functionality they're used to, or have it act like it does now
21:25<Timon>Also, channel history is limited to 30 channels (I arbitrarily limited it to that amount)
21:26<inman>i saw that.
21:26<inman>what do you mean like 'prev channel functionality they're used to'?
21:26<Timon>On my tv if I hit prev chan, it immedietly changes to the previous channel
21:27<inman>oh, that.
21:29<inman>i don't know what's wrong with my box, but i've never been able to use mpeg4. always loses frames at almost any quality level.
21:29<Timon>Whats your hardware?
21:29<Timon>I find I lose frames in mpeg4 if the channel signal is fuzzy
21:31<inman>smp athlon 2k, 1600x1200 res, 320x480 capture, 512meg o' lovin and 160mbit scsi
21:31<Timon>Hmm, I don't think you have a fast enough system :-)
21:32<Timon>have you run top while its encoding in mpg4?
21:32<inman>of course
21:33<Timon>what was the cpu at?
21:33<inman>depends on what my settings are
21:33<Timon>did it ever peg?
21:33<inman>always.
21:33<inman>it is a processing power issue, of course.
21:33<Timon>wonder why. Mine doesn't peg
21:34<Timon>And I have a lowly 1700+ athlon
21:34<inman>i'm running a UP kernel right now, for testing.
21:35<inman>i want myth to run acceptably on a single chip, but i can't for the life of me get decent quality out of mpeg4.
21:36<Timon>hres my settings: 480x480 MPEG-4 bit 3100 Max Quality: 2 Min Qual 15 Max Dif 3 scale bit rate is checked
21:36<Timon>Audio: 32000, 7 quality
21:37<inman>it stutters once per second.
21:38<inman>backend is using 80-88% of cpu and frontend is using 10-15%.
21:38<Timon>Hmmm, mine only does it if the signal is bad
21:38<inman>i'm using digital cable.
21:38<Timon>what if you just have it record? Whats the usage look like?
21:38<inman>maybe the signal is bad, though.
21:39<Timon>Is your signal fuzzy? Mine only stutters on a fuzzy pic
21:39<inman>just recording yields 80-98% cpu
21:40<Timon>wow
21:40<inman>seems at about 95% average
21:40<Timon>have you tried compiling it with optimizations for the athlon cpu?
21:40<inman>of course.
21:40<Timon>And?
21:40<inman>that's what's running now.
21:41<Timon>I bet its an issue with the chipset on your mobo. Cause I run at a lesser cpu usage, on a slower chip, at a higher resolution
21:41<inman>i know, everyone seems to. :-)
21:41<inman>it's entirely possible that it's the mobo, since i don't think there are many dual-athlon myth'ers running around,
21:41<inman>but it seems pretty unlikely to me.
21:42<Timon>It might be an issue with the chipset drivers the kernel has. Cause that mobo is pretty new
21:43<inman>it's really a shame, because i find that mpeg4 smooths out the picture quality so it looks much nicer than rtjpeg.
21:43<inman>i could try not using the 768mp kernel options, i guess.
21:43<inman>maybe turn off the -slow-myth compile option...?
21:44<Timon>hehe, I bet its the -slow-myth-for-andy directive :-)
21:44<inman>yeah, i must have "missed" that cvs-commit message. ;-)
21:44<yebyen>heh
21:45<yebyen>what's the best way of tracking cvs commit messages?
21:45<yebyen>does it take a mailing list, or is there a sane way to do it from the cvs program itself
21:45<Timon>get ont he mailing list
21:45<Timon>err get on the
21:45<yebyen>hm
21:46<yebyen>i was hoping for a cvs log command of some kind
21:48<inman>`cvs --help-commands`
21:48<yebyen>i've been through the man page and whatnot
21:48<yebyen>best I can find is cvs log -d"<=[date of last checkout]"
21:49<yebyen>there's supposed to be a -S command, based on the manual, that filters the output of cvs log to only show files with a commit
21:50<yebyen>but that doesn't seem to exist in practice
21:50<yebyen>HE
21:50<yebyen>H
21:53-!-nziarek_ [~nathanzia@CPE-24-167-222-150.wi.rr.com] has joined #MythTV
21:53<Timon>inman: In 15 words or less how would you describe the new previous channel functionality?
21:54<nziarek_>anyone else feel that the "All Programs" box is worthless (as it looks now) as far as reading show name and show title information?
21:54<Timon>All Programs? Wheres that box?
21:54<nziarek_>in the playback section, when you select "All Programs" instead of an individual show name
21:55<inman>Timon: is this for the settings help?
21:55<Timon>inman: Yeah
21:55<inman>hmmm.
21:55<Timon>nziarek_: I dunno, its kinda handy. Kinda not :-)
21:56<inman>doh
21:56<inman>Channel history in which multiple keypresses travel back in time. ;-)
21:56<yebyen>heh
21:56<moegreen>nziarek_: What data should it show
21:56<mdz_>Chutt: I have a recurring backend crash when resolving conflicts in 0.8; do you know of any bugs there?
21:57<Timon>thanks inman
21:57<inman>yep
21:57<inman>it may make sense to form an analogy to the brower's back button.
21:57<inman>eg.
21:57<inman>Channel history with a "back" button.
21:58<Timon>Bah, users can figure that out :-)
21:58<inman>people may be more familiar with the semantics of the browser than those of a stack...
21:58<Timon>Ok, then I'll use with a back button
21:59<nziarek_>moegreen, Timon: i like the feature, and I think it is showing all of the relevant data, just that with the show titles taking up a good deal of space, the 'Show Name - "Episode Title"' bit tends to get cut off. not a critique of the programming, but of the design
21:59<Timon>Well, there is a limited amount of space on the screen
22:00<-- rcaskey(~rcaskey@adsl-154-30-48.asm.bellsouth.net) has left #mythtv ("Client exiting")
22:00<moegreen>nziarek_: so you think drop the title or subtitle?
22:00-!-rcaskey [~rcaskey@adsl-154-30-48.asm.bellsouth.net] has joined #mythtv
22:01<nziarek_>moegreen: i just can't imagine which would be acceptable :) i really do feel like there should be a design way around it
22:02<nziarek_>i am just stuck right now. you want to make the fonts large enough that you can read them on the TV, but at the same time, you've got very little real estate
22:03<nziarek_>Chutt: you didn't do any experiments when working on the original GUI on acceptable font sizes, did you?
22:05<inman>you don't want to show the episode title below the show name? you could indent it a bit.
22:05<moegreen>nziarek_: Take confort in the fact that once you're done with that and MythWeather, and the program finder, you can hit up the conflict screen (which should be easy) http://www.untzuntz.com/mythtv/testing/conflict1.jpg
22:05<inman>i'm not sure i'm understanding the problem; is it simply that you're out of horizontal space?
22:05<Chutt>mdz, not that i recall
22:05<Chutt>nziarek, it really depends on the individual person's setup
22:06<nziarek_>moegreen: i just want to do it right :)
22:06<nziarek_>Chutt: yeah, that is what I am discovering :)
22:06<inman>Timon: i kinda like the idea of myth helping users to time-travel. :-)
22:06<Timon>hahaa
22:06<Chutt>how big the screen is, how far away you sit from it, etc
22:06<poptix>Chutt: perhaps an #ifdef is in order for the MPEG Hardware line to say 'Detected MPEG Hardware but no V4L2'
22:06<rkulagow_>moegren: another unknown weather type, zip code 11358, Friday is "193", "PM Drizzle"
22:07<nziarek_>inman: on the playback screen, where you can select either the show title or "all programs" i am running out of vertical space
22:07<inman>vertical?
22:07<Chutt>poptix, it doesn't even know the card exists without v4l2
22:07<moegreen>rkulagow_: thanks
22:07<Chutt>poptix, what _would_ work would be for the ivtv driver to support the v4l1-compat layer that's already written
22:07<rkulagow_>NP
22:08<inman>Chutt: yep.
22:08<Chutt>if i have time i'll look into modifying the driver to do that
22:08<Chutt>should be easy
22:08<nziarek_>inman: yeah, you have this: All Programs The Simpsons - "Treehouse of Horror" 1/23 4:30pm
22:08<nziarek_>just a long string
22:09<inman>nziarek: i'm suggesting that you put Treehouse on the line below. make each item in the list be on 2 lines.
22:09<Chutt>shouldn't it just be 'All Programs The Simpsons 1/23 4:30pm'
22:09<Chutt>?
22:09<inman>that's another perfectly acceptable solution, imo.
22:09<Chutt>doesn't it only show the subtitle if it's in episode mode?
22:09<Chutt>the subtitle's shown below, so it's no big deal if it's not in the list
22:10<moegreen>Chutt: it shows only the subtitle (if available) when you select the show, the All Programs shows both (until the cutoff where it puts the ...)
22:10<inman>Chutt: one thing i hate about tivo is that it's hard to find subtitles. i'd like to configure myth to be able to show the subtitle in each line, but i can do that in "episode mode".
22:10<Chutt>ah
22:10<nziarek_>Chutt: maybe that isn't such a bad idea
22:11<Chutt>inman, right
22:11<inman>so, just make more use of the fact that "all programs" has alternate semantic meaning.
22:11<nziarek_>moegreen: what do you think about that?
22:11<moegreen>nziarek_: I'm going to commit a change which will allow for more colors in the listboxes. Instead of the <active> and <inactive> tags in the <listarea>, it will now be <fcnfont name="active_font" function="active"></fcnfont>
22:12<moegreen>nziarek_: of what? the subtitle being on a second line?
22:12<mdz_>Chutt: I have not managed to catch it in the debugger yet
22:12<mdz_>but it seems to happen immediately after selecting the preferred program
22:12<nziarek_>moegreen: or not there at all? i can see it both ways
22:12<mdz_>at that point the frontend hangs before it has even redrawn the conflict screen, and the backend is gone
22:13<mdz_>I haven't noticed any pattern in when it happens and when it does not, maybe a race
22:13<Chutt>mdz, might be worth it to try and reproduce with current cvs
22:13<Chutt>i've fixed some races like that
22:13<mdz_>Chutt: I can't even reproduce with 0.8 :-)
22:13<moegreen>nziarek_: the 'name' is the name of a previously defined font. I'll email you a copy of the new ui.xml file I have for blue
22:13<Chutt>i don't recall if that particular one was anything i've seen before
22:13<nziarek_>moegreen: thanks, every time you add something, though, i want to redesign it all :)
22:14<moegreen>heh, well...this is just some extra colors :) (mainly for the conflict res. screen)
22:15<Chutt>if people can let me know if current cvs breaks stuff for them, that'd be excellent, btw =)
22:15<Chutt>i'm not entirely sure i got all the locking updated
22:16<rcaskey>this show about the all gay school is somewhat disturbinng
22:17<Chutt>mdz, the problem with the pvr-250 card is that it doesn't respond to v4l-1 ioctls at all
22:17<ply>how usable is myth now?
22:17<Chutt>and if v4l2 isn't installed properly, it won't probe
22:17<Timon>ply: Very
22:17<mdz_>Chutt: yeah, I saw your message after I'd sent mine
22:17<ply>Timon: thanx.
22:18<inman>Chutt: cvs has been working fine for me this evening.
22:18<mdz_>Chutt: what does it do when it gets a v4l1 ioctl? it doesn't return an error?
22:18<mdz_>the stuff in videosource.cpp prints an error when any v4l1 ioctl fails
22:18<Chutt>mdz, not exactly sure
22:19<Chutt>inman, since my last checkin?
22:19<ply>i was wondering if u can use actisys 200L module to control a cable box.
22:19<inman>i seem to be current and have been since i've been testing timon's patch, which has been for a couple hours now.
22:19<Chutt>ok, cool
22:19<Chutt>i'm just about to apply that =)
22:20<Timon>Don't apply it
22:20<inman>Chutt: don't take my testing as very complete, but there doesn't seem to be anything obviously wrong.
22:20<mdz_>ply: yes, you can
22:20<Chutt>timon, you can send in an update :p
22:20<mdz_>ply: er, sorry. I mean 'u' can
22:20<Chutt>inman, well, i'm fairly certain i got most of the common case
22:20<Chutt>inman, i'm just not sure i got the uncommon =)
22:20<ply>is it built in?
22:20<inman>Chutt: better to get some more testing by folks with multiple tuners, etc.
22:20<ply>how bout a radion 7000
22:20<Timon>Yeah, I'm re-working it a slight bit. So users can decide if they want it to work the traditional previous channel way, or the new way
22:20<Chutt>inman, i've been testing with pip and multiple tuners
22:21<Chutt>well, for the past 20 minutes at least
22:21<Chutt>timon, i'd really rather not have yet another config option
22:21<Chutt>for something as trivial as how the previous channel stuff works =)
22:21<Timon>Ok, then you want me to just leave it like it is?
22:21<inman>how about defaulting to the simple method and using a db option that people can tweak by hand?
22:22<Timon>Well, I would prefer that prev channel change the channel right away. But inman wants a channel history
22:22<Chutt>meaning?
22:23<inman>a stack of channels such that hitting 'h' twice jumps back 2 channels.
22:23<Chutt>and why can't it to both?
22:23<Chutt>do
22:23<Timon>Meaning that in Channel History mode, it will take .75 seconds to change channel. In Traditional mode, it will change channel instantly and there will be no history
22:24<inman>Timon: that .75 figure will probably need adjustment for remotes, etc. it may be too long, it may be too short.
22:24<inman>Timon: it's very long for a keyboard user. :-)
22:24<Timon>inman: Yeah, but since myth is designed to be used by a remote . . . : - )
22:24<inman>Timon: have you tested it with a remote?
22:24<Timon>inman: I found .75 was not to long, and not to short
22:24<inman>okay
22:25<Timon>yup. I have a remote on my desk
22:25<Timon>.5 was WAY to short
22:26<Timon>Chutt: Its your call. Commit like it is (Once I send in an update for some minor tweaks), or wait for the config option stuff to allow it to function like a traditional previous channel button and the new way.
22:28<Chutt>ok, i'll wait for the config option
22:28<Timon>ok
22:28<rcaskey>how much of that time is really the remote?
22:29<Timon>Most of it
22:29<Timon>Giving the user enough time to hit the button and decide where to stop
22:29<rcaskey>my remote here seems to take almost no time (i'm just nuding the volume up & down)
22:30<rcaskey>may...be... .25 here for the remote response time
22:30<Timon>rcaskey: Most of the time is used by the user deciding.
22:31<Timon>rcaskey: I'll tell you what, apply the patch and change the 500 (I had the wrong value on my second patch) to 250 and tellm ehow that works for you
22:31<rcaskey>ohhh, I must have misunderstodo
22:31<rcaskey>err misunderstood
22:33-!-hfb [] has quit ["Huh! Wasn't me!"]
22:36<Chutt>timon, i've just applied your hide osd stuff
22:36<Chutt>haven't checked it in, of course
22:36<Timon>Sweet. Thanks!
22:37<Chutt>it's patch time
22:37<Chutt>i changed it a tiny bit
22:37<Timon>How so?
22:37<Chutt>just made it return after it hid the containers
22:37<Chutt>so the entire function doesn't need to be in the else
22:37<Timon>Oh, I see ok. Makes sense
22:37<Chutt>just a small simplification
22:38<poptix>Chutt: i was referring to the issue where setup detects the card, but not the inputs
22:38<Timon>Cool!
22:38<Chutt>poptix, it can't detect the card and not the inputs
22:38<poptix>Chutt: *shrug* it did =p
22:38<Chutt>it's just letting you choose from all the existing /dev/video* entries
22:38<Chutt>not testing if something's there
22:38<poptix>Chutt: well
22:39<poptix>alright, the selection 'hardware mpeg' under a non-v4l2 setup, might want to show a 'V4L2 not detected'
22:39* poptixshrugs
22:39<Chutt>aaah
22:39<Chutt>ok
22:39<Chutt>that makes sense
22:39<poptix>the issue i had, and the issue that guy on the list had
22:39<poptix>same thing
22:39<Chutt>you can select 'hardware mpeg' for any card now =)
22:39<poptix>yeah
22:40<Chutt>not that it'll do anything, of course
22:40<Chutt>but =)
22:40<Chutt>poptix, it's going to be an issue until v4l2 is in an official, stable kernel, unfortunately
22:40<poptix>if nothing else, it'll give an indicator that 'hey, i need to fix v4l2'
22:40<Chutt>yeah
22:40<Chutt>i'll do that
22:40<Chutt>just make that selection be something else
22:41<poptix>changing the #warning to an #error wouldn't be good, because you don't necessarily need v4l2 for most setups
22:41<Chutt>right
22:42<poptix>indicating 'yes, this is the hardware mpeg selection, but no, it won't work because i didn't find v4l2' seems to be the best way of pointing people in the right direction.
22:43<poptix>as for that whine we're experiencing
22:43<poptix>you said the old msp driver fixed it for you, right?
22:43<Chutt>the pre-msp ivtv driver
22:43<poptix>hmm
22:43<Chutt>if i load that, and use the /dev/ivtv device to cat from
22:43<Chutt>no whine
22:43<Chutt>if i then unload that driver, and load the current one (without loading the msp3400 module)
22:43<Chutt>no whine
22:46<poptix>hmm.
22:46<Chutt>indicating that the tiny bit of msp init code in the old driver is correct
22:47<Chutt>and the init code in the new msp3400 driver is not
22:47-!-nziarek_ [] has quit [Read error: 104 (Connection reset by peer)]
22:48<Chutt>pblue_away, around?
22:49<poptix>Chutt: i might try to move that tiny bit of code over to the new driver =p
22:49<Chutt>i looked at it, and wasn't quite sure what's going on
22:49<Chutt>since the old stuff is just sending binary data to over the i2c bus
22:49<poptix>well, neither am I
22:49<poptix>doesn't mean i can't muck with it.
22:49<Chutt>and i wasn't willing to fuck with it
22:49<Chutt>without understanding what it was doing
22:54<poptix>Chutt: you hear that openbsd lost that DARPA grant?
22:54<Chutt>nope
22:54<Chutt>but then, i'm not really interested in the bsds
22:55<Chutt>ah, on /.
22:55<poptix>well, i saw it elsewhere
22:55<poptix>but yeah, i guess it's hit /. too
22:55<poptix>i'm glad we're not wasting money on it =p
22:56<Chutt>instead, people should be wasting money on me
22:56<poptix>yes
22:56<thor_>Chutt, if we made MythDialogBox have an XMLParse() and an xmldata , it could automate all the paintEvent stuff ... add sockets and signals to UITypes, and this would go a lot faster for others in the future.
22:56<poptix>i don't like the idea of giving $2m to a lunatic to fund word on an OS that isn't even SMP aware
22:56<poptix>word/work
22:56<Chutt>thor, perhaps, yeah
22:57<poptix>(whose primary FTP server is an old solaris box)
22:57<thor_>.... sounds like poptix is not a huge openbsd fan ....
22:58<Timon>I've never used open bsd. But I like FreeBSD
22:58<moegreen>thor_: if you take a look at what I had in the playbackbox.cpp before I created all the uitypes stuff - you'll see this is already saving you a lot of work
22:58<moegreen>:)
22:58<Timon>Chutt: Want me to sent the final patch to you or the list? (Btw, I decided to take out the traditional vs channel history type of functionality)
22:58<Chutt>either's fine
22:59<Timon>err send
22:59<Chutt>doesn't really matter to me
22:59<Chutt>but then, it's not my mail server =)
22:59<Chutt>but the patch is small
22:59<Chutt>so no worries
23:00<Chutt>huh, membership levels to mythtv-dev has been surprisingly stable
23:00-!-PeteCool [~pete@modemcable019.237-201-24.mtl.mc.videotron.ca] has joined #mythtv
23:00<Chutt>hovering right around 760 people for awhile now
23:00<Chutt>-users is up to almost 700 now, though =)
23:00<thor_>moegreen, yeah ... absolutely. Don't get me wrong, the stuff works really well.
23:01<Morph>Chutt: 680 of those addresses are me.
23:01<Chutt>heh
23:01<Timon>haha
23:01<Chutt>multiple posting addresses? :
23:01<Chutt>err, :p
23:01<Morph>Umm..yes..
23:01<rcaskey>27 belong to Tejanos living in my basement
23:01<PeteCool>Chutt, moegreen: I don't see the delete popup after coming back from playing a recording (it's fine when I'm deleting from the playbackbox, though)
23:01<moegreen>thor_: i'm going to be committing this change to UIListType - you'll be fine if you're not using the EnableForceColor() function
23:02<thor_>But loadWindow(), updateBackground() and the PaintEvent() looks like we are redoing (almost) the same thing in a few different places.
23:02<PeteCool>I just realized when I talked about this before I didn't say it was after playback
23:02<Chutt>petecool, ah, ok
23:02<thor_>... no problem here
23:02<Chutt>which theme?
23:02<rcaskey>Some guy gave me a call because of the jacket I bought on ebay. Apprently he just calls people based off of MSRP and not on sale price ;) He tried to sell me a pair of shoes that were like three times as much as the coat I bought :)
23:02<PeteCool>Chutt: sasquatch and iulius... havent used blue/liquid in ages, I'll try it now
23:03<Chutt>just to make sure it's not a theme bug
23:03<moegreen>thor_: updateBackground() is drawing to the actual background, so the images there don't need to be re-alpha blended - it's for speed
23:03<Timon>Chutt: Sent to the dev list
23:03<Chutt>moegreen, but that could be done automatically
23:03<Chutt>moegreen, if there was a dialog type or something
23:03<Chutt>same with the painting, i think
23:03<moegreen>Chutt: oh ... this is reference to the earlier converstation
23:03<thor_>yeah .. I would keep it, but out in MythDialogBox
23:03<Chutt>right
23:04<Chutt>the paintEvent could just be iterate over all the containers, see if they intersect, draw em
23:04<Chutt>assuming the container has its positioning rect and stuff, of course
23:04<PeteCool>Chutt: blue has the same problem
23:04<Chutt>ok
23:06<PeteCool>Chutt: that's with "prompt on playback exit" and "prompt at end of recording" enabled - both show it
23:06<Chutt>well, prompt on playback exit should be in the video
23:06<Chutt>and i don't ever recall that not working
23:07<PeteCool>hmm, that wasn't very clear of me... I meant when I'm deleting from that menu (playback exit) and that moment (end of file)
23:07<PeteCool>there's an unshaded area where the popup should be, though
23:07<Chutt>ok
23:08<thor_>moegreen ... you've already done 90% of the work ... I'd just like to stick LoadWindow(), parseContainer(), parseFont(), paintEvent(), updateBackground(), etc. into the MythDialog object.
23:08<Chutt>thor, maybe make a MythThemedDialog object
23:08<Chutt>just to keep it a tad separated =)
23:09<thor_>yup .... although I don't want to do anything with moegreen's blessing ...
23:09<thor_>s/with/sithout/
23:09<moegreen>make it happen :)
23:12-!-froggy19 [~fryguy@12-246-43-18.client.attbi.com] has joined #mythtv
23:12-!-febisfebi [~jimbob@12-231-127-22.client.attbi.com] has joined #mythtv
23:13<yebyen>woo mythtv
23:13<mdz_>I've been looking around for someplace local where I can pick up a tv-out card that will work for mythtv, but haven't had any luck so far
23:14<mdz_>this RMA is going to take weeks probably
23:14<_shad>poor mdz :)
23:14<_shad>just buy a 35" monitor
23:14<PeteCool>and return it when it's done
23:14<mdz_>heh
23:14<mdz_>that's bigger than my TV
23:15<mdz_>and I would have no place to put it
23:15<_shad>replace your tv :P
23:18<yebyen>HEH
23:20<yebyen>mythweather is freakin awesome
23:21<yebyen>too bad there's no weather to speak of
23:21<yebyen>or this doppler deal would look cooler
23:23<yebyen>at least there's no snow in the forecast :X
23:23<moegreen>rkulagow_: are you around?
23:27* rcaskeyhrmms and needs to take some pictures of himself this weekend
23:29<Chutt>thor, any reason for the lcd stuff to be in the mythmusic section?
23:31<moegreen>Chutt: conflict resolution is done (hopefully I didn't break anything :)
23:31<Chutt>sweet
23:31<moegreen>well - I broke iulius :)
23:32<thor_>Chutt ... which mythmusic section ?
23:32<Chutt>well, there's LCD settings in mythmusic's settings file
23:32<Chutt>is all
23:32<Chutt>so they're in this settings patch
23:33<Chutt>would probably make more sense to put em in the general section of the main bit
23:33<moegreen>is there (supposed) to be a way to cancel a scheduled recording from the conflict screen? Or do you have to find the program and do it from there?
23:33<thor_>Ahhhh ... no .... I did that before I realized that getSettings() went to the database as well as the local file
23:34<thor_>rip away
23:34<Chutt>moegreen, yeah, you have to find it and do it from there
23:35<moegreen>hmmm...looks like I'll have to fix something - mainly scrolling - oops
23:35<Chutt>heh
23:37-!-rcaskey [] has quit ["Client exiting"]
23:38<Chutt>hrm
23:38<Chutt>wonder how hard it'd be to rework the setup stuff
23:39<Chutt>only thing is, i'm out of room in the top level menu
23:39<yebyen>hm
23:39<Chutt>so i need to make an automatic "more" item
23:39<yebyen>Chutt: do you use a windowmanager with mythtv?
23:39<Chutt>generally kwin
23:39<yebyen>i'm using sawfish, works perfectly with mythtv but when I throw mplayer into the mix, it hates me.
23:40<yebyen>after mplayer quits, mythtv doesn't have focus until I go over and click on it with the mouse
23:40<yebyen>HEH
23:40-!-febisfebi [] has quit ["Client exiting"]
23:40<moegreen>Chutt: is this backend setup you're talking about?
23:40<Chutt>no
23:40<Chutt>frontend =)
23:40<Chutt>i want to move 'setup' to the top level
23:40<Chutt>then have a 'general' section
23:40<Chutt>with the theme and other shared settings
23:41<moegreen>Chutt: how about making some of the options 'Advanced'?
23:41<Chutt>then a 'TV setup' 'Music setup' 'Weather setup', etc
23:41<yebyen>Chutt: kwin require any configuration to work perfectly? Maybe I'll switch..
23:41<Chutt>yebyen, just has to be in click-to-focus mode
23:41<Chutt>moegreen, unsure of how to accomplish that
23:42<moegreen>hmmm...i see. For the weather setup - should I just setup mythweather to respond to a command line to go right to setup, or are you thinking you want to bring it in totally?
23:43<Chutt>i'll set it up to do that
23:43<Chutt>but yeah
23:43<Chutt>each module is going to have to respond to 'mythweather --configure' or something like that
23:43<Chutt>which'll just run the setup bits
23:43<moegreen>well ... looks like my work is pretty much finished with respect to that :)
23:44<froggy19>i'm having trouble getting mythTV working right.
23:44<froggy19>actually it's not a problem with mythTV i don't think
23:44<froggy19>but ALSA or something
23:47<froggy19>is there any better documentation for it than on http://www.alsa-project.org ?
23:47<Chutt>not that i'm aware of
23:47<froggy19>it might be a problem with the fact that i've got on-board sound and the cmedia chip (cmipci module)
23:48<froggy19>i can get alsamixer to pass-through the microphone, but when I mute it i can't get it to play using `aplay /dev/dsp`
23:49<Chutt>got capture set to whatever input you're using
23:49<Chutt>got the 'igain' or 'captur' at a non-zero volume level
23:49<froggy19>yes I hit space on it
23:49<froggy19>and it put CAPTUR around the microphone (the one I mute/unmute and it turns it on/off)
23:50<mdz_>Chutt: do you know if the nvidia binary drivers work with non-nvidia branded cards with the geforce2 mx gpu?
23:50<Chutt>there's sometimes a separate mixer
23:50<Chutt>mdz, there's non-nvidia branded cards?
23:50<froggy19>would igain be in the conf file?
23:50<mdz_>like this: http://www.compusa.com/products/product_info.asp?product_code=293736&pfp=BROWSE
23:50<mdz_>"Champion"
23:50<mdz_>from "Mad Dog Multimedia" :-)
23:51<PeteCool>mdz_: I don't know of any nvidia-branded retail video cards
23:51<PeteCool>mdz_: they all work the same
23:51<Chutt>froggy19, should be in alsamixer
23:51<mdz_>hmm
23:51<froggy19>how do I change igain?
23:51<Chutt>mdz, nvidia doesn't make cards
23:51<mdz_>oh
23:51<PeteCool>mdz_: actually they are most of the time the reference board with a different (noisier) cooler
23:51<Chutt>froggy19, if it exists
23:52<Chutt>just something to check for
23:52<Chutt>mdz, i'm pondering making space not 'accept' the current dialog
23:52<Chutt>in settings
23:52<mdz_>Chutt: in favor of what? navigating to the 'next' button?
23:52<Chutt>yup
23:53<thor_>I'd vote for that
23:53<mdz_>is there anything left which is using configurationdialog itself, and not configurationwizard?
23:53<Chutt>nope
23:53<thor_>(not that I have any delusions of democracy)
23:53<Chutt>which is why i think it can be done
23:53<mdz_>yeah, sounds sane
23:53<Chutt>do the list boxes need that?
23:53<Chutt>they just use the dialog, but i think they have an override anyway
23:53<mdz_>yeah
23:54<mdz_>wait, hmm
23:54<mdz_>have to look at it
23:54<mdz_>yeah, they use configurationdialog::exec
23:54<Chutt>and i can always just leave the space override for that widget
23:54<mdz_>they could be made to work differently though
23:55<mdz_>I'm not really happy with the way they work right now anyway
23:55<Chutt>i mainly just want text entry to work normally
23:55<mdz_>ah
23:55<mdz_>you crazy people and your keyboards
23:55<thor_>I can put space back into MythRemoteLineEdit
23:56<froggy19>also the other problem I was having was I couldn't get the Television input working on my ATI Wonder VE. The composite worked fine, but not the Television
23:56<mdz_>the various listboxes which don't actually store the selection anywhere could be reworked to be a regular list box with a binding for enter
23:56<yebyen>heh
23:56<Chutt>froggy19, are you setting the tuner type when you load the bttv/tuner modules?
23:56<froggy19>not manually
23:57<froggy19>it was crashing when I did
23:57<froggy19>er seg fault
23:57<froggy19>and I couldn't remove it from the modules list
23:57<froggy19>so I couldn't reboot without powering down manually
23:58<Chutt>mdz, well, for now i'll just remove the ignore for space from the line edit
23:58<mdz_>I think I made them configuration dialogs because it let me reuse code for loading the stuff from the db
23:58<mdz_>yeah, that seems fine
23:58<froggy19>tuner is listed in the lsmod list