Back to Home / #mythtv / 2007 / 03 / Prev Day | Next Day
#mythtv IRC Logs for 2007-03-24

---Logopened Sat Mar 24 00:00:30 2007
00:25|-|adante [] has quit [Read error: 104 (Connection reset by peer)]
00:25|-|adante [] has joined #mythtv
00:32|-|jd99 [] has quit ["Leaving"]
00:38<knowledgejunkie>xris: can you spare a couple of minutes to talk about the from yesterday?
01:00<xris>knowledgejunkie: I'm around now for a bit, so sure
01:02<knowledgejunkie>xris: thanks
01:03<knowledgejunkie>xris: wondering about how to get icon/xmltvid data submitted for uk_rt channels - i.e. CSV format of a submission, and whether the services server can actually handle this yet
01:04<xris>it'll get submitted, but I won't be approving things until there are more submissions
01:04<xris>easier to just have users run it on their own
01:05<knowledgejunkie>xris: when I run the script, it exitrs without allowing me to submit anything
01:05<knowledgejunkie>xris: yet when I run a lookup by xmltvid, it reports only 23/260 as available
01:05<xris>knowledgejunkie: wipe out the channel icon data from your database
01:06<xris>the new system doesn't just keep track of xmltvid or whatever. also does dvb id
01:07<knowledgejunkie>xris: i know that - shouldn't it populate the other fields using the icon as the key, but still allow part submissions
01:07<xris>if you already have channel icons, it won't try to look them up.
01:07<xris>it has no way to know if the icon on disk matches the one on the server.
01:07<xris>it's not like there's an ID system or anything for the filenames.
01:08<knowledgejunkie>xris: I thought it would match the URL and XMLTVID
01:08<xris>there is no URL stored anywhere
01:10<knowledgejunkie>xris: so how does the script give you an icon for a matched channel?
01:10<xris>download URL is stored in multiple pieces in the db on the server.
01:11<xris>but there's no URL stored on the mythbox
01:11<xris>just a file path
01:11<knowledgejunkie>xris: I meant the URL on the server - the icon is surely the link for everything else to key to
01:12<xris>but there's no way to match up the local file to anything on the server.
01:15<knowledgejunkie>xris: I see. So, in the found channels submit to the server, what does $icon contain?
01:16<xris>record id in the server's database
01:16<xris>and a few other things, but that id is the most meaningful
01:16<knowledgejunkie>xris: Ah
01:17<xris>it's basically the icon id, and then some info about the channel (callsign, xmltvid, dvb ids, etc)
01:17<xris>only way to get the icon id is to get it from the server.
01:17<xris>lyngsat isn't the only source.
01:20<xris>btw, if you want to rescan, you can use --rescan and it'll ignore existing icon mappings.
01:22<knowledgejunkie>xris: As I've been maintaining the XMLTV uk_rt grabber channels and icon listings for the past couple of years, what I guess I'm asking is whether there is an easy way to get those icon:xmltvid mappings onto the server
01:22<knowledgejunkie>xris: sorry for delay - batteries in keyboard just expired
01:23<xris>I imported as much of the existing uk iconmap stuff as I could find, but it was only callsign stuff.
01:23<xris>if you get me a list of xmltvids and icon urls, I can probably write an importer script, though
01:24|-|jshadias [] has joined #mythtv
01:25<knowledgejunkie>xris: what format would you like it in? You can very easily awk the listing I keep up-to-date at
01:25<knowledgejunkie>xris: nearly all are lyngsat
01:26<xris>I can work with that format
01:26<knowledgejunkie>xris: awesome
01:27<knowledgejunkie>xris: the other obvious UK data is linking those icons to DVB serviceids, and to DVB channel names
01:28<xris>I don't touch channel names
01:28<knowledgejunkie>xris: I guess the existing callsign stuff does some of the latter
01:29<xris>yeah. would like to avoid callsigns if possible
01:29<xris>too many overlaps around the world
01:29<knowledgejunkie>xris: exactly
01:30<xris>xmltvid is imho the best thing to use, but there are a number of people who don't use anything with xmltvids (e.g. pure EIT data sources)
01:30<knowledgejunkie>xris: what are the minimum number of identifiers? xmltvid and then dvb/atsc serviceid?
01:31<xris>what are the field names in that file?
01:31<xris>the middle ones
01:31<knowledgejunkie>xris: all XMLTV users should have XMLTVIDs, and OTA DVB/ATSC users should have serviceids or atscsrcids
01:31|-|czth__ [i=dbrobins@nat/microsoft/x-913806d98b1df708] has joined #mythtv
01:31<xris>for dvb: transportid, networkid, serviceid
01:32<knowledgejunkie>xris: the numeric field?
01:32<xris>I assume the other is callsign or channame
01:32<knowledgejunkie>xris: that's an abstract reference to the listings file name on the Radio Times site
01:33<knowledgejunkie>xris: the third field is either what I decide to call the channel :) or what the Radio Times does
01:33<knowledgejunkie>xris: We have no such thing as callsign in the UK
01:33<knowledgejunkie>xris: it's just a channel identifier really
01:33<xris>good enough
01:34<xris>just wanted to know what to use for variable names when extracting. heh
01:34<knowledgejunkie>xris: but it's what XMLTV will use when setting up channels for a channel name
01:34|-|daMaestro [n=jon@fedora/damaestro] has joined #mythtv
01:36|-|jshadias [] has quit ["Leaving"]
01:38|-|kali67 [] has joined #mythtv
01:39<knowledgejunkie>xris: Line 328, does the script return 1 if xmltvid lookup fails?
01:40<xris>returns an array with the error message and $error set to true
01:42<knowledgejunkie>xris: just wondering if I could use the error code to count for an unsuccessful lookup via another script
01:43<xris>it's not a code. just says "the first field in this array contains an error message"
01:43|-|kali67 [] has quit []
01:45<knowledgejunkie>could it return an exit code in addition to that?
01:46<xris>it's not an exit code.
01:46<xris>easy enough to retrieve the data from a query, though. it's just csv
01:46<xris>what lookup are you trying to do?
01:47<knowledgejunkie>I was after an exit code when running --xmltvid {string} - I could count all failed lookup as one I needed data for
01:48|-|czth_ [i=dbrobins@nat/microsoft/x-d87f0e274dc55eb2] has quit [Read error: 110 (Connection timed out)]
01:48<xris>or.. you could delete all of your channel icon data from the db and rerun now that I've imported all of your xmltvids
01:48<knowledgejunkie>ooh ok
01:49<xris>I actually imported all of the UK xmltvid entries from the xmltv source, but you've apparently added the icon info since I grabbed the source.
01:50<knowledgejunkie>the URL I posted earlier was current CVS
01:50<knowledgejunkie>not latest release :)
01:52<knowledgejunkie>just running script now
01:54<knowledgejunkie>ouch - returned info about 37 of 119 channels.
01:55<xris>oh, probably no perfect matches
01:56<xris>it imported a ton of them
01:56<xris>more than 200
01:56<knowledgejunkie>I should have 6 iconless channels, and 8 radio channels that are not supported via XMLTV
01:56<knowledgejunkie>the rest should all be
01:58<xris>I think there's a bug in the lookup code. hang on.
01:59|-|n33o [] has left #mythtv []
01:59<knowledgejunkie>e.g. BBC Four - has a XMLTVID and the script returns a single choice - so it should be a direct hit. It seems to be using fuzzy name matching, not xmltvid
02:01<xris>ok, rerun the scanner script
02:01<xris>had a bug in the lookup code on the server
02:02<knowledgejunkie>better - returned info about 73 of 119 channels.
02:02<xris>still not that many
02:04<knowledgejunkie>SELECT count(xmltvid) FROM `channel` where xmltvid != ''; returns 105 records
02:05<knowledgejunkie>105 xmltvid + 6 none + 8 unsupported = 119 channels
02:05<xris>it should have pulled down all of them
02:06<xris>unless there were that many non-lyngsat channels
02:06<knowledgejunkie>lemme check
02:06<xris>everything there should be on the server now
02:07<xris>ah, crap
02:07<xris>hang on
02:08<xris>ok, rerun your scan
02:10<knowledgejunkie>crap, MFDB is running atm so has re-populated my icons after I just cleared them
02:10<xris>haha. oops
02:11<knowledgejunkie>there looks to be a problem with MFDB too regarding icons - it is the culprit for the path breakage I blamed your script for yesterday :)
02:11<xris>yeah, I saw you mention that earlier. heh
02:12<knowledgejunkie>MFDB has nearly finsihed
02:13<knowledgejunkie>is stuarta the MFDB dev?
02:14<knowledgejunkie>there seems to be an issue with IconData::UpdateSourceIcons
02:14<xris>no clue
02:15<knowledgejunkie>righty, MFDB has finished
02:17<xris>how many icons detected this time?
02:17<knowledgejunkie>hey - returned info about 97 of 119 channels. looks promising as there are 8 with xmltvids but no icon
02:19<xris>I was accidentally excluding xmltvids with dashes in them
02:19<xris>should all be fixed now
02:19<knowledgejunkie>animal planet has a clash
02:20<xris>let it go through anyway
02:20<xris>the "force" part of that should work now.
02:20<xris>and it should submit the parts that don't conflict
02:21<knowledgejunkie>how does the server order fuzzy matches?
02:22<xris>probably alphabetically
02:22<knowledgejunkie>i have exact matches on and yet they are result #39. Shouldn't they be first?
02:23<knowledgejunkie>like, if match==exact, list first, then fuzzies
02:23<xris>dunno. this script is just a proof of concept, though
02:23<xris>the real code should go into the frontend, mythtv-setup and/or mythweb.
02:23<xris>so you can actually SEE the icons it's picking
02:24<knowledgejunkie>omg - E! Entertainment - only 400 matches!
02:24<xris>yeah. fix that one quick. heh
02:25<knowledgejunkie>submitted :)
02:26<knowledgejunkie>Imported 103 channel icons into your MythTV database.
02:26<knowledgejunkie>lemee see what the missing 2 are
02:27<xris>ok, most/etc approved
02:28<knowledgejunkie>2 missing both have xmltvids and icons in XMLTV channel_ids
02:28<knowledgejunkie>which weren't approved?
02:28<knowledgejunkie>missing #1:|482|CBBC|
02:28<knowledgejunkie>missing #2:|228|QVC|
02:29<xris>I don't see any pending for those
02:29<knowledgejunkie>didn't ask me about them
02:29<xris>resyncing with the lyngsat website
02:31<xris>qvc is already in the db
02:32<xris>as is the other
02:32<xris>they should have auto-assigned
02:36<xris>knowledgejunkie: what do you mean by "missing"?
02:40<knowledgejunkie>xris: they have an xmltvid in the DB, but the script didn't grab their icons
02:40<xris>it should have
02:40<xris>the server shows icons for them
02:41<knowledgejunkie>i'll run the script again, now you've accepted some of my submissions from the last run
02:41<knowledgejunkie>heh - returned info about 109 of 119 channels.
02:44<knowledgejunkie>it missed "1Xtra" which I manually searched for, then reported "Imported 103 channel icons into your MythTV database."
02:45<xris>no xmltvid for that one, but the dvb codes are in the system now
02:45<knowledgejunkie>still missed CBBC and QVC
02:46<xris>wtf. and those are in the system cleanly
02:46<xris>no prompt or errors or anything?
02:49<xris>if it finds a solid xmltvid match, it just imports the icon straight in
02:49<knowledgejunkie>aw crap, they're hidden
02:50<knowledgejunkie>that'll be it
02:50<xris>sounds like the UK users are all set up, then. heh
02:51<knowledgejunkie>i think it should be an option - --importhidden, or something
02:52<knowledgejunkie>MythTV/XMLTV uk_rt users of the UK thank you greatly
02:52<knowledgejunkie>so do I - thanks for the effort and 'getting it sorted' :)
02:52<xris>I just want to provide a good user experience. :)
02:53<knowledgejunkie>well, let's keep things in perspective here :)
02:53<knowledgejunkie>i mean, I actually have to run the script manually...
02:55<xris>like I said, it's just a proof of concept
02:55<knowledgejunkie>xris: how's the callsign/channum matching stuff going for mythweb?
02:56<xris>it can't go into mythweb until I have a way to tell the backend to download something
02:56<knowledgejunkie>it's a pretty good proof of concept then - it works brilliantly
02:56<xris>I'm trying to get mythweb separated from the mythtv filesystem as much as possible... no more requirements of mounting your recordings dir on your webserver, etc.
02:57<knowledgejunkie>everything through mythbackend directly?
02:59<xris>or some other backend-like util
02:59<xris>can already access music/video/recording files that way, but there's no way to tell the backend to go out and grab a file.
03:35|-|beavis [] has joined #mythtv
03:53|-|xris [] has quit ["Leaving."]
04:16|-|knowledgejunkie [n=knowledg@unaffiliated/knowledgejunkie] has left #mythtv ["I want my, I want my, I want my MythTV"]
04:50|-|daMaestro [n=jon@fedora/damaestro] has quit ["Leaving"]
05:46|-|gbee [] has joined #mythtv
06:55|-|aevil [] has joined #mythtv
07:29|-|aevil^aw [] has joined #mythtv
07:31|-|aevil [] has quit [Read error: 110 (Connection timed out)]
07:54|-|robthebob [] has joined #mythtv
08:23|-|boardom [] has joined #mythtv
09:09|-|jd86 [n=jdio@pdpc/supporter/student/jd86] has joined #mythtv
09:11|-|jd86 [n=jdio@pdpc/supporter/student/jd86] has quit [Remote closed the connection]
09:48|-|boardom [] has quit [Read error: 104 (Connection reset by peer)]
09:56|-|clever_ [] has joined #mythtv
10:01|-|clever [] has quit [Read error: 110 (Connection timed out)]
10:13|-|pac1 [n=pac1@] has joined #mythtv
10:25|-|butterfly__ [] has joined #mythtv
10:25|-|butterfly__ [] has left #mythtv ["Leaving"]
10:31|-|festr___ [] has joined #mythtv
10:31<festr___>anyone using G450 matrox with mythtv?
11:01<sigger_>ok, mythmusic's telnet link has two commands. the rest are on the way.
11:03<sigger_>Captain_Murdoch, it appears we're able to call GUI functions from within the GUI thread's customEvent. I suspect you're recollection you may have done the "impossible keypress" because tv playback is not a separate thread and it would have locked up the telnet control.
11:04<sigger_>er, I suspect your recollection was correct, that is.
11:04|-|stuarta [n=stuarta@unaffiliated/stuarta] has joined #mythtv
11:09|-|jd86 [n=jdio@pdpc/supporter/student/jd86] has joined #mythtv
11:15|-|aevil^aw [] has quit [Remote closed the connection]
11:22<Captain_Murdoch>but does it lock the gui thread if you run something that takes a while from within customEvent?
11:23<Captain_Murdoch>ie, the sleep(10) test I mentioned.
11:29|-|toad_ [n=toad@pdpc/supporter/student/toad-with-underline] has joined #mythtv
11:29|-|toad_ [n=toad@pdpc/supporter/student/toad-with-underline] has left #mythtv ["Konversation terminated!"]
11:40|-|billk [] has joined #mythtv
11:42|-|mca_ [] has joined #mythtv
11:42|-|mca [] has quit [Read error: 104 (Connection reset by peer)]
11:43|-|toad_ [n=toad@pdpc/supporter/student/toad-with-underline] has joined #mythtv
11:43<toad_>has anyone tried running myth with -fbounds-check recently?
11:46|-|toad_ [n=toad@pdpc/supporter/student/toad-with-underline] has left #mythtv ["Konversation terminated!"]
11:51|-|gnome42 [] has joined #mythtv
11:57|-|stuarta [n=stuarta@unaffiliated/stuarta] has quit ["leaving"]
11:58|-|rtsai1111 [] has joined #mythtv
11:59|-|onixian [] has quit [Read error: 110 (Connection timed out)]
12:01|-|Dave123 [] has quit ["Leaving"]
12:06|-|rtsai [] has quit [Read error: 60 (Operation timed out)]
12:07|-|hatlevip [] has joined #mythtv
12:17|-|mca_ [] has quit ["BitchX-1.1-final -- just do it."]
12:20|-|mca [] has joined #mythtv
12:47|-|rtsai [] has joined #mythtv
12:56|-|rtsai1111 [] has quit [Read error: 60 (Operation timed out)]
13:29|-|cattelan [] has joined #mythtv
13:29|-|perlmonkey [] has joined #mythtv
13:30|-|perlmonkey [] has left #mythtv []
13:43|-|onixian [] has joined #mythtv
13:51|-|beavis [] has quit [Connection timed out]
13:52|-|beavis [] has joined #mythtv
13:56|-|xris [] has joined #mythtv
14:17<sigger_>Captain_Murdoch, putting sleep(10) in the gui thread's customEvent does prevent the network command code from getting control back..
14:39|-|Kritter [] has joined #mythtv
14:45<Captain_Murdoch>I'm not wondering about the network control code, I'm wondering about the whole gui thread.
14:47<Captain_Murdoch>if customEvent is fired from the GUI thread, then it would seem that any action that blocks inside customEvent would hold up the GUI thread from doing other things (such as sending other events or updating the screen, etc.. that's what I think thinking happened when I was testing that code originally.
14:48|-|daMaestro [n=jon@fedora/damaestro] has joined #mythtv
14:49|-|daMaestro [n=jon@fedora/damaestro] has quit [Read error: 104 (Connection reset by peer)]
14:51<sigger_>oh I gotcha will test
14:54<sigger_>yes, a sleep in the gui thread means gui not getting updated for the duration of the sleep. certainly makes sense; that thread is prevented from getting back to the event loop.
14:59<sigger_>but as I think about it, the KeypressEvent stuff is just as much a part of the event loop as customEvent, so they would have the same effect it seems.
15:13|-|fiber0pti [] has joined #mythtv
15:13|-|fiber0pti [] has left #mythtv []
15:49|-|brianw_ changed nick to JoeyJoeJo
15:49|-|hatlevip [] has quit [Remote closed the connection]
15:54|-|tominglis [] has joined #mythtv
15:56|-|tominglis [] has left #mythtv ["Konversation terminated!"]
16:10|-|jd86 [n=jdio@pdpc/supporter/student/jd86] has quit [Remote closed the connection]
16:23|-|tominglis [] has joined #mythtv
16:24<tominglis>hi guys, does anyone know why my remote control works basically (for navigation et al) with a clean feisty and 0.20 install and no lirc / lirc config files?
16:24<tominglis>it is nebula electronics digitv rev b
16:38<xris>tominglis: /topic
16:39<tominglis>yeah, i know but nobody in there seems to know, and i thought you might? :)
16:40<tominglis>sorry to disrupt your conversation :)
16:43|-|billk [] has quit ["Leaving"]
16:51|-|tominglis [] has quit ["Konversation terminated!"]
16:56|-|stoffel [] has joined #mythtv
17:19|-|Inssomniak [] has joined #mythtv
17:21|-|Inssomniak [] has left #mythtv ["Leaving"]
17:35|-|maxxer [] has joined #mythtv
17:36|-|maxxer [] has left #mythtv ["hypeeeeeee"]
17:39|-|packetscan [] has quit [SendQ exceeded]
18:05|-|knowledgejunkie [n=knowledg@unaffiliated/knowledgejunkie] has joined #mythtv
18:06|-|stoffel_ [] has joined #mythtv
18:20|-|stoffel [] has quit [Read error: 110 (Connection timed out)]
18:25|-|packetscan [] has joined #mythtv
18:30|-|Dave123 [] has joined #mythtv
18:39|-|beavis [] has quit ["Verlassend"]
19:10|-|doc_ [] has joined #mythtv
19:11<doc_>hi there, would be possible to have an open connection with socket and send every event from backend to it ?
19:37|-|stoffel_ [] has quit ["leaving"]
20:02|-||egacy2 [n=b@] has joined #mythtv
20:06|-|jams [] has quit [Read error: 104 (Connection reset by peer)]
20:07|-|jams [] has joined #mythtv
20:12|-|rbellamy [n=rbellamy@pdpc/supporter/silver/rbellamy] has joined #mythtv
20:23|-|akty [] has joined #mythtv
20:26|-|akty [] has left #mythtv []
20:35|-||egacy2 [n=b@] has quit []
20:42|-|doc_ [] has quit ["boumf"]
21:20|-|MoRpHeUz [] has joined #mythtv
21:28|-|MoRpHeUz [] has left #mythtv ["Leaving..."]
21:47<Captain_Murdoch>sigger: yeah, that makes sense. I can't remember what then. there was a reason I did it this way, I remember having issue with something I think, but can't remember what. I threw 90-95% of the network control code together in a 10-hour overnight coding session one weekend, so it's no wonder I don't remember what the issue was. :)
21:48|-|rg535748 [] has joined #mythtv
21:49|-|rg535748 [] has left #mythtv []
21:50|-|robthebob [] has quit [No route to host]
22:01|-|purserj [n=purserj@] has quit [Read error: 54 (Connection reset by peer)]
22:01|-|purserj [] has joined #mythtv
23:34|-|xian__ [] has joined #mythtv
23:45|-|onixian [] has quit [Read error: 110 (Connection timed out)]
---Logclosed Sun Mar 25 00:00:18 2007