:: #mythtv

Daily chat history

Current users (75):

alan`, aloril, amessina, Anssi, arescorpio, Beirdo, brfransen, buu, caelor, Casper82, cesman, ChanServ, Chutt_, clever, coling, cybrNaut, dekarl, ElmerFudd, esperegu, fetzerch, gary_buhrmaster, ghoti, gregL, GreyFoxx, Hydr0p0nX, J-e-f-f-A, jab416171, jams_, jheizer, jnylen, jpabq_, jpharvey, jst, jwhite, jya, Kevin`, knightr, kormoc, kurre2, len, len__, moparisthebest, MythBuild, MythLogBot, nephyrin, peper03, poptix, pppingme, purserj_, rhpot1991, rich0, robink, rOOb, seld, Sharky112065, sheedy-away, skd5aner, sl1ce_1g, sphery, sraue, stuarta, stuartm, suffice, superm1, taylorr, tgm4883, toeb, tonsofpcs, tris, unforgiven512, wagnerrp, XDS2010_, xris, zentec, _charly__
Monday, February 9th, 2015, 00:00 UTC
[00:00:25] dekarl: btw, the example at MSDN has a mix of open and closed tags, too
[00:01:01] stuartm: dekarl: I've overlooked the need to do replacement of special xml chars, there's a function in utility.js for that purpose
[00:01:09] dekarl: ok, so the xml headers wasn't enough for vlc
[00:01:47] stuartm: dekarl: the Microsoft documentation leaves a lot to be desired, I found it severely lacking, imprecise and generally confusing
[00:03:14] dekarl: maybe its another missing feature in vlc, I'll test with WMP and then hit sack
[00:04:14] joki (joki! has joined #mythtv
[00:09:48] stuartm: so STARTTIME doesn't really work as advertised?
[00:10:53] stuartm: actually, I can image why it might not, deriving a byte offset from a time based value is extremely difficult, it can't really work very well unless the bitrate is a) known and b) constant
[00:11:44] stuartm: to seek to STARTTIME it would need to download then decode until it reached that point, which would be slow
[00:12:16] dekarl: stuartm: WMP 12 a) you must not use the xml utf-8 header (breaks the file), b) utf-8 works
[00:13:10] stuartm: figures they'd use an XML format which isn't actually XML
[00:13:14] dekarl: I've not tested STARTIME, yet. but i looked at the documentation looking for an ENDITME
[00:13:32] stuartm: this is MS after all, standards don't apply to them
[00:13:48] dekarl: looks like ENDTIME is STARTTIME+DURATION
[00:14:42] dekarl: yes, STARTTIME is limited. "can be used only with stored, on-demand content that has been indexed."
[00:17:29] stuartm: dekarl: to me xspf is the more interesting format, no WMP support, but it supports stuff like album art for music which is a definite plus – lacks a built in cutlist type feature, although vlc has extensions to the format that do add that
[00:17:31] stuartm:
[00:18:22] stuartm: again though, uses time in seconds instead of byte range
[00:18:32] dekarl: stuartm, I concur. ASX was just for feature parity with mythweb, now it works good enough :)
[00:19:00] dekarl: we should do so too
[00:19:05] Warped (Warped!~Warped@ has quit (Quit: ChatZilla [Firefox 35.0.1/20150122214805])
[00:19:14] dekarl: use the PCR or so to cut
[00:19:21] dekarl: but thats future stuff
[00:20:25] dekarl: also we got a map of time<->byte now, might as well use it
[00:23:14] stuartm: dekarl: we can support the cutlist on the server side instead, just serve up the video skipping over the byte ranges, worth a try anyway
[00:23:33] stuartm: thought about that for upnp
[00:23:41] dekarl: but only for MPEG2TS?
[00:24:17] dekarl: might just demux and serve a video and an audio stream via RTP (using standards and such)
[00:24:29] dekarl: the use plain old RTSP trick play
[00:24:29] stuartm: yes, obviously can't do it for containers where the discontinuities wouldn't be expected, or where headers are only at the start of the file
[00:26:08] dekarl: demuxing and just sending one audio/video stream might be nice for the use case I heard lately "I transcode so the recording fits the network pipe to the frontend" (aka WiFi / Powerlinelan / etc)
[00:26:13] stuartm: I've been reading the RTSP docs this weekend, when I have time I'll implement a RTSP server, but initially only supporting raw tcp/udp, RTP will have to wait until later
[00:26:52] dekarl: would also take care of uncommon containers for some generic clients
[00:27:19] phuff (phuff!~phuff@unaffiliated/phuff) has quit (Ping timeout: 245 seconds)
[00:27:31] stuartm: upnp optionally supports rtsp, so that's my main motivation – it's a reasonably simple standard
[00:27:44] superm1 (superm1!uid4318@ubuntu/member/superm1) has quit (Quit: Connection closed for inactivity)
[00:27:47] dekarl: would implementing a simple RTSP server be towards replacing the file streaming via a custo protocol?
[00:28:29] stuartm: RTP isn't that bad either, but it's treading very similar ground to the websocket stuff I just finished, and I wouldn't mind doing something different for a while
[00:29:35] stuartm: dekarl: no, that's certainly not my goal, too complicated, the existing protocol has evolved to work through a lot of trial and error, repeating that level of tweaking with a new protocol doesn't appeal to me
[00:29:46] stuartm: it ain't broke
[00:31:32] stuartm: someone might be inclined to take whatever work I do and attempt that in the future, but I don't really want to be a punching bag for users who find it doesn't work for them
[00:31:55] dekarl: maybe the QT5 can of worms is enough for a day :)
[00:32:18] dekarl: ahh configure, why is libcec 2.1 not better then 1.6?
[00:33:04] dekarl: When I manually fiddle with the numbers in the configure script it nicely switches between enabled and disabled.
[00:35:22] stuartm: btw, remembered another of the advantages of QT5 – uses chromium (basically) for the webbrowser widget, chromium on linux supports Netflix natively (or so I've heard)
[00:36:02] dekarl: I vaguely remember hearing that netflix might be easier to support with qt5, too
[00:37:22] dekarl: looks like we didn't detect libcec2 before either, maybe the change to configure in isn't cosmetic after all ;)
[00:42:02] dekarl: stuarta/superm1/tgm4883: ^- I plan to merge the configure change for libcec to fixes/0.27 soonish unless someone objects. I removes the broken CEC from 12.04lts, but doesn't add CEC to 14.04 (or later) as we don't support that yet. But its good enough to fix
[00:46:56] stuartm: heh, Geordie's react badly to the news that Newcastle Brown Ale is going to become a real beer with natural, instead of artificial, ingredients
[00:47:16] stuartm: bet it will taste like shit though
[00:47:26] stuartm: still taste even
[00:48:56] dekarl: sturatm: didn't see that coming ;) as its contained and I understand the configure change, there are two success reports. I'm tempted to just push it
[00:49:12] dekarl: oh my, time to hit sack... ^- stuartm
[00:50:31] stuartm: dekarl: I was unlikely to get around to looking at it for some time
[00:50:40] stuartm: and without the hardware ...
[00:51:53] dekarl: I can send you a RPi1 :D
[00:53:19] dekarl: unless you miss other side of the CEC pipe
[00:56:52] stuartm: dekarl: I thought libcec was just for the pulse 8 devices?
[00:58:17] stuartm: of course I appreciate the joke about the RPi, when you get mythfrontend working on there maybe I'll reconsider ;)
[00:58:46] stuartm: a lot later than I realised, going to bed now
[01:03:48] gary_buhrmaster: stuartm: It is my understanding that in *theory* libcec was 100% device agnostic. However, it is sort of like VDPAU, in that it was written/supported by a vendor (pulse8) and other vendors have chosen (or not) to come along.
[01:16:39] sheedy-away (sheedy-away!sheedy@kde/developer/sheedy) has quit (Ping timeout: 272 seconds)
[01:18:50] tgm4883: dekarl: sounds good to me
[01:27:44] sheedy-away (sheedy-away!sheedy@kde/developer/sheedy) has joined #mythtv
[02:16:41] andreaz (andreaz! has quit (Read error: Connection reset by peer)
[02:22:30] amessina (amessina!~amessina@unaffiliated/amessina) has quit (Quit: Konversation terminated!)
[02:40:30] gary_buhrmaster (gary_buhrmaster!~gtb@2601:9:4980:1843:222:4dff:fe51:6728) has quit (Remote host closed the connection)
[03:11:54] peper03 (peper03!~peper03@mythtv/developer/peper03) has quit (Ping timeout: 265 seconds)
[03:12:11] peper03_ (peper03_!~peper03@mythtv/developer/peper03) has joined #mythtv
[03:12:18] peper03_ is now known as peper03
[03:30:55] arescorpio (arescorpio! has joined #mythtv
[04:46:01] fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has quit (Ping timeout: 264 seconds)
[05:25:37] gigem: stuarta: I haven't looked at the code yet, but if it's the code I'm thinking of, it's the Watch List support. I don't recall the scheduler doing anything with invalid datetimes.
[05:35:07] phuff (phuff!~phuff@unaffiliated/phuff) has joined #mythtv
[05:47:15] phuff1 (phuff1! has joined #mythtv
[05:47:16] phuff (phuff!~phuff@unaffiliated/phuff) has quit (Read error: Connection reset by peer)
[05:51:03] arescorpio (arescorpio! has quit (Excess Flood)
[06:12:53] phuff1 (phuff1! has quit (Ping timeout: 246 seconds)
[06:24:58] robink (robink!~quassel@unaffilated/robink) has quit (Ping timeout: 245 seconds)
[06:26:49] robink (robink!~quassel@unaffilated/robink) has joined #mythtv
[07:16:48] FabriceMG (FabriceMG! has joined #mythtv
[07:21:17] lomion0815 (lomion0815!5bd9374e@gateway/web/freenode/ip. has joined #mythtv
[07:24:31] lomion0815: Good morning! The VideoGallery is not deleting DVD folder structures recursively. Is there any Mythtv code already in place which deletes a whole directory? I could reuse it then for patch to be written!
[07:41:02] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has joined #mythtv
[07:48:45] gary_buhrmaster (gary_buhrmaster!~gtb@2601:9:4980:1843:222:4dff:fe51:6728) has joined #mythtv
[07:54:26] fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has joined #mythtv
[07:55:30] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has quit (Ping timeout: 252 seconds)
[08:09:36] Chutt_ (Chutt_! has quit (Read error: Connection reset by peer)
[08:10:21] Chutt_ (Chutt_! has joined #mythtv
[08:12:03] gary_buhrmaster: gigem: Note that the ceton patch moves to using mythdownloadmanager. As I recall there is a (as yet undetermined/unfixed) leak in mythdownloadmanager in some scenarios. I know I never tested for a long enough time period that I would have seen a leak, nor do I know if the HLS case mentioned for needing the mythsingledownload is special in some way. Just something to be aware of, I guess.
[08:12:35] joki (joki! has quit (Ping timeout: 252 seconds)
[08:13:11] sl1ce_1g (sl1ce_1g! has joined #mythtv
[08:14:47] jya (jya!~jya@mythtv/developer/jya) has quit (Ping timeout: 252 seconds)
[08:16:59] sl1ce (sl1ce! has quit (Ping timeout: 252 seconds)
[08:17:46] joki (joki! has joined #mythtv
[08:17:52] jya (jya!~jya@mythtv/developer/jya) has joined #mythtv
[09:10:55] Merlin83b (Merlin83b! has joined #mythtv
[09:12:26] xris (xris!~xris@mythtv/developer/xris) has quit (Quit: Goodbye.)
[09:14:16] xris (xris! has joined #mythtv
[09:25:55] stuarta: i'm wondering if i should just make a rule, that if i recorded something more than 5 years ago, and haven't watched it yet, i should delete it....
[10:04:05] stuartm: stuarta: . . . 49c57085468d
[10:04:14] stuartm: if anyone can explain that to me ...
[10:05:26] stuarta: well i can tell you what the code does, or you after why it was disabled?
[10:07:22] stuarta: interesting, my master (dev) backend doesn't have that fix, yet it still recorded fine
[10:09:08] stuartm: wonder if it's related to what the driver supports, I recently learned that some of the T2 hardware doesn't support auto-switching from T to T2 when pointed at a T2 signal so the driver needs to be told whether it's T or T2
[10:09:39] stuartm: chances are that I've updated the kernel recently, albeit a minor point release, but the driver could have been affected
[10:10:19] stuartm: stuarta: I know what the code does, I'm not sure why it was disabled nor why it worked before with it disabled
[10:10:59] stuarta: very good questions
[10:11:03] stuarta: -s
[10:12:47] stuartm: if it's the driver then we should hear from other people who are using 0.27 and have been affected
[10:14:10] stuarta: . . . 2b4c3b6ed39c
[10:14:20] stuartm: I don't want to backport to 0.27 until it's been tested by others in master, there may have been a good reason for disabling it, although the define makes me think that something got lost from configure
[10:14:26] stuarta: 5yrs that code has been there
[10:14:58] stuarta: have you made any kernel changes recently?
[10:15:23] stuartm: stuarta: actually, first change says it all – it's failing to detect it in configure
[10:15:45] stuarta: wonder why
[10:17:27] stuartm: dekarl: you changed the configure check 3 days ago
[10:17:33] stuartm: and it no longer works
[10:18:15] stuartm: I searched configure for HAVE_FE_CAN_2G_MODULATION, forgot to drop the HAVE_
[10:18:23] stuartm: which is why I missed it until now
[10:18:28] stuarta: doh
[10:19:04] stuartm:
[10:19:57] stuartm: note the removal of "return 1"
[10:21:09] stuartm: fwiw, I think all kernels should have support by now so we could perhaps drop the check and raise the minimum DVB API version instead
[10:21:13] dekarl: stuartm, meh yes just saw your commit
[10:22:52] dekarl: although I'm not sure what the return code is about
[10:23:05] dekarl: I did not find a spot where *our* configure even checks for the return code
[10:23:27] stuartm: #if (LIBXML_VERSION > 20600) #error libxml2 version to old
[10:23:33] stuartm: looks suspicious
[10:23:50] stuarta: ah, now i understand why my dev backend didn't break, i didn't rerun configure
[10:24:33] stuartm: whereas I had to re-run it to build against QT5
[10:24:47] dekarl: the checks are basic "does it compile without error or not?" checks
[10:24:50] gary_buhrmaster: stuartm: Well, at least you no longer need to be baffled regarding the why. There is that.
[10:24:51] stuarta: so potentially it wasn't qt5 in this case
[10:25:18] stuarta: gary_buhrmaster: you are up early today!
[10:26:13] dekarl: can I see config.ep around the check when it works/fails?
[10:26:23] stuartm: gary_buhrmaster: indeed
[10:26:26] gary_buhrmaster: stuarta: up would imply down. I have not quite managed that yet.
[10:26:34] stuarta: hahaha
[10:26:59] dekarl: ohh, its an enum, not a preprocessor define, so that's why
[10:27:48] dekarl: for libxml2 its a define. I can fix the test
[10:28:06] stuarta: teamwork++
[10:28:20] stuartm: dekarl:
[10:29:07] dekarl:
[10:29:53] dekarl: stuartm, looks good. so that is that
[10:30:50] stuartm: dekarl: might have noticed a problem if the warning appeared in the configure output
[10:31:27] stuartm: as it was, there's the "DVB-S2 support no" buried in the middle there
[10:32:39] stuartm: stuarta: still need to fix recording schedule creation under QT5
[10:34:30] stuarta: stuartm: is that frontend or backend do you think?
[10:35:57] stuartm: dekarl:
[10:37:01] stuartm: stuarta: frontend, it's a change in behaviour around QDateTime – previously if you specified a datetime of 00-00–00T00:00:00 it was considered !valid but !null
[10:37:16] stuartm: in QT5 it's consider to be a NULL datetime
[10:37:26] stuarta: i'll spin up a qt5 based instance
[10:38:23] stuartm: so when we try to insert 00-00–00T00:00:00 into the database to represent an unknown date, the QT mysql driver interprets it as NULL and errors because the column is NOT NULL
[10:38:26] dekarl: stuartm, ty its the same over here. strange
[10:39:54] dekarl: so this doesn't work for your case? . . . data.cpp#L31
[10:41:46] stuartm: dekarl: it's a bit flag so #if (!(FE_CAN_2G_MODULATION & 0x10000000))
[10:43:49] dekarl: hmm, we are comparing if the mask is the expected mask :)
[10:44:11] dekarl: but its not finding the enum :/
[10:45:08] dekarl: need to figure how to test if FE_CAN_2G_MODULATION exists *in the preprocessor*
[10:45:38] stuartm: I'm starting to think that configure check was always broken
[10:45:46] dekarl: I'll back the change out.
[10:46:08] dekarl: The test worked in the sense of "does the compiler recognize FE_CAN_2G_MODULATION or fail?"
[10:46:09] stuartm: dekarl: right of course
[10:46:43] dekarl: so it did something completely different then intended, but it did actually lead to the expected result
[10:48:12] stuartm: dekarl: in which case the original check should have worked, ifndef (FE_CAN_2G_MODULATION)
[10:48:44] stuartm: which is why I'm wondering whether it ever really worked
[10:49:56] stuartm: if FE_CAN_2G_MODULATION was defined it should have been enabled the DVB-S2 code, irrespective of whether the define contained the correct value – we know the define exists in frontend.h
[10:50:40] gary_buhrmaster: My recollection is FE_CAN_2G_MODULATION was added in/about DVB version 5.0. So a simpler check of #if DVB_API_VERSION > 5 || DVB_API_VERSION == 5 && DVB_API_VERSION_MINOR >= 0 (all defines)
[10:51:16] gary_buhrmaster: might be easier. Would have to double check the DVB_API_VERSION number.
[10:51:31] stuartm: gary_buhrmaster: yep, that's what I suggested earlier, I'm not sure why we're checking for the define
[10:51:46] stuarta: i like a cunning plan
[10:53:07] jheizer_ (jheizer_!~jheizer@2601:d:3a00:1d58:38c3:db48:9535:cdb4) has quit (Read error: Connection reset by peer)
[10:53:15] Roklobsta (Roklobsta! has joined #mythtv
[10:53:37] gary_buhrmaster: stuartm: Especially since (in theory) the actual value of the enum might change. I dislike future gotchas.
[10:54:12] stuarta: in practice is shouldn't because that is really bad for ABI compatibility
[10:54:47] lomion0815: Hi, did you read post this morning? The VideoGallery is not deleting DVD folder structures recursively. Is there any Mythtv code already in place which deletes a whole directory? I could reuse it then for patch to be written!
[10:55:34] gary_buhrmaster: stuartm: especially in "our" configure code (which has confounded me with its results more times that I like to admit).
[10:56:28] dekarl: gary_buhrmaster: I just test for FE_CAN_2G_MODULATION != 0 now
[10:56:58] jheizer (jheizer!~jheizer@2601:d:3a00:1d58:38c3:db48:9535:cdb4) has joined #mythtv
[10:58:36] gary_buhrmaster: stuarta: I agree, but never say never. And while many/most are more careful now, I do have the burn marks from ABI breakage in the kernel.
[11:01:28] stuartm: lomion0815: there probably is, but I can't think of where it might be atm
[11:03:05] stuartm: dekarl: now the libxml test
[11:03:18] stuartm: where are we using libxml? qt has xml parsing built in
[11:04:25] dekarl: stuartm, the xml test should be ok. its a preprocessor define
[11:04:42] stuartm: dekarl: yes, but it's backwards
[11:04:59] stuartm: if VERSION > 1234 then ERROR – TOO OLD
[11:06:18] stuartm: the end result might be the desired outcome, but it seems all wrong
[11:06:39] stuartm: unless 20600 is a magic number
[11:06:42] dekarl: bah, it was me wrote turned it around from a noop to the wrong way around
[11:07:05] dekarl: I hate magic numbers (even if its 0 and 1)
[11:10:25] dekarl: that will teach me to stay away from drive by fixes for some hours...
[11:11:36] ** stuarta offers coffee **
[11:14:28] stuartm: ok, libxml2 is a dependency of libbluray
[11:20:51] stuarta: might need that ;-)
[11:25:32] dekarl1 (dekarl1!~dekarl@mythtv/developer/dekarl) has joined #mythtv
[11:26:07] stuartm: yip
[11:26:46] stuartm: it's an optional dep, but some bluray metadata is in xml format, without libxml it can't be read
[11:28:00] stuartm: currently installed version of libxml2 is 2.9.1
[11:28:36] dekarl (dekarl!~dekarl@mythtv/developer/dekarl) has quit (Ping timeout: 246 seconds)
[11:28:56] lomion0815 (lomion0815!5bd9374e@gateway/web/freenode/ip. has quit (Quit: Page closed)
[12:00:52] Roklobsta: i deleted a tonne (imperial) of recordings this evening, it took quite a while to free up the space. Why does MBE take to long to delete recordings?
[12:00:53] stuarta: gary_buhrmaster: i checked tunnel offer a couple of months ago, they are blocking outbound smtp which is unfortunate, but your message makes me think it's worth chasing osuosl again
[12:02:01] Roklobsta: is there a process that takes queued up recordings-to-be-deleted and processes them once every 10 secs?
[12:02:14] stuarta: yes, housekeeper iirc
[12:02:36] stuarta: so it takes even longer if you set keep deleted recordings to 7 days
[12:02:40] stuartm: Roklobsta: by default recordings aren't deleted immediately, this gives you a chance to change your mind and undelete them
[12:02:52] Roklobsta: oh
[12:03:08] stuartm: there's a 10–15 minute delay
[12:03:41] stuartm: like the recycle bin on your desktop
[12:03:46] Roklobsta: problem is the recoding pool was 100% full and we needed to make space for new recordings Right Now.
[12:04:08] stuartm: Roklobsta: it will delete them immediately if the space is required by the recorder
[12:04:13] stuarta: you can change to the deleted recordings gropu and delete them
[12:04:15] Roklobsta: ok
[12:04:33] Roklobsta: i understand.
[12:05:04] stuartm: also check that you haven't enabled the "Slow Deletes" option, that exists because some filesystems can't delete very large files without locking up
[12:05:13] Roklobsta: 550GB free now
[12:05:25] stuarta: that's quite a bit
[12:05:25] Roklobsta: yep, mfe or mbe?
[12:05:29] stuartm: so instead of deleting the entire file in one go, it deletes it in chunks
[12:05:48] Roklobsta: we record a lot of crap, never watch live tv
[12:05:51] stuartm: backend
[12:06:40] Roklobsta: mythweb is brialliant for finding programs i might like to see, especially one offs or small series.
[12:10:51] dekarl1: Roklobsta: . . . _source.html may do what you want. Not sure though
[12:10:54] dekarl1 is now known as dekarl
[12:13:51] dekarl: also might come in handy... --olderthan=<last year> --not-set-to-keep
[12:26:05] _charly___ is now known as _charly_
[12:48:00] SteveGoodey (SteveGoodey! has joined #mythtv
[13:16:24] moparisthebest (moparisthebest!~mitb@unaffiliated/moparisthebest) has quit (Ping timeout: 264 seconds)
[13:20:02] Jordack (Jordack! has joined #mythtv
[13:25:47] moparisthebest (moparisthebest!~mitb@gateway/tor-sasl/moparisthebest) has joined #mythtv
[13:31:49] SteveGoodey (SteveGoodey! has quit (Quit: Konversation terminated!)
[13:31:49] Tobbe5178 (Tobbe5178!~asdf@2001:2002:d9d4:ce7e:1c2b:2d14:5c8a:a6b8) has quit (Read error: Connection reset by peer)
[13:48:47] stuarta: stuartm: just reading through . . . d4c5bc205a2b AFAIR there are QHostAddress functionality to handle all the variations wrt ipv4 & ipv6 handling
[13:50:42] stuartm: stuarta: yes, but only for ip addresses
[13:51:07] stuartm: it doesn't understand hostnames AFAIK
[13:51:11] stuarta: nothing for urls?
[13:51:16] ** stuarta wonders about QUrl **
[13:52:14] stuarta: think QUrl does what we need
[13:52:29] stuarta:
[13:52:34] stuartm: yeah I think you're right
[13:53:03] stuarta:
[13:55:48] SteveGoodey (SteveGoodey! has joined #mythtv
[13:56:33] stuarta: grrr. /me adds packaging to cgit
[14:05:31] stuarta: why is this so difficult....
[14:07:43] moparisthebest (moparisthebest!~mitb@gateway/tor-sasl/moparisthebest) has quit (*.net *.split)
[14:12:39] Jordack (Jordack! has quit ()
[14:16:57] SteveGoodey (SteveGoodey! has quit (Quit: Konversation terminated!)
[14:17:43] moparisthebest (moparisthebest!~mitb@unaffiliated/moparisthebest) has joined #mythtv
[14:37:24] dblain (dblain!~dblain@mythtv/developer/dblain) has quit (Ping timeout: 264 seconds)
[14:47:50] phuff (phuff!~phuff@unaffiliated/phuff) has joined #mythtv
[14:49:16] SteveGoodey (SteveGoodey! has joined #mythtv
[15:14:27] phuff (phuff!~phuff@unaffiliated/phuff) has quit (Ping timeout: 245 seconds)
[15:42:50] stuartm: heh, RPi 2 reboots itself if exposed to bright light
[15:44:07] jheizer: Saw that. Pretty damn funny.
[15:45:18] dekarl: its a feature! on sunny days, go out and play :)
[15:46:20] stuarta: wtf....
[15:46:58] stuartm: for those that are interested –
[15:47:17] jheizer: Same time I can't be too mean. My very first gen rpi is probably the most stable server/computer in my house.
[16:08:09] Roklobotomy (Roklobotomy! has joined #mythtv
[16:10:19] Roklobsta (Roklobsta! has quit (Ping timeout: 255 seconds)
[16:13:55] dblain (dblain!~dblain@ has joined #mythtv
[16:13:55] dblain (dblain!~dblain@mythtv/developer/dblain) has joined #mythtv
[16:13:55] dblain (dblain!~dblain@ has quit (Changing host)
[16:22:06] stuartm: wonder if anyone makes racks for RPis
[16:24:03] stuartm: something up from
[16:25:15] Merlin83b (Merlin83b! has quit (Quit: Leaving)
[16:26:27] stuartm: crap, CSP nonces can't be used with inline event handlers
[16:26:38] stuartm: well that was a waste of time :(
[16:35:10] gary_buhrmaster: stuartm: re: racks of pi. There is raspberrycolocation which has a custom solution for high(?) density racks. Some info at:
[16:35:25] gary_buhrmaster: stuartm: Do you have designs on a 500 RPi rack for your basement?
[16:36:02] jheizer: Few other cool simple ones too like . . . PiTower1.JPG
[16:36:56] jheizer: or if you like legos there are a few designs out there
[16:38:49] FabriceMG (FabriceMG! has quit (Quit: FabriceMG)
[16:42:47] stuartm: gary_buhrmaster: not quite that many, but I am thinking about something neat and scalable – perhaps no more than 4–5 total
[16:43:50] jab416171 (jab416171! has quit (Max SendQ exceeded)
[16:44:20] jab416171 (jab416171! has joined #mythtv
[17:27:19] superm1 (superm1!uid4318@ubuntu/member/superm1) has joined #mythtv
[17:47:39] buu: Why would you want multiple rpis =|
[18:01:30] stuartm: buu: well they aren't very powerful, and frankly I wouldn't want to try running multiple vms on one, easier to have one dedicated to a single task
[18:04:46] buu: stuartm: Why not just have a single $100 64bit machine =/
[18:05:56] stuartm: a $100 machine is going to be considerably larger than the footprint of even 5 Pis
[18:06:33] stuartm: you can get more powerful small form factor devices, but they tend to be a lot more expensive
[18:07:50] stuartm: and besides, what I'm mostly envisioning them being used for it building and testing arm builds, can do that so easily or reliably on a non-arm device
[18:08:02] stuartm: the build part yes, the testing, no
[18:09:32] gigem: gary_buhrmaster: The Ceton patch fails when built with Qt4. I'm going to investigate that before trying with Qt5.
[18:09:56] sheedy-away is now known as sheedy
[18:17:52] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has joined #mythtv
[18:38:04] Jordack (Jordack! has joined #mythtv
[18:47:48] stuartm: gigem: did you apply the pre-requisite patch mentioned in the ceton ticket?
[18:49:16] clever: i could see some specialized highly parallel jobs running on the 3d core, its basicaly a 200mhz 192 core system
[18:49:51] clever: and that would spread across multiple systems easily, since your task already has to be highly parallel, but would a desktop GPU beat it in TFLOP per watt?
[18:53:34] gigem: gary_buhrmaster: No, I read your comment from yesterday as ignore that patch altogether. I take it you only meant the ABI bump part. I'll retry in a while.
[18:57:17] esperegu (esperegu! has joined #mythtv
[18:59:50] esperegu: Hi. I suppose this is for this channel since I heard this was being worked on in 0.28. I have a question regarding input connections. I thought that normally I could add a second connection for a card since it uses a unicable switch with 2 lnb positions but now I only see one option per card. Has this been changed recently and if so how this should now be done?
[19:09:30] gary_buhrmaster: gigem: Sorry, right. I meant no (independent) ABI bump required, but the referenced patch itself is required (it adds the required functionality that the ceton patch uses).
[19:11:31] gary_buhrmaster: gigem: None of my patches actually include ABI bump code, even when required, since they may live in trac for an extended period before review. However, they all include comments to the committer to bump the ABI. I had thought (before I checked) that the patch to mythdownloadmanager needed a ABI bump. It does not.
[19:12:33] gary_buhrmaster: gigem: Another of my mythdownloadmanager patches did require a ABI bump, but I never completed (adequate to me) testing, so it is still stashed somewhere and never submitted. I was confused. And apparently passed on the confusion. Sorry.
[19:14:29] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has quit (Ping timeout: 246 seconds)
[19:14:58] gary_buhrmaster: gigem: I mentioned it (the fantasy ABI bump) because as a comment in my (often long) ticket documentation, the bump has occasionally been missed. I was getting ahead of myself. Sorry.
[19:19:24] gary_buhrmaster: gigem: However, to be fair, I have not tried to compile MythTV on Qt4 with those patches recently. It is entirely possible that some bitrot has occurred. If it fails to compile, please paste the output and I will take a look.
[19:19:51] dekarl: esperegu: are you talking about the DiSEqC setup? Every DVB-S frontend has one input, then you start building your DiSEqC tree from that one input.
[19:20:18] esperegu: dekarl: I have that unicable stuff.
[19:20:46] esperegu: dekarl: it shows only 4 entries on the input connections page (one for each card)
[19:20:47] dekarl: esperegu: isn't that DiSEqC with a a shared wire?
[19:20:52] esperegu: dekarl: yes
[19:21:23] dekarl: esperegu: and that input is connected to a universal lnb?
[19:21:28] esperegu: dekarl: when I select the entries on teh input connections page I can select the A/B for the LNB but it still only gives me one entry
[19:22:10] esperegu: the cards are connected to unicable and unicable has one lnb under it
[19:22:41] esperegu: then on connection page I see the 4 cards and they all connected to #1
[19:22:41] ** MythLogBot **
[19:22:58] esperegu: but I cannot add more
[19:23:34] esperegu: (I thought that was how it was. I think some of the recent updates changed the way it is handled)
[19:26:36] esperegu: (I use mythbuntu ppa's)
[19:38:21] esperegu: dekarl: maybe it works differently now, maybe it only works on a clean install or maybe the changes were not made properly for unicable or maybe I am just to tired ;-)
[19:39:00] dekarl: i've never seen dvb-s in action, so I can only guess :/
[19:46:46] gigem: esperegu: I made the changes that could very well have broken what you're tyring. When I asked for testers, no one had any idea how such cards were supposed to work, so I guessed and had no way to test it. If can explain in layman's terms how it is supposed to work, I'll try to fix it. Also, in case you don't know yet, cards can no longer have multiple inputs. If your card really does have some form of
[19:46:48] gigem: multiple inputs, you need to configure multiple cards using the save device and then configure each to use the appropriate input. Perhaps that something you need to do in your situation.
[19:52:02] esperegu: gigem: where do I specify that 'save device' ?
[19:53:17] gigem: esperegu: Sorry, fat fingers. That should be 'same device'.
[19:53:57] gary_buhrmaster: stuartm: btw, for your rack mounted (datacenter) RPi2's, it is not uncommon for commercial buildings (at least in the US) to have xenon flashing fire alarms. So, if you have a fire alarm (real or test), all your RPi2s reset.......
[19:55:25] esperegu: gigem: will mythtv still be able to know it cannot use those simultaniously?
[19:55:35] gigem: esperegu: Yes.
[19:55:46] gary_buhrmaster: stuartm: To be fair, the usual solution (of all imagined problems) of duct tape can fix it. Electrical tape for those of a better persuasion (although, as been shown, you can make a bridge with duct tape; not so much electrical tape)
[19:56:01] gigem: gary_buhrmaster: Still fails with http status 500. It might be this evening before I can dig any deeper.
[19:56:17] esperegu: gigem: k. will try that. guess I need to remove everything first then and recreate it all.. thx
[19:56:46] rich0 (rich0!~quassel@gentoo/developer/rich0) has quit (Remote host closed the connection)
[19:57:41] gigem: esperegu: You're fee to, but you shouldn't need to delete anything. The update is supposed to keep working configurations working. If it doesn't, it's a bug.
[19:57:53] gigem: esperegu: s/fee/free/
[19:58:06] gary_buhrmaster: gigem: If you can get a (raw) tcpdump, I might be able to figure it out, but since I do not have a ceton device, I will probably need to see a working (and non-working) tcpdump.
[20:02:31] gigem: I'll see what I can do.
[20:03:26] gary_buhrmaster: gigem: I am actually going to be otherwise occupied until later this evening, so I, too, might not be able to get to it until later.
[20:04:45] rich0 (rich0!~quassel@gentoo/developer/rich0) has joined #mythtv
[20:05:12] stuartm: xenon flashing alarms seem like the worst idea ever, trying to escape a burning building in the dark? Let's help you out by searing your retinas leaving you still trapped in the building, but now also blind
[20:07:34] esperegu: gigem: when I add it twice it does not show it twice...
[20:08:34] esperegu: gigem: however I see 4 entries in the database table capturecard
[20:12:07] dekarl: stuartm, in the webfrontend upcoming recordings, do you have sensible values for Rule Type? I get the translated string for ruletype=0
[20:12:42] stuartm: dekarl: yeah, I broke that a while back, keep putting off working on the fix
[20:12:48] dekarl: in the recording rule list I get the correct text, but its untranslated
[20:13:28] dekarl: stuartm, ok. not a priority, just been wondering about that (is it just me?)
[20:14:19] sheedy is now known as sheedy-away
[20:21:46] gigem: esperegu: Okay. I think I know why. Fixing it will require a change I'd hoped to not need until later. I'll try to do it tonight.
[20:24:49] rich0 (rich0!~quassel@gentoo/developer/rich0) has quit (Ping timeout: 252 seconds)
[20:26:18] stuartm: dekarl: one services class RecordingInfo uses the enum value, the RecRule class uses the String value, and so the code to translate it doesn't like getting feed an integer (enum) when it expects a string
[20:28:49] stuartm: proper fix is to convert over all the other services endpoints to return the enum values, but I'd prefer to do that properly with the enum support stuff in QT5 – hence why I'm pushing QT5 right now :)
[20:29:34] esperegu: gigem: k. thx
[20:29:48] rich0 (rich0!~quassel@gentoo/developer/rich0) has joined #mythtv
[20:38:21] Roklobotomy (Roklobotomy! has quit (Quit: KVIrc 4.2.0 Equilibrium
[20:51:58] Steve-Goodey (Steve-Goodey! has joined #mythtv
[20:52:41] andreaz (andreaz! has joined #mythtv
[20:58:50] stuartm: dekarl: so thinking about it I came up with a fix that is both a hack and yet elegantly simple
[21:04:58] jheizer (jheizer!~jheizer@2601:d:3a00:1d58:38c3:db48:9535:cdb4) has quit (Ping timeout: 245 seconds)
[21:06:25] jheizer (jheizer!~jheizer@2601:d:3a00:1d58:c5ae:1b5e:4b29:b55a) has joined #mythtv
[21:15:18] stuartm: dekarl: can you confirm that this patch fixes the translation of those strings –
[21:33:39] Roklobsta (Roklobsta! has joined #mythtv
[21:44:35] SteveGoodey (SteveGoodey! has quit (Quit: Konversation terminated!)
[22:00:14] Steve-Goodey (Steve-Goodey! has quit (Quit: Konversation terminated!)
[22:02:05] dekarl: stuartm, nice patch, will try it
[22:03:04] stuartm: some of the strings come from the backend itself, and we don't currently load the translations for the backend, hence untranslated strings
[22:04:38] stuartm: I need to experiment with extracting strings from qsp files, in theory I believe lupdate should just work once it's told the qsp files contain translatable strings
[22:05:19] dekarl: etiher that or a trick like the one for the themestrings
[22:06:29] stuartm: that technically shouldn't be necessary since we're using the qscript translation stuff, but we'll see how it goes
[22:06:42] stuartm: still need to fix the category colour issue
[22:07:21] stuartm: and recording status translation bug
[22:08:28] dekarl: I think the category colour issue is a bigger project
[22:09:08] dekarl: judging from the lists of categories that jny len and myself maintain for our guide data its huge :)
[22:15:32] stuartm: well to start with I'd just substitute the current english named css classes with around 30 or so named "Category1", "Category2" etc, then create a mapping system of "local category names > Category1"
[22:16:13] stuartm: that way the translators aren't picking the colours, they simply decide which categories to group together
[22:19:56] stuartm: I'd use the exact same system for the frontend guide, so the two would share the same mappings to keep the work down
[22:21:00] Jordack (Jordack! has quit ()
[22:22:08] dekarl: also the consistency might be appreciated
[22:22:14] stuartm: throw in an algorithm to automatically assign unmapped categories to different groups ...
[22:26:11] dekarl: basically the 5 bit of some kind of hash for 32 categories?
[22:30:31] amessina (amessina!~amessina@unaffiliated/amessina) has joined #mythtv
[22:31:49] dekarl: tgm4883: should the packaging depend on dcraw? "Dcraw support will not be included in MythGallery" from . . . ADING.txt.gz
[22:32:44] _charly__ (_charly__! has joined #mythtv
[22:33:49] _charly_ (_charly_! has quit (Ping timeout: 250 seconds)
[22:35:01] tgm4883: dekarl: there should be a build-dep on it
[22:35:06] tgm4883: probably a dep as well
[22:36:28] tgm4883: dekarl: I also wonder about other things, such as dvb-s2 support
[22:36:49] tgm4883: or libxml2, or ASI support
[22:38:09] dekarl: IIUIC dvb-s2 support is in the kernel depending on kernel version
[22:42:24] Roklobsta (Roklobsta! has quit (Quit: KVIrc 4.2.0 Equilibrium
[22:43:32] dekarl: stuartm, latest + loading the translation => (SendResponse) HTTPRequest::SendResponse(xml/html) () :404 Not Found
[22:43:59] dekarl: try to access the root: (ExtractMethodFromURL) ExtractMethodFromURL(end) :  : /
[22:44:49] stuartm: just loading the translation breaks it?
[22:45:22] stuartm: that seems ... improbable
[22:48:20] stuartm: dekarl: check that backend_index.qsp has been installed
[22:48:57] dekarl: It appears as if backend_index.qsp is missing, sounds like that might be it (beat me to it)
[22:50:17] stuartm: it's in the repo, backend_index.html was renamed
[22:50:26] dekarl: ahh ok, I'll fix it
[22:52:02] dekarl: that build step might as well fail if it doesn't find a file :/
[22:53:39] superm1 (superm1!uid4318@ubuntu/member/superm1) has quit (Ping timeout: 265 seconds)
[22:54:10] dekarl: now I forgot why I built in the first place ;)
[22:54:11] dekarl: I now see the correct and translated Rule Type
[22:57:16] andreaz (andreaz! has quit (Read error: Connection reset by peer)
[22:57:40] superm1 (superm1!uid4318@ubuntu/member/superm1) has joined #mythtv
[22:59:17] len (len! has joined #mythtv
[22:59:24] len__ (len__! has joined #mythtv
[23:09:56] arescorpio (arescorpio! has joined #mythtv
[23:11:54] superm1 (superm1!uid4318@ubuntu/member/superm1) has quit (Read error: Connection reset by peer)
[23:37:55] gigem: gary_buhrmaster: I just sent you an email with some pcaps. It looks to me like Qt and consequently MythDownloadManager don't like the "302 Found" replies the Ceton gives for POSTs.
[23:43:32] superm1 (superm1!uid4318@ubuntu/member/superm1) has joined #mythtv
[23:58:25] joki (joki! has quit (Ping timeout: 256 seconds)
[23:58:37] Hydr0p0nX (Hydr0p0nX!~hydr@ has quit (Ping timeout: 245 seconds)
[23:59:10] Hydr0p0nX (Hydr0p0nX!~hydr@ has joined #mythtv

IRC Logs collected by BeirdoBot.
Please use the above link to report any bugs.