| Monday, December 3rd, 2018, 00:21 UTC | ||
| [00:21:23] | Hydr0p0nX (Hydr0p0nX!~Hydr0p0nx@66-191-154-200.dhcp.mtgm.al.charter.com) has quit (Read error: Connection reset by peer) | |
| [00:22:13] | Hydr0p0nx (Hydr0p0nx!~Hydr0p0nx@66-191-154-200.dhcp.mtgm.al.charter.com) has joined #mythtv | |
| [00:35:48] | nephyrin (nephyrin!~neph@2601:600:817f:f46a:e41d:6f5c:e3be:b8ad) has quit (Read error: Connection reset by peer) | |
| [00:41:56] | nephyrin (nephyrin!~neph@2601:600:817f:f46a:e41d:6f5c:e3be:b8ad) has joined #mythtv | |
| [01:36:06] | amessina (amessina!~amessina@unaffiliated/amessina) has quit (Remote host closed the connection) | |
| [01:36:10] | jheizer (jheizer!~jon@c-73-51-93-177.hsd1.il.comcast.net) has quit (Ping timeout: 250 seconds) | |
| [01:36:33] | amessina (amessina!~amessina@unaffiliated/amessina) has joined #mythtv | |
| [01:45:13] | jheizer (jheizer!~jon@c-73-51-93-177.hsd1.il.comcast.net) has joined #mythtv | |
| [01:45:13] | Mode for #mythtv by ChanServ!ChanServ@services. : +v jheizer | |
| [01:57:32] | blahdodo (blahdodo!~blahdodo@69.172.190.157) has quit (Quit: Bye bye) | |
| [01:57:52] | blahdodo (blahdodo!~blahdodo@69.172.190.157) has joined #mythtv | |
| [02:20:34] | amessina (amessina!~amessina@unaffiliated/amessina) has quit (Remote host closed the connection) | |
| [02:21:03] | amessina (amessina!~amessina@unaffiliated/amessina) has joined #mythtv | |
| [05:23:27] | brfranse- (brfranse-!~brfransen@47.42.27.251) has quit (Ping timeout: 240 seconds) | |
| [05:33:19] | amessina (amessina!~amessina@unaffiliated/amessina) has quit (Quit: Konversation terminated!) | |
| [06:12:21] | brfransen (brfransen!~brfransen@47.42.27.251) has joined #mythtv | |
| [07:09:10] | Steve-Goodey (Steve-Goodey!~steve@2a00:23c5:7da3:4501:7a84:3cff:fedf:a99) has joined #mythtv | |
| [07:09:11] | Mode for #mythtv by ChanServ!ChanServ@services. : +v Steve-Goodey | |
| [08:57:33] | willcooke (willcooke!~willcooke@ubuntu/member/willcooke) has joined #mythtv | |
| [09:00:10] | SteveGoodey (SteveGoodey!~steve@2a00:23c5:7da3:4501:7a24:afff:fe9d:c233) has joined #mythtv | |
| [09:00:10] | Mode for #mythtv by ChanServ!ChanServ@services. : +v SteveGoodey | |
| [09:00:23] | Warped_ (Warped_!~Warped@2600:1702:b40:851f:2972:cf7d:334c:7b62) has joined #mythtv | |
| [09:00:45] | Warped_ (Warped_!~Warped@2600:1702:b40:851f:2972:cf7d:334c:7b62) has quit (Client Quit) | |
| [09:01:13] | zbot (zbot!~supybot@c-67-161-9-155.hsd1.ca.comcast.net) has quit (Ping timeout: 250 seconds) | |
| [09:01:50] | gary_buhrmaster (gary_buhrmaster!~gtb@2601:647:4802:4c3:222:4dff:fe51:6728) has quit (Ping timeout: 268 seconds) | |
| [09:01:50] | Warped_ (Warped_!~Warped@2600:1702:b40:851f:2972:cf7d:334c:7b62) has joined #mythtv | |
| [09:03:04] | Warped (Warped!~Warped@unaffiliated/warped) has quit (Ping timeout: 268 seconds) | |
| [09:05:59] | Warped_ (Warped_!~Warped@2600:1702:b40:851f:2972:cf7d:334c:7b62) has quit (Client Quit) | |
| [09:06:08] | Merlin83b (Merlin83b!~Daniel@office.34sp.com) has joined #mythtv | |
| [09:07:41] | Warped (Warped!~Warped@unaffiliated/warped) has joined #mythtv | |
| [09:16:37] | Steve-Goodey (Steve-Goodey!~steve@2a00:23c5:7da3:4501:7a84:3cff:fedf:a99) has quit (Quit: Konversation terminated!) | |
| [09:25:23] | zbot (zbot!~supybot@c-67-161-9-155.hsd1.ca.comcast.net) has joined #mythtv | |
| [09:31:13] | gary_buhrmaster (gary_buhrmaster!~gtb@c-67-161-9-155.hsd1.ca.comcast.net) has joined #mythtv | |
| [10:45:31] | tris (tris!tristan@camel.ethereal.net) has quit (Ping timeout: 260 seconds) | |
| [10:48:19] | tris (tris!tristan@camel.ethereal.net) has joined #mythtv | |
| [10:58:17] | frobnic (frobnic!fb@b2b-130-180-11-14.unitymedia.biz) has quit (Read error: Connection reset by peer) | |
| [11:03:01] | frobnic (frobnic!fb@b2b-130-180-11-14.unitymedia.biz) has joined #mythtv | |
| [12:03:37] | Steve-Goodey (Steve-Goodey!~steve@2a00:23c5:7da3:4501:7a84:3cff:fedf:a99) has joined #mythtv | |
| [12:03:37] | Mode for #mythtv by ChanServ!ChanServ@services. : +v Steve-Goodey | |
| [15:53:18] | ghoti (ghoti!ghoti@dev-160.experiencepoint.com) has quit (Read error: Connection reset by peer) | |
| [15:53:51] | ghoti (ghoti!~paul@dev-160.experiencepoint.com) has joined #mythtv | |
| [15:56:20] | aloril_ (aloril_!~aloril@mobile-access-5d6a9e-141.dhcp.inet.fi) has quit (Ping timeout: 250 seconds) | |
| [15:57:12] | gregl (gregl!~greg@cpe-24-194-253-7.nycap.res.rr.com) has quit (Ping timeout: 250 seconds) | |
| [15:57:48] | aloril_ (aloril_!~aloril@mobile-access-5d6a9e-141.dhcp.inet.fi) has joined #mythtv | |
| [15:57:48] | gregl (gregl!~greg@cpe-24-194-253-7.nycap.res.rr.com) has joined #mythtv | |
| [16:48:03] | hampton: | I'm stumped. If I explicitly force the cache control attribute to PreferNetwork just before calling QNetworkAccessManager::get, Qt will still return the cached version of the themes file. If I force it to AlwaysNetwork then it returns the new version from themes.mythtv.org. |
| [16:48:10] | hampton: | Maybe because there's no "Expires" header on the downloaded file? |
| [16:51:25] | hampton: | The date on the cached file is Fri, 30 Nov 2018 23:16:20 GMT. |
| [17:02:04] | stuarta: | is there an e-tag on the response? |
| [17:03:05] | stuarta: | hampton: more to the point, can you capture the conversation with tcpdump, so we can analyze the request & response |
| [17:06:08] | stuarta: | but *please* write it to a file so we can look at it more easily later |
| [17:06:52] | stuarta: | tcpdump -w <filename> |
| [17:07:04] | stuarta: | and the relevant filter |
| [17:08:06] | hampton: | I'll do that this afternoon. I'm filtering on anything to themes.mythtv.org. |
| [17:09:34] | hampton: | The headers I see are Date, Server, Last-Modified, Accept-Ranges, Content-Length, Vary(?), Connection, and Content-Type. |
| [17:10:26] | hampton: | Saving my current trace. |
| [17:12:41] | ** stuarta loves wireshark ** | |
| [17:13:50] | hampton: | I used it a lot back in the day. I know I wrote a plugin for it, but I can't remember what protocol now. |
| [17:14:07] | stuarta: | :) |
| [17:15:20] | hampton: | https://pastebin.com/GK2dUaFK |
| [17:20:10] | stuarta: | hmmm, doesn't shed much light. pcap be better |
| [17:21:52] | hampton: | That's a pcapng file. Need to save it as raw. |
| [17:22:57] | hampton: | That was a trace where I forced AlwaysNetwork so I could see the headers on the returned themes.zip file. The reassembled file is in frame 132. |
| [17:23:31] | stuarta: | tbh, it's the headers (req & resp) i'm more interested in |
| [17:23:43] | stuarta: | like, does it do an IMS request? |
| [17:27:43] | hampton: | IMS request? The GET is in frame 4. |
| [17:32:21] | stuarta: | which has no IMS header |
| [17:32:45] | stuarta: | and the response header has no e-tag |
| [17:36:20] | gigem: | stuarta, peterbennett: I reopened #12902 which you'd both commented on. It's the single biggest issue that has caused my MAF (mom acceptance factor) to plummet after I put Shields on all of her TVs and she started using MythTV for live TV too. |
| [17:36:20] | ** MythLogBot https://code.mythtv.org/trac/ticket/12902 ** | |
| [17:40:00] | hampton: | Agreed, no IMS or ETag. Why no Expires header on the response? |
| [17:40:08] | peterbennett: | gigem: This is a vaapi problem, so how does it affect the shield? |
| [17:41:21] | peterbennett: | gigem: It looks like that ticket strayed from its original subject? |
| [17:41:31] | hampton: | I'm pulling the qt source so I can go look see how the qnetworkdiskcache decides whether to return the cached version or request a new version from the network. Its clearly doing the wrong thing in this case. |
| [17:46:25] | gigem: | peterbennett: It's not a vaapi problem. It's a problem where the decoder thread detects a speed change and tries to change the deinterlacer in the opengl video output thread. It can happen with anytime opengl is used for video rendering. |
| [17:47:24] | peterbennett: | gigem: So it is with using Standard decode for the mpeg2 because of the nvidia issue? |
| [17:47:37] | gigem: | Yes. |
| [17:48:14] | peterbennett: | gigem: Can you get a sample video extract? |
| [17:48:27] | peterbennett: | I have to go now |
| [17:48:29] | gigem: | Because of the decode corruption on one channel and the ff/rew pauses, I have her using ffmpeg decoding. I use it most of the time myself right now. |
| [17:48:37] | ** peterbennett leaves ** | |
| [17:49:31] | stuarta: | hampton: expires is good for caching |
| [17:51:21] | stuarta: | ok, interesting, we aren't returning any cache-control headers |
| [17:52:12] | stuarta: | yet, the request has cache-control: none |
| [17:52:21] | stuarta: | s/none/no-cache |
| [17:53:06] | stuarta: | i wonder if a `Cache-Control: must-revalidate` is required by Qt |
| [17:53:08] | gigem: | peterbennett: You went through that all with the original reporter. It's not a problem specific to any video sample. It's decoder thread vs. video output thread issue. The decoder thread is trying to do something that can only be done in the output thread. To reproduce it, tune live TV to a normally interlaced channel and wait. It might take several hours but it will evenutally happen. |
| [17:53:09] | Merlin83b (Merlin83b!~Daniel@office.34sp.com) has quit (Quit: Leaving) | |
| [17:55:03] | ** stuarta ponders if apache can be dropped in favour of nginx on that host now ** | |
| [17:56:34] | stuarta: | hampton: let me know what you find out from the Qt sources |
| [17:56:45] | stuarta: | curious to know what Qt is expecting here |
| [17:56:58] | stuarta: | especially since it actually downloaded a new copy of the file |
| [18:10:34] | hampton: | It only downloaded the new copy when I changed myth to force AlwaysNetwork, If I leave it at the default of PreferNetwork it never contacts themes.mythtv.org. |
| [18:12:57] | stuarta: | ah |
| [18:15:41] | ** stuarta heads out ** | |
| [18:17:55] | willcooke (willcooke!~willcooke@ubuntu/member/willcooke) has quit (Quit: Do your hobbies) | |
| [18:57:45] | amessina (amessina!~amessina@unaffiliated/amessina) has joined #mythtv | |
| [19:47:16] | amessina_ (amessina_!~amessina@unaffiliated/amessina) has joined #mythtv | |
| [19:48:01] | amessina (amessina!~amessina@unaffiliated/amessina) has quit (Ping timeout: 260 seconds) | |
| [19:59:34] | hampton: | Grumble, grumble. Qt appears to be doing an RFC2616 calculation to figure out the expiration date since the download doesn't have one. Basically the freshness period for the file becomes however long was the file valid before you downloaded it, divided by ten. The last 29 themes was modified on 22 Sep 2018 and I downloaded it 68 days later on 30 Nov 2018, therefore I see that it will be valid for 7 days from 30 Nov. |
| [20:05:01] | hampton: | Myth attempts to download the themes.zip file when you first enter the theme chooser. I'm thinking that if the user opens the menu and selects 'Refresh Downloadable Themes' that needs to force qt to bypass its cache. It doesn't, currently. It just requests a download which gets served out out the cache again. |
| [20:16:47] | gigem: | Regarding #12902, I'm testing a crude workaround that simply doesn't touch the deinterlacer when ChangeSpeed() is called from the decoder thread. I think stuarta was on the right track about in comment 17 about using signals and slots. Who knows that part of mythplayer best? Would all of FileChangedCallback() or just the part between Pause() and Play() need to be run in a slot? |
| [20:16:47] | ** MythLogBot https://code.mythtv.org/trac/ticket/12902 ** | |
| [20:47:11] | peterbennett: | gigem: I think I have a better solution. Getting ready to test it |
| [20:47:27] | peterbennett: | gigem: see my comment in teh tiocket |
| [20:47:46] | hampton: | Gotta love a one word fix to a problem that took half a day to track down. :-) |
| [20:49:00] | peterbennett: | gigem: I think maybe the bug can be created by watching a program at a speedup and crossing over the end of one program to teh next |
| [21:02:16] | peterbennett: | gigem: No that does not re-create it |
| [21:30:16] | gigem: | peterbennett: That could work too. I just need something that works as the low MAF has been causing increasing tension. It's been particularly frustrating because she refuses to use the workarounds I provided (cable box and HDHomeRun app) for live Tv. "It's too confusing to do all that. I just want to watch TV!" |
| [22:09:39] | Steve-Goodey (Steve-Goodey!~steve@2a00:23c5:7da3:4501:7a84:3cff:fedf:a99) has quit (Quit: Konversation terminated!) | |
| [22:12:37] | peterbennett: | gigem: See patch on ticket. I tested with OpenGL Slim, I could not recreate the problem so my change has not been tested. |
| [22:13:52] | peterbennett: | gigem: I tested Live TV with several transitions and also with speedup .9 and everything worked, |
| [22:22:18] | gigem: | peterbennett: I'm about to build with your patch. My hack worked for about 2 hours without failing. On average, it takes about 4–5 hours with transitions every 30 minutes. |
| [22:23:06] | peterbennett: | gigem: I could not see what triggers that callback, it did not get called in my tests\ |
| [22:27:27] | gigem: | I've never looked myself. I just know live TV program boundaries have always been extremely fragile. |
| [22:28:21] | peterbennett: | gigem: I assume it needs an interlaced station |
| [22:32:39] | SteveGoodey (SteveGoodey!~steve@2a00:23c5:7da3:4501:7a24:afff:fe9d:c233) has quit (Quit: Konversation terminated!) | |
| [22:37:43] | gigem: | peterbennett: Yes. I'd been using HLN as my test, but Frotier seems to have just dropped it. BBC World News America is what I'm using now. They usually change programs every 30 minutes and sometimes even every 15 minutes. |
| [22:38:37] | gigem: | I forgot. You have Comcast, right? That means most of your channels are 720p, then. |
| [22:39:00] | gigem: | I have a pretty even mix of 720p and 1080i. |
IRC Logs collected by
BeirdoBot.
Please use the above link to report any bugs.