| *** tpb has joined #timvideos | 00:00 | |
| CarlFK | tumbleweed: what should re-assign the network dev names? | 02:21 |
|---|---|---|
| CarlFK | https://github.com/CarlFK/video-stack-deploy/blob/pxe-toucheup/roles/dhcp-server/handlers/main.yml command: "/bin/systemctl restart networking.service" | 02:22 |
| tpb | Title: video-stack-deploy/main.yml at pxe-toucheup · CarlFK/video-stack-deploy · GitHub (at github.com) | 02:22 |
| CarlFK | not that I guess | 02:22 |
| tumbleweed | CarlFK: as I said, doing this without rebooting is actually pretty hard | 02:22 |
| tumbleweed | you want to ifdown the interfaces that are going to be renamed | 02:22 |
| tumbleweed | then run the udev thing to rename them | 02:23 |
| tumbleweed | then bring them up | 02:23 |
| CarlFK | oh, missed that it was hard. | 02:23 |
| *** karan has joined #timvideos | 03:53 | |
| xfxf | tumbleweed: i'm about to use 2x of the setups we used at LCA for another conference. laptops will be offline, i.e. connected direct to each other using a crossover cable, no internet. any caveats re the ansible stuff i should be aware of? | 04:47 |
| CarlFK | xfxf: you can use a crossover but it isn't needed with gig-e | 05:00 |
| xfxf | indeed, sorry, old habit of saying crossover | 05:02 |
| xfxf | oh i have a new format json file for that conf if you feel like trying to create a show up on our cloud veyepar? | 05:03 |
| xfxf | i was gonna have a crack later tonight | 05:03 |
| xfxf | it's more sane then the last one i gave you :P | 05:03 |
| CarlFK | I'll look, but the last one scared me | 05:39 |
| *** cr1901_modern has quit IRC | 06:07 | |
| *** cr1901_modern has joined #timvideos | 06:08 | |
| *** karan_ has joined #timvideos | 06:16 | |
| *** karan has quit IRC | 06:18 | |
| *** karan_ is now known as karan | 06:18 | |
| tumbleweed | xfxf: so, one of the caveats was that we never solved network secrets, so I just had those checked out on my laptop | 06:42 |
| tumbleweed | but that doesn't matter, if you have no internet | 06:42 |
| tumbleweed | if you use the static hosts config, like we did at LCA, you don't need internet | 06:42 |
| xfxf | I'm just fiddling with the systemd scripts manually for now and will redeploy later | 06:46 |
| xfxf | It's been a while since I've used ansible in anger; you pushed config changes from your laptop? | 06:47 |
| xfxf | (we use puppet at work, which runs an agent and pulls) | 06:47 |
| xfxf | one fun caveat I just noticed is if the hdmi2usb ingest can't start for whatever reason, the USB module is constantly removed and inserted into the kernel | 06:48 |
| xfxf | I wonder if that has any fun side effects if left running for hours | 06:48 |
| xfxf | in theory the kernel should be solid enough | 06:48 |
| CarlFK | tumbleweed: I want to test the debian stuff, but ansible_distribution: Debian didn't seem to do anything. and scares me. maybe ansible stuff uses that | 06:59 |
| CarlFK | I am thinking I should change it to ... distro: debian, and replace the places you used ansible_distribution | 07:00 |
| tumbleweed | xfxf: it seemed to not kill the kernel | 07:04 |
| tumbleweed | but yes, I was initially skeptical too :P | 07:04 |
| tumbleweed | CarlFK: ansible_distribution is a magic variable that says it discovered it was running on debian | 07:05 |
| CarlFK | tumbleweed: right. Im running ansible on my ubuntu laptop and want to setup a debian install. | 07:06 |
| tumbleweed | CarlFK: right, so poin tyour laptop at a debian box, and it should just work | 07:06 |
| CarlFK | it detects the target? | 07:07 |
| tumbleweed | yes | 07:07 |
| CarlFK | weird | 07:07 |
| tumbleweed | thats' what ansible_distribution is | 07:07 |
| tumbleweed | why is that weird? it's a useful feature, isn't it? | 07:08 |
| CarlFK | I fiugred it was set as part of the ansible package | 07:09 |
| tumbleweed | there are other things one wants to discover | 07:10 |
| CarlFK | but I guess you can pip or git get it.. so never mind that | 07:10 |
| tumbleweed | grep for ansible_ and you'll see we're using a few | 07:10 |
| tumbleweed | https://docs.ansible.com/ansible/playbooks_variables.html#information-discovered-from-systems-facts | 07:11 |
| tpb | Title: Variables Ansible Documentation (at docs.ansible.com) | 07:11 |
| xfxf | tumbleweed: to confirm sorry, w/ ansible you pushed from your laptop every time there was a config change? | 07:18 |
| tumbleweed | xfxf: yes and no :P | 07:19 |
| tumbleweed | during the conference, yes | 07:19 |
| tumbleweed | during the hackfest, we'd reimage from scratch regularly, to be sure that it was still possible | 07:19 |
| tumbleweed | (and so that I didn't have to think about backwards compatibility) | 07:20 |
| *** karan has quit IRC | 08:53 | |
| xfxf | ugh, ok, thumbs down to the blackmagic web presenter | 10:00 |
| xfxf | all audio inputs are muxed together, you can't mute HDMI audio when you're doing XLR or RCA in for example, and XLR/RCA sounds muddy and horrible | 10:01 |
| xfxf | forums say the same thing | 10:01 |
| xfxf | 'will be fixed in a firmware update' | 10:01 |
| xfxf | more like they released something that wasn't finished :/ | 10:01 |
| xfxf | video quality looks good though, even 1080p downscaled to its 720p | 10:01 |
| xfxf | tumbleweed: good info, thx | 10:01 |
| xfxf | https://usercontent.irccloud-cdn.com/file/rMafAAve/20170305_205935.jpg | 10:08 |
| xfxf | the optional panel looks cool. But the only buttons that work are 1 and 2 (cut between inputs) | 10:09 |
| xfxf | hdmi2usb is significantly more functional than a shipped product, there you go | 10:09 |
| xfxf | oh herm, actually, this thing might have a serial console too, just its usb vid/pid isn't recognised by any linux drivers | 10:11 |
| xfxf | at least there's one thing that's not bound to any drivers (presents as UVC video + a seperate audio device, works with vocto with our existing ingest scripts fine) | 10:12 |
| mithro | xfxf: dump the output of lsusb -vv somewhere | 10:22 |
| xfxf | https://usercontent.irccloud-cdn.com/file/WY9TNFeH/lsusb.txt | 10:23 |
| xfxf | the USB cable plugged into the back appears to present the webcam/audio devices, the USB cable plugged into the front you can see in the pic above presents a device nothing binds to | 10:28 |
| xfxf | i assume it'll be what blackmagic get us to use when there's a firmware update. as yet, no firmware updates | 10:28 |
| xfxf | interestingly too, i don't need to do the gstreamer audio delay thing in our ingest scripts when using the USB audio on this... it's already perfectly in sync | 10:29 |
| xfxf | thaytan_: did you tell me what that option was re sync that i should try? i forget if you did. i can try it now | 10:30 |
| thaytan_ | setting a common base-time of 0 everywhere | 10:33 |
| thaytan_ | so packet timestamps actually become wall clock times from the shared clock | 10:33 |
| xfxf | right, ta | 10:50 |
| mithro | xfxf: So you going to pull it apart? | 10:50 |
| xfxf | will look that up | 10:50 |
| xfxf | mithro: i planned to but right now it's useless to me, so i'm considering sending it back - it's $800 i could use elsewhere | 10:50 |
| xfxf | i'll see when this supposed firmware update comes out | 10:50 |
| xfxf | i'm half wondering whether this is stuff is toggable if i can get a console to it somehow, either via that USB port or jtag/serial pins on the board somewhere | 10:51 |
| mithro | xfxf: Have you plugged it into a Windows/Mac computer? | 10:51 |
| xfxf | yes | 10:51 |
| xfxf | blackmagic don't provide any software or firmware for it (yet) | 10:51 |
| xfxf | just comes up as a usb webcam + audio-in-only sound card | 10:52 |
| xfxf | herm comparing the output of this to the output of the hdmi2usb - the hdmi2usb has screen tearing on moving images, both on the usb encoder + hdmi output. is that known? | 10:53 |
| mithro | xfxf: What do you mean tearing and how are you testing it? | 10:53 |
| xfxf | i'm playing a movie through it | 10:54 |
| xfxf | like vsync tearing if you've ever played a video game that suffers from it | 10:54 |
| xfxf | lines through the image when there's fast motion | 10:54 |
| mithro | xfxf: I don't trust that test | 10:54 |
| xfxf | well the blackmagic sitting next to it isn't doing it at all | 10:54 |
| xfxf | same source | 10:54 |
| xfxf | actually, let me test more, maybe the video card (i.e. HDMI source) just sucks on this thing | 10:55 |
| mithro | xfxf: There are plenty of ways it could be happening | 10:55 |
| xfxf | actually, nm, it's the video card/HDMI source, just the way i had it hooked up was deceptive | 10:58 |
| xfxf | hilariously, i just swapped the HDMI in/out from the hdmi2usb to the blackmagic | 10:58 |
| xfxf | the blackmagic isn't doing HDMI out in a way my monitor likes | 10:58 |
| xfxf | it appears to provide its own EDID data to the computer, though, it's showing as blackmagic | 10:58 |
| xfxf | will do a EDID dump if you want later | 10:58 |
| xfxf | plug cables back into hdmi2usb, works perfectly | 10:59 |
| xfxf | i think there's zero way i'm taking this thing with me to NZ on wed... hdmi2usb is way better | 10:59 |
| xfxf | i assume they'll release a firmware update which will help - it's already failed out a bunch of ways already from 30 mins of testing :/ | 11:00 |
| *** karan has joined #timvideos | 11:56 | |
| *** karan__ has joined #timvideos | 12:48 | |
| *** karan has quit IRC | 12:50 | |
| *** karan__ is now known as karan | 12:50 | |
| *** gk_1wm_su has joined #timvideos | 13:36 | |
| *** gk_1wm_su has quit IRC | 13:43 | |
| cr1901_modern | karan: You will need to wait to talk to mithro due to the time difference | 14:52 |
| *** rohitksingh has joined #timvideos | 15:43 | |
| *** rohitksingh has quit IRC | 15:55 | |
| *** karan has quit IRC | 16:10 | |
| *** rohitksingh has joined #timvideos | 16:12 | |
| *** rohitksingh has quit IRC | 16:31 | |
| *** karan has joined #timvideos | 16:32 | |
| tumbleweed | ho hum. in the keynote venue at SCaLE, they only have DVI capture | 17:52 |
| tumbleweed | (apparently turned into VGA at the back of the room, for capture into tho hardware mixer thing) | 17:53 |
| *** thaytan has joined #timvideos | 18:09 | |
| *** ChanServ sets mode: +v thaytan | 18:09 | |
| *** thaytan_ has quit IRC | 18:13 | |
| *** karan has quit IRC | 18:23 | |
| *** karan has joined #timvideos | 19:03 | |
| *** karan_ has joined #timvideos | 19:06 | |
| *** karan has quit IRC | 19:09 | |
| *** karan has joined #timvideos | 19:09 | |
| *** karan_ has quit IRC | 19:10 | |
| CarlFK | tumbleweed: gross ;) | 19:20 |
| CarlFK | We (you!) need to get something going like CCC has. they get money from somewhere. | 19:21 |
| tumbleweed | CarlFK: haven't managed to have a good chat with them yet (they're busy, and it seems like a 1-man show, + volunteers) | 19:22 |
| tumbleweed | the one man is ensconsed in a corner, fixing shit ;P | 19:22 |
| tumbleweed | https://www.dropbox.com/sc/04uhvyvqxb555vz/AADMPMdSwxZlfWvbdk23VoLea | 19:23 |
| CarlFK | nice fab work | 19:36 |
| tumbleweed | looks kinda painful to build, but they built them | 19:36 |
| CarlFK | how interested are you in the whacky results of debian pxe server weirdness? like one laptop client works fine, a 2nd (that worked fine with ubuntu) struggles to tftp - it takes minutes to GET the first file, that boots prints the selinux Peter Avalon? message, then times out trying to chain the next | 19:38 |
| CarlFK | I 'want' to dig into it, but I really don't have time. | 19:39 |
| tumbleweed | don't know how much I can dig into that from here | 19:42 |
| tumbleweed | but I'll help | 19:42 |
| CarlFK | I can give you ssh access to the server and you can watch logs while I send pxe stuff with it | 19:57 |
| CarlFK | really just knowing that you care bumps up my ... something | 19:57 |
| CarlFK | I may just blame this hp laptop | 20:07 |
| CarlFK | it has ubuntu installed (only relevant to track down where this next came from: | 20:08 |
| CarlFK | apt-get install grub-ipxe | 20:08 |
| CarlFK | reboot, grub menu, ipxe | 20:08 |
| tumbleweed | yeah. Sometimes chaining ipxe from pxe is useful | 20:08 |
| CarlFK | screen clears, caps lock light is flashing, box hung | 20:08 |
| CarlFK | It did pxe /install a few times yeterday | 20:09 |
| CarlFK | so it can work. just doesn't now. but in a very "local hardware/firmware" sort of way. | 20:10 |
| CarlFK | so not sure I want to spend time looking for problems on the server side | 20:10 |
| tumbleweed | yeah. I could care less about shit like that :P | 20:10 |
| CarlFK | good. scratch that off my list | 20:11 |
| *** puck`` is now known as puck | 20:19 | |
| *** thaytan_ has joined #timvideos | 21:19 | |
| *** thaytan has quit IRC | 21:24 | |
| *** puck is now known as puck`` | 21:31 | |
| *** karan is now known as karan_off | 22:23 | |
| *** karan_off has quit IRC | 22:23 | |
| CarlFK | tumbleweed: sigh.. I had to poke... wget http://boot.ipxe.org/ipxe.lkrn (so latest stable I guess) and get: | 23:05 |
| CarlFK | "waiting for link-up..." Permission denied http://ipxe.org/0227603c | 23:06 |
| tpb | Title: iPXE - open source boot firmware [err:022760] (at ipxe.org) | 23:06 |
| CarlFK | hmm, looking closer, looks like it gives up on the wired and tries the wifi ... ignore all this shit. | 23:07 |
| tumbleweed | yeah | 23:07 |
| xfxf | tumbleweed: my guess is if somebody spent a lot of time building those boxes, they're going to be somewhat emotionally attached to using them | 23:08 |
| xfxf | why would they make something in 2017 with VGA only :( | 23:09 |
| * tumbleweed just sat in a talk where it hardly ever managed to keep sync :( | 23:10 | |
| tumbleweed | but mostly, they seem to be doing a pretty good job (at least from what I can see on the projector) | 23:10 |
| xfxf | i do wish more of these conferences were working on collective solutions | 23:25 |
| xfxf | cross-pollination def better now then they were 5 years ago but everybody is cooking their own solution | 23:25 |
| xfxf | s/then they were/then it was/ | 23:26 |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!
