#timvideos

Merry Xmas everyone!

Merry Christmas :D

cr1901_modern: \o/

cr1901_modern: I have an adapter which goes from the Numato<->Digilent style PMOD spacing and high speed layout

cr1901_modern: I'll send you some when I get them made

mithro: Tyvm :). I got an RGB LED module, and I got a breadboard module

Title: PMOD Pins - Google Sheets (at docs.google.com)

My plans for the breadboard module are to put an ECJ*-series chip on it and add Ethernet to MimasV2

_florent_: LiteEth works with the ECJ* PHYs?

ENC*

Merry Christmas to you too mithro and everyone else

cr1901_modern: http://numato.com/100base-t-ethernet-expansion-...

Title: 100BASE-T Ethernet Expansion Module (at numato.com)

cr1901_modern: You could save a bunch of those pins by using RGMII rather than MII

tumbleweed: http://openocd.org/2016/12/openocd-0-10-0-rc1-r...

Title: OpenOCD 0.10.0-rc1 release candidate is out « Open On-Chip Debugger (at openocd.org)

mithro: \o/

tumbleweed: Are you in Hamburg now?

mithro: on my way. I arrive tomorrow

sitting at DTW right now

I suspect you'll sleep a lot better than I will :P

tumbleweed: I have sleeper bed :-P

Internet is a bit spotty as we go through tunnels though

those get kinda expensive on planes

ah, yes, that is a downside

brb

they have planes that go though tunnels?

it usually doesn't end well

tumbleweed: I have this ton of things ... should I post this to some list or ... this: http://paste.ubuntu.com/23684139/

Title: Ubuntu Pastebin (at paste.ubuntu.com)

it is kinda 1/2 baked, but I suspect 1/2 the things have quick replies.

one big sweeping question I have: does it make sense to try and maintain a single playbook that can drive the 3 or 4 .... flavors? (like icecast / youtube, Debian/Ubuntu) or just cut/paste stuff back and forth

I think a lot of those need to be talked through

and many of them are one-line answers :)

nohter one... is there something like ansible-hub that is a collection of playbooks others have used?

CarlFK: Ansible Galaxy

like if I want to setup pxe installs using dnsmasq

Title: Ansible Galaxy | Find, reuse, and share the best Ansible content (at galaxy.ansible.com)

runs back and clcks

Allows you to upload and download roles

presumably, like every other config management system, it's full of terrible playbooks that nobody should ever use

that said, all software is duct tape and string

Look at the number of downloads, the creator, etc. There's some gold, but also some complete rubbish

oh, even popular stuff can be terrible

https://galaxy.ansible.com/debops/dnsmasq/ deeb bop ah boo bop!

Title: Ansible Galaxy | debops.dnsmasq (at galaxy.ansible.com)

my personal bugbear is how chef cookbooks handle package installation. It's usually fragile and terribly insecure

arguably that's because it's hard to do it right

presumably, like everything on the internet :p

also, modular general purpose libraries are often not better than a simple re-implementation yourself

ansible is simple enough that for simple tasks, reimplementation doesn't seem so bad

CarlFK: quick thoughts from your paste

line 3: wasn't a requirement for debconf. So we probably want to spin up an xterm with a tail -f in it

there's a big overlap between debian & ubuntu, so I don't think the bits are as seperable as you are saying

mainly what repos to add I think - like testing/ppa

line 31: debconf has opsis boards, so again not a requirement

actually you probably want production mode rather than jtag?

line 33: so that you can lock the screen, and keep fiddlers out

I hardly remember the details of what board / what condidtions

line 39: lol ubuntu

lol - well yeah

line 44: debconf usually has a full blown mirror. But yes, that packgae is never harmful, so trivial enough to add

46-53: yeah, we haven't got fancy there. That's all manual

56: cu is the venerable unix serial terminal emulator. It's simple and ubiquitous

67: also manual. But I think debconf video team would like to avoid having room names configured anywhere if possible. If you can keep it down to integers, you may be able to get away without per-conference config

70: I know, right? wtf

lol

I seem to remember your config does that, too

mine put it in dv/$room/cutlist.log

which is still not per-day

dv/$room/$date/$time.ts

you want it per day?

it makes things a little easier when manually rsyncing files around

because you have to remember to rsync both the day and the cutlist

I am surprised you rsync a single day

I think veyepar will work with it either way, and I can't think of any reason not to put it in the same dir as the .ts files

if you ran out of disk space, you may have deleted yesterday on the local machine

so?

and you wouldn't want to accidentally delete files on the central server

I guess you usually aren't doing rsync --delete

ah, but you don't want all the junk from after the end of the day, either

you don't? only 1/2 kidding

I have been trying to hard to script that stuff so I an never doing rsync. now I have like 6-30 get/grab/sync/getme/push/pull/pop.sh scripts floating around

yeah, we used to have a dc-move-recordings script, that was nice

but when everything isn't nice and package-managed / configuratino-managed, you find yourself doing a lot of ad-hoc rsyncing

my readme says "use your file system skills." maybe I should pick one of my scripts and reference it in my docs so I will start using it all the time

exactly, rsync is so easy

tumbleweed: #22/23 Where is this set? when: ansible_distribution == 'Debian'

I saw you added that, but not sure ... how it is used? triggered, filtered, somethinged

Well, that was an interesting family visit... had a bad reaction to *something*/couldn't stop coughing/sneezing till I left. 3/10 wouldn't do again.

so, that task only executes on Debian

mithro: Prob out of commission for tonight lol (for when you wake up)

tumbleweed: oh, ansible_distribution is ... auto system something

yeah

ah, so that handles my "tag with deb" thing, right ?

I guess so

I wasn't sure what the thing wsa

paddatrapper: getting anywhere with the fx2 stuff?

cr1901_modern: Okay, I assume you haven't gotten any further with the SPI flash

mithro: Nope :(. I'll work on it through tomorrow

long tunnels.... 64 bytes from ham04s01-in-f4.1e100.net (216.58.213.228): icmp_seq=9791 ttl=49 time=308803 ms