| Time |
S |
Nick |
Message |
| 07:43 |
|
aditsu |
I never made a makefile, I feel clean :) |
| 10:15 |
|
pdurbin |
heh |
| 10:22 |
|
prologic |
testing creating a docker swarm/cluster with factory atm |
| 10:23 |
|
pdurbin |
It would be highly interesting to me if someday someone who is trying to install our app asks for a Docker image. That day hasn't come yet. |
| 10:24 |
|
prologic |
it will |
| 10:24 |
|
prologic |
:) |
| 10:24 |
|
prologic |
even if I have to ask for it :P |
| 10:24 |
|
prologic |
aha |
| 10:25 |
|
pdurbin |
what they do ask for are packages for our app so they can yum or apt-get install it |
| 10:26 |
|
prologic |
ahh |
| 10:26 |
|
prologic |
stil stuck in that mindset |
| 10:26 |
|
prologic |
yeah give it time |
| 10:26 |
|
prologic |
I say a few more months |
| 10:26 |
|
prologic |
andt they'll want to: docker run it |
| 10:26 |
|
pdurbin |
time will tell |
| 10:27 |
|
pdurbin |
recently there was interest in a VMWare image: https://github.com/IQSS/dataverse/issues/2280 |
| 10:28 |
|
prologic |
http://i.imgur.com/T1rN4Z3.png |
| 10:29 |
|
prologic |
sweet :) |
| 10:29 |
|
prologic |
creating a Docker cluster -- just works (tm) |
| 10:29 |
|
prologic |
:P |
| 10:29 |
|
prologic |
with this factory.yml https://gist.github.com/1f7011d207a15945d2f7 |
| 10:30 |
|
pdurbin |
prologic: are you running ratpoison? |
| 10:30 |
|
prologic |
ratpoison? no |
| 10:31 |
|
prologic |
XFCE 4.12 |
| 10:31 |
|
pdurbin |
ok. I saw a rat with an x on it |
| 10:39 |
|
prologic |
ahh |
| 10:39 |
|
prologic |
https://github.com/prologic/factory |
| 10:39 |
|
prologic |
well creating a docker swarm cluster works really nicely |
| 10:46 |
|
pdurbin |
is a swarm cluster a bunch of docker machines? |
| 10:47 |
|
prologic |
yeap |
| 10:47 |
|
prologic |
with discovery |
| 10:48 |
|
pdurbin |
and if one of the docker machines goes down? |
| 10:48 |
|
prologic |
and launch strategies |
| 10:48 |
|
prologic |
the containers can move to other hosts I believe |
| 10:48 |
|
prologic |
I'm not sure how far along some of the feature sare |
| 10:48 |
|
pdurbin |
automatically? |
| 10:48 |
|
prologic |
but I'm only just playing with swarm a bit more now |
| 10:48 |
|
prologic |
good question :) |
| 10:49 |
|
prologic |
autodock-migrate could do it :P |
| 10:49 |
|
prologic |
better build it :) |
| 10:49 |
|
pdurbin |
:) |
| 10:49 |
|
pdurbin |
what's a launch strategy? |
| 10:49 |
|
prologic |
I'm working on autodock-autoscale atm though |
| 10:50 |
|
prologic |
I think I'm using the wrong terminology here |
| 10:50 |
|
prologic |
but it has several strategies for launching containers |
| 10:50 |
|
prologic |
binpack is one I remember |
| 10:50 |
|
prologic |
where it'll spawn containers as tightly as possible on one host |
| 10:50 |
|
prologic |
the other is spread |
| 10:50 |
|
prologic |
https://docs.docker.com/swarm/ |
| 10:51 |
|
pdurbin |
"The spread and binpack strategies compute rank according to a node’s available CPU, its RAM, and the number of containers it is running. The random strategy uses no computation. It selects a node at random and is primarily intended for debugging." https://docs.docker.com/swarm/scheduler/strategy/ |
| 10:51 |
|
pdurbin |
gotcha |
| 10:51 |
|
prologic |
cool! |
| 10:51 |
|
prologic |
Docker itself is actually swarm aware |
| 10:51 |
|
prologic |
d'uh |
| 10:51 |
|
prologic |
so docker ps on a connected cluster |
| 10:52 |
|
prologic |
will actually list where containers are running |
| 10:52 |
|
prologic |
I see no migration strategies though at this time |
| 10:52 |
|
prologic |
so perhaps I *will* build autodock-migrate |
| 10:52 |
|
pdurbin |
:) |
| 10:52 |
|
prologic |
I might rename autodock-autoscale to just autodock-scale |
| 10:52 |
|
prologic |
remove the 'auto' dupe |
| 10:53 |
|
pdurbin |
"Autodocks, roll out!" |
| 10:53 |
|
prologic |
haha! |
| 10:53 |
|
prologic |
I love it! |
| 10:53 |
|
prologic |
help me come up with a name for a PaaS based on this |
| 10:54 |
|
prologic |
atm I just have this lame autodock-paas repo going on with that curl | bash -s setup |
| 10:54 |
|
pdurbin |
could use tricider |
| 10:54 |
|
prologic |
Also it's worth nothing that hipache is not the only viable frontend/load-balancer |
| 10:56 |
|
prologic |
I can also (or someone can) create an autodock-redx (https://github.com/rstudio/redx) |
| 10:56 |
|
prologic |
as another viable option |
| 10:56 |
|
pdurbin |
searchbot: lucky hipache |
| 10:56 |
|
searchbot |
pdurbin: https://github.com/hipache/hipache |
| 10:56 |
|
prologic |
tricider? |
| 10:57 |
|
pdurbin |
hɪ'pætʃɪ |
| 10:57 |
|
prologic |
hmm |
| 10:57 |
|
prologic |
one thing with factory |
| 10:57 |
|
prologic |
I can't decide how much of docker-machine's cli interface it should wrap around |
| 10:57 |
|
prologic |
thoughts? |
| 10:57 |
|
pdurbin |
tricider is what we're using for https://github.com/pdurbin/addressbookmvc/issues/11 |
| 10:58 |
|
prologic |
http://i.imgur.com/mXqd4YP.png |
| 10:58 |
|
pdurbin |
aditsu: I just added a "pro" argument for one of your ideas |
| 10:59 |
|
prologic |
oh |
| 11:16 |
|
prologic |
actually you know... |
| 11:16 |
|
prologic |
I just realized how bloody useful factory really is |
| 11:17 |
|
pdurbin |
the next step is world domination |
| 11:17 |
|
prologic |
you can have a repo with factory.yml docker-compose.yml and a bunch of dirs with Dockerfile(s) for various services/containers/etc |
| 11:18 |
|
prologic |
factory up && eval $(factory env) && docker-compose up |
| 11:18 |
|
prologic |
hah |
| 11:18 |
|
prologic |
but that leads me to my previous question |
| 11:18 |
|
prologic |
how much of docker-machine/docker-compose should I wrap up |
| 11:19 |
|
pdurbin |
prologic: maybe you can give me an account on this PaaS so I can play around |
| 11:19 |
|
prologic |
an account you say? |
| 11:19 |
|
prologic |
there are no accounts :P |
| 11:19 |
|
prologic |
it's all automatic :) |
| 11:19 |
|
prologic |
but you have a DO account right? |
| 11:19 |
|
pdurbin |
heroku has accounts |
| 11:21 |
|
prologic |
I'll walk through this myself fully first |
| 11:21 |
|
prologic |
when I'll get you to try it out yeh? |
| 11:24 |
|
pdurbin |
sounds good. thanks |
| 12:28 |
|
prologic |
pdurbin, http://asciinema.org/a/23ujlplez9yyulmqojinw9ei4 |
| 12:42 |
|
aditsu |
prologic: you have a gopher server?!? :O |
| 12:43 |
|
prologic |
yes of course I do |
| 12:43 |
|
prologic |
why? |
| 12:43 |
|
prologic |
gopher://daisy.shortcircuit.net.au/ |
| 12:43 |
|
prologic |
go check it out :) |
| 12:43 |
|
prologic |
I keep daily content on there too |
| 12:55 |
|
aditsu |
I saw it in your screenshot |
| 12:56 |
|
aditsu |
"of course"? nobody uses gopher |
| 12:57 |
|
prologic |
says you :P |
| 12:57 |
|
aditsu |
it doesn't even work in my browser |
| 12:59 |
|
aditsu |
actually, I tried 3 different browsers now |
| 12:59 |
|
aditsu |
ah, works with lynx (not very surprisingly) |
| 13:02 |
|
prologic |
you need to install OverbiteFF plugin |
| 13:02 |
|
prologic |
then Firefox will happily access gopher:// uris |
| 13:03 |
|
prologic |
also elinks, lynx and curl |
| 13:03 |
|
prologic |
There used to be a Chrome extensions; but recently Chrome has made it impossible to build extensions that create custom protocol handlers |
| 13:04 |
|
aditsu |
your website link says: 3404 Not Found: Resource not found! error.host 0 . |
| 13:06 |
|
aditsu |
I have elinks compiled without gopher support; it still works... by automatically running lynx >_< |
| 13:09 |
|
prologic |
yeah! |
| 13:14 |
|
prologic |
anyway |
| 13:14 |
|
prologic |
I actually quite like Gopher really |
| 13:14 |
|
prologic |
I wish more people would use it :) |
| 13:16 |
|
pdurbin |
heh. "says you" |
| 13:17 |
|
pdurbin |
aditsu: cool. the tricider admin link seems to work for you |
| 13:18 |
|
aditsu |
gopher is so... text :p |
| 13:18 |
|
aditsu |
pdurbin: yeah |
| 13:18 |
|
pdurbin |
mmm, sweet, sweet text |
| 13:20 |
|
prologic |
and what's wrong with text? |
| 13:20 |
|
prologic |
you're using such a thing now :P |
| 13:20 |
|
prologic |
IRC! |
| 13:20 |
|
prologic |
and IRC is even older than Gopher |
| 13:20 |
|
prologic |
so you're all a bunch of old farst :) |
| 13:20 |
|
pdurbin |
prologic: are you into NNTP too? |
| 13:21 |
|
prologic |
sadly no |
| 13:21 |
|
prologic |
remember I really don't like forums :) |
| 13:21 |
|
prologic |
NNTP is the predecessor to forums/lists right? |
| 13:22 |
|
prologic |
aditsu, I also wrote my own Gopher daemon :P |
| 13:22 |
|
prologic |
https://github.com/prologic/cgod |
| 13:22 |
|
pdurbin |
I guess reddit is the new NNTP :) |
| 13:22 |
|
prologic |
fully pluggable with apps, sessions and all the whizz bang :) |
| 13:22 |
|
aditsu |
prologic: so hardcore :p |
| 13:22 |
|
prologic |
pdurbin, yeah |
| 13:22 |
|
prologic |
haha |
| 13:22 |
|
prologic |
and ofc it's written in circuits |
| 13:22 |
|
prologic |
so it's super fast ;) |
| 13:22 |
|
pdurbin |
I imagine mailing lists are older than NNTP but I'm not sure. |
| 13:23 |
|
prologic |
unlike many Gopherd(s) that are/were inetd based with a forking concur ency model uggh |
| 13:23 |
|
pdurbin |
mailing lists are 40 years old, I read in a recent paper |
| 13:23 |
|
prologic |
yeah I think NNTP outdates mailing lists |
| 13:24 |
|
aditsu |
I'm using IRC with a graphical client, and there's no replacement for IRC (but stackexchange chat is a good contender) |
| 13:25 |
|
pdurbin |
highly recommended: Mailing Lists: Why Are They Still Here, What's Wrong With Them, and How Can We Fix Them? http://people.csail.mit.edu/axz/mailinglists.html http://people.csail.mit.edu/axz/papers/mailinglists.pdf |
| 13:26 |
|
pdurbin |
aditsu: it would be nice if stackexchange chat were open source |
| 13:27 |
|
aditsu |
what about gitter then? |
| 13:27 |
|
prologic |
well I'm using a graphical client too |
| 13:27 |
|
prologic |
HExchat |
| 13:27 |
|
prologic |
but it's still text |
| 13:28 |
|
prologic |
I don't really like gitter all that much |
| 13:28 |
|
prologic |
it's trying to be slack for github repos |
| 13:28 |
|
pdurbin |
I'm quite happy to use https://irc.gitter.im but I don't believe the server side is open source. |
| 13:28 |
|
prologic |
yeah that's one of the things I like about Slack |
| 13:28 |
|
pdurbin |
prologic: what's wrong with gitter? it's great! |
| 13:28 |
|
prologic |
is it's XMPP/IRC gateway inteegrations |
| 13:28 |
|
prologic |
so you don't have to use their app or webapp |
| 13:28 |
|
prologic |
well nothing now that I knwo about it's irc integration |
| 13:28 |
|
prologic |
did not know that :) |
| 13:29 |
|
aditsu |
my irc is text with fonts, colors, links and little images too :p |
| 13:29 |
|
prologic |
I actually use bitlbee for many IM protocols |
| 13:29 |
|
prologic |
Twitter, Google Talk, etc |
| 13:29 |
|
prologic |
Colloquy? |
| 13:29 |
|
aditsu |
chatzilla |
| 13:29 |
|
pdurbin |
prologic: so you admit that gitter is awesome? |
| 13:29 |
|
prologic |
ahh |
| 13:30 |
|
prologic |
it's okay now :) |
| 13:30 |
|
prologic |
I'll try out the irc interface later on |
| 13:30 |
|
pdurbin |
heh |
| 13:30 |
|
prologic |
I setup a gitter for circuits |
| 13:30 |
|
prologic |
people are strange though |
| 13:30 |
|
prologic |
quick to jump on the next new thing :P |
| 13:30 |
|
prologic |
*cough* |
| 13:31 |
|
pdurbin |
people should have stayed on gopher rather than jumping on the web |
| 13:32 |
|
prologic |
lol |
| 13:32 |
|
prologic |
absolutely! |
| 13:32 |
|
prologic |
the internet at large would be a much better place |
| 13:32 |
|
prologic |
we'd not be wasting time on Facbeook and the likes |
| 13:32 |
|
prologic |
haha |
| 13:32 |
|
pdurbin |
aditsu: ok, so maybe prologic is an old fart |
| 13:32 |
|
prologic |
I wonder what a social network would look like in Gopherspace |
| 13:32 |
|
prologic |
haha |
| 13:32 |
|
prologic |
joking :) |
| 13:32 |
|
pdurbin |
:) |
| 13:32 |
|
aditsu |
gopherface |
| 13:33 |
|
pdurbin |
golang has a lot of gopher stuff going on |
| 13:33 |
|
aditsu |
I used to waste a lot of time on facebook, then I deleted my account |
| 13:33 |
|
prologic |
it does :) |
| 13:37 |
|
pdurbin |
the biggest problem with gitter is no easy data liberation. but they have an API so it shouldn't be too tough to write a downloader |
| 13:37 |
|
pdurbin |
prologic: would that be a good circuits project? |
| 13:39 |
|
aditsu |
isn't gitter open-source? |
| 13:39 |
|
prologic |
depends |
| 13:39 |
|
prologic |
is this a once-only task? |
| 13:39 |
|
pdurbin |
well, maybe you'd run it every week or so to get a backup. to liberate your data |
| 13:40 |
|
aditsu |
oh, apparently not (or not completely) |
| 13:40 |
|
prologic |
ahh yeah |
| 13:40 |
|
prologic |
so you could probably write a circuits daemon and periodocally poll the api end point |
| 13:40 |
|
prologic |
but requests (the library) might be a better fit as a simple cron job |
| 13:41 |
|
aditsu |
and btw, I hate mailing lists but I like forums |
| 13:42 |
|
prologic |
other way around for me |
| 13:42 |
|
prologic |
at least gmail's ui is nice to use with filters, etc |
| 13:42 |
|
prologic |
so I filter mailing lists into labels which are easy to follow and access |
| 13:43 |
|
prologic |
*a i *n |
| 13:43 |
|
prologic |
is a nice keyboard set of shortcuts |
| 13:43 |
|
prologic |
:) |
| 13:43 |
|
prologic |
select all, ignore, select none |
| 13:43 |
|
pdurbin |
here's the last one I wrote: https://github.com/pdurbin/data-liberation-shapado |
| 13:44 |
|
prologic |
oooh gawd man |
| 13:44 |
|
prologic |
learn Python quick! |
| 13:44 |
|
prologic |
I really can't stand reading Perl |
| 13:44 |
|
prologic |
uggh ;) |
| 13:46 |
|
pdurbin |
heh |
| 13:46 |
|
pdurbin |
maybe sivoais will appreciate it |
| 13:47 |
|
prologic |
lol |
| 13:47 |
|
prologic |
ahh it's fine |
| 13:47 |
|
prologic |
but Python *is* more readable code :) |
| 13:48 |
|
pdurbin |
I'm not picky. |
| 13:48 |
|
prologic |
it's what you know |
| 13:48 |
|
prologic |
:) |
| 13:48 |
|
pdurbin |
is written in Perl and it works great |
| 13:48 |
|
pdurbin |
not that I wrote it |
| 13:48 |
|
prologic |
heh |
| 13:49 |
|
prologic |
let's not have a Perl vs Python thingy :) |
| 13:49 |
|
prologic |
Perl is fine for Perl programmers :) |
| 13:49 |
|
prologic |
haha |
| 13:49 |
|
pdurbin |
at this point I guess I know Java best |
| 13:49 |
|
prologic |
that's true |
| 13:50 |
|
pdurbin |
but I'd be fine working on a gitter data liberator in python. whatever |
| 13:50 |
|
prologic |
yeah sure |
| 13:50 |
|
prologic |
be kinda similar to your shapado one |
| 13:51 |
|
pdurbin |
I try not to but data into any system that doesn't let me get it out. |
| 13:51 |
|
pdurbin |
put* |
| 13:51 |
|
prologic |
yeah fair enough |
| 13:53 |
|
prologic |
think it's bedtime |
| 13:53 |
|
pdurbin |
nice chatting with you |
| 16:47 |
|
sivoais |
I like that you wrote tests for the shapado script. I don't do that enough when I am scraping |
| 16:48 |
|
sivoais |
scraping is tiring, so I try to limit the services I use :-P |
| 16:48 |
|
sivoais |
and keep things that have open specs |
| 16:49 |
|
sivoais |
I absolutely lurv reading specs. especially IETF RFCs |
| 16:49 |
|
pdurbin |
oh. I didn't even look at that old code. I assumed there was an api! |
| 23:14 |
|
pdurbin |
yeah, there is an api. or at least the data is retrievable in JSON format. good enough |
