| Time |
S |
Nick |
Message |
| 00:34 |
|
pdurbin |
prologic: frustrated by Docker questions? |
| 00:35 |
|
prologic |
haha |
| 00:35 |
|
prologic |
there was one ont he docker-user mailing list |
| 00:35 |
|
prologic |
it started out as a post with several points |
| 00:35 |
|
prologic |
all negative |
| 00:35 |
|
prologic |
and ending with |
| 00:35 |
|
prologic |
"what's up with that?" |
| 00:35 |
|
prologic |
I was brave enough to "try" to answer it |
| 00:35 |
|
prologic |
admittedly not very well haha |
| 00:36 |
|
prologic |
Here's a link to the thread: https://groups.google.com/forum/#!topic/docker-user/TgCGntvTAjs |
| 00:36 |
|
prologic |
The thing is though (aside from IPV6 isues) |
| 00:36 |
|
prologic |
a lot of these complaints and kind of silly |
| 00:36 |
|
prologic |
and IHMO outside of the bound/goals of Docker in the first place |
| 00:36 |
|
prologic |
Why should I care about what IP Address my conatiner happens to have? |
| 00:37 |
|
prologic |
and bi-directional linking of containers just to solve "configuration issues" of antiquated software or software that has poor discoverability/configurability |
| 00:37 |
|
prologic |
so oh well :) |
| 00:38 |
|
prologic |
I guess Apache Mesos is designed to be distributed/p2p somehow? |
| 00:38 |
|
prologic |
but needing upfront ip addressing to configure master/slave relationships? wtf? |
| 00:39 |
|
prologic |
broadcast discovery ftw? |
| 00:39 |
|
prologic |
client connects to server |
| 00:39 |
|
prologic |
server determines ip of client and connects back |
| 00:39 |
|
prologic |
there are lots of ways of solving bidirecitonal communications between N or more processes |
| 00:39 |
|
prologic |
</rant> |
| 00:40 |
|
pdurbin |
yeah, not very well. :) |
| 00:40 |
|
prologic |
haha |
| 00:40 |
|
pdurbin |
but good that you aren't having these problems |
| 00:40 |
|
prologic |
oh well |
| 00:40 |
|
prologic |
I just get a bit tired of negativity sometimes |
| 00:40 |
|
pdurbin |
you've embraced the zen of docker or something |
| 00:41 |
|
prologic |
there are a lot of folk that demend so much of open source software/hardware |
| 00:41 |
|
prologic |
and expect the hard slogging developers to bend to their every will |
| 00:41 |
|
prologic |
and as you and I both know |
| 00:41 |
|
prologic |
it's one of those things |
| 00:41 |
|
prologic |
a new tech comes along |
| 00:41 |
|
prologic |
and we all (even I'm guilty) abuse it for all it's worth |
| 00:42 |
|
prologic |
bending it into things it was never meant for |
| 00:42 |
|
prologic |
I'm seeing far too often now folk trying to use Docker as a lightweight hypervisor-style full paravirtualed vm |
| 00:42 |
|
prologic |
and it's just not that |
| 00:42 |
|
pdurbin |
for a long time people have cared what ip addresses their boxes have :) |
| 00:43 |
|
prologic |
isn't it time we moved on from that? |
| 00:43 |
|
prologic |
what differences does it make whether my service has 10.0.0.1 or 192.168.0.1 |
| 00:43 |
|
prologic |
as long as it's service get exposed to the outside world |
| 00:43 |
|
prologic |
obviously you have to have an incoming/outgoing point in your infrastructure |
| 00:43 |
|
prologic |
but that's normally at the host level |
| 00:43 |
|
prologic |
not at the container level |
| 00:43 |
|
prologic |
you *don't* expose a container to the outside world directly |
| 00:43 |
|
prologic |
you're not meant to |
| 00:43 |
|
pdurbin |
I hear you. I'm just saying... old habits die hard. |
| 00:44 |
|
prologic |
yes |
| 00:44 |
|
prologic |
they do :) |
| 00:45 |
|
prologic |
anyway |
| 00:45 |
|
prologic |
what's new? :) |
| 00:45 |
|
prologic |
see my little fibonacci webapp? |
| 00:45 |
|
prologic |
kinda neat idea |
| 00:45 |
|
pdurbin |
I clicked it from my phone but couldn't read it so well. |
| 00:46 |
|
prologic |
ahh |
| 00:46 |
|
prologic |
yeah phones kinda suck a bit like that |
| 00:46 |
|
prologic |
heh |
| 00:46 |
|
prologic |
or maybe gist.github.com isn't so Mobile friendly |
| 00:47 |
|
pdurbin |
oh, I see, the question was altered but used to be about fib: http://stackoverflow.com/questions/30844887/for-loop-inside-a-flask-function-internal-server-error |
| 00:49 |
|
prologic |
oh how silly |
| 00:49 |
|
prologic |
wonder why the content was removed |
| 00:49 |
|
pdurbin |
people are weird |
| 00:50 |
|
pdurbin |
anyway, we had our first conference, which was fun: https://twitter.com/hashtag/dataverse2015 |
| 00:50 |
|
pdurbin |
our users are very nice |
| 00:50 |
|
pdurbin |
it might help that a lot of them are librarians :) |
| 00:51 |
|
prologic |
ahh |
| 00:51 |
|
prologic |
very nice :) |
| 00:56 |
|
prologic |
so re addressbookmvc |
| 00:57 |
|
prologic |
what next should we build as a demo |
| 00:57 |
|
prologic |
or should we expand on this |
| 00:57 |
|
prologic |
add in support for some kind of non-file-based db? |
| 01:05 |
|
pdurbin |
prologic: it's been such a whirlwind lately I haven't even tried to get your implementation running! |
| 01:07 |
|
prologic |
ahh |
| 01:07 |
|
prologic |
well it's pretty easy to get it running :) |
| 01:07 |
|
prologic |
build the image and run :) |
| 01:11 |
|
* pdurbin |
forgets how to build an image |
| 01:12 |
|
pdurbin |
Unable to find image 'prologic/addressbookmvc:latest' locally |
| 01:15 |
|
prologic |
docker build -t prologic/addressbookmvc . |
| 01:16 |
|
prologic |
docker run -i -t -p 3000:3000 --rm prologic/addressbookmvc |
| 01:18 |
|
pdurbin |
it's downloading something! thanks. meanwhile I spun it up the old fashioned way. nice validation. search doesn't seem to be working |
| 01:20 |
|
pdurbin |
prologic: when I click "edit" next to a contact I added the previously existing contact is edited |
| 01:24 |
|
prologic |
no search works |
| 01:24 |
|
prologic |
but it's not very good :) |
| 01:24 |
|
prologic |
few bugs eh? |
| 01:24 |
|
prologic |
:) |
| 01:24 |
|
prologic |
I'll fix them tonight I guess |
| 01:33 |
|
pdurbin |
the docker commands worked fine. thanks |
| 01:38 |
|
* pdurbin |
merges https://github.com/pdurbin/addressbookmvc/pull/7 :) |
| 01:38 |
|
prologic |
hehe nps :) |
| 01:38 |
|
prologic |
I kind of like Github's builtin editor |
| 01:38 |
|
prologic |
for quick 'n dirty changes |
| 01:40 |
|
pdurbin |
I like it but I have two email addresses registered with GitHub and I'd like to be able to pick and choose which one I made the online edit from. |
| 01:45 |
|
prologic |
ahh |
| 01:45 |
|
prologic |
you have multiple GH identities? |
| 01:46 |
|
pdurbin |
no, just the one. multiple email addresses associated with it. work email vs personal email |
| 01:46 |
|
prologic |
ahh right |
| 01:47 |
|
prologic |
at least you can route notifications to different email addresses per organization |
| 01:47 |
|
prologic |
which is nice |
| 01:47 |
|
pdurbin |
yeah, I'm thinking more of the commit history |
| 01:49 |
|
pdurbin |
oh, nice! custom routing at https://github.com/settings/notifications |
| 01:49 |
|
pdurbin |
described at https://github.com/blog/1204-notifications-stars |
| 01:54 |
|
prologic |
yeap :) |
| 01:54 |
|
pdurbin |
not quite what I was talking about but useful nonetheless :) |
| 01:58 |
|
prologic |
yeah I figured |
| 01:59 |
|
prologic |
somewhat related :) |
| 02:06 |
|
pdurbin |
prologic: for non-file-based DB aditsu is using postgres and I'd like to use postgres as well (right now I'm using derby) |
| 03:34 |
|
prologic |
hmm |
| 03:34 |
|
prologic |
I'm not really sure whether that's such a great idea ihmo |
| 03:34 |
|
prologic |
because it's a pretty large and complex dependency un and of itself |
| 03:34 |
|
prologic |
but if you guys are going down that path |
| 03:34 |
|
prologic |
I'll follow suit |
| 03:35 |
|
prologic |
Are either of you planning on using an ORM asw well |
| 03:35 |
|
prologic |
or straight DB API over SQL? |
| 07:03 |
|
aditsu |
complex? haha, just install postgres, create a db and run a script |
| 07:03 |
|
aditsu |
I'm using a database library that I wrote, it's not really an ORM but much more convenient than straight DB API |
| 08:12 |
|
prologic |
lol |
| 08:12 |
|
prologic |
no it's complex :) |
| 08:12 |
|
prologic |
okay maybe complex isn't the right word |
| 08:12 |
|
prologic |
hard(er) |
| 08:12 |
|
prologic |
anyway guess I'll fix up a few bugs with mine otnight |
| 08:12 |
|
prologic |
and think about a database |
| 08:12 |
|
prologic |
I'll pick postgres too |
| 08:12 |
|
prologic |
but ofc I'm going to do all this via Docker :) |
| 08:12 |
|
prologic |
(even though you *could* also do it byhand but uggh) |
| 10:25 |
|
pdurbin |
prologic: I'm using JPA as an ORM |
| 12:35 |
|
prologic |
so you’re all using ORM(s) |
| 12:35 |
|
prologic |
okay |
| 12:44 |
|
pdurbin |
I mean, we could define the spec a bit more: https://github.com/pdurbin/addressbookmvc/issues/4 |
| 12:44 |
|
pdurbin |
it's not like an ORM is required |
| 12:44 |
|
pdurbin |
some sort of persistence is required, I'd say |
| 12:44 |
|
pdurbin |
prologic: your use of a file for persistence may not scale very well :) |
| 13:24 |
|
prologic |
how will it not scale pdurbin ? :) |
| 13:29 |
|
pdurbin |
prologic: sorry. will it? |
| 19:41 |
|
pdurbin |
https://confluence.atlassian.com/display/JIRAKB/Java+8+Improves+JIRA+Performance+Dramatically |
| 20:40 |
|
aditsu |
me: "it's not really an ORM", prologic: "so you’re all using ORM(s)" :) |
| 21:02 |
|
prologic |
pdurbin, just curious as to why you think files on a file system don't/can't scale? :) |
| 21:03 |
|
prologic |
AFIK and IHMO it's more to do with concurrency and how many read/write operations you can support |
| 21:03 |
|
prologic |
right? |
| 21:03 |
|
prologic |
e.g: SQLite doesn't scale very well with writes; it blocks quite badly; but is quite fast for many reads |
| 21:19 |
|
pdurbin |
prologic: I assumed you were storing all the contacts in a single file. Should be fine if you're using multiple files. |
| 21:21 |
|
aditsu |
you'd still need indexes |
| 21:22 |
|
pdurbin |
yeah |
| 21:26 |
|
prologic |
yes you're quite right |
| 21:27 |
|
prologic |
as it grows in size searching won't scale without some kind of indicies |
| 21:27 |
|
prologic |
but not that hard to implement using files either and a library or two :) |
| 21:27 |
|
prologic |
e.g: https://pypi.python.org/pypi/Whoosh/ |
