18:02:59 <dbite> #startmeeting training-labs
18:03:00 <openstack> Meeting started Mon Aug  1 18:02:59 2016 UTC and is due to finish in 60 minutes.  The chair is dbite. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:03:01 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
18:03:03 <openstack> The meeting name has been set to 'training_labs'
18:03:09 <julen> hi dbite
18:03:11 <dbite> hello
18:04:12 <dbite> julen: Lets discuss your patch :)
18:04:24 <dbite> hopefully Roger joins us soon
18:04:40 <julen> should we wait for him?
18:06:43 <dbite> Not sure if he is joining us today. I guess he is busy today :)
18:07:19 <julen> well.. we can always go on and he reads the logs whenever he joins
18:07:29 <dbite> #frontend testing
18:07:36 <dbite> I agree
18:07:43 <dbite> #topic frontend testing
18:07:44 <julen> by the way... how do you read the log before the meeting is finished?
18:08:11 <dbite> ohh, there is a link for that, just a second let me fetch it
18:10:39 <dbite> http://eavesdrop.openstack.org/meetings/training_labs/2016/
18:11:02 <dbite> Anything from your side for this patch?
18:11:21 <dbite> http://eavesdrop.openstack.org/meetings/training_labs/2016/training_labs.2016-08-01-18.02.log.txt
18:11:26 <julen> yes
18:11:42 <dbite> Sorry, this should be the right link, just hit refresh and it should give you updated logs :)
18:11:51 <dbite> So, tell me about the patch
18:12:05 <julen> (I didn't know those files were written on the fly :) )
18:12:13 <julen> so... so far so good
18:12:20 <dbite> Ok,
18:12:45 <dbite> Sorry I have been zoned out in the last couple of days. I will review your patch asap (mostly tomorrow :))
18:12:47 <julen> for the first steps, the only challenge was to find out the syntax of the POST data to pass either wget or curl
18:13:37 <julen> it works fine with test-once.sh or running it directly from anywhere
18:14:00 <dbite> Ok, I guess this should technically run independently :)
18:14:02 <julen> now, the challenge I'm facing is dealing with the javascript part for starting the instance
18:14:16 <dbite> julen: I have a few nitpicks in there. Ill publish the comments after the meeting.
18:14:28 <dbite> go on.
18:15:17 <julen> so, I took a look to the source code of horizon and I got a little lost when dealing with the javascript part
18:15:53 <dbite> hmm, the JS code from the browser you mean?
18:15:56 <julen> my best attempt is commented out at the end of the script I uploaded, if someone wants to take a look...
18:16:01 <dbite> Or from the repository?
18:16:07 <julen> browser
18:16:22 <rluethi> hey all, sorry for being late, was travelling
18:16:37 <dbite> How about keeping the scope of the first commit to just login and confirm that Horizon is working?
18:16:44 <dbite> rluethi: No worries :)
18:16:49 <dbite> Good to see you
18:17:11 <julen> when filling in the data for the instance (name, flavor, network, ...) and clicking in Launch, it submits some post data, but I think it's not enough to be reproduced with wget
18:17:15 <rluethi> dbite: I agree, I'd be happy to have a limited test, it would have been useful to us before.
18:17:20 <julen> hi rluethi! :)
18:17:49 <rluethi> hi julen, I saw your updated patch.
18:18:23 <julen> rluethi: thanks for the comments! :)
18:18:36 <julen> rluethi: and? is it ok now?
18:19:24 <rluethi> julen: I think it is okay. There are some style issues, but they are not hard requirements.
18:20:03 <rluethi> julen: for instance, your code doesn't use the standard headers all our other client-side scripts use.
18:20:28 <julen> rluethi: do you mean, the "source"s?
18:20:33 <rluethi> julen: our code is also more verbose and tells the user much more about what it is currently doing.
18:20:52 <rluethi> julen: yes, and the setopt and all that jazz.
18:21:17 <rluethi> julen: you hardcode things all the other scripts have in one central config file.
18:21:20 * dbite Jazz!!!
18:21:43 <rluethi> dbite: chill, dbite :)
18:21:58 <dbite> julen: I have added some inline comments for the same. Its better not to hardcode. If the user changes the IP/Subnet then it becomes a nightmare for the user.
18:22:25 <dbite> rluethi: Yes :)
18:22:25 <julen> dbite: true, that's a good point
18:23:54 <julen> can I put some conditional, so that if the osbash stuff is found, it uses the controller address, and if not, uses by default?
18:23:59 <rluethi> julen: there's a ton of details I don't want to annoy you with. We try to follow style guides used by other OpenStack projects (devstack, openstack-manuals) to provide a familiar look for other contributors.
18:24:01 <julen> this would make it more standalone
18:24:49 <dbite> julen: Nope, you could still make it stand alone without hardcoding
18:24:55 <julen> rluethi: I would be happy to learn
18:25:07 <rluethi> julen: it is a test-script for a training-labs specific setup, I am not sure how much being more standalone can actually buy us.
18:25:12 <dbite> sourcing the configuration files does not create dependencies :)
18:26:19 <rluethi> julen: for instance, note how we echo what we are about to do, e.g. "Configuring file X.", using that tense, ending with a period.
18:26:27 <dbite> julen: We should ship this script as a part of training-labs
18:26:45 <rluethi> julen: these are details, but our scripts are pretty consistent in this regard as well.
18:27:01 <julen> I was thinking, of having the possibility of copying the file to the compute1 and executing it from there, without config files. But well.. There's not likely many cases where someone would do that...
18:27:03 <rluethi> dbite: absolutely, it is a very useful regression test.
18:27:49 <rluethi> julen: we should able to run it on the host as well (provided it is Linux or OS X).
18:27:51 <dbite> julen: Just imagine a real deployment, why would the horizon dashboard be on the same network and accessible from compute node?
18:28:02 <rluethi> julen: at least I don't see I reason right now why it shouldn't.
18:28:06 <julen> rluethi: hmm.. never thought about the format of the comments before
18:28:44 <julen> dbite: good point
18:29:29 <rluethi> julen: or there is this bit "[ -n "$FOO" && do_something "]; we always use "if [ -n "$FOO" ]; then\n    do_something".
18:29:47 <dbite> rluethi: julen Are you guys satisfied with this? I guess we should take this off over email or docs channel after the meeting :)
18:29:57 <rluethi> julen: it's not better, it's just a convention to make code look more consistent.
18:30:07 <rluethi> no issues.
18:30:14 <rluethi> we can move on.
18:30:39 <dbite> julen: As of thinking about the JS debugging/understanding, I think we should try to search in the source of horizon's repository over the web console.
18:31:11 <dbite> next topic :)
18:31:33 <julen> rluethi: sure... I just never thought about those things. If you help me spot those little issues, I definitely work on them
18:32:36 <dbite> julen: thats part of learning to becoming a pragmatic programmer ;_
18:32:40 <dbite> *;)
18:32:50 <dbite> #topic parser
18:32:59 <julen> dbite: I tried with tcpdump and firefox extensions, but didn't manage to find out how it goes
18:33:58 <dbite> julen: Ill try to investigate the source of horizon tomorrow in the day and see if we could get some good clues in there :). Also asking the devs directly would save a lot of time. Just shoot an email over openstack-dev ML
18:34:26 <dbite> rluethi: My apologizes I havent been able to cleanup and push my current local work for the parser. I have full priority for it this week.
18:34:48 <dbite> rluethi: Lost track of time last week and its already Monday.
18:34:53 <rluethi> dbite: great. I pushed a new version of the python port.
18:34:54 <dbite> :|
18:35:05 <dbite> rluethi: Awesome, lets move on to python port then
18:35:09 <dbite> #topic python port
18:35:37 <dbite> rluethi: I need to take another look at the python port. Did you manage to do cool git stuff finally?
18:36:10 <rluethi> dbite: not sure about "cool", but I nuked the old branch and pushed a new, updated one.
18:36:45 <rluethi> the port is not done, but it has almost caught up (again) with osbash.
18:36:51 <dbite> rluethi: -f to the rescue, lets have a pair programming session sometime soon
18:37:00 <dbite> julen: Ofcourse, Ill invite you too :)
18:37:20 <julen> dbite: I'm in :)
18:37:26 <rluethi> dbite: sure
18:37:52 <dbite> rluethi: Ill link you to a video which explains why it is really important to preserve the history of your source
18:38:12 <dbite> even if its bad, buggy or there are mistakes in there ... but I am sure you know this already ;)
18:38:55 <rluethi> dbite: ...says the guy who kept posting new histories for his PoCs. That's rich coming from you :).
18:40:03 <dbite> rluethi: I could be sarcastic ;) ... but its not for the repositories where others are working with me. And ofcourse I am not perfect, I am learning these finer details in the last few months ...
18:40:24 <dbite> Any interesting updates?
18:41:16 <rluethi> dbite: I've been busy with the python port and fixing the Windows batch files. That's all.
18:44:05 <dbite> rluethi: Ok, lets move on then
18:44:14 <dbite> #topic any other buisness?
18:44:22 <dbite> Anything else?
18:44:49 <dbite> From my side there are some house keeping/boiler plate work which I will be carrying out this week.
18:45:13 <dbite> Additionally I have full focus on training labs for the next couple of months.
18:45:34 <julen> dbite: cool :)
18:45:39 <rluethi> any progress on the mascot?
18:46:12 <dbite> rluethi: Not yet. Should take another look, I got lost into different emails for the same :|
18:46:20 <dbite> I guess, thats the housekeeping work ;)
18:47:26 <dbite> rluethi: julen Lets call it a day then :)
18:47:38 <rluethi> dbite: k
18:47:41 <dbite> Additionally I plan to start attending the Docs meeting and install guides meeting.
18:47:42 <julen> dbite: agree
18:47:55 <dbite> #endmeeting