=== schwuk_away is now known as schwuk
=== schwuk is now known as schwuk_away
=== schwuk_away is now known as schwuk
[11:36] <gary_poster> Hey rick_h.  Thank you for your reviews.  Made a reply because I'm still not resolved about the valueChange thing.  Tell me what you think.
[11:37] <rick_h> gary_poster: ok, looking now
[11:39] <rick_h> gary_poster: ignore me then. I could have sworn I twice opened all the collapsed sections of the diff and did a search in the file for that. 
[11:40] <gary_poster> rick_h, lol, cool, thanks.  will do the rest of the changes. :-)
[11:40] <rick_h> gary_poster: when I saw you added it I did a full page search looking for it, but I must have typo'd something stupid ... /me goes for more coffee
[11:40] <gary_poster> :-) np!
[11:57] <hatch> good morning
[12:04] <rick_h> morning hatch, early for you isn't it?
[12:04] <hatch> incredibly 
[12:04] <hatch> my flight was delayed so I got in late, jet lagged and all that
[12:06] <hatch> nothing like taxiing out to take off and they saying there has been an issue and they need to go back to the gate haha
[12:06] <rick_h> doh!
[12:08] <hatch> I'll be pretty much useless today so I put in a swap day request
[12:08] <hatch> rick_h, http://tympanus.net/Development/TripleViewLayout/ for the details tabview 
[12:09] <rick_h> nope, wtf is with that messing with the scroll bar?
[12:09] <rick_h> and you can't jump from first to last
[12:10] <hatch> the articles are different lengths
[12:10] <rick_h> right, but I can't click on the scrollbar to move up/down?
[12:10] <hatch> oh what the heck
[12:11] <hatch> well the UI is pretty cool, clearly this code woudln't fly hah
[12:11] <rick_h> :P
[12:12] <hatch> ok going to try and get some more sleep! I'll try and make it back for the call
[12:55] <benji> bac: I reviewed https://code.launchpad.net/~bac/charmworld/readme/+merge/184387 
[12:55] <bac> thanks benji
[12:55] <benji> bac: do you know anything about Huw's branch?
[12:56] <bac> on friday he said he had a few changes from review to incorporate but ran out of time
[12:56] <rick_h> benji: he didn't get it done. He ran into some issues on the build side I think. 
[12:56] <bac> benji: i assume he returned home, so i'm unsure when he'll get to it
[12:56] <rick_h> yea, when we talked friday I got the impression he'd start back on his Tues, our Monday night
[12:57] <bac> benji: which is a bit of a drag since i leaves my branch in limob
[12:57] <bac> limbo
[12:57] <benji> bac: yeah
[12:58] <benji> bac: your head is on "Add lp branch link to API results" are you really looking forward to that one or can I take it real quick
[12:59] <gary_poster> bac, is there anything you can do to land Huw's thing as a temporary step?  IOW is it reasonable to take over, and then land your branch?
[12:59] <gary_poster> jujugui, I need a second review of https://codereview.appspot.com/13385045/ when someone can
[13:00] <benji> gary_poster: I'll take a look
[13:00] <bac> gary_poster: sure i can do that.  i think he should be mighty close.  not sure what kind of build issues he saw as the mods to get the makefile running were trivial.  makes me wonder if i'm missing something.
[13:00] <gary_poster> thank you benji
[13:00] <gary_poster> bac cool.  just an idea to unblock you.
[13:12] <adeuring> abentley: I'm wokring on the exodus for fixed ES mappings. "bin/migrations prepare-upgrade" works fine, but when I run "bin/migrations opgrade", I get this error http://paste.ubuntu.com/6083418/
[13:13] <abentley> adeuring: looking.
[13:15] <bac> benji: the css handles the markdown types with extensions of 'md', 'markdown', or 'mkd'.  everything else is treated as plaintext. i think we're making up the rules as we go along so we could limit the extension to those three plus '' or '.txt'
[13:16] <benji> bac: sounds good; I assume markdown will render plain text in a reasonable way (or are we rendering it at all?)
[13:16] <bac> benji: yes, the md is rendered very nicely
[13:16] <rick_h> bac: I think there was talk of having the proofing tools limit to md? and it's noted as policy in https://juju.ubuntu.com/docs/authors-charm-policy.html
[13:16] <bac> thanks rick_h
[13:17] <rick_h> bac: benji we copied the JS markdown -> html converter from charmworld if I recall
[13:17] <bac> rick_h: i wonder if they'll want a bundle proofing tool?
[13:18] <rick_h> bac: yea, we should bring that up to marcoceppi_ and jcastro as I think it makes sense. 
[13:18] <rick_h> especially as they might 'grow' in the next release
[13:18] <abentley> adeuring: I think the problem is that on your box, "charms" is a real index, while on production machines, it's an alias of an index.
=== marcoceppi_ is now known as marcoceppi
[13:18] <bac> rick_h: that page says .md, .txt, or .rst.  we don't support nice display of .rst
[13:18] <bac> for charms or bundles
[13:18] <adeuring> abentley:  ah, thanks.
[13:18] <rick_h> bac: no, we don't support rst->html but it was a requested feature. Just low priority and never got to it
[13:19] <marcoceppi> rick_h: bac: how aggressive of a linting tool?
[13:19] <rick_h> marcoceppi: yea, that's the question I guess. Go through bundles and defined what we want to make sure is there. Right now it's just what, the deployer file, icon, and readme?
[13:20] <bac> marcoceppi: probably just to verify the expected files are present and the bundles.yaml file is parsable yaml
[13:20] <bac> rick_h: that is correct
[13:20] <jcastro> I am fine with whatever guys
[13:20] <rick_h> marcoceppi: yea, +1 on bac there. And then we could grow it as required. Though I think it'd be preferable to have one lint/proof tool 
[13:20] <marcoceppi> rick_h bac: that sounds good to me. I only ask because there are some applications that add additional fields to the deployer that can be ignored but shouldn't prevent it from being used
[13:20] <bac> bundles.yaml, icon.svg, and some sort of readme.[md, txt, '']
[13:20] <jcastro> ~charmers are doing .md for most things
[13:21] <jcastro> but we have some .rst and .txt in there. 
[13:21]  * marcoceppi pushes for everything markdown
[13:21] <sinzui> We are going to switch everything to md though?
[13:21] <jcastro> no
[13:21] <jcastro> some people like hazmat prefer rst, but I didn't care enough to argue
[13:21] <marcoceppi> I think it should be suggested as the default format, but not required
[13:21] <jcastro> I prefer everything in the project be markdown. :)
[13:21] <sinzui> jcastro, I think I gave you a list of the rst files. I thought your intent was to change to md
[13:22] <rick_h> marcoceppi: sure, when you create a template for a new one start with a .md file 
[13:22] <jcastro> sinzui: I changed them all I think
[13:22] <jcastro> but like, we don't tell people "your readme is RST you suck"
[13:22] <rick_h> I'd bet nearly everyone sticks with what came 'out of the box'
[13:22] <jcastro> if we were to say "use markdown for everything" I would also not mind
[13:22] <marcoceppi> rick_h: did you want the linting tool to live in charm-tools, or elsewhere?
[13:22]  * marcoceppi can imagine a charm create --bundle command
[13:22] <bac> jcastro: right now, the only format we display in a rich rendering is md
[13:22] <jcastro> bac: it seems to me that you guys have better things to do than support RST. :)
[13:22] <rick_h> marcoceppi: well I'd think that the same proof/lint tool can detect a charm vs a bundle and check appropriately?
[13:23] <marcoceppi> rick_h: I would think so, yes
[13:23] <rick_h> marcoceppi: so it'd be the same proof tool we use now
[13:23] <bac> jcastro: absolutely
[13:23] <jcastro> I could just propose we do markdown-only for everything?
[13:23] <rick_h> marcoceppi: whereever that lives
[13:23] <marcoceppi> rick_h: charm-tool
[13:23] <marcoceppi> charm-tools*
[13:23] <jcastro> hey sinzui, can you confirm that we did in fact move every README to md? 
[13:23] <jcastro> I am 95% sure I did that already
[13:23] <jcastro> but it wouldn't hurt
[13:24] <sinzui> jcastro, I can. I lost the query when staging when tits up. I need to recreate it
[13:25] <jcastro> I'd be happy to finish converting them if there's any stragglers
[13:25] <jcastro> I am like president of the pure Markdown society
[13:28] <rick_h> marcoceppi: so should I file both the bundle proof and bundle --create as bugs in charm-tools?
[13:29] <marcoceppi> rick_h: yes please! if you could also include the directory structure of what is needed in a bundle for either (or an example of a bundle) I can make sure the proper templates are there
[13:29] <rick_h> bac: do we have a sample bundle charm that's in the store I can link to?
[13:31] <gary_poster> benji, lol, thanks, will address :-)
[13:31] <benji> :)
[13:31] <bac> rick_h: yes, search for 'precise' and there are two at the bottom
[13:32] <sinzui> jcsackett, hangout?
[13:33] <rick_h> bac: oooh, sweet
=== hazmat` is now known as hazmat
[13:43] <gary_poster> I have another branch--very small--ready for review by one person: https://codereview.appspot.com/13621043
[13:45] <benji> gary_poster: I guess I can put off getting a cup of coffee for a couple of minutes
[13:46] <gary_poster> benji, heh, ok thanks
[13:49] <hatch> has anyone ever had a deepEqual fail but not report the diff?
[13:51] <rick_h> hatch: normally the diff is unsuable for me
[13:51] <hatch> yeah this is really irritating me
[14:09] <benji> bac: your head is on "Add lp branch link to API results" are you really looking forward to that one or can I take it real quick
[14:10] <bac> benji: off with his head
[14:10] <benji> heh
[14:10] <bac> my head
[14:13] <bac> sinzui, abentley: do you know what tarmac does if i have an approved MP with a dependency on a pre-req branch that has not landed?
[14:14] <abentley> bac: It complains, and I believe it marks the MP "need-review".
[14:15] <abentley> bac: You have to land the prereq first.  Or you can do a new MP that includes all branches and land that.
[14:16] <bac> abentley: i wonder which huw would consider ruder?  :)
[14:16] <bac> the cleanest thing would be for me to cause his branch to get landed...
[14:18] <abentley> bac: If his branch shouldn't land in its current form, you could make changes and then "re-submit", and use your own branch as the source branch.
[14:23] <bac> abentley: could i just resubmit my MP removing his as a pre-req?  it would then land with both branches included?
[14:24] <abentley> bac: Yes, you could.  That's like doing a new MP that includes all branches.
[14:25] <bac> abentley: that seems easiest.  thanks.
[14:29] <hatch> rick_h, FINALLY found the issue
[14:29] <hatch> ugh
[14:33] <rick_h> hatch: oh, I thought it was working as designed. I've not seen it work better
[14:33] <hatch>       // when doing a deep equals on an object which contains a LazyModelList
[14:33] <hatch>       // call toArray() on it first else it will fail and it won't tell you why.
[14:33] <hatch>       attrs.relations = attrs.relations.toArray();
[14:33] <hatch> rick_h, ^^ there are my notes on it
[14:53] <bac> benji: i've resubmitted my MP.  the one at https://code.launchpad.net/~bac/charmworld/readme/+merge/184590 contains huw's work, my changes, my fixes to huw's branch, and my fixes in response to your original review.  sadly it is huge
[14:54] <benji> bac: I think that is the right way to go, despite the sadness.  I'll undertake a review soonish.
[14:56] <abentley> bac: Could you ping me when your MP goes to "Approved"?  I am hacking on a new Jenkins for you.
[14:56] <Makyo> Oops, reboot for updates.
[14:57] <bac> abentley: will do
[15:03] <abentley> sinzui: Could you please add juju-gui-bot to ce-orange-squad so that it can download the credentials?
[15:04] <sinzui> abentley, done
[15:04] <abentley> sinzui: Thanks.  Confirmed working.
[15:12] <rick_h_> irc went boom, back now
[15:42] <benji> bac: the diff at https://code.launchpad.net/~bac/charmworld/readme/+merge/184590 doesn't look right
[15:44] <bac> benji: in what way?
[15:45] <benji> bac: it seems to be missing some of Huw's changes, particularly the template changes
[15:46] <benji> bac: "The diff has been truncated for viewing"
[15:46] <benji> that's what got me, once I downloaded the diff I see what I was expecting now
[15:47] <bac> ok
[15:50] <gary_poster> jujugui call in 10
[15:50] <rick_h_> I miss all of our devices ringing off at the same moment
[15:54] <hatch> do you? do you really?
[15:54] <hatch> *TRAIN SOUND*
[15:59] <gary_poster> jujugui call in 2
[16:01] <gary_poster> bcsaller, call now
[16:16] <hatch> rick I'll need you to pm me the link
[16:16] <hatch> I'm not on my typical acct right now
[16:16] <hatch> rick_h_, ^
[16:16] <rick_h_> hatch: rgr
[16:22] <luca__> can someone explain scale up and scale out?
[16:47] <hatch> bcsaller, can you hop into our call?
[16:47] <rick_h_> bcsaller: ping, got a sec to chat databindings?
[16:47] <hatch> hah
[16:47] <bcsaller> yeah
[16:47] <rick_h_> bcsaller: https://plus.google.com/hangouts/_/1f48bd83e28661eebe401649dbb7d4bfa8666331?hl=en
[16:54] <hatch> ok I'm stepping away from the computer, just ping if you need and I'll get back as soon as I can
=== schwuk is now known as schwuk_away
[17:15] <benji> bac: I'm finally done with https://code.launchpad.net/~bac/charmworld/readme/+merge/184590
[17:15] <benji> I included a CSS patch to fix a small issue.
[17:15] <bac> thanks benji.  i'll look at it when i finish my cake
[17:17] <benji> I'll eat my cake when I finish driving my Lamborghini to get my free massage.
[17:54] <Makyo> Our hacking file seems a little stale.  lbox, python-gf, and lags aren't found in standard repos or the PPAs specified.
[18:07] <bac> thanks for the patch benji.  it needs to be applied to bundle.pt too
[18:08] <benji> ah; I'm glad you knew that
[18:08] <bac> benji: the cake was superb, too.  from a gas station near the outer banks.
[18:09] <benji> being from Tennessee, it would be unsafe to try gas station cake
[18:10] <bac> abentley: i'm about to mark my MP 'approved'.  stand back.
[18:11] <bac> benji: the techroline adds a special something
[18:11] <benji> heh
[18:37] <abentley> bac: Sorry, was lunching.
[18:38] <bac> abentley: it just got rejected due to unreviewed revisions.  i'll fix that now and mark it approved
[18:46] <gary_poster> *just* over 400 lines, so two reviewers needed: https://codereview.appspot.com/13477044
[18:46]  * gary_poster goes to get some lunch
[18:51] <bac> hey abentley, how do i get around this problem?  have huw added to the team?  https://code.launchpad.net/~bac/charmworld/readme/+merge/184590
[18:53] <abentley> bac: I guess so.  I haven't seen that before.  Maybe ask rick about it, since they've collaborated on branches, I think?
[18:53] <bac> hey rick_h_ ^^
[18:53] <rick_h_> looking, I read the error that same way. 
[18:54] <rick_h_> when I did stuff though I'd just take his branch, pull it down, push my own new one, and there was only one branch, no pre-req
[18:54] <rick_h_> bac: from the sounds of things you labeled his as a pre-req branch vs just taking his branch over?
[18:54] <bac> rick_h_: basically the same here.  no pre-req on this MP
[18:55] <bac> rick_h_: originally, then i resubmitted without the pre-req
[18:55] <rick_h_> bac: hmm, no idea then. I never hit this when I'd bzr colo-branch my-copy && bzr merge lp:huw's copy && bzr push ~rharding/my-copy && lbox submit
[18:55] <bac> but he'll still be an author on an unmerged revision
[18:56] <bac> rick_h_, abentley: at this point perhaps i should just manually push lp:charmworld
[18:57] <abentley> rick_h_: did you always merge his branch into yours, never branch from his branch as a starting point?  I suspect that Tarmac only looks at mainline revisions.
[18:58] <rick_h_> abentley: correct, I'd always create my own branch, pull his changes in, and push it was my own branch. 
[18:58] <rick_h_> abentley: it would end up marking his branch as merged and seemed to work so never thought to do it another way
[18:58] <abentley> rick_h_: pull his changes in or merge them in?  Big difference in how it looks in "bzr log".
[18:59] <rick_h_> abentley: bzr merge lp:~huw....
[18:59] <abentley> rick_h_: Cool.
[18:59] <rick_h_> bac: so I'd just do that. Create a fresh branch from trunk, merge in the current work you've got there. Self review ok, link to the old review/comments, and try to submit that. 
[19:00] <rick_h_> bac: unless you're sure it's safe/testing is going to be ok. 
[19:00] <abentley> rick_h_: Good point.  That should work around it.
[19:02] <abentley> bac: pushing is an option, and I've done that before, so I'd be a hypocrite to tell you you can't.  But I've also screwed things up by pushing, even though I was trying to be cautious.
[19:02] <bac> abentley: yeah, i'd rather let the machinery do the work.  i'm following rick_h_'s suggestion to see what happens.
[19:06] <bac> ok, tarmac should be chewing on it soon
[19:06] <rick_h_> woot
[19:06] <bac> no whammies!
[19:07] <rick_h_> hah, I always think that stuff
[19:07] <rick_h_> make lint ... 'no whammies! "boom!"'
[19:12] <abentley> sinzui: AFAICT, Launchpad is lying to my trigger script.
[19:12] <abentley> sinzui: My browser gets a total_size of 1, but my trigger script gets 0, when I access https://api.launchpad.net/devel/~juju-jitsu/charmworld/trunk?ws.op=getMergeProposals&status=Approved
[19:15]  * sinzui is looking
[19:16] <abentley> sinzui: The branch has landed now, so a total_size of 0 is now correct.
[19:16] <sinzui> understood.
[19:17] <sinzui> As long as non-private branches are involved, anon will be fine. The Omnibus and security branches are the only MP factors that should be a problem 
[19:17] <bac> abentley, rick_h_: same result -- https://code.launchpad.net/~bac/charmworld/readme-mega/+merge/184661
[19:18] <bac> tries something else
[19:21] <sinzui> abentley, looking at the juju-project, I think I get sane results. https://api.launchpad.net/devel/juju-project?ws.op=getMergeProposals&status=Approved Have you eliminated timing and caching from scenario?
[19:22] <abentley> sinzui: No, I have not eliminated caching from this scenario.  I think it is very likely that caching is the culprit.
[19:23] <abentley> sinzui: But I think a minutes-long cache on an API is Just Wrong.
[19:25] <abentley> bac: I've looked at the tarmac code, and I see that considers the author of every revision.  I don't understand how this worked for rick, since huw would have been the author of some revisions.
[19:26] <sinzui> abentley, I don't see any cache rules in the api response. Squid could be caching.
[19:27] <abentley> sinzui: Maybe it was Squid.  Something has to be caching; I was consistently getting one response in my browser and a different response from my script.
[19:28] <sinzui> abentley, non-anonymous wont be cached. I think we always assume authenticated users are working with non-public mutable information
[19:29] <abentley> sinzui: I was hitting reload in my browser, so it would have been sending cache-disabling headers.  But both were accessing the same URL.
[19:29] <sinzui> ah, right
[19:30] <bac> abentley: in that case, for the long term the best solution is to add huw to the team
[19:31] <sinzui> Yes, we want how added to the team.
[19:32] <abentley> bac: Sounds like it.  I don't plan to reproduce that behaviour in Jenkins.
[19:32] <bac> sinzui: can you add him to the orangemen?
[19:33] <sinzui> Lp's franken-lander required people to be a member of a team composed of drivers, former drivers, and people listed on a wiki page who signed an agreement
[19:33] <sinzui> I will
[19:34] <abentley> sinzui: So for copyright assignment reasons, we might want to restrict it to ~canonical for now.
[19:34] <sinzui> abentley, yes, that is easiest
[19:42] <sinzui> bac: abentley I added huwshimi to ~ce-orange-squad to get him into ~juju-jitsu
[19:42] <bac> yay you
[19:42] <sinzui> I really don't understand how his work was merged and landed before? Maybe he was removed from a team?
[19:43]  * bac tries again.  crosses fingers.  fetches beverage.
[19:43] <bac> sinzui: only explanation i can think of
[19:44] <sinzui> oh, bac, abentley: I don't think he has any contributions to charmworld. Most of his work with orange was juju-gui
[19:44] <abentley> sinzui: Oh.
[19:50] <sinzui> abentley, I see I can append bogus args to the api query. maybe we can append a timestamp to ensure non-cached content: &_ignore=1234
[19:52] <abentley> sinzui: I'm investigating using PS's jenkins-launchpad plugin now.  It seems to be under heavy use and active development.
[19:52] <sinzui> goody
[19:54] <gary_poster> thank you very much for review and qa rick_h_.  I'm inclined to treat it as a ~400 liner. :-) Are you comfortable being the only reviewer on that one, or should I pursue another?
[19:55] <gary_poster> given your general preference alone, I'm expecting "another" :-)
[19:55] <bac> sinzui, abentley, rick_h_: after adding huw to the team the original mega branch landed.
[19:56] <sinzui> good.
[19:56] <gary_poster> thank you for moving that forward, bac.  cool.
[19:56] <bac> gary_poster: moved it forward inches at a time
[19:56] <gary_poster> :-)
[20:11] <benji> bac: I made it harder than it needed to be before realizing the error of my ways, does this http://bazaar.launchpad.net/~benji/charmworld/add-lp-branch/revision/390?start_revid=390 satisfy the "Add lp branch link to API results" card, as far as you know?
[20:14] <bac> benji: do we have a test that actually shows what is being returned for the API call?
[20:15] <benji> I'll look.
[20:17] <benji> bac: yep: charmworld/views/tests/test_api.py:test_results_match
[20:18] <bac> cool
[20:18] <bac> such a descriptive name.  bet i wrote that.  :)
[20:23] <benji> heh
[20:39] <rick_h_> gary_poster: sure thing from here. 
[20:39] <gary_poster> cool thanks rick_h_ 
[20:40] <benji> bac: the easiest review you'll do this month: https://code.launchpad.net/~benji/charmworld/add-lp-branch/+merge/184680
[20:41] <bac> fastest too
[20:42] <hatch> hey my fitbit started working again
[20:42] <hatch> of course after all the walking in London hah
[20:42] <hatch> rick_h_: how's the thing coming?
[20:43] <rick_h_> hatch: it's coming, EOD now and will try to get things finished tomorrow. 
[20:43] <hatch> coolio
[20:44] <rick_h_> hatch: according to my fitbit, london was a 29.99mi week for me 
[20:45] <rick_h_> sun - friday, guess sat was another 2
[20:45] <hatch> haha nice and I probably walked more because of the days on either end....so yikers!
[20:45] <rick_h_> hatch: yea, that doesn't count the biking days really
[20:45] <hatch> so back now to sitting on my a** all week
[20:46] <hatch> lol
[20:46] <rick_h_> hatch: wheee!
[20:46] <hatch> haha, actually this morning I walked to get some groceries....had to keep at least SOME activity going :)
[20:48] <hatch> I should really get you and gary on my fitbit
[20:48] <hatch> not that I'll ever be able to get close to gary haha
[21:16] <gary_poster> Short review adding some nice UX cleanups: https://codereview.appspot.com/13368051
[21:16] <gary_poster> I only need 1 person (< 200 lines!)
[22:27] <huwshimi> Morning