=== uru_ is now known as urulama
[06:50] <rogpeppe1> mornin' all
[07:33] <urulama> robbiew: morning
[07:33] <urulama> robbiew: sorry ... meant rogpeppe1
[07:33] <rogpeppe1> urulama: yo!
[07:34] <urulama> rogpeppe1: i might have empty slot and continue reviewing your PR so that you're not blocked
[07:34] <rogpeppe1> urulama: thank you v much. if it's too much, i will split it up like i should have done in the first place :-)
[07:35] <urulama> rogpeppe1: i just have internal/v4/stats.go and tests, no problem
[07:59] <rogpeppe1> frankban: hiya
[07:59] <rogpeppe1> urulama, bac, jrwren, frankban: https://github.com/juju/charmstore/pull/53
[08:01] <frankban> rogpeppe1: morning, on it
[08:15] <rogpeppe1> frankban: thanks
[08:16] <rogpeppe1> "Nice! It's great that go stdlib supports this out of the box."
[08:16] <rogpeppe1> frankban: it wasn't immediately obvious how to take advantage of it for our use case...
[08:19] <frankban> rogpeppe1: yeah, I see you required to mimic a FileSystem (and a File)
[08:19] <rogpeppe1> frankban: yeah.
[08:22] <frankban> rogpeppe1: done
[08:22] <rogpeppe1> frankban: ta!
[08:25] <rogpeppe1> frankban: and os.FileInfo...
[08:26] <frankban> rogpeppe1: indeed
[09:28] <rogpeppe1> frankban: fancy pairing on some charmstore stuff?
[09:30] <frankban> rogpeppe1: sounds good, IIRC this week I will be working on the charmstore
[09:30] <rogpeppe1> frankban: cool
[09:30] <rogpeppe1> frankban: also, i'd quite like a review of https://github.com/juju/charmstore/pull/52 but i realise it's more sprawling than it should be
[09:30] <rogpeppe1> frankban: https://plus.google.com/hangouts/_/canonical.com/gogogo?authuser=1 ?
[09:32] <frankban> rogpeppe1: I'm there
[10:03] <urulama> rogpeppe1: sorry, got sucked up in meetings ... no time for review till evening :(
[10:03] <rogpeppe1> urulama: np. frankban is on it.
[10:03] <urulama> rogpeppe1: ok, great, it can proceed! *happy now*
[10:03] <rogpeppe1> urulama: :-)
[10:03] <rogpeppe1> urulama: how's it going?
[10:03] <urulama> intense
[10:04] <urulama> but progress being made
[10:04] <rogpeppe1> cool
[10:04] <urulama> identity, authorization, different core parts, billing ... a lot of issues at least being started on and documented and agreed on initial plan
[10:30] <frankban> rogpeppe1: review done
[10:30] <rogpeppe1> frankban: brill, thanks
[10:31] <rogpeppe1> frankban: am just making the changes
[10:31] <frankban> rogpeppe1: cool
[11:28]  * frankban lunches
[12:30] <jrwren> rogpeppe1: I'll rev you if you rev me: https://github.com/CanonicalLtd/charmstore-charm/pull/1
[12:30] <jrwren> rogpeppe1: j/k, I'll rev you either way.
[12:38] <rogpeppe1> jrwren: looking
[12:44] <frankban> rogpeppe1: ready when you are
[12:44] <rogpeppe1> frankban: just reviewing jrwren's PR
[12:45] <frankban> rogpeppe1, jrwren: I also need reviews for https://github.com/juju/charmstore/pull/54
[12:59] <rogpeppe1> jrwren: reviewed
[13:04] <jrwren> Is +1 an alias for LGTM ?
[13:10] <rogpeppe1> jrwren: i think so
[13:11] <rogpeppe1> frankban: reviewed
[13:20] <rogpeppe1> frankban: hmm, the 'bot doesn't look too happy: http://ci.jujugui.org:8080/job/charmstore-merge/
[13:46] <kadams54> rick_h__: so… are there any exceptions to having two reviews? I just finished QAing a PR from huw that is just an updated image file…
[14:10] <rogpeppe1> kadams54, rick_h__: i also have a bit of a problem with the two review thing. when i've been pairing with frankban, obviously i can't get a review from him, but then there's only one other available reviewer.
[14:10] <rogpeppe1> jrwren: https://github.com/juju/charm/pull/35
[14:13] <rogpeppe1> well, with jenkins off line, i guess we can't land anything anyway, so it doesn't make much difference
[14:17] <jrwren> jenkins is down again?
[14:18] <jrwren> ugh
[14:20] <rogpeppe1> jrwren: yeah
[14:20] <jrwren> looking
[14:40] <jrwren> rogpeppe1 and all: jenkins it building again
[14:40] <rogpeppe1> jrwren: lol. we've just bypassed it :-)
[14:41] <jrwren> Hopefully your bypass doesn't crash the lander.
[14:41] <frankban> rogpeppe1: http://ci.jujugui.org:8080/job/charmstore/60/
[14:51] <Makyo> jujugui call in 9
[14:59] <hatch> jujugui call now
[15:01] <Makyo> jcsackett, you in today?
[15:10] <hatch> I think that huws cards in landing have landed
[15:10] <frankban> rogpeppe1: gogogo in a minute
[15:10] <rogpeppe1> frankban: am back in gogogo
[15:11] <hatch> frankban there is a card in Project 1 with my head on it about adding a 'deleted' flag to machines and containers.....didn't you already do this?
[15:13] <frankban> hatch: no
[15:14] <hatch> frankban you just did services and units? 
[15:14] <hatch> I'm trying to decide what card to pick up next
[15:14] <frankban> hatch: I didn't add a deleted flag to anything
[15:14] <hatch> wasn't there something you did about removing services? 
[15:15] <frankban> hatch: if you are going to do that, I'd suggest to also add a isGhost flag
[15:15] <frankban> hatch: it was about renaming services, not removing them
[15:15] <frankban> hatch: we discussed about a possible solution for entities removal, and we agree don the db flags
[15:15] <hatch> ohh ok so the removal stuff wrt the ecs is still all not done
[15:16] <frankban> hatch: AFAICT yes
[15:28] <hatch> frankban ok thanks
[15:42] <hatch> Makyo Did you forget to add a release tag?
[15:42] <hatch> :)
[15:42] <hatch> hi luca 
[15:42] <luca> hi hatch 
[15:43] <Makyo> hatch, https://github.com/juju/juju-gui/releases/tag/1.1.1
[15:43] <hatch> interesting, I don't have that tag.....hmmm
[15:43] <luca> hatch: sup
[15:43] <hatch> doh I forgot to fetch
[15:43] <hatch> carryon....sorry :)
[15:44] <hatch> luca oh just workin away on this darn machine view.....wish it would just be done already
[15:44] <luca> hatch: me and you both bud
[15:44] <luca> hatch: me and you both
[15:44] <hatch> wish the designs would stop changing for a day or so....
[15:44] <hatch> :P
[15:44] <hatch> lol
[15:44] <luca> hatch: rofl
[15:44] <luca> hatch: don’t make me send another revision!
[15:45] <hatch> hahaha, "make you"? You mean you don't already HAVE another ready to go?
[15:45] <hatch> lol
[15:49] <hatch> jcsackett didn't you do something wrt the uncommitted units recently? https://bugs.launchpad.net/juju-gui/+bug/1352973 If you're still in that area you might want to see this bug
[15:52] <jcsackett> hatch: i've moved on, but i can tackle that next--can you assign me to the card?
[15:52] <hatch> sure will do
[15:53] <jcsackett> hatch: thansk.
[15:53] <jcsackett> er, thanks.
[15:53] <hatch> jcsackett there is a util method for taking that id and getting the real name 
[15:53] <jcsackett> hatch: cool.
[15:53] <hatch> so should be easyish to do
[15:53] <hatch> created
[16:26] <hatch> hey fabulous 
[16:26] <hatch> 1 more mo :)
[16:26] <fabulous> yes
[16:27] <fabulous> Still waiting for my exact end date :(
[16:27] <fabulous> hi everyone
[16:55] <rogpeppe1> jrwren, Makyo: https://github.com/juju/charmstore/pull/55
[17:22] <kadams54> guihelp: where did things land on the uncommitted indicators? We're using text characters for those now, right?
[17:22] <kadams54> hatch: ^
[17:23] <hatch> kadams54 I think some are using png and some is a character
[17:23] <kadams54> What's the character being used?
[17:23] <hatch> look at the mv stuff it's a character there
[17:25] <kadams54> &deg;
[17:26] <hatch> nope
[17:26] <kadams54> hatch: Well that's what it is in machine-token.handlebars :-)
[17:27] <hatch> hmm check in the css 
[17:28] <hatch> kadams54                     content: "\00B0";
[17:28] <hatch> from juju-inspector.less
[17:28] <hatch> it might be that we have to use one one place and one ther other
[17:28] <kadams54> Yeah
[17:28] <kadams54> That's likely the hex equivalent of &deg;
[17:28] <kadams54> I also found a .uncommitted-circle in mixins.less
[17:29] <hatch> ahh right it s
[17:29] <kadams54> Makyo: The ticket I'm working on means truncating long names. Trying to figure out if that needs to happen in the JS or if it can happen via CSS in the charm SVG.
[17:29] <hatch> kadams54 css has an elipsis overflow property
[17:30] <Makyo> kadams54, hatch if that works well with SVG, then we should be good.
[17:30] <kadams54> Makyo: as best I can tell, text-overflow: ellipsis doesn't work in SVG-land. Unless I'm missing something?
[17:30] <hatch> well it's container needs a defined width for sure
[17:30] <hatch> does it?
[17:30] <kadams54> Yeah, I set a width on the container
[17:30] <kadams54> width, overflow: hidden, and text-overflow: ellipsis
[17:30] <kadams54> But they didn't seem to take. I noticed that some CSS stuff works for SVG elements, others do not.
[17:31] <kadams54> border: 1px solid red, for example, does not work.
[17:31] <Makyo> kadams54, hatch It might not, unfortunately.  May need to truncate the actual name, add an ellipsis (and then the pending indicator if necessary down the line)
[17:31] <kadams54> But changing the font-size doese
[17:31] <Makyo> Yeah, not everything's applicable with svg elements.
[17:31] <hatch> boo
[17:32] <Makyo> There isn't really a concept of inline/block for example.
[17:32] <hatch> calculating the width of the text is going to be a pita
[17:32] <kadams54> I think that's a bad route.
[17:32] <Makyo> May just need to pick a number of characters to fit on the name of a service.
[17:32] <kadams54> If we can't truncate in CSS, the next best option, IMO, is to truncate at a certain character count
[17:33] <kadams54> hatch: "Guidelines from UX were to truncate names with ellipses but keep the indicator at the end whether or not truncated."
[17:33] <hatch> so it'll always be ...? even if it's not truncated?
[17:33] <kadams54> That's from the ticket. Does that mean they want the indicator moved to the end of the service name, or can I leave it at the beginning, where it currently is?
[17:33] <Makyo> Yeah, calculating the width of the text wouldn't really work with variable-width text.
[17:33] <Makyo> hatch, no, only when it's truncated.  If it's eg: 15 or more characters, truncate to 13, add ellipsis.
[17:33] <kadams54> hatch: No, it'll only have the ellipsis if we truncate, which would happen after X characters
[17:34] <hatch> I suppose we would then need to err on the side of caution and do at least a single extra character because of the variable width char set
[17:34] <Makyo> Sure
[17:35] <Makyo> Just need to find a good threshold.
[17:35] <hatch> yeah might be tough cross browser/os hah
[17:35] <Makyo> m is always going to be the widest character, so a service named mmmmmmmmmmm is a good place to start.
[17:35] <Makyo> hatch, we're loading webfonts.
[17:35] <hatch> yeah good point
[17:35] <hatch> right, but osx/ubuntu render fonts very differently :)
[17:36] <Makyo> They shouldn't in SVG-land, otherwise they wouldn't pass W3C spec.
[17:36] <hatch> huw and I ran into that when doing the deg icon for uncommitted
[17:36] <Makyo> Right, but in dom land.
[17:36] <hatch> point to point in svg will be the same, but I'm pretty sure fonts are a separate case
[17:36] <Makyo> Okay.
[17:36] <kadams54> http://www.opera.com/docs/specs/presto25/svg/cssproperties/
[17:37] <kadams54> Opera-specific, but still…
[17:37] <kadams54> Not only does SVG not support all HTML CSS properties, but in SVG land some are different.
[17:37] <kadams54> background-color becomes fill
[17:37] <Makyo> Yes, because fill works differently.
[17:38] <Makyo> If you have a non-flat polygon, there are fill-rules you have to worry about.
[17:39] <kadams54> It looks like there's an overflow property, but it only applies to certain elements (i.e., not text or tspan).
[17:39] <kadams54> And it's still not text-overflow: ellipsis
[17:40] <kadams54> Looks like I'll have to truncate in JS
[17:40] <hatch> boo urns!
[17:41] <kadams54> Wow, with an all-m service name… truncate after 7 characters.
[17:41] <kadams54> Ouch.
[17:42] <Makyo> kadams54, maybe hunt around for something that's a bit more intelligent about truncating in JS?  Like, a library of wide/narrow characters that can help?
[17:42] <hatch> kadams54 yikes....
[17:42] <hatch> Makyo that would be pretty highly dependant on the font
[17:43] <hatch> kadams54 you COULD render it to the dom, measure it, then truncate really fast :)
[17:43] <kadams54> Yeah, that's the standard workaround if you absolutely *must* have character width.
[17:44] <kadams54> Make in hidden div, populate it with the text, get the width, then delete it.
[17:44] <kadams54> But I hate that.
[17:44] <hatch> have to be a hidden svg in this point :) because of scaling haha
[17:44] <kadams54> Yup
[17:44] <kadams54> Like I said…
[17:45] <Makyo> Lets not get too in the weeds here.  Can we come up with a prototype or two along different lines of functionality?
[17:45] <Makyo> Just a smash-and-bash here's-what-we-could-do thing.
[17:45] <kadams54> https://github.com/jprichardson/d3-measure-text
[17:45] <kadams54> Yeah, I'll see what I can hack up shortly.
[17:49] <hatch> oh cool little script there
[17:50] <hatch> still falls into what you 'dont' want to do :)
[19:02] <hatch> kadams54 hey how goes the battle?
[19:02] <kadams54> Truncation on character count works fine. Playing with that d3-measure-text plugin now to see how easy/hard it is to get something going with that.
[19:04] <hatch> right - but doesn't that turn wordpress into wordpre.. ? :)
[19:05] <kadams54> Yes, it does. I think that's the intent?
[19:06] <kadams54> Well, actually, I went with a character limit of 10. So wordpress is untruncated
[19:06] <jcsackett> when did we stop showing "new" in the charm results in GUI?
[19:06] <kadams54> But "ubuntu-mirror" becomes "ubuntu-mi…"
[19:07] <hatch> kadams54 yeah imho that's just not acceptable 
[19:07] <hatch> I'd be interested to see how the size calculation bit works
[19:07] <kadams54> hatch: what's unacceptable?
[19:07] <hatch> jcsackett not sure I follow?
[19:08] <hatch> kadams54 well maybe if we had it show the real name in an 'alt' on hover ?
[19:08] <jcsackett> hatch: in the left hand browser. the editorial view. it used to have "featured", "recommended", and "new".
[19:08] <hatch> looking
[19:08] <hatch> jcsackett scroll down?
[19:08] <kadams54> hatch: it shows the full name in the inspector.
[19:08] <hatch> it's showing for me
[19:09] <jcsackett> hatch: i have scrolled down. in development, on my machine, i'm not seeing "new".
[19:09] <jcsackett> weird.
[19:09] <hatch> kadams54 right - but people usually do things like 'mysql-foobar' and 'mysql-baz' so if it truncates off the differentiator then that's no good
[19:09] <hatch> know what i mean?
[19:09] <hatch> jcsackett what about on jujucharms? or comingsoon?
[19:10] <kadams54> hatch: sounds to me like your issues is more with truncating at all rather than with how it's implemented (char count vs. pixel width) - is that accurate?
[19:10] <jcsackett> does comingsoon still run the most recent build? don't we have some qa.something.something url for that?
[19:11] <jcsackett> or am in conflating juju gui and charmworld urls?
[19:11] <hatch> jcsackett conflating :)
[19:12] <jcsackett> so, comingsoon i'm seeing the normal thing.
[19:12] <hatch> ohhh wait a second comingsoon is behind...
[19:12] <jcsackett> that's what i thought.
[19:13] <jcsackett> so, i'm assuming this is something someone landed; b/c it's now "recommended" and "other"; i'm just wondering when we did that.
[19:13] <hatch> interesting the version it's on isn't even in the develop tree....
[19:14] <hatch> jcsackett I'm just checking locally as well
[19:14] <hatch> jcsackett ok on develop I still see featured/popular/new
[19:14] <hatch> on both flags
[19:15] <hatch> jcsackett are you sure you don't have a search result?
[19:15] <hatch> ?text= in the addybar
[19:16] <jcsackett> oh bloody hell, yes i do.
[19:16] <jcsackett> thanks, hatch.
[19:16] <jcsackett> was getting quite confused. :p
[19:16] <hatch> hahaha
[19:16] <hatch> ahh it happens 
[19:16] <hatch> :D
[19:20] <kadams54> bbiab; off to pick the kiddos up from school
[19:20] <hatch> lata
[19:20] <hatch> brb rebooting
=== uru_ is now known as urulama
[19:32] <hatch> jcsackett so do you have any idea what this new comingsoon url is?
[19:32] <jcsackett> no idea.
[19:32] <jcsackett> i believe there was an email from bac.
[19:32] <jcsackett> i am digging for it now.
[19:32] <jcsackett> but i haven't found anything yet.
[19:33] <hatch> yeah I just did a quick search as well...no luck
[19:41] <rick_h__> kadams54: yes, using a character for uncommitted
[19:42] <rick_h__> howdy jujugui how goes?
[19:42] <hatch> hey rick_h__ 
[19:42] <hatch> tis trucking along well
[19:42] <jrwren> going, going going...
[19:42] <jcsackett> hola rick_h__, how's germany?
[19:42] <rick_h__> it's ok, to warm for my tastes
[19:43] <rick_h__> but I'm feeling a bit toasty after dinner and drinks
[19:43] <rick_h__> so how goes the release plans?
[19:45] <hatch> I think it's done...
[19:45] <rick_h__> orly?
[19:45] <hatch> last I heard this am Makyo was doing the charm
[19:45] <rick_h__> cool
[19:45] <hatch> https://github.com/juju/juju-gui/releases
[19:45] <hatch> the gui release is done for sure...
[19:45] <rick_h__> woot
[19:45] <rick_h__> ok, will watch out for a charm update. 
[19:46] <rick_h__> how goes things overall?
[19:46] <hatch> doesn't look the charm update is done yet
[19:46] <rick_h__> k
[19:46] <hatch> good good - found some bugs this morning doing a quick qa
[19:46] <rick_h__> bugs in the release?
[19:46] <hatch> no in mv stuff
[19:46] <rick_h__> or in MV?
[19:46] <rick_h__> :/ wheeee
[19:47] <hatch> tried to do some testing on a live env with my vagrant image and it was a pita so I switched to a vm which is now lagging and driving me nuts
[19:47] <hatch> (can't win today) 
[19:47] <hatch> :)
[19:47] <rick_h__> heh :(:(
[19:48] <hatch> there are so many possible interactions with mv
[19:48] <hatch> we will need some serious checklist
[19:49] <rick_h__> yea, definitely
[19:49] <hatch> ohhh yeah, you use a synology NAS right?
[19:49] <hatch> you should make sure you update it asap
[19:50] <rick_h__> heh, yea it's offline atm
[19:50] <hatch> ahh ok good
[19:50] <rick_h__> I need to put it on the new router and hadn't gotten around to it
=== uru_ is now known as urulama
[20:02] <hatch> ahh well in this case lazyness? has possibly saved you a big headache :D
[20:03] <hatch> The rosetta spacecraft should be ready to get into orbit around a comet tomorrow 
[20:18] <jcsackett> what's going on with synology?
[20:19] <hatch> jcsackett randsomware
[20:19] <jcsackett> ?
[20:19] <hatch> there apparently is a hole in an old version where some dickheads could hack into the system and would encrypt your drives
[20:19] <hatch> then they would ask for bitcoins to unlock it
[20:21] <hatch> rick_h__ jcsackett  figured you'd get a big kick out of this https://twitter.com/pgte/status/496750687037169664
[20:27] <jcsackett> hatch: that code is terrifying.
[20:28] <hatch> haha tis
[20:40] <Makyo> hatch, ping
[20:42] <hatch> ping
[20:42] <hatch> er pong
[20:44] <Makyo> I think I figured it out, but just to be sure, the precise and trusty charms are the same except for their branch, right?
[20:44] <hatch> correct
[20:44] <Makyo> Cool, thanks.
[20:44] <hatch> np 
[20:44] <Makyo> Trusty charm is up, getting precise one updated.
[20:44] <hatch> that reminds me, I should make the trusty charm for my ghost charm at some point
[20:45] <urulama> jrwren: hi there, how's it going?
[20:51] <hatch> guess not well... lol
[20:56] <urulama> :)
[22:32] <hatch> kadams54 you around?
[22:35] <jrwren> urulama: I missed ya.
[22:58] <huwshimi> Morning
[22:58] <hatch> ahoy!
[22:59] <hatch> I assigned you a card today huwshimi a bug from your auto-place branch
[22:59] <huwshimi> hatch: Ah right, thanks
[23:00] <hatch> blame whomever did the qa
[23:00] <hatch> :)
[23:00] <huwshimi> hatch: Can you give me a bit more on how to reproduce?
[23:01] <huwshimi> hatch: Never mind
[23:01] <hatch> nope that's all the steps
[23:01] <hatch> ok
[23:01] <hatch> :)
[23:01] <huwshimi> hatch: It's got an attached bug :)
[23:01] <hatch> oh haha, yeah
[23:02] <Makyo> hatch, can't get functional tests to pass on the precise charm, keeps failing on test_legacy_server.  Any ideas?
[23:03] <Makyo> (need to figure out how to run just that one, since a full run takes forever.
[23:03] <Makyo> )
[23:03] <hatch> Makyo sorry I have no idea, the last release (I think I did) went through without a hitch, is there any details about what the failure is caused by?
[23:04] <hatch> legacy_server sounds like pyjuju which maybe we can just delete for now
[23:04] <Makyo> Cannot connect to the environment.  Checking again on a fresh env
[23:04] <Makyo> hatch, yeah, not really sure what's up
[23:09] <Makyo> hatch, if I don't get this sorted this evening, I'm out tomorrow through Monday.  Can you at the very least delegate finishing it?  GUI is released, trusty charm is released, just the precise charm needs some love
[23:10] <hatch> Makyo yeah sure I have to take off shortly so can you send me an email with the outcome
[23:10] <Makyo> hatch, will do
[23:14] <huwshimi> hatch: Any ideas on how to list to the changes to units for a particular service? Here we just do a data-bind: https://github.com/juju/juju-gui/blob/develop/app/templates/scale-up.handlebars#L8 but I need to display the committed state and be able to pluralise the lable etc.
[23:14] <hatch> hmm
[23:14] <hatch> looking
[23:16] <hatch> huwshimi see service-overview.js
[23:16] <hatch> line.....381
[23:16] <hatch> er line 400
[23:17] <hatch> when the units change this gets called
[23:17] <hatch> try there
[23:17] <huwshimi> hatch: Ah great!
[23:45] <hatch> bbl