[08:43:13 CDT(-0500)] <Justin_o> fluid-everyone: anyone have some time to provide some feedback for a user implementing uio http://infusion-users.961009.n3.nabble.com/Infusion-users-Our-Test-Website-td3106231.html
[09:24:30 CDT(-0500)] <heidi_> thanks michelled for the treats
[09:24:40 CDT(-0500)] <michelled> np
[09:24:58 CDT(-0500)] <jhung> am I missing yummy stuff again?
[09:27:55 CDT(-0500)] <heidi_> jhung birthday wafer cookies mm
[09:28:12 CDT(-0500)] <heidi_> jhung get my pull req comments?
[09:29:15 CDT(-0500)] <jhung> Heidi: not yet. I'll try to get to it this afternoon.
[09:30:01 CDT(-0500)] <heidi_> jhung it includes styling stuff... i went ahead and did that for jameswy while i was at it
[09:30:16 CDT(-0500)] <jhung> Aw thanks Heidi.
[09:30:25 CDT(-0500)] <heidi_> np
[09:41:33 CDT(-0500)] <heidi_> Justin_o anastasiac forget... are we retiring the rust theme?
[09:41:51 CDT(-0500)] <anastasiac> heidi_, that's what I was told. It's buggy
[09:41:53 CDT(-0500)] <heidi_> and is debug theme used... ever?
[09:42:04 CDT(-0500)] <anastasiac> I have no idea what the debug theme is for!
[09:42:17 CDT(-0500)] <heidi_> ditto - there's not much in there
[09:43:44 CDT(-0500)] <huslage> this is high tech. http://logio.org/
[09:44:00 CDT(-0500)] <Justin_o> heidi_: i'm not sure... about the debug theme.. i guess it's to help you when using fss containers or something... for the rust theme i think we were going to depricate it
[09:44:04 CDT(-0500)] <Justin_o> does that sound right anastasiac
[09:44:22 CDT(-0500)] <anastasiac> sounds right
[09:45:15 CDT(-0500)] <heidi_> nothing in webapp uses the debug.css file
[09:45:24 CDT(-0500)] <Justin_o> huslage: that looks pretty neat
[09:45:45 CDT(-0500)] <Justin_o> heidi_: yah.. i would think it's just something that you would use when you are developing.. and then remove
[09:46:20 CDT(-0500)] <huslage> Justin_o: yeah. or when you're running some large-scale deal
[09:46:45 CDT(-0500)] <huslage> node rocks
[09:46:50 CDT(-0500)] <huslage> wish i knew js anymore
[09:47:05 CDT(-0500)] <Justin_o> huslage: can't wait till we get to use node.js for kettle
[09:47:08 CDT(-0500)] <colinclark> we can teach you
[09:47:18 CDT(-0500)] <huslage> what's kettle again Justin_o ?
[09:47:23 CDT(-0500)] <huslage> ok colinclark
[09:47:29 CDT(-0500)] <huslage> i'll come up for a week and bootcamp
[09:47:32 CDT(-0500)] <huslage> hehe
[09:47:35 CDT(-0500)] <colinclark> wicked
[09:48:05 CDT(-0500)] <Justin_o> kettle is a javascript server side we have been working on.. we started it for fluid engage, but were using rhino for javascript
[09:48:11 CDT(-0500)] <Justin_o> colinclark could probably explain better
[09:48:41 CDT(-0500)] <colinclark> you've got it, Justin_o
[09:48:56 CDT(-0500)] <colinclark> a declarative framework for server-side development using Infusion
[09:50:35 CDT(-0500)] <huslage> ah cool
[09:50:42 CDT(-0500)] <huslage> GPII type thing?
[09:50:49 CDT(-0500)] <huslage> or for GPII
[09:51:48 CDT(-0500)] <colinclark> well, we'll use it for the GPII
[09:51:57 CDT(-0500)] <colinclark> but we first built Kettle on top of the JVM a couple years ago
[09:51:59 CDT(-0500)] <colinclark> for our museum project
[09:52:27 CDT(-0500)] <colinclark> which reminds me that I need to talk to you about our working demo of Fluid Engage
[09:53:46 CDT(-0500)] <Justin_o> michelled: we can chat about 1.4 a bit now if you like
[09:54:56 CDT(-0500)] <heidi_> fyi Justin_o anastasiac i've closed http://issues.fluidproject.org/browse/FLUID-4184 with won't fix and noted that we're deprecating rust
[09:55:08 CDT(-0500)] <michelled> Justin_o: that would be great
[09:55:36 CDT(-0500)] <michelled> Justin_o: do you know what is still blocking the 1.4 release?
[09:57:41 CDT(-0500)] <Justin_o> michelled: okay so from what i'm working.. 1) finishing up unit tests for live preview stuff, there's more work that could go in to make it better, but they probably aren't blockers. It still needs code review too 2) ToC still having problems with the rendering.. I'm debugging with Harris. It seems that even the old style of renderer component tree has the issue. This also will need code review
[09:58:01 CDT(-0500)] <huslage> ok colinclark
[09:58:23 CDT(-0500)] <Justin_o> michelled: we should double check with heidi_ and cindyli with what they have left, but i think they were close to winding down if not done already
[09:58:41 CDT(-0500)] <Justin_o> michelled: i think mlam also had something else to address for uploader, which shouldn't be too much work
[09:59:05 CDT(-0500)] <michelled> colinclark: have you managed to get a start on !important yet?
[09:59:06 CDT(-0500)] <Justin_o> michelled: also not sure where colinclark is at with the !important stuff
[09:59:12 CDT(-0500)] <colinclark> no
[09:59:25 CDT(-0500)] <heidi_> michelled Justin_o me: UIO in IE7 styling issue, FSS themes clean up.... and also going to start testing how components work with UIO
[10:01:11 CDT(-0500)] <michelled> like putting in the UIO demos in the demo portal
[10:01:15 CDT(-0500)] <cindyli> and it's already in project repo, the removing of the linearization, Justin_o
[10:01:15 CDT(-0500)] <michelled> is there a JIRA for that?
[10:01:27 CDT(-0500)] <Justin_o> cindyli: thanks
[10:01:32 CDT(-0500)] <cindyli> np, Justin_o
[10:01:39 CDT(-0500)] <Justin_o> michelled: i don't think so.. i guess it should have been done with the demo work
[10:02:09 CDT(-0500)] <michelled> cindyli: could you make that JIRA and then fix it
[10:02:27 CDT(-0500)] <michelled> basically right now in the demo portal there's a broken link to the old demo
[10:02:32 CDT(-0500)] <Justin_o> cindyli: please at that one to the bug parade list too
[10:02:34 CDT(-0500)] <cindyli> michelled: about "putting in the UIO demos in the demo portal"? sure, i can do that
[10:02:41 CDT(-0500)] <cindyli> sure, Justin_o
[10:02:46 CDT(-0500)] <Justin_o> cindyli: thanks
[10:02:51 CDT(-0500)] <cindyli> np
[10:04:20 CDT(-0500)] <michelled> colinclark: I'm thinking when you're out of your meeting we should talk about pairing on the !important
[10:04:58 CDT(-0500)] <michelled> also, we have some outstanding pull requests that need to get in before some of the last minute stuff can be done like fixing copyrights
[10:10:26 CDT(-0500)] <cindyli> michelled: is this the uportal demo? src/webapp/integration-demos/uportal/html/portal.html
[10:10:33 CDT(-0500)] <cindyli> that i should fix?
[10:10:55 CDT(-0500)] <michelled> no cindyli - unless it's also broken
[10:11:00 CDT(-0500)] <cindyli> michelled: it's now broken from the project repo, which is not supposed to be
[10:11:20 CDT(-0500)] <cindyli> probably just a path issue
[10:11:28 CDT(-0500)] <michelled> it was working yesterday - did it break with one of the pushes I did?
[10:11:42 CDT(-0500)] <cindyli> michelled: not sure
[10:11:47 CDT(-0500)] <michelled> cindyli: it's working on the build sit
[10:11:58 CDT(-0500)] <michelled> http://build.fluidproject.org/infusion/integration-demos/uportal/html/portal.html
[10:12:24 CDT(-0500)] <cindyli> interesting, not on my localhost, ok. i will find out
[10:12:40 CDT(-0500)] <michelled> cindyli: what I was referring to was the demo portal: http://build.fluidproject.org/infusion/demos/
[10:12:51 CDT(-0500)] <michelled> if you try to view the uio demo you get an error
[10:13:00 CDT(-0500)] <cindyli> ah, ic. thanks, michelled
[10:13:03 CDT(-0500)] <michelled> since it doesn't exist anymore
[10:13:19 CDT(-0500)] <michelled> jameswy: actually, we probably need you to weigh in with what we should do on that page
[10:13:50 CDT(-0500)] <jameswy> michelled: The first or second?
[10:13:57 CDT(-0500)] <michelled> do we link to a particular version of UIO from the demo portal? or should we be showing all three?
[10:14:07 CDT(-0500)] <michelled> second link
[10:15:22 CDT(-0500)] <jameswy> Mm. Good question. I think the demo portal should be updated to reflect the "family" of UIO, as we do with the family of inline edits. Each one of those should link to the default demo that we've been working on on 1.4.
[10:16:44 CDT(-0500)] <jameswy> The integration demos should probably remain as integration demos for the time being? We have plans to revisit those areas (re: fluidproject.org, docs redesign)
[10:17:55 CDT(-0500)] <michelled> makes sense, thanks jameswy
[10:18:15 CDT(-0500)] <michelled> cindyli: do you need more detail or is that enough to go on?
[10:19:24 CDT(-0500)] <cindyli> michelled: jameswy, let me check how inline edit demos are presented
[10:21:14 CDT(-0500)] <cindyli> jameswy: rich and simple inline edit have their own boxes on the demo protal main page. are we going to do the same for the 3 ui options demos?
[10:22:52 CDT(-0500)] <cindyli> jameswy: if so, we need the designs of the icons and descriptions for 3 boxes
[10:24:03 CDT(-0500)] <jameswy> cindyli: Mm. Actually, let's overload the UIO box with links to the three demos. e.g., "Fat panel demo | Full page w/ preview demo | Full page w/out preview demo"-- something to that effect.
[10:24:25 CDT(-0500)] <cindyli> ic, jameswy. i will try. thanks
[10:24:29 CDT(-0500)] <heidi_> jameswy coal theme (med contrast greyscale) and slate theme (low contrast greyscale) not useful in UIO? lizna say anything about their usefulness?
[10:25:04 CDT(-0500)] <jameswy> jhung, can you help cindyli out with some layout on that if she needs it?
[10:26:43 CDT(-0500)] <jameswy> heidi_: Not not useful, but not our primary goal either. Lizna's position is that those four high contrast themes are the most important, followed by a number of other high contrast themes (pink-black, green-black, etc.) that we're not putting into 1.4. Nothing on "mild" themes.
[10:26:51 CDT(-0500)] <jhung> jameswy: Perhaps. Maybe late this afternoon.
[10:27:28 CDT(-0500)] <heidi_> jameswy still something worth keeping up to date in FSS tho? useful... to someone?
[10:28:26 CDT(-0500)] <jameswy> heidi_: Maybe. Remind me why they're there in the first place?
[10:28:50 CDT(-0500)] <jameswy> jhung: Thanks! Should be pretty straightforward I think.
[10:40:37 CDT(-0500)] <heidi_> jameswy i'm not sure why they're there.. i guess jacob thought they'd be useful
[10:40:46 CDT(-0500)] <jhung> jessm: fyi,. I posted meeting notes from yesterday's Decapod meeting here: http://wiki.fluidproject.org/display/fluid/Decapod+Weekly+Meeting+Notes
[10:41:06 CDT(-0500)] <jessm> jhung: thanks!
[10:43:20 CDT(-0500)] <colinclark> michelled: We should probably have a look at this http://glazman.org/JSCSSP/
[10:43:32 CDT(-0500)] <colinclark> My sense is that we're best, for the issue of !important, to use a real parser
[10:43:40 CDT(-0500)] <Bosmon2> That does look interesting, colinclark
[10:44:10 CDT(-0500)] <colinclark> The more I think about it, the more I think using any kind of token replacement scheme is a poor idea
[10:44:20 CDT(-0500)] <colinclark> Bosmon2: As background, we have the issue of !important declarations that used to be in the FSS
[10:44:30 CDT(-0500)] <colinclark> but have been removed to make the lives of our FSS users less hellish
[10:44:47 CDT(-0500)] <colinclark> UIO, however, requires a bit of a hammer in some cases to affect the page correctly
[10:44:57 CDT(-0500)] <Bosmon2> We will need to repackage their code somewhat to make it meet our standards
[10:45:02 CDT(-0500)] <colinclark> Our idea, at least as a stopgap for 1.4...
[10:45:03 CDT(-0500)] <Bosmon2> It is just a heap of globals, even without a protecting closure
[10:45:17 CDT(-0500)] <colinclark> is to generate a new stylesheet from our existing themes, with !important in the places it needs to be
[10:45:21 CDT(-0500)] <colinclark> I think this can actually be done statically
[10:45:28 CDT(-0500)] <Bosmon2> It also seems to be written to some non-standard profile of JS
[10:45:30 CDT(-0500)] <colinclark> In other words, once per release
[10:45:49 CDT(-0500)] <Bosmon2> Uses the "const" keyword...
[10:45:49 CDT(-0500)] <colinclark> Bosmon2: What are you seeing that is non-standard?
[10:45:52 CDT(-0500)] <colinclark> ah
[10:46:21 CDT(-0500)] <colinclark> I guess it's supported in Firefox
[10:46:32 CDT(-0500)] <colinclark> What I'd really like to do, I think, is run this code as part of a build process
[10:46:42 CDT(-0500)] <Bosmon2> Sounds like a good idea
[10:46:44 CDT(-0500)] <colinclark> which means it needs to be implemented in Rhino (today) and V8 (some day)
[10:46:54 CDT(-0500)] <colinclark> Likely if it's in Firefox, it's in Rhino
[10:47:00 CDT(-0500)] <colinclark> but V8, not so much
[10:47:07 CDT(-0500)] <Bosmon2> Looks like it is supported in Rhino
[10:47:38 CDT(-0500)] <Bosmon2> Ok, as long as we are not expecting to run this stuff alongside Infusion in a browser, we will be ok
[10:47:52 CDT(-0500)] <colinclark> Yeah, Rhino 1.6R6 supports it
[10:48:07 CDT(-0500)] <colinclark> I'll have to check what version of Rhino we run inside Ant for our build system
[10:48:15 CDT(-0500)] <Bosmon2> It is at least as recent as that
[10:48:23 CDT(-0500)] <Bosmon2> I remember we had to make our own patched version that was even more recent
[10:48:40 CDT(-0500)] <colinclark> For Kettle, yes
[10:48:46 CDT(-0500)] <colinclark> I think not for our build system, though
[10:48:55 CDT(-0500)] <Bosmon2> Well, even this "recent" release dates from 2007
[10:49:02 CDT(-0500)] <Bosmon2> Rhino moves about as fast as its namesake
[10:49:19 CDT(-0500)] <Bosmon2> And not on a "charging" day
[10:49:58 CDT(-0500)] <Bosmon2> As fast as a snoozing Rhino....
[10:50:05 CDT(-0500)] <colinclark>
[10:50:31 CDT(-0500)] <colinclark> I can't even remember where our build system stashes its Rhino jar
[10:50:35 CDT(-0500)] <colinclark> I'll have to check the documentation
[10:50:48 CDT(-0500)] <colinclark> Looks like V8 does support const, too
[10:50:58 CDT(-0500)] <colinclark> so if we ever switch to the Node-based build I dream of, we'll be okay
[10:51:27 CDT(-0500)] <Bosmon2> Oh wow
[10:51:32 CDT(-0500)] <Bosmon2> There actually have been Rhino releases lately
[10:51:45 CDT(-0500)] <Bosmon2> 1.7R3 earlier this month
[10:52:35 CDT(-0500)] <colinclark> Ah, look, our build system stuffs the Rhino jar right there in Git
[10:52:48 CDT(-0500)] <colinclark> or perhaps that's just for JSLint
[10:53:25 CDT(-0500)] <Bosmon2> It's the first release in more than 2 years!
[10:53:32 CDT(-0500)] <colinclark> wow
[10:53:38 CDT(-0500)] <colinclark> Maybe it's insanely fast
[10:53:40 CDT(-0500)] <colinclark>
[10:53:49 CDT(-0500)] <Bosmon2> They say it is slightly faster....
[10:54:08 CDT(-0500)] <Bosmon2> "Rhino has seen some performance improvements since the last release which are most notable when running benchmarks or the test suite."
[10:54:16 CDT(-0500)] <Bosmon2> As fast as a slowly trotting Rhino
[10:55:34 CDT(-0500)] <colinclark> Interestingly, the guy who wrote this is a co-author of the CSS Variables spec
[10:55:35 CDT(-0500)] <colinclark> so
[10:55:39 CDT(-0500)] <colinclark> his parser supports them
[10:56:50 CDT(-0500)] <Bosmon2> I didn't hear about those before
[10:56:53 CDT(-0500)] <Bosmon2> Sounds dangerous
[10:57:36 CDT(-0500)] <colinclark> I guess maybe it hasn't really gone anywhere
[10:57:37 CDT(-0500)] <colinclark> http://disruptive-innovations.com/zoo/cssvariables/
[10:57:48 CDT(-0500)] <colinclark> "Mixins" are the thing that landed in WebKit fairly recently
[10:58:12 CDT(-0500)] <heidi_> jameswy i've created this for now: http://issues.fluidproject.org/browse/FLUID-4311 re: slate/coal/uioptions
[10:58:23 CDT(-0500)] <Bosmon2> How do "Mixins" relate to "CSS Variable" (if at all)?
[10:58:28 CDT(-0500)] <Bosmon2> s
[10:58:43 CDT(-0500)] <colinclark> I guess this is a reasonable overview of them...
[10:58:44 CDT(-0500)] <colinclark> http://johanbrook.com/design/css/webkit-css-variables-mixins-nesting/
[10:58:54 CDT(-0500)] <colinclark> I remember reading the slide deck he's referring to on the subway, awhile ago
[10:59:00 CDT(-0500)] <colinclark> but it's all a blur
[10:59:16 CDT(-0500)] <Bosmon2> It really is
[10:59:24 CDT(-0500)] <Bosmon2> I think the existing CSS model confused quite enough people
[10:59:34 CDT(-0500)] <colinclark> I guess this does look like the same variable proposal from our parser author
[10:59:48 CDT(-0500)] <colinclark> I can't quite figure out, though, if any of it has actually landed
[11:00:18 CDT(-0500)] <Bosmon2> Although I guess something like "Variables" is something that working with the FSS convinced us we needed for quite some time
[11:00:33 CDT(-0500)] <colinclark> I think they probably are quite necessary
[11:00:49 CDT(-0500)] <colinclark> It looks like some effort has been made to put them into WebKit...
[11:01:49 CDT(-0500)] <Bosmon2> Without "Variables" we are doomed to endless repetition of the same material... although putting this parser into our build system would be a way to start to head this off
[11:02:03 CDT(-0500)] <colinclark> No apparent action on the Mozilla front: https://bugzilla.mozilla.org/show_bug.cgi?id=442864
[11:02:11 CDT(-0500)] <Bosmon2> Perhaps we could even jig up the parser to spit out a non-variable expansion of "variable-aware code" for older browsers
[11:02:41 CDT(-0500)] <colinclark> I think that's probably quite possible
[11:03:28 CDT(-0500)] <heidi_> colinclark is that webkit stuff in the current ff? mixins nesting etc? looks neat
[11:03:39 CDT(-0500)] <colinclark> No
[11:03:39 CDT(-0500)] <Bosmon2> Ah, mixins looks like something much more reasonable than I was fearing
[11:03:47 CDT(-0500)] <colinclark> It looks to me, heidi_, like they haven't yet landed anywhere
[11:03:56 CDT(-0500)] <heidi_> ah, still just ideas
[11:04:04 CDT(-0500)] <colinclark> heidi_: Beyond that
[11:04:14 CDT(-0500)] <colinclark> One gets the sense that they'll show up any day now
[11:04:19 CDT(-0500)] <colinclark> In WebKit
[11:04:24 CDT(-0500)] <heidi_> interesting
[11:04:29 CDT(-0500)] <colinclark> as I mentioned above, it doesn't seem like there's been any work on that front from Mozilla
[11:04:41 CDT(-0500)] <colinclark> I guess this perhaps points us in an interesting potential direction...
[11:04:45 CDT(-0500)] <colinclark> For 1.5
[11:04:55 CDT(-0500)] <Bosmon2> If only Microsoft would just adopt WebKit and be done with it
[11:05:05 CDT(-0500)] <colinclark> where we could work with this parser to build a kind of "super SASS light" for FSS
[11:05:23 CDT(-0500)] <colinclark> Where FSS authors could work with the emerging specs for variables and mixins
[11:05:45 CDT(-0500)] <colinclark> and then we could generate old style sheets as needed
[11:06:03 CDT(-0500)] <heidi_> yeah them seem like really neat/powerful features.
[11:06:12 CDT(-0500)] <colinclark> I think 1.5 could be an interesting milestone for us
[11:06:12 CDT(-0500)] <Justin_o> Bosmon2: do you have time to chat about the ToC rendering issue
[11:06:24 CDT(-0500)] <Bosmon2> colinclark - I think it will be
[11:06:25 CDT(-0500)] <Bosmon2> Justin_o - yes
[11:06:27 CDT(-0500)] <heidi_> esp in terms of accessibility and swappable css and such
[11:06:29 CDT(-0500)] <colinclark> I'm slowly gestating a proposal for a new status for IE 6 and 7 for us
[11:06:47 CDT(-0500)] <Justin_o> Bosmon2: could you take a look at my FLUID-4209 branch
[11:07:06 CDT(-0500)] <colinclark> I figure such a proposal should wait until after we actually release 1.4 and can take a breather
[11:07:06 CDT(-0500)] <colinclark> b
[11:07:18 CDT(-0500)] <colinclark> but after long chats with Justin_o and michelled about IE, the gears are turning
[11:07:43 CDT(-0500)] <colinclark> Okay, so for the short run, michelled
[11:07:43 CDT(-0500)] <heidi_> like... ditch IE6/7 or something else?
[11:07:58 CDT(-0500)] <colinclark> Well, I hate to show my cards before I have a reasonable proposal written up
[11:08:08 CDT(-0500)] <heidi_> fair enough
[11:08:24 CDT(-0500)] <colinclark> but I'm thinking that we might want to more deeply embrace progressive enhancement and graceful degradation
[11:08:25 CDT(-0500)] <colinclark> out of the box
[11:08:27 CDT(-0500)] <colinclark> for all components
[11:08:50 CDT(-0500)] <colinclark> That might give us some room to better reflect the changing state of browsers
[11:08:59 CDT(-0500)] <heidi_> colinclark any thoughts on supporting browsing on mobile? is that a special thing anymore or do mobile browsers deal with our stuff well already
[11:09:12 CDT(-0500)] <colinclark> It depends what we mean by mobile
[11:09:19 CDT(-0500)] <colinclark> Smart mobile is generally pretty good
[11:09:23 CDT(-0500)] <colinclark> except that interaction is quite different
[11:09:33 CDT(-0500)] <colinclark> drag and drop, say, isn't very meaningful on a touch screen, for the most part
[11:09:50 CDT(-0500)] <colinclark> not-so-smart mobile is another issue altogether
[11:10:08 CDT(-0500)] <colinclark> which would be well-served by a greater use of progressive enhancement
[11:10:18 CDT(-0500)] <Bosmon2> Justin_o - looking at your branch
[11:10:24 CDT(-0500)] <Bosmon2> Does it work?
[11:10:36 CDT(-0500)] <colinclark> So, michelled...
[11:10:38 CDT(-0500)] <colinclark> you there?
[11:10:38 CDT(-0500)] <Bosmon2> Your tree looks a bit odd in that it doesn't begin with "children" - but that might be ok
[11:11:03 CDT(-0500)] <Bosmon2> Justin_o - there's no reason to work with a completely expanded tree btw
[11:11:15 CDT(-0500)] <Bosmon2> All I was suggesting was that you should avoid the use of the protoComponent expanders
[11:11:27 CDT(-0500)] <Bosmon2> The repetition and condition expanders, in particular
[11:18:19 CDT(-0500)] <Justin_o> Bosmon2: hmm.. okay..
[11:18:38 CDT(-0500)] <Justin_o> Bosmon2: i made it fully expanded and static just for getting an idea of what the structure would look like
[11:18:45 CDT(-0500)] <Bosmon2> Justin_o - that's good
[11:18:49 CDT(-0500)] <Justin_o> Bosmon2: in either case though, i'm still running into the same issue
[11:18:51 CDT(-0500)] <Bosmon2> So - what do you observe with this tree?
[11:19:01 CDT(-0500)] <Bosmon2> Justin_o - It couldn't possibly be the same issue
[11:19:03 CDT(-0500)] <Justin_o> so same problem exactly
[11:19:16 CDT(-0500)] <Justin_o> still getting the double humans
[11:19:18 CDT(-0500)] <Bosmon2> By definition, that is impossible - since the problem last time was that you had the wrong tree
[11:19:44 CDT(-0500)] <Bosmon2> Ok, in some ways it could be the same problem
[11:19:49 CDT(-0500)] <Bosmon2> Since you still have the wrong tree
[11:20:06 CDT(-0500)] <Bosmon2> A big issue is that you can't reuse the id "items:" in the way that you have
[11:20:49 CDT(-0500)] <Justin_o> Bosmon2: yes. harriswong and I tried switching that i think, but had the same issue
[11:20:57 CDT(-0500)] <Bosmon2> But it has certainly helped a lot to have the static tree written out like this, since these problems can be more obvious
[11:20:58 CDT(-0500)] <Justin_o> we did something like items1:
[11:21:14 CDT(-0500)] <Justin_o> Bosmon2: yes for sure...
[11:21:31 CDT(-0500)] <Bosmon2> Justin_o - you should go back to using items1: etc. for a start
[11:21:46 CDT(-0500)] <Bosmon2> And then see what happens
[11:21:58 CDT(-0500)] <Justin_o> Bosmon2: okay.. let me quickly whip that up and will push it up
[11:22:02 CDT(-0500)] <Bosmon2> Justin_o - thanks
[11:22:45 CDT(-0500)] <colinclark> Bosmon2: Just a quick check-in about FLUID-4285
[11:22:50 CDT(-0500)] <colinclark> That ball is your court now, right?
[11:23:06 CDT(-0500)] <Bosmon2> colinclark - certainly, yes - thanks
[11:23:08 CDT(-0500)] <colinclark> ok, cool
[11:23:14 CDT(-0500)] <colinclark> I'll review it again whenever you get a chance to make a fix
[11:23:26 CDT(-0500)] <colinclark> in the meantime, I'll have a quick look at the JIRA you filed for me
[11:23:30 CDT(-0500)] <Bosmon2> Immediately after TTEA, I expect
[11:23:42 CDT(-0500)] <Bosmon2> Thanks, colinclark - you mean the one about reconstructing the reasons behind the "DOM injection failure"?
[11:23:46 CDT(-0500)] <colinclark> yes, that one
[11:23:50 CDT(-0500)] <Bosmon2> cool
[11:23:55 CDT(-0500)] <colinclark> and then michelled and I will probably talk through our !important injection approach here for a little bit
[11:24:03 CDT(-0500)] <colinclark> and then maybe a sandwich
[11:24:05 CDT(-0500)] <colinclark>
[11:34:12 CDT(-0500)] <Justin_o> Bosmon2: i've updated my branch, the issue still remains though
[11:34:41 CDT(-0500)] <Bosmon2> Ok Justin_o - thanks, I will look at it now
[11:34:48 CDT(-0500)] <Justin_o> Bosmon2: thanks
[11:35:20 CDT(-0500)] <Justin_o> there's also a simpleStaticTree there too.. which is a pretty barebones case, that harris set up, to show the issue
[11:44:19 CDT(-0500)] <jameswy> heidi_: Super, thanks!
[11:46:02 CDT(-0500)] <michelled> colinclark: I'm here now
[11:48:51 CDT(-0500)] <colinclark> Hi michelled
[11:49:04 CDT(-0500)] <colinclark> I don't know how much of the long, somewhat indirect part of the conversation you saw from earlier
[11:49:08 CDT(-0500)] <colinclark> To summarize really quickly...
[11:49:14 CDT(-0500)] <colinclark> 1. I fired you a link to a JavaScript-based CSS parser
[11:49:21 CDT(-0500)] <michelled> just looking at that now
[11:49:29 CDT(-0500)] <colinclark> 2. Bosmon2 was intrigued by it
[11:49:35 CDT(-0500)] <colinclark> 3. It got me to talking about CSS variables and mixins
[11:49:41 CDT(-0500)] <colinclark> 4. heidi_ was intrigued by that
[11:49:56 CDT(-0500)] <michelled>
[11:49:59 CDT(-0500)] <colinclark> 5. I speculated on a future vision for FSS that involved embracing modern CSS standards
[11:50:06 CDT(-0500)] <colinclark> perhaps enabled by this parser
[11:50:26 CDT(-0500)] <colinclark> 6. We took a detour into me hinting at my vague IE proposal
[11:50:53 CDT(-0500)] <colinclark> 7. Then cycled back to the point. Let's consider this parser as an option for doing !important injection in our style sheets
[11:50:55 CDT(-0500)] <colinclark> statically
[11:50:59 CDT(-0500)] <colinclark> and as part of the build process
[11:51:18 CDT(-0500)] <michelled> I like the idea of doing this at build time
[11:51:20 CDT(-0500)] <colinclark> The more I contemplated the prospect of some kind of "super simple" token replacement, the more it seemed evil to me
[11:51:31 CDT(-0500)] <colinclark> Certainly we could implement something in all of an hour
[11:51:45 CDT(-0500)] <colinclark> but the FSS would be permanently tattooed with odd comment tokens or whatever
[11:51:52 CDT(-0500)] <colinclark> which would be a source of incredible brittleness
[11:52:12 CDT(-0500)] <colinclark> So, I guess the term "build time" is a bit misleading
[11:52:19 CDT(-0500)] <colinclark> What I was, off the top of my head, thinking is this:
[11:52:44 CDT(-0500)] <colinclark> We want UI Options to work correctly, straight out of a git clone
[11:52:52 CDT(-0500)] <colinclark> So, currently, UI Options is in some way broken
[11:52:59 CDT(-0500)] <colinclark> because of the absence of !important
[11:53:21 CDT(-0500)] <colinclark> If we generated these UIO-specific sheets when someone wants a custom build of Infusion, it would be fine for non-contributing users
[11:53:38 CDT(-0500)] <heidi_> Justin_o fyi http://issues.fluidproject.org/browse/FLUID-4312
[11:53:45 CDT(-0500)] <colinclark> but for anyone who wanted to just grab the latest and greatest and have it work fully, they'd have a problem
[11:53:49 CDT(-0500)] <colinclark> I guess Johnny is doing this now
[11:53:52 CDT(-0500)] <michelled> colinclark: so are you thinking of a commit hook?
[11:53:56 CDT(-0500)] <colinclark> He just cloned Infusion and is working on it
[11:54:04 CDT(-0500)] <colinclark> michelled: That's actually pretty interesting
[11:54:10 CDT(-0500)] <colinclark> I was thinking, first, of a new target in our build scripts
[11:54:18 CDT(-0500)] <colinclark> one that isn't run as part of doing a build
[11:54:22 CDT(-0500)] <colinclark> but is run at "development time"
[11:54:35 CDT(-0500)] <colinclark> meaning, when you change an FSS theme, you run the target
[11:54:49 CDT(-0500)] <colinclark> The process would be configured by some little blob of JSON
[11:55:09 CDT(-0500)] <colinclark> identifying the themes that need to be generated
[11:55:10 CDT(-0500)] <michelled> it's an extra step for the developer and I'm guessing it would get forgotten
[11:55:24 CDT(-0500)] <michelled> for example our release notes still say we are using jquery 1.4
[11:55:35 CDT(-0500)] <colinclark> well, we haven't released yet
[11:55:42 CDT(-0500)] <colinclark> michelled: You're right
[11:55:49 CDT(-0500)] <colinclark> it's prone to forgetfulness
[11:55:50 CDT(-0500)] <colinclark> as am I
[11:55:57 CDT(-0500)] <colinclark> "What is your alternative proposal?"
[11:56:07 CDT(-0500)] <colinclark> I think I've developed air quote insanity
[11:56:11 CDT(-0500)] <colinclark> and I blame Bosmon2 entirely
[11:56:13 CDT(-0500)] <michelled> lol
[11:56:47 CDT(-0500)] <Bosmon2> "Is that so!"
[11:56:56 CDT(-0500)] <colinclark> jessm sent me an article from the National Post, lamenting the fall of civilization. It was one of those typically things, blaming the kinds for using bad grammar and "air quotes"
[11:57:06 CDT(-0500)] <michelled> I can't remember what we finally figured out about commit hooks and git - I'm so in the habit of putting in a JIRA number that I don't think I'd hit it
[11:57:20 CDT(-0500)] <Bosmon2> Did it mention the "lavender of the subjunctive"?
[11:57:26 CDT(-0500)] <colinclark> lol
[11:57:27 CDT(-0500)] <colinclark> no
[11:57:44 CDT(-0500)] <michelled> Justin_o: what's the deal with commit hooks and git?
[11:57:47 CDT(-0500)] <colinclark> Justin_o: What is your opinion of what michelled is calling "commit hooks" in the universe of Git and Github
[11:57:50 CDT(-0500)] <colinclark> lol
[11:58:05 CDT(-0500)] <colinclark> at least michelled remembered the punctuation, unlike me
[11:58:22 CDT(-0500)] <Justin_o> michelled, colinclark: what did you want to know
[11:58:36 CDT(-0500)] <michelled> do they exist?
[11:58:45 CDT(-0500)] <colinclark> we are in a post-question mark world
[11:58:45 CDT(-0500)] <Justin_o> github has post commit hooks
[11:58:49 CDT(-0500)] <colinclark> "post question mark"
[11:58:56 CDT(-0500)] <colinclark> Justin_o: Where do they run?
[11:59:04 CDT(-0500)] <colinclark> On Github's server?
[11:59:18 CDT(-0500)] <Justin_o> colinclark: yes.. githubs serve.. actually i guess they are post receive hooks
[11:59:25 CDT(-0500)] <colinclark> okay, so that won't work
[11:59:37 CDT(-0500)] <Justin_o> but they trigger actions outside of their server, not in
[11:59:38 CDT(-0500)] <colinclark> unless you want to convince github to install Rhino and a JVM, michelled
[11:59:44 CDT(-0500)] <colinclark> Justin_o: Can you elaborate?
[11:59:57 CDT(-0500)] <Justin_o> colinclark: for example our commits e-mail list
[11:59:58 CDT(-0500)] <colinclark> Nothing like trying to convince Ruby fanatics to install a JVM
[12:00:10 CDT(-0500)] <colinclark> what we want is a pre-commit hook, anyway
[12:00:20 CDT(-0500)] <Justin_o> colinclark: so that you have to do yourself
[12:00:27 CDT(-0500)] <michelled> that would be ok
[12:00:30 CDT(-0500)] <colinclark> Because I think michelled is advocating for a script to actually change the user's commit
[12:00:45 CDT(-0500)] <Justin_o> colinclark, michelled: what is that for?
[12:00:58 CDT(-0500)] <colinclark> Justin_o: for evil science
[12:01:02 CDT(-0500)] <michelled> for generating the !important stylesheets
[12:01:09 CDT(-0500)] <colinclark> yes, exactly
[12:01:28 CDT(-0500)] <Justin_o> exciting
[12:01:46 CDT(-0500)] <michelled> colinclark: as long as it was run prior to a push to the project repo we'd be fine
[12:01:50 CDT(-0500)] <Bosmon2> yes well.... isn't this actually insane?
[12:01:56 CDT(-0500)] <Justin_o> but you'd have to make sure people actually used it, which may be the harder part
[12:01:59 CDT(-0500)] <colinclark> I think it may well be
[12:02:13 CDT(-0500)] <Bosmon2> I assumed you were just bringing up the possibility Socratically
[12:02:19 CDT(-0500)] <Justin_o> Bosmon2: this time it's not me being insane
[12:02:29 CDT(-0500)] <colinclark> If people have to be made to use it
[12:02:35 CDT(-0500)] <colinclark> I think it probably isn't what michelled is looking for
[12:02:47 CDT(-0500)] <michelled> what do you mean 'made to use it'?
[12:02:59 CDT(-0500)] <Bosmon2> Surely we're not actually seriously proposing this... as opposed to just putting this into our process which makes builds of Infusion
[12:03:04 CDT(-0500)] <michelled> once people install it, wouldn't it happen automatically?
[12:03:23 CDT(-0500)] <colinclark> Bosmon2: Well, we're considering everything at the moment
[12:03:36 CDT(-0500)] <colinclark> you can see my point above about UIO being "broken" out of the box if we do it at build time
[12:03:56 CDT(-0500)] <Bosmon2> Yes
[12:04:00 CDT(-0500)] <Bosmon2> How soon do we actually need to do this?
[12:04:02 CDT(-0500)] <Bosmon2> Surely not for 1.4?
[12:04:07 CDT(-0500)] <colinclark> Surely so!
[12:04:11 CDT(-0500)] <Bosmon2> glarg
[12:04:11 CDT(-0500)] <colinclark>
[12:04:18 CDT(-0500)] <colinclark> It's the last big blocker for UIO
[12:04:42 CDT(-0500)] <colinclark> I'd be willing to buy an argument that we can really only feasibly do it at build time...
[12:04:44 CDT(-0500)] <colinclark> but it's really not ideal
[12:04:52 CDT(-0500)] <colinclark> neither is having some code in the repo that is generated
[12:04:56 CDT(-0500)] <colinclark> and runs the risk of falling out of date
[12:04:59 CDT(-0500)] <Bosmon2> ok
[12:05:02 CDT(-0500)] <Bosmon2> So how about this
[12:05:07 CDT(-0500)] <colinclark> which I imagine is at the heart of michelled and your concerns
[12:05:14 CDT(-0500)] <Bosmon2> Set up a little server which does the transformation in itself
[12:05:20 CDT(-0500)] <Bosmon2> And spits back an appropriately munged CSS file
[12:05:32 CDT(-0500)] <Bosmon2> Then refer to that URL which we will maintain stable, inside the UIO checkout
[12:05:41 CDT(-0500)] <colinclark> I seem to remember Justin_o and I talking about something like this
[12:06:09 CDT(-0500)] <colinclark> So at build time, we'd need to change this URL
[12:06:15 CDT(-0500)] <colinclark> to the generated, local versions
[12:06:27 CDT(-0500)] <colinclark> otherwise we're in the business of running a CDN
[12:06:34 CDT(-0500)] <Bosmon2> yes
[12:07:02 CDT(-0500)] <colinclark> It's a big awkward
[12:07:08 CDT(-0500)] <colinclark> we'd be modifying a real source file at build time
[12:07:13 CDT(-0500)] <Justin_o> colinclark, michelled, Bosmon2: just an aside, how will this work if we start concatenating our css files?
[12:07:22 CDT(-0500)] <colinclark> Justin_o: That's a good question
[12:07:28 CDT(-0500)] <colinclark> we could ignore it, saying it's out of scope for 1.4
[12:07:30 CDT(-0500)] <colinclark> which it most definitely is
[12:07:48 CDT(-0500)] <Bosmon2> colinclark - how is this different from modifications from our current build process?
[12:07:52 CDT(-0500)] <colinclark> but I actually think there's something more interesting underneath all of this
[12:08:04 CDT(-0500)] <colinclark> Bosmon2: I can't think of any case where the build actually changes code
[12:08:11 CDT(-0500)] <colinclark> aside from smashing it together
[12:08:35 CDT(-0500)] <colinclark> "code" in this case, I guess, being a bunch of <link> tags
[12:09:33 CDT(-0500)] <Bosmon2> colinclark - doesn't it change includes in demos, etc. to refer to "InfusionAll.js" or some such?
[12:09:49 CDT(-0500)] <Bosmon2> Or was this just a brief chimera in the Engage days.....
[12:11:29 CDT(-0500)] <colinclark> No, that does exist
[12:11:33 CDT(-0500)] <colinclark> but it's only used for QA purposes
[12:11:43 CDT(-0500)] <colinclark> and I think we have something of a distrust for it
[12:11:48 CDT(-0500)] <Bosmon2> Why is that?
[12:12:01 CDT(-0500)] <colinclark> Well, it was a pretty noble effort by the King
[12:12:17 CDT(-0500)] <colinclark> to reduce the mind-numbing process of testing full builds
[12:12:26 CDT(-0500)] <colinclark> but it's just a wild regexp
[12:12:36 CDT(-0500)] <colinclark> Justin_o: I dunno, do you trust it?
[12:12:43 CDT(-0500)] <colinclark> I might just be paranoid
[12:12:49 CDT(-0500)] <colinclark> (don't tell anyone)
[12:12:56 CDT(-0500)] <Justin_o> colinclark: i trust it as long as we are careful with our naming
[12:13:02 CDT(-0500)] <colinclark> lol
[12:13:15 CDT(-0500)] <Justin_o> basically if you were to name the demo's js file the same as one of our main component js files, then it will get removed
[12:13:43 CDT(-0500)] <colinclark> It somehow excludes paths with the word "demo" in it or something?
[12:14:18 CDT(-0500)] <colinclark> So I guess there's another question we could consider.
[12:14:39 CDT(-0500)] <colinclark> Do we really care if one feature of UIO is not working straight out of the repo?
[12:14:54 CDT(-0500)] <colinclark> I guess it would have to be QA'd only against a build
[12:15:08 CDT(-0500)] <colinclark> The nightly Infusion, I think, is built before deployment
[12:15:09 CDT(-0500)] <colinclark> "built"
[12:15:19 CDT(-0500)] <colinclark> which is where we'd do QA
[12:15:41 CDT(-0500)] <colinclark> if you were a UIO developer or contributor, you'd just have to remember to run a build before working on this particular feature
[12:15:48 CDT(-0500)] <colinclark> which is, currently, the theme-swapping feature of UIO
[12:16:00 CDT(-0500)] <cindyli1> huslage: hi
[12:16:09 CDT(-0500)] <huslage> hi cindyli1
[12:17:00 CDT(-0500)] <cindyli1> huslage: can u help to change the mount of the image gallery nightly build. let me dig out the info
[12:17:28 CDT(-0500)] <huslage> sure
[12:18:41 CDT(-0500)] <cindyli1> huslage: http://build.fluidproject.org/image-gallery/image-gallery.php is now pointing to "/home/continuum/apache-continuum-1.4.0/data/working-directory/22/image-gallery.php" at server "forge.fluidproject.org"
[12:19:35 CDT(-0500)] <cindyli1> huslage: can u point it to ""/home/continuum/apache-continuum-1.4.0/data/working-directory/22/image-gallery/image-gallery.php". Note the "image-gallery" folder in btw "22" and the script name
[12:19:57 CDT(-0500)] <huslage> ok
[12:20:27 CDT(-0500)] <cindyli1> thanks, huslage
[12:21:10 CDT(-0500)] <huslage> ok fixed i think cindyli1
[12:21:27 CDT(-0500)] <colinclark> hey huslage
[12:21:30 CDT(-0500)] <colinclark> how is that actually implemented?
[12:21:36 CDT(-0500)] <huslage> it's an apache config
[12:21:36 CDT(-0500)] <colinclark> some kind of apache rewrite rule somewhere?
[12:21:38 CDT(-0500)] <colinclark> ok
[12:21:42 CDT(-0500)] <huslage> an alias
[12:21:47 CDT(-0500)] <colinclark> right
[12:21:58 CDT(-0500)] <huslage> Alias /image-gallery /home/continuum/apache-continuum-1.4.0/data/working-directory/22/image-gallery/
[12:22:03 CDT(-0500)] <colinclark> so this is the kind of stuff i'm thinking we might want to "cloudize"
[12:22:11 CDT(-0500)] <colinclark> you asked me for some requirements
[12:22:12 CDT(-0500)] <huslage> that's harder to cloudize
[12:22:18 CDT(-0500)] <colinclark> which I still haven't had time to write
[12:22:22 CDT(-0500)] <huslage> slacker
[12:22:25 CDT(-0500)] <colinclark>
[12:22:25 CDT(-0500)] <colinclark> ]
[12:22:43 CDT(-0500)] <colinclark> but I think increasingly we need think about how to self-manage and self-provision our build server
[12:22:51 CDT(-0500)] <huslage> fair enough.
[12:22:56 CDT(-0500)] <colinclark> so that we can make it easier for people with shell access to register builds
[12:22:56 CDT(-0500)] <colinclark> a
[12:22:59 CDT(-0500)] <colinclark> where they go,
[12:23:00 CDT(-0500)] <colinclark> etc.
[12:23:03 CDT(-0500)] <colinclark> it'll be awesome
[12:23:06 CDT(-0500)] <colinclark> I'm sure of it
[12:23:11 CDT(-0500)] <huslage> hudson
[12:23:12 CDT(-0500)] <colinclark> once we figure out what "it" is
[12:23:18 CDT(-0500)] <huslage> buildbot?
[12:25:35 CDT(-0500)] <huslage> so colinclark
[12:26:18 CDT(-0500)] <huslage> continuum "builds" things like this image-gallery demo by running 'install.sh'
[12:26:29 CDT(-0500)] <huslage> it checks out and runs this script correct?
[12:27:21 CDT(-0500)] <huslage> but infusion runs an ant job or something
[12:28:02 CDT(-0500)] <cindyli1> huslage: colinclark is off for lunch or something else
[12:28:07 CDT(-0500)] <huslage> he's lame
[12:28:09 CDT(-0500)] <huslage> that guy
[12:28:11 CDT(-0500)] <huslage> sheesh
[12:28:16 CDT(-0500)] <cindyli1>
[12:28:53 CDT(-0500)] <cindyli1> huslage: problem with repointing the url
[12:28:59 CDT(-0500)] <cindyli1> http://build.fluidproject.org/image-gallery/image-gallery.php is throwing js error
[12:28:59 CDT(-0500)] <huslage> ?
[12:29:00 CDT(-0500)] <Bosmon2> http://en.wikipedia.org/wiki/Colin's_Sandwich
[12:29:47 CDT(-0500)] <huslage> this? The requested URL /image-gallery/uploader.php was not found on this server.
[12:30:17 CDT(-0500)] <cindyli1> huslage: oh, the script name should be "image-gallery.php"
[12:30:37 CDT(-0500)] <cindyli1> huslage: if u look into folder "/home/continuum/apache-continuum-1.4.0/data/working-directory/22"
[12:31:00 CDT(-0500)] <cindyli1> "image-gallery.php" is underneath folder "image-gallery", which has a sibling folder "infusion"
[12:31:20 CDT(-0500)] <huslage> ugh
[12:31:20 CDT(-0500)] <huslage> ok
[12:31:23 CDT(-0500)] <cindyli1> huslage: with the new pointer, i suspect "infusion" folder becomes inacessible
[12:31:29 CDT(-0500)] <huslage> yes cindyli1
[12:31:37 CDT(-0500)] <cindyli1> ya, that's the problem
[12:31:52 CDT(-0500)] <huslage> should it be under /image-gallery/infusion?
[12:32:26 CDT(-0500)] <cindyli1> no, "image-gallery" and "infusion" should be siblings
[12:32:30 CDT(-0500)] <cindyli1> with the current design
[12:32:39 CDT(-0500)] <huslage> isn't that gonna break the other stuff?
[12:32:48 CDT(-0500)] <huslage> on this server
[12:33:08 CDT(-0500)] <cindyli1> no, i think, working-directory/22 is solely for image gallery
[12:33:13 CDT(-0500)] <huslage> no i know
[12:33:14 CDT(-0500)] <cindyli1> it uses its own infusion
[12:33:38 CDT(-0500)] <huslage> but the problem is that i'm not pointing to /22 now…i'm pointing to 22/image-gallery/
[12:34:03 CDT(-0500)] <huslage> and i can't alias /infusion to 22/infusion because that'd break other things not the server
[12:34:36 CDT(-0500)] <cindyli1> what do u think alias "working-directory/22/image-gallery/image-gallery.php" to "working-directory/22/image-gallery.php"
[12:34:42 CDT(-0500)] <cindyli1> would that help, huslage
[12:34:49 CDT(-0500)] <huslage> you don't alias to a file, you alias to a directory
[12:35:09 CDT(-0500)] <huslage> hm
[12:35:19 CDT(-0500)] <cindyli1> well, symbolic link
[12:35:29 CDT(-0500)] <huslage> hm
[12:35:35 CDT(-0500)] <huslage> lemme try something
[12:36:04 CDT(-0500)] <cindyli1> thanks
[12:40:32 CDT(-0500)] <huslage> arg
[12:42:38 CDT(-0500)] <huslage> so cindyli1 why does the form submit to uploader.php?
[12:43:27 CDT(-0500)] <cindyli1> huslage: to upload the images into server
[12:43:38 CDT(-0500)] <huslage> but there isn't an uploader.php
[12:43:48 CDT(-0500)] <cindyli1> it submits to uploader-server.php
[12:43:54 CDT(-0500)] <huslage> not mine
[12:44:00 CDT(-0500)] <cindyli1> in "22"?
[12:44:12 CDT(-0500)] <huslage> yes
[12:44:13 CDT(-0500)] <huslage> http://build.fluidproject.org/image-gallery/uploader.php
[12:44:28 CDT(-0500)] <huslage> <form action="uploader.php" method="post" enctype="multipart/form-data" class="fl-progEnhance-basic">
[12:44:35 CDT(-0500)] <cindyli1> ah, ic
[12:44:43 CDT(-0500)] <huslage> yay i found a booooh
[12:44:44 CDT(-0500)] <huslage> booog
[12:44:56 CDT(-0500)] <cindyli1> that submit action can be removed.
[12:45:10 CDT(-0500)] <cindyli1> it's actually thru ajax call
[12:45:23 CDT(-0500)] <huslage> well. breaks
[12:45:52 CDT(-0500)] <cindyli1> that's the cause?
[12:46:05 CDT(-0500)] <huslage> i fixed the infusion thing
[12:46:08 CDT(-0500)] <huslage> i think
[12:46:16 CDT(-0500)] <cindyli1> thanks. trying
[12:46:32 CDT(-0500)] <huslage> infusion is at image-gallery/infusion now
[12:46:42 CDT(-0500)] <huslage> you may need to reconfigure something i dunno
[12:48:11 CDT(-0500)] <cindyli1> i don't see image-gallery/infusion at working-directory/22/
[12:50:41 CDT(-0500)] <cindyli1> is there a way to get it work with "image-gallery" and "infusion" siblings, rather than "infusion" being a sub-folder of "image-gallery"?
[12:51:26 CDT(-0500)] <cindyli1> huslage?
[12:51:40 CDT(-0500)] <huslage> no
[12:51:52 CDT(-0500)] <huslage> not given the requirements of other things on that vhost
[12:52:42 CDT(-0500)] <cindyli1> okay. thanks, huslage
[12:53:10 CDT(-0500)] <Justin_o> Bosmon2: are you still around?
[12:59:59 CDT(-0500)] <cindyli1> Justin_o: colinclark, as huslage has pointed out, image gallery nightly/hourly builds would not work if "image-gallery/" script folder and "infusion" were pulled out as siblings
[13:00:30 CDT(-0500)] <Justin_o> cindyli1: really... why is that?
[13:00:39 CDT(-0500)] <Justin_o> i guess i missed it somewhere here
[13:01:05 CDT(-0500)] <cindyli1> as huslage said, it woulld mess up with other stuff on the server
[13:01:23 CDT(-0500)] <cindyli1> Justin_o: colinclark, another solution is to change the setting of "infusion" in "image-gallery-settings.ini"
[13:01:45 CDT(-0500)] <Bosmon2> Justin_o - I am here
[13:02:35 CDT(-0500)] <Justin_o> Bosmon2: hello.. did you have a chance to look over the ToC rendering thing again.. i figure you were busy with that !important conversation
[13:09:51 CDT(-0500)] <jhung> heidi, I'm going to take a look at your comments for Pull request 87 now.
[13:10:15 CDT(-0500)] <heidi_> jhung k lemee know if anything needs clarifying
[13:10:23 CDT(-0500)] <jhung> k
[13:11:02 CDT(-0500)] <huslage> fluid-everyone you really need to make sure that your builds are self-contained within continuum. Multiple directories is kind of a kludge right now, so having things inside one directory tree would make things work better.
[13:11:42 CDT(-0500)] <heidi_> huslage fluid builds? more context...
[13:11:59 CDT(-0500)] <anastasiac> huslage, can you elaborate a bit?
[13:13:05 CDT(-0500)] <huslage> heidi_: for instance this image-gallery contains a private infusion build…it should really live inside the directory tree for image-gallery (in lib or something), but it's currently a directory above the code. this creates a namespace collision in the URL mapping that really can't exist.
[13:13:32 CDT(-0500)] <Bosmon2> Justin_o - I am running your branch now
[13:13:51 CDT(-0500)] <Bosmon2> It fails on test 10, "linear headings" since the tree references a part of the model that doesn't exist?
[13:14:41 CDT(-0500)] <cindyli1> huslage: thanks for the help. a bit more: can u change http://build.fluidproject.org/image-gallery/image-gallery.php pointing back to "/home/continuum/apache-continuum-1.4.0/data/working-directory/22/image-gallery.php"
[13:14:44 CDT(-0500)] <Justin_o> Bosmon2: i didn't bother fixing up the unit tests yet
[13:15:00 CDT(-0500)] <huslage> you moved it cindyli1 ?
[13:15:03 CDT(-0500)] <Justin_o> Bosmon2: so i'm not sure.. i'll take a quick look.. so far i've just been using the demo to see it
[13:15:11 CDT(-0500)] <Justin_o> it's in the standalone demos directory
[13:15:13 CDT(-0500)] <cindyli1> not yet, in process, huslage
[13:15:14 CDT(-0500)] <Bosmon2> ok
[13:15:19 CDT(-0500)] <huslage> k
[13:15:25 CDT(-0500)] <cindyli1> thanks
[13:15:45 CDT(-0500)] <huslage> ok done cindyli1
[13:15:56 CDT(-0500)] <cindyli1> great. thanks
[13:16:48 CDT(-0500)] <Justin_o> Bosmon2: i think that error in the unit tests is because i have a static tree there.. so it makes a ton of assumptions about the model.. again this was just for quick testing and knowing how the tree should end up looking
[13:21:49 CDT(-0500)] <Justin_o> jameswy: uio question? should the fat panel close if you hit the esc key?
[13:23:14 CDT(-0500)] <jameswy> Justin_o: that's an issue heidi_, colinclark, and some others of us were chatting about last week. There's a bit of discussion about what key is most appropriate to close the panel.
[13:24:13 CDT(-0500)] <Justin_o> okay.. so at this point is that a feature we won't be supporting for 1.4?
[13:25:15 CDT(-0500)] <jameswy> If it's not difficult, I'd like it implemented to user test. But if there's no time for it, then no worries--I can user test the opposing case.
[13:27:08 CDT(-0500)] <Justin_o> jameswy: I don't think it would be too hard to capture some sort of keyboard event. i guess we need to decide which one.. though.. i guess it's really something that we need for the slidingPanel subcomponent as opposed to ui options in particular
[13:29:07 CDT(-0500)] <Justin_o> jameswy, heidi_, cindyli1, anastasiac: any of you know if we will be dropping the panel version of uio from the demos
[13:29:16 CDT(-0500)] <Justin_o> i guess the one that uses that still would be the sakai mock up
[13:29:58 CDT(-0500)] <jameswy> Justin_o: The panel version of UIO?
[13:31:01 CDT(-0500)] <heidi_> jameswy at the office, we're talking about replacing it with fat panel... it's the old implementation now
[13:31:24 CDT(-0500)] <jameswy> heidi_: Gotcha.
[13:31:38 CDT(-0500)] <anastasiac> actually, jameswy and heidi_, it is a fat panel now
[13:31:47 CDT(-0500)] <anastasiac> its styling is a bit wonky
[13:32:00 CDT(-0500)] <heidi_> ah, good to know
[13:32:35 CDT(-0500)] <colinclark> i'm back
[13:32:37 CDT(-0500)] <colinclark> and catching up on the logs
[13:32:43 CDT(-0500)] <colinclark> looks like quite the conversation on builds
[13:33:06 CDT(-0500)] <colinclark> damn colinclark and his lunch-eating
[13:33:13 CDT(-0500)] <colinclark> so unreasonable
[13:37:13 CDT(-0500)] <colinclark> huslage: you're a crazy dreamer
[13:37:28 CDT(-0500)] <colinclark> we wish the world worked this way
[13:38:38 CDT(-0500)] <colinclark> cindyli1: Can you summarize what you've done to the Image Gallery to get it working, if anything?
[13:39:39 CDT(-0500)] <colinclark> heidi_: What's "activation highlighting?"
[13:40:02 CDT(-0500)] <heidi_> mmm... when you click on something or press a button etc?
[13:40:23 CDT(-0500)] <cindyli1> colinclark: just get the image gallery builds work by modifying the shell script executed by continuum to change the "infusion" setting @ image-gallery-setting.ini, from "../infusion" to "infusion"
[13:40:24 CDT(-0500)] <colinclark> oh wait
[13:40:27 CDT(-0500)] <colinclark> this is from Justin_o
[13:40:41 CDT(-0500)] <colinclark> sorry, heidi_. I was asking about FLUID-4313
[13:40:44 CDT(-0500)] <colinclark> which Justin_o filed
[13:40:48 CDT(-0500)] <heidi_> i thought you might be css quizzing me
[13:40:51 CDT(-0500)] <colinclark> about "activation highlighting"
[13:40:57 CDT(-0500)] <colinclark> heidi_: you pass!
[13:41:01 CDT(-0500)] <colinclark> +1
[13:41:02 CDT(-0500)] <heidi_>
[13:41:48 CDT(-0500)] <Bosmon2> Hi Justin_o - I have found the cause of your issue
[13:42:09 CDT(-0500)] <Bosmon2> In the case that the renderer doesn't find any component in its current container, it will fall back to look for one in a parent
[13:42:31 CDT(-0500)] <Bosmon2> colinclark, you are quite right in asserting that the renderer should be burned to the ground
[13:42:54 CDT(-0500)] <Bosmon2> Justin_o - so you will need to give the links at different levels different ids as well as giving different ids to all of their containers
[13:43:09 CDT(-0500)] <Justin_o> oh really
[13:43:24 CDT(-0500)] <Justin_o> Bosmon2: okay..
[13:43:50 CDT(-0500)] <Justin_o> Bosmon2: so i think i can probably just go back to the old protostyle component tree then too
[13:43:56 CDT(-0500)] <Justin_o> sorry.. that came out wrong
[13:44:05 CDT(-0500)] <Justin_o> i think i can still use the new protoComponent style then too
[13:44:12 CDT(-0500)] <Bosmon2> Yes
[13:44:23 CDT(-0500)] <Bosmon2> It will make the tree slightly shorter
[13:44:39 CDT(-0500)] <colinclark> the old is new
[13:44:59 CDT(-0500)] <Bosmon2> "The new is merely the old that has been thoroughly forgotten!"
[13:45:32 CDT(-0500)] <colinclark> cindyli1: Okay, so for the daily build, infusion is now inside the image gallery's clone?
[13:45:35 CDT(-0500)] <colinclark> Does git like this okay?
[13:45:45 CDT(-0500)] <colinclark> I don't quite even know how these builds work
[13:45:47 CDT(-0500)] <cindyli1> colinclark: so far so good
[13:45:49 CDT(-0500)] <colinclark> maybe Justin_o or huslage does
[13:46:04 CDT(-0500)] <colinclark> but it seems odd to me to clone one repo inside a clone for another
[13:46:09 CDT(-0500)] <colinclark> and then expect it to update itself properly
[13:46:18 CDT(-0500)] <colinclark> but perhaps we blast the clones entirely
[13:46:21 CDT(-0500)] <colinclark> and reclone them?
[13:46:39 CDT(-0500)] <Bosmon2> blast the clones!
[13:46:43 CDT(-0500)] <Bosmon2> This marks the end of the Clone Wars
[13:46:52 CDT(-0500)] <colinclark> They were terrible movies, anyway
[13:46:55 CDT(-0500)] <cindyli1> lol
[13:47:04 CDT(-0500)] <Justin_o> Bosmon2: at least they had a guy from Toronto in them
[13:47:12 CDT(-0500)] <Justin_o> colinclark: ^
[13:47:23 CDT(-0500)] <cindyli1> colinclark: yes, fresh clone every time.
[13:47:31 CDT(-0500)] <colinclark> Does it make me cool, as far as nerds go, to be making Star Wars jokes?
[13:47:32 CDT(-0500)] <cindyli1> i agree that "it seems odd to me to clone one repo inside a clone for another"
[13:47:37 CDT(-0500)] <colinclark> I feel especially cool
[13:47:50 CDT(-0500)] <colinclark> which is why I mentioned that huslage is a bit of a dreamer, cindyli1
[13:48:06 CDT(-0500)] <Bosmon2> Because you suspect he is not the only one?
[13:48:09 CDT(-0500)] <cindyli1>
[13:48:15 CDT(-0500)] <colinclark> Bosmon2:
[13:48:27 CDT(-0500)] <colinclark> I guess it's the roughest edge in my idea for self-managed builds, which I was mentioning to michelled at lunch
[13:48:35 CDT(-0500)] <colinclark> dependencies across repositories really is an issue with git
[13:48:44 CDT(-0500)] <colinclark> maybe submodules can actually be made to work?
[13:48:59 CDT(-0500)] <Bosmon2> colinclark - might we suspect that they wouldn't work?
[13:49:08 CDT(-0500)] <colinclark> I thought Justin_o was of that mind, yes
[13:49:18 CDT(-0500)] <colinclark> I think I will propose switching back to SVN
[13:49:24 CDT(-0500)] <colinclark> I'm sure that will be popular
[13:49:38 CDT(-0500)] <Bosmon2> That can accompany our switch to backbone + jQuery templates
[13:50:02 CDT(-0500)] <colinclark>
[13:50:09 CDT(-0500)] <colinclark> Along with our Star Wars jokes, we will be so cool!
[13:50:17 CDT(-0500)] <Justin_o> Bosmon2, colinclark: submodules work in there own special way, which isn't the way that externals worked in svn
[13:50:44 CDT(-0500)] <Bosmon2> Justin_o - "can you elaborate?"
[13:50:47 CDT(-0500)] <Bosmon2> Justin_o - "huh?"
[13:51:22 CDT(-0500)] <colinclark> such a wacky channel today!
[13:51:28 CDT(-0500)] <Bosmon2> hahaha
[13:51:37 CDT(-0500)] <colinclark> michelled: while we're waiting for Justin_o to get back from his phone call
[13:51:41 CDT(-0500)] <colinclark> let's make a decision
[13:51:49 CDT(-0500)] <colinclark> So, I think most of the ideas are insane
[13:52:09 CDT(-0500)] <colinclark> Bosmon2's "server + HTML rewriting" solution is appealing from a maintenance perspective
[13:52:13 CDT(-0500)] <colinclark> but otherwise insane
[13:52:41 CDT(-0500)] <colinclark> My "keep two copies of the same thing in the repo, one of which has been autogenerated by a script that everyone will forget to run" idea is insane
[13:53:03 CDT(-0500)] <Bosmon2> I'm really eagerly awaiting which of these ideas is nominated as NOT insane
[13:53:13 CDT(-0500)] <colinclark> I guess we'll have to pick our poison
[13:53:16 CDT(-0500)] <colinclark> I'm mixing metaphors
[13:53:21 CDT(-0500)] <colinclark> Pick our mania?
[13:53:34 CDT(-0500)] <Bosmon2> poison our meat?
[13:53:51 CDT(-0500)] <Bosmon2> The "git commit hooks" idea is also entirely insane
[13:53:56 CDT(-0500)] <colinclark> yes
[13:53:58 CDT(-0500)] <colinclark> The last insane idea is to just leave UIO half-baked in the repo
[13:54:06 CDT(-0500)] <colinclark> and ask people to run a build if they are developing it
[13:56:43 CDT(-0500)] <Justin_o> Bosmon2: with submodules you don't get the code right away.. you have to make sure if you trigger them to pull in their resources
[13:56:56 CDT(-0500)] <huslage> man. go to the bathroom and you become a dreamer?
[13:56:58 CDT(-0500)] <Bosmon2> Justin_o - that's right - but is it a problem to trigger it?
[13:56:59 CDT(-0500)] <Justin_o> should be no "if" up there ^^
[13:57:08 CDT(-0500)] <Justin_o> well you have to remember to do this
[13:57:14 CDT(-0500)] <Bosmon2> Justin_o - isn't this running in a script?
[13:57:21 CDT(-0500)] <anastasiac> Justin_o, I just created http://issues.fluidproject.org/browse/FLUID-4314 - not sure if it's bug-parade material?
[13:57:31 CDT(-0500)] <Justin_o> well it's not that hard, but definitely harder than just being there
[13:57:45 CDT(-0500)] <Justin_o> if you have someone who is new to your code base and doesn't realize they are there.. it maybe confusing
[13:57:48 CDT(-0500)] <huslage> colinclark: why not just have a build of infusion that all of the things on build reference?
[13:57:53 CDT(-0500)] <huslage> or are there customizations?
[13:58:14 CDT(-0500)] <Bosmon2> Justin_o - I find plenty of projects checked in to github that have submodules....
[13:58:19 CDT(-0500)] <Bosmon2> It shouldn't necessarily be completely outlandish
[13:58:26 CDT(-0500)] <Justin_o> Bosmon2: plus with javascript i find it a bit annoying whenever i have to do something that resembles "compiling"
[13:58:29 CDT(-0500)] <colinclark> huslage: that's usually the way it works
[13:58:34 CDT(-0500)] <colinclark> (bathrooms and dreaming)
[13:58:36 CDT(-0500)] <michelled> colinclark: of all the insane ideas I favour the last one
[13:58:37 CDT(-0500)] <huslage> lol
[13:58:45 CDT(-0500)] <colinclark> lemme catch up
[13:58:45 CDT(-0500)] <Justin_o> Bosmon2: yes.. jquery ui i think uses them
[13:58:46 CDT(-0500)] <huslage> i will support whatever you decide
[13:58:49 CDT(-0500)] <colinclark> everyone is insane, all at once
[13:58:51 CDT(-0500)] <huslage> just trying to make sense of it
[13:58:53 CDT(-0500)] <Bosmon2> Justin_o - git submodule init resembles "compiling"?
[13:59:14 CDT(-0500)] <huslage> i can't make sense of the senseless
[13:59:22 CDT(-0500)] <Bosmon2> Can you make hap of the hapless?
[13:59:23 CDT(-0500)] <Justin_o> Bosmon2: only in the sense that i have to trigger something to get my code to work
[13:59:28 CDT(-0500)] <huslage> but having /infusion point to n-versions isn't feasible.
[13:59:28 CDT(-0500)] <Justin_o> it doesn't work out of the box
[13:59:35 CDT(-0500)] <Justin_o> or out of the clone as it is
[13:59:45 CDT(-0500)] <huslage> haps are obsolete Bosmon2
[13:59:46 CDT(-0500)] <huslage> sorry
[13:59:53 CDT(-0500)] <Bosmon2> How about MAD HAPS?
[13:59:57 CDT(-0500)] <colinclark> okay, my head is spinning
[13:59:58 CDT(-0500)] <Bosmon2> I hear that those are still pretty current
[14:00:07 CDT(-0500)] <colinclark> Let's briefly untangle these threads
[14:00:11 CDT(-0500)] <colinclark> michelled and Bosmon2: You two first
[14:00:18 CDT(-0500)] <Bosmon2> On which issue?
[14:00:24 CDT(-0500)] <colinclark> So, I think while I was dreaming in the bathroom, I just realized the weakness of the server idea
[14:00:25 CDT(-0500)] <Bosmon2> Submodules, or insane plans for !important?
[14:00:29 CDT(-0500)] <colinclark> !importnat
[14:00:30 CDT(-0500)] <huslage> rofl
[14:00:41 CDT(-0500)] <huslage> E_NO_NAT
[14:00:51 CDT(-0500)] <colinclark> So the !important server would only ever serve up some "official" version of the UIO sheets
[14:00:54 CDT(-0500)] <Bosmon2> One weakness of the server idea is that you would need to POST the contents you wanted transformed... it would be a bit of a mess
[14:00:59 CDT(-0500)] <colinclark> yeah
[14:01:00 CDT(-0500)] <colinclark> right
[14:01:02 CDT(-0500)] <colinclark> exactly
[14:01:06 CDT(-0500)] <Bosmon2> Not entirely impossible
[14:01:08 CDT(-0500)] <colinclark> no
[14:01:12 CDT(-0500)] <Bosmon2> But you would need to POST them, and then inject them
[14:01:24 CDT(-0500)] <Bosmon2> It would be rather unstable
[14:01:26 CDT(-0500)] <colinclark> So, let's investigate michelled's purportedly least insane option
[14:01:26 CDT(-0500)] <Bosmon2> Not to say, slow
[14:01:27 CDT(-0500)] <huslage> slow for the lusers
[14:01:36 CDT(-0500)] <colinclark> You'd clone Infusion
[14:01:46 CDT(-0500)] <colinclark> and out of the box, UIO wouldn't quite successfully switch styles properly
[14:01:48 CDT(-0500)] <huslage> infusion written in infusion
[14:01:51 CDT(-0500)] <Bosmon2> .
[14:01:54 CDT(-0500)] <colinclark> due to missing !importants
[14:01:58 CDT(-0500)] <colinclark> you'd have to make a build
[14:01:59 CDT(-0500)] <Bosmon2> Sorry, that was from the CATTT
[14:02:04 CDT(-0500)] <colinclark> what a good cat!
[14:02:19 CDT(-0500)] <colinclark> reminding us of the importance of punctuation!
[14:02:23 CDT(-0500)] <Bosmon2>
[14:02:26 CDT(-0500)] <colinclark> okay
[14:02:32 CDT(-0500)] <colinclark> so, michelled and Bosmon2
[14:02:37 CDT(-0500)] <colinclark> how insane is this insanity?
[14:02:42 CDT(-0500)] <Bosmon2> It seems pretty insane
[14:02:55 CDT(-0500)] <Bosmon2> How about another option - we hack the parsing codebase so it can run nicely in the browser
[14:04:07 CDT(-0500)] <michelled> why would we want to do that?
[14:04:15 CDT(-0500)] <huslage> colinclark: can you restate the problem so that we can get back to sanity?
[14:04:25 CDT(-0500)] <huslage> i feel we've gone off into neverneverland
[14:04:35 CDT(-0500)] <Bosmon2> UIO could then transform its own stylesheets on the fly
[14:04:41 CDT(-0500)] <colinclark> huslage: Well, this issue seems related...
[14:04:48 CDT(-0500)] <colinclark> since we're using the word "build"
[14:04:52 CDT(-0500)] <huslage> (isn't it always)
[14:05:00 CDT(-0500)] <colinclark> but actually is totally separate from your issue with the builds
[14:05:09 CDT(-0500)] <colinclark> so michelled, Bosmon2 and I are trying to figure out a strategy...
[14:05:12 CDT(-0500)] <colinclark> here's what we're doing
[14:05:23 CDT(-0500)] <colinclark> Our UI Options component has the ability to re-theme a page
[14:05:27 CDT(-0500)] <colinclark> the user comes along, picks a theme
[14:05:38 CDT(-0500)] <colinclark> and UI Options will replace the page's look and feel with their chosen them
[14:05:48 CDT(-0500)] <colinclark> Interestingly, we've found that there is occasionally the need for a CSS "hammer"
[14:06:03 CDT(-0500)] <colinclark> So if the implement's base theme overrides a bunch of styles
[14:06:10 CDT(-0500)] <colinclark> with great specificity
[14:06:14 CDT(-0500)] <colinclark> we have to beat them
[14:06:38 CDT(-0500)] <colinclark> the last resort technique we need in a handful of cases is to ensure that the theme has an !important declaration on some rules
[14:07:04 CDT(-0500)] <colinclark> In the last versions of Infusion, we sort of awkwardly shipped all of our Fluid Skinning System themes with these !importants baked in
[14:07:10 CDT(-0500)] <colinclark> Which, as you can imagine, drove our users nuts
[14:07:16 CDT(-0500)] <Bosmon2> colinclark - one interesting issue, is how we know WHEN we are meant to apply our !important?
[14:07:19 CDT(-0500)] <colinclark> since they'd have to get in an arms race with us if they wanted to override something
[14:07:22 CDT(-0500)] <Bosmon2> Is this actually an option within UIOptions?
[14:07:35 CDT(-0500)] <colinclark> Bosmon2: We know that the user has requested a theme
[14:07:45 CDT(-0500)] <colinclark> So we know that UI Options' version of the theme has to win at all costs
[14:07:56 CDT(-0500)] <colinclark> so really we just need, in its most basic form, a special version of each of themes
[14:07:59 CDT(-0500)] <colinclark> built to win
[14:08:01 CDT(-0500)] <Bosmon2> Interesting...
[14:08:08 CDT(-0500)] <Bosmon2> But when are the OTHER versions of the themes applied?
[14:08:10 CDT(-0500)] <huslage> colinclark: can't there just be a list of what's !important living somewhere that's canonical and planned?
[14:08:17 CDT(-0500)] <huslage> other than in the code
[14:08:24 CDT(-0500)] <huslage> so you can deprecate it?
[14:08:41 CDT(-0500)] <colinclark> huslage: See, the thing is at some point, !important has to get into the sheets
[14:08:42 CDT(-0500)] <colinclark> we have to win
[14:08:50 CDT(-0500)] <colinclark> So, really, there are two different types of themes
[14:08:57 CDT(-0500)] <colinclark> A theme people use and customize
[14:09:06 CDT(-0500)] <colinclark> and a theme UI Options uses to smack down the other themes
[14:09:23 CDT(-0500)] <Bosmon2> I see
[14:09:25 CDT(-0500)] <colinclark> We can have some configuration file that lists which types of rules need important on them
[14:09:34 CDT(-0500)] <colinclark> but somewhere, some code has to put them there
[14:09:35 CDT(-0500)] <Bosmon2> So, it is really only UIOptions' use of the themes which needs to be crushing
[14:09:38 CDT(-0500)] <colinclark> or we can cut and paste everything
[14:09:40 CDT(-0500)] <colinclark> which is horrific
[14:09:45 CDT(-0500)] <colinclark> Bosmon2: yes, exactly
[14:09:48 CDT(-0500)] <Bosmon2> If people are using the themes "directly", they expect them to be civilized?
[14:09:50 CDT(-0500)] <Bosmon2> ok
[14:09:55 CDT(-0500)] <colinclark> If you think about it, UIO is what !important was made for
[14:10:01 CDT(-0500)] <colinclark> things like UIO, that is
[14:10:10 CDT(-0500)] <colinclark> Bosmon2: yes
[14:10:19 CDT(-0500)] <colinclark> heidi_ and Justin_o took !important out of all our themes
[14:10:23 CDT(-0500)] <colinclark> in order to make them civilized
[14:10:25 CDT(-0500)] <Bosmon2> ok - so, how about the option of transforming the CSS on the fly, using a hacked version of the parser?
[14:10:29 CDT(-0500)] <colinclark> but now UIO needs its sledgehammer back
[14:10:35 CDT(-0500)] <colinclark> Bosmon2: We could definitely do that
[14:10:36 CDT(-0500)] <Bosmon2> I have the feeling that UIoptions is going to have to need to parse CSS in the medium term in any case
[14:10:39 CDT(-0500)] <Bosmon2> So we should probably get used to it
[14:10:43 CDT(-0500)] <colinclark> I suspect that is actually the case
[14:10:53 CDT(-0500)] <colinclark> michelled and I were talking about exactly this on the way to lunch
[14:11:00 CDT(-0500)] <colinclark> before we started talking about the other issue
[14:11:04 CDT(-0500)] <colinclark> so, Bosmon2 has a new proposal
[14:11:05 CDT(-0500)] <Bosmon2> Our "clearClashingStyles" etc. is clearly the tip of a very deep iceberg
[14:11:16 CDT(-0500)] <colinclark> michelled: we suck it up and dynamically generate sheets on the fly
[14:11:26 CDT(-0500)] <colinclark> presumably inlining them at the bottom or top of the document
[14:11:33 CDT(-0500)] <colinclark> so they really, really will win
[14:12:08 CDT(-0500)] <colinclark> thoughts, michelled?
[14:12:14 CDT(-0500)] <colinclark> I know you're distracted by insanity in other channels
[14:12:17 CDT(-0500)] <colinclark> the real kind of insanity
[14:13:33 CDT(-0500)] <colinclark> huslage: if you're still around in a sec, let's talk about the other build issue
[14:13:39 CDT(-0500)] <colinclark> once michelled has a chance to weigh in here
[14:13:45 CDT(-0500)] <huslage> i'm not going anywhere. this is too insane to miss
[14:13:49 CDT(-0500)] <colinclark>
[14:14:05 CDT(-0500)] <huslage> seems like you really want an infusion proxy service to some degree.
[14:14:25 CDT(-0500)] <michelled> colinclark: it seems like a lot to bite off at this point in the release
[14:14:27 CDT(-0500)] <colinclark> for these style sheets, you mean?
[14:14:39 CDT(-0500)] <colinclark> michelled: is it much more to bite off than doing it statically?
[14:14:49 CDT(-0500)] <michelled> how much work will it be to hack the parser?
[14:14:54 CDT(-0500)] <colinclark> ah, yes
[14:14:57 CDT(-0500)] <colinclark> the parser sort of sucks
[14:15:04 CDT(-0500)] <michelled> also, I didn't check, are the licenses compatible? can we hack it?
[14:15:05 CDT(-0500)] <colinclark> its use of const will cause IE to barf
[14:15:11 CDT(-0500)] <colinclark> yeah, it's Mozilla tri-licensed
[14:15:15 CDT(-0500)] <colinclark> MPL is happily compatible
[14:15:20 CDT(-0500)] <michelled> cool
[14:15:31 CDT(-0500)] <colinclark> Bosmon2: We'd have to actually fork this thing
[14:15:43 CDT(-0500)] <colinclark> Not like we haven't forked other similar parsers
[14:15:51 CDT(-0500)] <colinclark> michelled: I guess that's probably the biggest work difference
[14:16:07 CDT(-0500)] <colinclark> we'd have to isolate it to a namespace and make it work in IE
[14:16:41 CDT(-0500)] <Bosmon2> Yes
[14:16:50 CDT(-0500)] <Bosmon2> It wouldn't be too awful
[14:16:54 CDT(-0500)] <colinclark> michelled: "how long?"
[14:16:56 CDT(-0500)] <colinclark>
[14:17:07 CDT(-0500)] <colinclark> 43 minutes!
[14:17:14 CDT(-0500)] <Bosmon2> A couple of hours work, with a following wind....
[14:17:17 CDT(-0500)] <michelled> not 42?
[14:17:19 CDT(-0500)] <michelled>
[14:17:24 CDT(-0500)] <Bosmon2> I thought it was 49
[14:17:47 CDT(-0500)] <michelled> colinclark: talk me through the work flow just so I understand
[14:17:54 CDT(-0500)] <colinclark> Ok
[14:17:55 CDT(-0500)] <colinclark> So
[14:17:56 CDT(-0500)] <michelled> when would the parser actually come into action
[14:18:02 CDT(-0500)] <colinclark> User requests theme
[14:18:04 CDT(-0500)] <colinclark> model changes
[14:18:24 CDT(-0500)] <colinclark> New fancy component listens to change applier changes to the theme in the model
[14:18:39 CDT(-0500)] <colinclark> New fancy component finds the < link > tag for the specified theme
[14:19:02 CDT(-0500)] <colinclark> it has configuration for what sorts of CSS rules need to be Super Powered
[14:19:08 CDT(-0500)] <colinclark> background colours and images, I believe
[14:19:15 CDT(-0500)] <colinclark> so it parses the CSS
[14:19:17 CDT(-0500)] <huslage> i have a dumb newb question. what about something like a minifier that pre-compiles the CSS? so you put the onus on the server/website builder. make a PHP lib that can do it inline or something
[14:19:20 CDT(-0500)] <colinclark> searches for related rules
[14:19:30 CDT(-0500)] <huslage> or is that what we DON'T want to do?
[14:19:39 CDT(-0500)] <colinclark> huslage: no, you've got something
[14:19:46 CDT(-0500)] <colinclark> it was the suggestion Bosmon2 made earlier, I think
[14:19:55 CDT(-0500)] <huslage> ok. i thought so but couldn't tell
[14:21:15 CDT(-0500)] <michelled> colinclark: what about the next time the user uses the page - meaning they don't use UIO to set a setting but expect their settings to already be there
[14:21:18 CDT(-0500)] <Bosmon2> We can stuff the thing in a cookie or in some other place - or follow this strategy in conjunction with a server-based strategy
[14:21:28 CDT(-0500)] <Bosmon2> Right now our priority is to get a component that just actually works out of the box
[14:21:33 CDT(-0500)] <huslage> ok
[14:21:36 CDT(-0500)] <huslage> iterate
[14:21:38 CDT(-0500)] <huslage> got it
[14:21:42 CDT(-0500)] <huslage> i'm throwing out bombs
[14:21:49 CDT(-0500)] <colinclark> What's our limit on cookie size?
[14:21:52 CDT(-0500)] <colinclark> 4K?
[14:21:53 CDT(-0500)] <Bosmon2> 4k?
[14:21:55 CDT(-0500)] <Bosmon2> I believe so
[14:21:56 CDT(-0500)] <colinclark>
[14:22:02 CDT(-0500)] <colinclark> huslage: do you know, off the top of your head?
[14:22:07 CDT(-0500)] <huslage> 4k
[14:22:12 CDT(-0500)] <huslage> recommended by good is 2k
[14:22:13 CDT(-0500)] <colinclark> IE6 is probably the worst of them
[14:22:13 CDT(-0500)] <Bosmon2> Anyway, I don't think I was really serious about the cookies
[14:22:14 CDT(-0500)] <huslage> goog
[14:22:29 CDT(-0500)] <huslage> you could do html5 storage
[14:22:30 CDT(-0500)] <huslage> at worst
[14:22:33 CDT(-0500)] <colinclark> we could
[14:22:35 CDT(-0500)] <Bosmon2> In a serious production situation we would actually do this on the server, "getting into the CDN" business as you put it
[14:22:37 CDT(-0500)] <colinclark> but we'd have to back it with Flash for IE6
[14:22:40 CDT(-0500)] <huslage> db init is slow, but everything else isn't
[14:22:47 CDT(-0500)] <huslage> f that
[14:22:53 CDT(-0500)] <huslage> IE6 is dead
[14:22:57 CDT(-0500)] <huslage> long live IE9
[14:23:03 CDT(-0500)] <colinclark> IE8, really
[14:23:07 CDT(-0500)] <Bosmon2> 3.7% market share, unfortunately, doesn't quite count as being entirely dead
[14:23:11 CDT(-0500)] <huslage> no one supports IE6 anymore do they?
[14:23:17 CDT(-0500)] <colinclark> Since IE9 doesn't run on the platform everyone actually uses
[14:23:18 CDT(-0500)] <huslage> (ugh this again)
[14:23:31 CDT(-0500)] <colinclark> huslage: Unfortunately many large companies and governments do
[14:23:35 CDT(-0500)] <Bosmon2> Anyway, do we care if the "out of the box" component is a little slow?
[14:23:37 CDT(-0500)] <huslage> so use flash on ie6. we already have flash components, no?
[14:23:37 CDT(-0500)] <colinclark> Anyway, the whole IE angle...
[14:23:42 CDT(-0500)] <colinclark> Let's shelf it until we ship
[14:23:48 CDT(-0500)] <colinclark> After we ship, I have an idea
[14:23:54 CDT(-0500)] <colinclark> terrifying
[14:23:55 CDT(-0500)] <huslage> proxy
[14:24:04 CDT(-0500)] <huslage> varnish cache.
[14:24:06 CDT(-0500)] <Bosmon2> Each of these "transformed" CSS sheets should have a well-defined URL
[14:24:06 CDT(-0500)] <huslage> done
[14:24:11 CDT(-0500)] <huslage> node.js websockets
[14:24:40 CDT(-0500)] <colinclark> There's a lot of sense in what you and Bosmon2 are saying, huslage
[14:24:45 CDT(-0500)] <michelled> Bosmon2: I don't think that most people will do more then use the out of the box component. do you?
[14:24:47 CDT(-0500)] <huslage> it's open source. get google or amazon to host the service. they already run that crap anyway
[14:24:48 CDT(-0500)] <colinclark> but we are opening up a big can of worms this late
[14:25:13 CDT(-0500)] <Bosmon2> michelled - I thought we were just talking about an initial developer experience
[14:25:17 CDT(-0500)] <huslage> why are you people all so smart?
[14:25:28 CDT(-0500)] <Bosmon2> Someone who actually USES UIOptions will actually at least go to the trouble of making an Infusion build, right?
[14:25:41 CDT(-0500)] <Bosmon2> Which will, by our Insane Plan #3, actually produce a physical transformed CSS file
[14:26:05 CDT(-0500)] <huslage> Insane Plan #666 - Drupal, Wordpress, Joomla, etc modules to bake it in.
[14:26:05 CDT(-0500)] <colinclark> Which puts us in the position of having to do #3 and #4 simultaneously
[14:26:06 CDT(-0500)] <huslage> lol
[14:26:13 CDT(-0500)] <Bosmon2> CATTT is trying to get inside the PC again......
[14:26:14 CDT(-0500)] <colinclark> which agues for just doing #3 for now
[14:26:21 CDT(-0500)] <colinclark> Bosmon2: He has some kind of master plan
[14:26:31 CDT(-0500)] <colinclark> or he just wants to chew on a microprocessor
[14:26:35 CDT(-0500)] <colinclark> who knows with cats?
[14:26:39 CDT(-0500)] <Bosmon2> #3, you mean, "just putting it in the builder"?
[14:26:41 CDT(-0500)] <huslage> mmm chips
[14:26:44 CDT(-0500)] <colinclark> Bosmon2: yes
[14:26:49 CDT(-0500)] <Bosmon2> And "telling developers to lump it"
[14:26:54 CDT(-0500)] <colinclark> yep
[14:26:56 CDT(-0500)] <colinclark> users get the full experience
[14:26:59 CDT(-0500)] <Bosmon2> ok
[14:27:05 CDT(-0500)] <colinclark> it's just us who have to have a sucky time of it
[14:27:12 CDT(-0500)] <colinclark> I feel like there are so many things converging on the FSS
[14:27:13 CDT(-0500)] <huslage> The Infustion "we hate developers" release
[14:27:15 CDT(-0500)] <Bosmon2> That seems a reasonable tradeoff
[14:27:16 CDT(-0500)] <colinclark> concat and minify
[14:27:30 CDT(-0500)] <colinclark> possible embracing of this parser for supporting variables and mixins some day
[14:27:30 CDT(-0500)] <Bosmon2> I remember were were in this territory with Engage a lot
[14:27:35 CDT(-0500)] <colinclark> UIO generating CSS
[14:27:35 CDT(-0500)] <colinclark> etc.
[14:27:37 CDT(-0500)] <Bosmon2> And we suffered, how we suffered
[14:27:42 CDT(-0500)] <colinclark> woe is us
[14:27:45 CDT(-0500)] <huslage> sad face
[14:27:49 CDT(-0500)] <huslage> FIRST WORLD PROBLEMS
[14:27:51 CDT(-0500)] <colinclark> which reminds me of another issue i'm quite confused by
[14:27:53 CDT(-0500)] <colinclark> huslage:
[14:27:57 CDT(-0500)] <colinclark> touché
[14:28:01 CDT(-0500)] <huslage> http://swiffy.googlelabs.com/
[14:28:08 CDT(-0500)] <huslage> i'm confused by huslage too
[14:28:11 CDT(-0500)] <huslage> who is that guy?
[14:28:37 CDT(-0500)] <colinclark> the guy who wrote swiffy?
[14:28:45 CDT(-0500)] <heidi_> jhung sent a pull request your way re: text panel in IE7
[14:28:56 CDT(-0500)] <colinclark> Bosmon2: I notice that our demo instance of Engage is actually subtly dead
[14:29:13 CDT(-0500)] <jhung> heidi: okay
[14:29:24 CDT(-0500)] <colinclark> I was wondering about just getting huslage to reboot the JVM and its CouchDB and see if magic happnes
[14:29:34 CDT(-0500)] <colinclark> But it really isn't immediately explainable
[14:29:42 CDT(-0500)] <huslage> which?
[14:30:28 CDT(-0500)] <colinclark> huslage: The one running at fluidengage.org
[14:30:45 CDT(-0500)] <colinclark> There is an instance of Kettle and the Fluid Engage app running there in a JVM
[14:30:47 CDT(-0500)] * huslage looks it up on the handy dandy wikisite
[14:30:55 CDT(-0500)] <colinclark> as well as a CouchDB database, from which is pulls museum data
[14:31:14 CDT(-0500)] <huslage> ok
[14:31:15 CDT(-0500)] <colinclark> I guess it probably can't hurt to kick it over, just in the hope that it'll solve the problem
[14:31:30 CDT(-0500)] <colinclark> I can't imagine any of us have the time to triage it right now
[14:31:37 CDT(-0500)] <colinclark> but it is an app we like to have available to do demos with
[14:31:52 CDT(-0500)] <colinclark> In fact I discovered its borkedness while trying to show it to someone
[14:32:07 CDT(-0500)] <colinclark> michelled: Okay, #3, then?
[14:32:20 CDT(-0500)] <michelled> yes
[14:32:22 CDT(-0500)] <colinclark> With an eye towards making it suck less in 1.5?
[14:32:25 CDT(-0500)] <huslage> restarted colinclark
[14:32:27 CDT(-0500)] <huslage> magic
[14:32:28 CDT(-0500)] <colinclark> thanks, huslage
[14:32:29 CDT(-0500)] <michelled> makes sense to me
[14:33:05 CDT(-0500)] <colinclark> huslage: Too bad
[14:33:09 CDT(-0500)] <colinclark> It runs, but the same issue
[14:33:13 CDT(-0500)] <colinclark> It could well be a real error
[14:33:16 CDT(-0500)] <huslage> i didn't do it
[14:33:19 CDT(-0500)] <colinclark> no, i know
[14:33:21 CDT(-0500)] <huslage> promise
[14:33:22 CDT(-0500)] <huslage> not i
[14:33:27 CDT(-0500)] <colinclark> you're cool
[14:33:28 CDT(-0500)] <huslage> i didn't rm -rf anything
[14:33:29 CDT(-0500)] <huslage> really
[14:33:32 CDT(-0500)] <colinclark>
[14:33:38 CDT(-0500)] <colinclark> something to put on our to do list post-release
[14:33:53 CDT(-0500)] <colinclark> It's complaining that the artifact data feed is returning empty artifacts
[14:34:19 CDT(-0500)] <colinclark> If I remember, Couch is firewalled so I can't get at it without doing some kind of SSH tunnelling magic
[14:34:26 CDT(-0500)] <colinclark> so, for later
[14:34:32 CDT(-0500)] <colinclark> okay, so the other build issue, huslage
[14:34:48 CDT(-0500)] <colinclark> cindyli1: Did you get your change to the Image Gallery pushed into the project repo?
[14:34:52 CDT(-0500)] <colinclark> we're cool on that end?
[14:35:23 CDT(-0500)] <cindyli1> colinclark: not yet. i sent a pull request for the image gallery build shell script used by continuum: https://github.com/fluid-project/build-scripts/pull/1. can u or someone code review it? thanks.
[14:35:30 CDT(-0500)] <colinclark> yep, I'll look now
[14:35:59 CDT(-0500)] <cindyli1> thanks
[14:36:53 CDT(-0500)] <colinclark> cindyli1: I don't see the pull request
[14:37:15 CDT(-0500)] <cindyli1> colinclark: in build-scripts project, not "infusion"
[14:37:22 CDT(-0500)] <cindyli1> or not "image-gallery"
[14:37:25 CDT(-0500)] <colinclark> ahh
[14:37:40 CDT(-0500)] <colinclark> this one! https://github.com/fluid-project/build-scripts/pull/1
[14:37:44 CDT(-0500)] <colinclark> thanks, cindyli1
[14:37:47 CDT(-0500)] <cindyli1> exactly
[14:38:20 CDT(-0500)] <cindyli1> colinclark: i should probably remove that <form> submission that huslage discovered, which is now useless.
[14:38:27 CDT(-0500)] <cindyli1> that would be on "image-gallery"
[14:38:34 CDT(-0500)] <colinclark> that makes sense to me
[14:38:38 CDT(-0500)] <colinclark> whenever you get a chance
[14:38:47 CDT(-0500)] <cindyli1> ok, creating jira
[14:38:50 CDT(-0500)] <Bosmon2> colinclark - I'm trying to get the FileQueueView in your branch working again and finding a return of the "that is undefined" error
[14:39:02 CDT(-0500)] <colinclark> Bosmon2: I have it doing that currently
[14:39:04 CDT(-0500)] <Bosmon2> In this case, since it is expecting an argument called "that.options.uploaderContainer"
[14:39:28 CDT(-0500)] <colinclark> wait, where is this argument?
[14:39:41 CDT(-0500)] <Bosmon2> FileQueueView.js, line 190
[14:39:47 CDT(-0500)] <colinclark> cindyli1: I guess we should make sure that the single file upload is still POSTing correctly, after my mangling
[14:39:55 CDT(-0500)] <colinclark> Bosmon2: ?!?!
[14:39:57 CDT(-0500)] <colinclark> yack
[14:39:58 CDT(-0500)] <colinclark> lemme look
[14:40:02 CDT(-0500)] <colinclark> okay
[14:40:11 CDT(-0500)] <colinclark> 1. Review and pull cindyli1's request
[14:40:22 CDT(-0500)] <colinclark> 2. Look at my insanity in FileQueueView
[14:40:26 CDT(-0500)] <colinclark> 3. Talk to huslage about builds
[14:40:33 CDT(-0500)] <cindyli1> colinclark: aha, making sense. trying
[14:41:06 CDT(-0500)] <jhung> heidi: I'm looking at your comments for Uploader styling...
[14:41:16 CDT(-0500)] <colinclark> cindyli1 and her insanely good sed skillz
[14:41:21 CDT(-0500)] <colinclark> Lemme ask you one question, cindyli1
[14:41:27 CDT(-0500)] <colinclark> Why use sed to change this value on the build server
[14:41:37 CDT(-0500)] <colinclark> rather than changing it as the default in the Image Gallery?
[14:41:44 CDT(-0500)] <colinclark> Just to make life easier for the developer?
[14:42:06 CDT(-0500)] <jhung> You said the upload button text is small. It looks fine here - 1.0em (13px)
[14:42:35 CDT(-0500)] <cindyli1> the least is that would involve change to README. and justin and i figured developers, and maybe some users, would prefer to have "infusion" and "image-gallery" as sibling folders
[14:42:53 CDT(-0500)] <Bosmon2> colinclark - in terms of FLUID-4285 - you mentioned that you found a different set of tests failing too?
[14:43:03 CDT(-0500)] <Bosmon2> One where it complained about "container" being unexpected?
[14:43:12 CDT(-0500)] <colinclark> yes, that's right
[14:43:15 CDT(-0500)] <colinclark> Bosmon2 ^
[14:43:22 CDT(-0500)] <colinclark> cindyli1: hmm
[14:43:27 CDT(-0500)] <colinclark> I don't know which is better
[14:43:32 CDT(-0500)] <Bosmon2> colinclark ^?
[14:43:33 CDT(-0500)] <cindyli1> us too
[14:43:42 CDT(-0500)] <colinclark> So, the risk with this approach...
[14:43:59 CDT(-0500)] <colinclark> is that if a developer ever wanted to change the structure or naming of the gallery's .ini file
[14:44:06 CDT(-0500)] <colinclark> the build server would probably just start failing
[14:44:07 CDT(-0500)] <cindyli1> exactly
[14:44:11 CDT(-0500)] <colinclark> without any apparent reason
[14:44:17 CDT(-0500)] <colinclark> "Works on my system!"
[14:44:20 CDT(-0500)] <colinclark> that kind of thing
[14:44:35 CDT(-0500)] <cindyli1> and if some one changes even one character of that line
[14:44:43 CDT(-0500)] <colinclark> The downside to the other approach is that most developers with the image gallery would have to change the ini file and remember not to accidentally commit it
[14:44:46 CDT(-0500)] <colinclark> at risk of breaking the build
[14:46:03 CDT(-0500)] <colinclark> ok
[14:46:14 CDT(-0500)] <colinclark> so the issue is that we're trying to mount the Image Gallery at /image-gallery?
[14:46:18 CDT(-0500)] <huslage> unless there is a canonical build of infusion that i can point at
[14:46:23 CDT(-0500)] <huslage> yes colinclark
[14:46:29 CDT(-0500)] <colinclark> ok
[14:46:48 CDT(-0500)] <colinclark> So the reason I said you were a dreamer...
[14:47:02 CDT(-0500)] <colinclark> was that this type of cross-repo dependency is sort of complex to manage
[14:47:07 CDT(-0500)] <colinclark> SVN had it all sorted out for us
[14:47:12 CDT(-0500)] <colinclark> there were "externals"
[14:47:24 CDT(-0500)] <colinclark> where you could refer to another URL, and it would be pulled into a checkout automatically
[14:47:33 CDT(-0500)] <huslage> that exists in diff
[14:47:34 CDT(-0500)] <colinclark> so you could have these easily self-contained checkouts
[14:47:36 CDT(-0500)] <huslage> err git
[14:47:40 CDT(-0500)] <colinclark> Yeah, submodules
[14:47:43 CDT(-0500)] <colinclark> but they don't quite work the same way
[14:47:46 CDT(-0500)] <huslage> no
[14:47:48 CDT(-0500)] <colinclark> they require user intervention
[14:47:58 CDT(-0500)] <colinclark> which, in the case of a daily build server, seems fine to me
[14:48:09 CDT(-0500)] <colinclark> I guess we'll ping Justin_o about his concerns tomorrow
[14:48:29 CDT(-0500)] <huslage> is the image-gallery infusion a special build or just the regular repo?
[14:48:46 CDT(-0500)] <huslage> because isn't there already a daily build of infusion on this server?
[14:48:47 CDT(-0500)] <colinclark> It's a separate repo
[14:48:50 CDT(-0500)] <colinclark> and depends on Infusion
[14:49:02 CDT(-0500)] <colinclark> there is, I believe, already a daily build of infusion on this server
[14:49:11 CDT(-0500)] <colinclark> I guess we can think through whether a cross-build dependency is okay
[14:49:18 CDT(-0500)] <colinclark> Superficially, it might be just fine
[14:49:30 CDT(-0500)] <huslage> so image-gallery extends infusion?
[14:49:42 CDT(-0500)] <colinclark> No, image-gallery uses Infusion
[14:49:52 CDT(-0500)] <colinclark> It's just a little PHP app
[14:49:57 CDT(-0500)] <huslage> yeah
[14:50:01 CDT(-0500)] <colinclark> not unlike the sort of apps one of our users might build
[14:50:10 CDT(-0500)] <huslage> so what's wrong with pointing to the regular nightly build of infusion?
[14:50:23 CDT(-0500)] <Bosmon2> cindyli1, michelled, colinclark - in fixing up FLUID-4285, I am thinking that this diagnostic is exposing what is actually a genuine error in UIEnhancer
[14:50:52 CDT(-0500)] <Bosmon2> This is a diagnostic suggested by athena's initial experiences in using the framework - in that it will now complain at finding unexpected entries in a "subcomponent record"
[14:51:08 CDT(-0500)] <Bosmon2> In the checkout I have right now, this section appears at line 261 in UIEnhancer.js
[14:51:22 CDT(-0500)] <Bosmon2> settingsStore: {
[14:51:22 CDT(-0500)] <Bosmon2> type: "fluid.uiOptions.store",
[14:51:22 CDT(-0500)]

<Bosmon2> container: "

Unknown macro: {uiEnhancer}

.container",

[14:51:22 CDT(-0500)] <Bosmon2> createOnEvent: "onInitSettingStore"
[14:51:22 CDT(-0500)] <Bosmon2> }
[14:51:32 CDT(-0500)] <Bosmon2> As far as I can see, the "container" option is actually unnecessary here?
[14:52:24 CDT(-0500)] <colinclark> huslage: I'm not sure
[14:52:56 CDT(-0500)] <colinclark> Bosmon2: I guess michelled has a branch of something new
[14:53:15 CDT(-0500)] <colinclark> huslage: Off the top of your head, can you think of any reason why it's a problem?
[14:53:25 CDT(-0500)] <huslage> build breaks, demo breaks
[14:53:28 CDT(-0500)] <colinclark> What sorts of subtle, unnoticed problems could crop up if, say, one build fails
[14:53:31 CDT(-0500)] <colinclark> yeah
[14:53:39 CDT(-0500)] <colinclark> To be specific
[14:53:41 CDT(-0500)] <colinclark> Build fails
[14:53:47 CDT(-0500)] <colinclark> Demo succeeds, yet breaks
[14:54:00 CDT(-0500)] <colinclark> But I guess the thing about Infusion is that not much is really "built"
[14:54:03 CDT(-0500)] <huslage> yeah
[14:54:14 CDT(-0500)] <colinclark> So the other risk would be that Infusion fails, demo succeeds but because of stale code
[14:54:19 CDT(-0500)] <huslage> and the dependency might actually be good
[14:54:21 CDT(-0500)] <colinclark> but I guess that's unlikely to be a silent failure
[14:54:29 CDT(-0500)] <colinclark> if Infusion dies, everything should be considered suspect
[14:54:40 CDT(-0500)] <huslage> if you ran selenium tests or something against the demos, you'd know something bombed
[14:54:51 CDT(-0500)] <colinclark> How would the dependency be expressed, huslage?
[14:54:57 CDT(-0500)] <huslage> it wouldn't
[14:54:59 CDT(-0500)] <huslage> not formally
[14:55:07 CDT(-0500)] <colinclark> somehow, it would have to
[14:55:17 CDT(-0500)] <colinclark> in that we need to refer to a stable URL
[14:55:19 CDT(-0500)] <huslage> other than image-gallery knows the canonical location of the daily build
[14:55:34 CDT(-0500)] <colinclark> but remember that developers will develop locally with the image-gallery
[14:55:39 CDT(-0500)] <huslage> build.fluidproject.org/daily/infusion
[14:55:43 CDT(-0500)] <huslage> or something
[14:55:43 CDT(-0500)] <colinclark> Like, I have PHP enabled on my Mac
[14:55:56 CDT(-0500)] <colinclark> and I wouldn't want it talking to a version of Infusion on the build server
[14:56:01 CDT(-0500)] <michelled> Bosmon2: good catch - it's still a bug even in the 4207 branch
[14:56:21 CDT(-0500)] <huslage> colinclark: the install script should configure it, not sitting as a default
[14:56:27 CDT(-0500)] <huslage> sorry misunderstood you
[14:56:35 CDT(-0500)] <colinclark> okay
[14:56:42 CDT(-0500)] <colinclark> so essentially what cindyli1 has
[14:56:43 CDT(-0500)] <huslage> it does this now. 'sed's in the right place
[14:56:44 CDT(-0500)] <huslage> yeah
[14:56:56 CDT(-0500)] <huslage> that's reasonable
[14:56:59 CDT(-0500)] <Bosmon2> michelled - shall I fix it in 4285, which I guess will get pushed shortly, or do you want to do it in one of your branches?
[14:57:07 CDT(-0500)] <colinclark> huslage: I guess partially reasonable, at least
[14:57:10 CDT(-0500)] <huslage> it's unreasonable to have multiple "builds" of infusion.
[14:57:15 CDT(-0500)] <colinclark> The worry being this scenario
[14:57:26 CDT(-0500)] <colinclark> Say I get really excited about refactoring the Image Gallery
[14:57:30 CDT(-0500)] <colinclark> (this has happened

)
[14:57:46 CDT(-0500)] <colinclark> I rename and restructure all the variable in the config file
[14:57:50 CDT(-0500)] <michelled> Bosmon2: there is only the one branch and it's awaiting review - I'm ok with either but the reviewer might have to deal with a conflict if it's put into trunk first
[14:57:57 CDT(-0500)] <colinclark> Everything works really nicely on my local machine
[14:58:00 CDT(-0500)] <colinclark> but then fails on the build server
[14:58:10 CDT(-0500)] <colinclark> huslage: Something we have to just live with?
[14:58:17 CDT(-0500)] <huslage> yes
[14:58:24 CDT(-0500)] <huslage> i don't see why it would fail
[14:58:29 CDT(-0500)] <colinclark> I agree, by the way, that having zillions of copies of Infusion cloned is not awesome
[14:58:36 CDT(-0500)] <huslage> the point of CI is to find what broke anyway
[14:58:39 CDT(-0500)] <huslage> well one of the points of it
[14:58:39 CDT(-0500)] <michelled> Bosmon2: I suppose if you think your branch will go in first you will need to make the fix
[14:58:40 CDT(-0500)] <colinclark> Having switched to git, I've finally got rid of them from my local machine
[14:58:52 CDT(-0500)] <colinclark> The problem is when the mechanics of the CI itself breaks
[14:59:01 CDT(-0500)] <huslage> that's another problem
[14:59:03 CDT(-0500)] <Bosmon2> michelled - ok
[14:59:06 CDT(-0500)] <huslage> E_OUTOFSCOPE
[14:59:07 CDT(-0500)] <colinclark> which this script is essentially just a detail of how the build site builds
[14:59:29 CDT(-0500)] <huslage> yeah it's just a script to configure the site with the correct info
[14:59:44 CDT(-0500)] <huslage> it could just as easily be a '.ini' file instead of a script, but whatever works
[15:00:05 CDT(-0500)] <colinclark> So, I guess that is another feature of this self-managed build world I've been dreaming about
[15:00:07 CDT(-0500)] <colinclark> and not writing down
[15:00:12 CDT(-0500)] <colinclark> So, what I think we want, in the long run
[15:00:34 CDT(-0500)] <colinclark> Is the following developer experience:
[15:00:46 CDT(-0500)] <colinclark> 1. Developer says "hey, I want to share my cool thing with you on the build server"
[15:00:54 CDT(-0500)] <colinclark> 2. Developer asks committers
[15:00:57 CDT(-0500)] <colinclark> 3. Committers approve
[15:01:11 CDT(-0500)] <colinclark> 4. A git repo is set up on the build server for the developer
[15:01:25 CDT(-0500)] <colinclark> 5. Developer pushes their code up to this repo
[15:01:32 CDT(-0500)] <colinclark> 6. It contains a JSON "manifest" file
[15:01:38 CDT(-0500)] <colinclark> which describes the nature of what will be built
[15:01:43 CDT(-0500)] <colinclark> where it should be mounted to on the web server
[15:01:50 CDT(-0500)] <colinclark> how often it should be built
[15:01:52 CDT(-0500)] <colinclark> etc.
[15:02:06 CDT(-0500)] <colinclark> Equally, this could be all satisfied with a web app
[15:02:18 CDT(-0500)] <colinclark> which didn't require the user to do anything at the shell level to set up a build
[15:02:28 CDT(-0500)] <colinclark> standard convention for the name and location of an install script
[15:02:34 CDT(-0500)] <colinclark> all packaged up with the thing itself
[15:02:47 CDT(-0500)] <colinclark> Right now, it's all so fragmented
[15:02:47 CDT(-0500)] <huslage> hm
[15:02:53 CDT(-0500)] <colinclark> Am I insane again, huslage?
[15:03:00 CDT(-0500)] <huslage> what's the purpose of the approval process?
[15:03:11 CDT(-0500)] <colinclark> Just so that any random person can't push up a daily build of something
[15:03:28 CDT(-0500)] <huslage> you'd presumably get an account
[15:03:37 CDT(-0500)] <colinclark> yep, or something
[15:03:45 CDT(-0500)] <colinclark> I mean, my Git-ish-ness is really just fanciness
[15:03:52 CDT(-0500)] <colinclark> A web app and accounts would be fine
[15:04:09 CDT(-0500)] <colinclark> the real goal being to standardize builds to the point where nothing would ever need to be set up on the command line
[15:04:14 CDT(-0500)] <colinclark> specific to the thing being built
[15:04:17 CDT(-0500)] <huslage> maybe it should just be a post-commit hook on github which triggers a build
[15:04:19 CDT(-0500)] <colinclark> it would all be self-contained
[15:04:30 CDT(-0500)] <huslage> and you set up the thing to build by tracking that
[15:04:31 CDT(-0500)] <colinclark> huslage: That part is easy enough
[15:04:43 CDT(-0500)] <colinclark> I think the key is the setup part
[15:04:45 CDT(-0500)] <colinclark> yes
[15:04:48 CDT(-0500)] <huslage> ok
[15:05:03 CDT(-0500)] <colinclark> So, I guess we could break it down and make it more concrete
[15:05:11 CDT(-0500)] <colinclark> the Image Gallery and Infusion are probably nice examples
[15:05:18 CDT(-0500)] <colinclark> we know that the Image Gallery has a custom install script
[15:05:22 CDT(-0500)] <colinclark> as does, in a sense, Infusion
[15:05:26 CDT(-0500)] <colinclark> since it has to run its Ant wildness
[15:05:41 CDT(-0500)] <colinclark> We know that we want to mount both of these things at standard locations
[15:06:00 CDT(-0500)] <colinclark> Could we abstract this to the point where we had some script that just knew where to go look
[15:06:02 CDT(-0500)] <colinclark> to find this configuration file
[15:06:09 CDT(-0500)] <colinclark> in the thing being built, itself?
[15:06:23 CDT(-0500)] <huslage> i don't think it's too hard really.
[15:06:26 CDT(-0500)] <colinclark> I guess one obvious problem would be things like passwords for databases, etc.
[15:06:32 CDT(-0500)] <colinclark> but that's just something else entirely, I guess
[15:06:42 CDT(-0500)] <huslage> yeah. how often is that a problem for us?
[15:07:00 CDT(-0500)] <colinclark> I guess it was an issue for Fluid Engage
[15:07:06 CDT(-0500)] <colinclark> but I can't quite remember how we dealt with it
[15:07:11 CDT(-0500)] <colinclark> possibly we didn't

[15:07:17 CDT(-0500)] <heidi_> jhung sorry did you have a q about uploader feedback
[15:07:26 CDT(-0500)] <huslage> then there needs to be a script that can take db info as a parameter and do the right thing with it
[15:07:33 CDT(-0500)] <colinclark> yeah
[15:07:39 CDT(-0500)] <huslage> since this needs to be generic
[15:07:41 CDT(-0500)] <colinclark> where the db info was specified outside of the code itself
[15:07:50 CDT(-0500)] <huslage> yeah it's just configuration in the build server
[15:07:53 CDT(-0500)] <colinclark> So, again, the ideal situation would be this thing that sits and listens to a git repo
[15:08:00 CDT(-0500)] <colinclark> and then, say, automatically sets up Continuum accordingly
[15:08:10 CDT(-0500)] <colinclark> alternatively it might just be some generic scripts and a template or wahtever for continuum
[15:08:11 CDT(-0500)] <colinclark> or hudson
[15:08:16 CDT(-0500)] <colinclark> or wahtever CI system we want to use
[15:08:20 CDT(-0500)] <huslage> or bamboo or ....
[15:08:25 CDT(-0500)] <colinclark> right, exactly
[15:09:21 CDT(-0500)] <michelled> jhung: are you working on 4300?
[15:09:21 CDT(-0500)] <jhung> heidi: yeah. You said the upload button text is too small. On my screen it's 1em (inherited from 13px). It looks okay to me.
[15:09:29 CDT(-0500)] <jhung> michelled: yes.
[15:09:35 CDT(-0500)] <michelled> cool
[15:10:05 CDT(-0500)] <heidi_> jhung it was too small for me mac ff/safari
[15:10:12 CDT(-0500)] <heidi_> compared to the size of the other txt
[15:10:23 CDT(-0500)] <michelled> heidi_: did you take a look at anastasiac's 4218 request?
[15:10:32 CDT(-0500)] <heidi_> michelled awhile ago yes
[15:10:45 CDT(-0500)] <michelled> what about the new one - the new rounded buttons?
[15:11:06 CDT(-0500)] <heidi_> michelled i think that was just images being swapped, haven't looked
[15:11:27 CDT(-0500)] <jhung> heidi_: weird. It's not obviously smaller for me. It's the same size as the text elsewhere. I'll change it anyway to match the size of the "Add more" button, but even that is just 1.07em.
[15:11:46 CDT(-0500)] <heidi_> jhung i found setting the size to 1em fixed it for me
[15:11:55 CDT(-0500)] <heidi_> are you looking on a mac?
[15:11:59 CDT(-0500)] <michelled> heidi_: anastasiac wanted your opinion on whether they looked buttony enough

[15:12:10 CDT(-0500)] <heidi_> hehe okay let me take a peek
[15:12:33 CDT(-0500)] <jhung> heidi: no. I'll put it in if that's what you found works for FF on mac. I'll make note of that.
[15:12:55 CDT(-0500)] <heidi_> cool
[15:15:20 CDT(-0500)] <heidi_> anastasiac jhung the new button images are roundy and nice, but i'm wondering about hci/hc - the sizing different than y/b b/y? the text is pushed to the top instead of centred ... in demos/fss/themes/html/themes.html
[15:16:54 CDT(-0500)] <jhung> heidi_: I suppose the b/y and y/b buttons should be updated to match the rounded-ness of the HC and HCI themes.
[15:17:08 CDT(-0500)] <jhung> but text alignment will need to get fixed it seems.
[15:17:09 CDT(-0500)] <heidi_> jhung they are round
[15:17:19 CDT(-0500)] <heidi_> hc/hci is not aligned right
[15:17:24 CDT(-0500)] <jhung> ah
[15:17:28 CDT(-0500)] <colinclark> huslage: Shall I go ahead and push up cindyli1's changes to the Image Gallery build script?
[15:17:34 CDT(-0500)] <huslage> yup
[15:17:38 CDT(-0500)] <colinclark> I don't know if the server actually uses these straight out of git
[15:17:39 CDT(-0500)] <colinclark> or how that works
[15:17:40 CDT(-0500)] <colinclark> ok
[15:17:44 CDT(-0500)] <heidi_> jhung take a peek at FLUID-4218-roundButtonOutlines in a's github
[15:18:08 CDT(-0500)] <huslage> me either colinclark
[15:18:14 CDT(-0500)] <colinclark> lol
[15:18:15 CDT(-0500)] <colinclark> ok
[15:18:17 CDT(-0500)] <colinclark> done
[15:18:34 CDT(-0500)] <huslage> continuum is a circle of hell i'm not yet familiar enough with.
[15:18:38 CDT(-0500)] <huslage> it ain't baked
[15:18:41 CDT(-0500)] <colinclark> huslage: Let's figure out a plan and schedule for a build server rebirth
[15:18:43 CDT(-0500)] <heidi_> jhung ah it's padding issue. anastasiac padding-bottom for .fl-theme-hci .fl-widget-titlebar .fl-button-inner and hc looks off
[15:18:45 CDT(-0500)] <colinclark> You've got lots to do before then
[15:18:50 CDT(-0500)] <huslage> yeah colinclark
[15:18:53 CDT(-0500)] <colinclark> but within the next month or so, it'd be really great
[15:18:58 CDT(-0500)] <huslage> i'm gonna do a quick survey of the stuff out there
[15:19:02 CDT(-0500)] <colinclark> It has been driving me completely insane for years
[15:19:05 CDT(-0500)] <huslage> yes
[15:19:08 CDT(-0500)] <jhung> heidi_: okay cool. You going to put in a fix for that?
[15:19:12 CDT(-0500)] <colinclark> a spaghetti of shell scripts
[15:19:17 CDT(-0500)] <huslage> writing something seems stupid at this point
[15:19:24 CDT(-0500)] <colinclark> No point in writing something new, no
[15:19:33 CDT(-0500)] <colinclark> just pulling together some existing CI tools that do what we need them to
[15:19:42 CDT(-0500)] <huslage> yep
[15:19:47 CDT(-0500)] <colinclark> where a key criteria is putting an end to the proliferation of fancy custom build scripts, etc.
[15:19:48 CDT(-0500)] <heidi_> jhung anastasiac can update her pull request maybe? looks like just remove both those padding entries
[15:19:56 CDT(-0500)] <colinclark> though I may be the dreamer, there

[15:20:12 CDT(-0500)] <huslage> colinclark: you can't stop entropy. sorry.
[15:20:13 CDT(-0500)] <heidi_> ill add a comment to the req
[15:20:17 CDT(-0500)] <huslage> coders gonna code
[15:20:32 CDT(-0500)] <colinclark> Sure, but there can expectations about how coders code

[15:20:42 CDT(-0500)] <colinclark> Like not cutting and pasting shit
[15:20:51 CDT(-0500)] <huslage>

[15:20:51 CDT(-0500)] <colinclark> And actually bothering to make it accessible
[15:20:54 CDT(-0500)] <colinclark>

[15:21:00 CDT(-0500)] <huslage> you are correct sir
[15:21:35 CDT(-0500)] <Bosmon2> huslage - I assume that, having vowed to let entropy take its course, you will now proceed to lie down in your grave

[15:21:41 CDT(-0500)] <colinclark> lol
[15:21:52 CDT(-0500)] <huslage> Bosmon2: where do you think i am now?
[15:22:16 CDT(-0500)] <Bosmon2> I still see redundancy in your IRC messages
[15:22:22 CDT(-0500)] <Bosmon2> Vowels, the occasional space
[15:22:36 CDT(-0500)] <huslage> lol
[15:22:37 CDT(-0500)] <jhung> heidi_: k.
[15:22:44 CDT(-0500)] <huslage> kl;;jokcw8ofjklfnkobnkoqej
[15:22:48 CDT(-0500)] <Bosmon2> Better
[15:22:51 CDT(-0500)] <huslage> ah crap. it's happening again
[15:23:02 CDT(-0500)] <Bosmon2> More realistic output from the huslage CORPSE-O-FONE
[15:23:07 CDT(-0500)] <huslage> lol
[15:23:14 CDT(-0500)] <huslage> cat /dev/urandom
[15:23:35 CDT(-0500)] <Bosmon2> Now you're talking
[15:23:46 CDT(-0500)] <huslage> E_NOENTROPY
[15:23:48 CDT(-0500)] <huslage> oh no.
[15:24:09 CDT(-0500)] <Bosmon2> You're clearly not a VMS man
[15:24:18 CDT(-0500)] <Bosmon2> Or you would be DR. LIVINGSTONE_I_PRESUME
[15:24:18 CDT(-0500)] <huslage> ugh
[15:24:34 CDT(-0500)] <huslage> i think i have a CD of VMS 8.x somewhere
[15:31:01 CDT(-0500)] <anastasiac> heidi_, I don't think the problem with those button images is in the demo's css file. I see the same issue in the unit test file a well
[15:31:15 CDT(-0500)] <heidi_> anastasiac yeah it's in the theme files
[15:31:18 CDT(-0500)] <colinclark> I thoroughly hate git
[15:31:40 CDT(-0500)] <colinclark> hopefully someone will +1 my proposal to switch back to svn
[15:32:18 CDT(-0500)] <athena> hmm, interesting
[15:32:21 CDT(-0500)] <heidi_> why do you hate git!
[15:32:28 CDT(-0500)] <athena> so you're not a fan of the transition so far colinclark?
[15:32:30 CDT(-0500)] <heidi_> did it git-cha
[15:32:57 CDT(-0500)] <colinclark> athena, heidi_: See, the thing this has to be prefaced with...
[15:33:03 CDT(-0500)] <colinclark> is that I'm sort of clueless and short on patience
[15:33:14 CDT(-0500)] <athena> yeah i hear that
[15:33:16 CDT(-0500)] <colinclark> I wouldn't actually ever want to switch back to svn
[15:33:19 CDT(-0500)] <athena> we're considering moving to git
[15:33:26 CDT(-0500)] <colinclark> but I don't know that I'm all that satisfied with being here, either
[15:33:29 CDT(-0500)] <athena> though i'm not looking forward to having to spend time learning a new system
[15:33:31 CDT(-0500)] <colinclark> I want Magic!
[15:33:39 CDT(-0500)] <colinclark> I just find it inscrutable some times
[15:33:40 CDT(-0500)] <athena> ooh sign us up for that too
[15:33:54 CDT(-0500)] <colinclark> michelled, Justin_o, and Bosmon2 all seem quite confident with it
[15:34:09 CDT(-0500)] <athena> yeah Bosmon2 had positive things to say about it when we cornered him on that

[15:34:16 CDT(-0500)] <colinclark> I've managed to just get myself into a situation where I can push changes up from a local clone to my fork on github
[15:34:22 CDT(-0500)] <colinclark> and it's just completely incomprehensible to me
[15:34:33 CDT(-0500)] <colinclark> For some reason it thinks that I'd be losing history
[15:34:41 CDT(-0500)] <colinclark> but it just is being insane
[15:34:45 CDT(-0500)] <colinclark> like the rest of us in this channel
[15:34:54 CDT(-0500)] <colinclark> you excepted of course, athena

[15:34:55 CDT(-0500)] <athena> i definitely need to go read some tutorials or something
[15:34:59 CDT(-0500)] <colinclark> yeah
[15:35:01 CDT(-0500)] <athena> lol
[15:35:04 CDT(-0500)] * athena runs around crazily
[15:35:04 CDT(-0500)] <colinclark> me too
[15:35:06 CDT(-0500)] <colinclark>

[15:35:12 CDT(-0500)] <colinclark> okay, you're in
[15:35:53 CDT(-0500)] <colinclark> I guess with any complex tool it takes a while to build a model for how it works
[15:36:19 CDT(-0500)] <michelled> or you just find a way that works for you and stick to it like I do

[15:36:19 CDT(-0500)] <heidi_> thoughts on why it says you'd be losing history? that is werid
[15:36:35 CDT(-0500)] <michelled> I've got a handful of commands that I use and a couple work flows
[15:36:46 CDT(-0500)] <colinclark> heidi_: I'm really not sure. michelled thinks it's because I merged master to master with --no-ff
[15:37:02 CDT(-0500)] <colinclark> I just randomly recite commands and hope it works

[15:37:10 CDT(-0500)] <heidi_> lol
[15:37:12 CDT(-0500)] <anastasiac> ok, heidi_ and michelled, I've pushed that fix. that pull request (https://github.com/fluid-project/infusion/pull/89) should be good to go
[15:37:16 CDT(-0500)] <michelled> that was my theory until I saw your log compares colinclark
[15:37:21 CDT(-0500)] <heidi_> i've never used --no-ff, not sure what ti does
[15:37:24 CDT(-0500)] <colinclark> eek
[15:37:27 CDT(-0500)] <colinclark> it's very important

[15:37:30 CDT(-0500)] <heidi_> oyyy
[15:37:35 CDT(-0500)] <colinclark> but probably only for the opposite of what you do
[15:37:52 CDT(-0500)] <colinclark> merging stuff into the master from branches, you absolutely must use --no-ff
[15:38:03 CDT(-0500)] <heidi_> ahh, i don't ever merge to master
[15:38:12 CDT(-0500)] <colinclark> right
[15:38:23 CDT(-0500)] <colinclark> anyway, it has something to do with having reset back to a previous revision
[15:38:32 CDT(-0500)] <colinclark> then merging in some changes from the upstream master
[15:39:39 CDT(-0500)] <colinclark> athena: I think you'll dig git
[15:40:03 CDT(-0500)] <colinclark> I'm just being a dork, having not read the book closely enough
[15:40:52 CDT(-0500)] <michelled> heidi_: wanna take another glance at 4218 before I push it?
[15:41:00 CDT(-0500)] <heidi_> sure
[15:41:39 CDT(-0500)] <colinclark> So, michelled...
[15:41:44 CDT(-0500)] <heidi_> michelled_ A+ it's ready
[15:41:51 CDT(-0500)] <colinclark> do you recommend that one use --no-ff for pretty much any merge
[15:41:54 CDT(-0500)] <colinclark> except master:master?
[15:41:55 CDT(-0500)] <michelled> thx heidi_
[15:42:24 CDT(-0500)] <michelled> colinclark: I only use --no-ff when merged a branch into the master
[15:42:34 CDT(-0500)] <colinclark> when else does one merge?
[15:42:37 CDT(-0500)] <michelled> otherwise I prefer the fast forward
[15:42:42 CDT(-0500)] <colinclark> when merging changes in from upstream/master
[15:42:43 CDT(-0500)] <colinclark> when else?
[15:42:51 CDT(-0500)] <colinclark> I'm just trying to think of other workflows
[15:42:56 CDT(-0500)] <michelled> sometimes when merging branches together
[15:43:01 CDT(-0500)] <colinclark> ok, so
[15:43:08 CDT(-0500)] <colinclark> you've got one branch, foo
[15:43:14 CDT(-0500)] <colinclark> and, say, heidi_'s got another branch, bar
[15:43:21 CDT(-0500)] <colinclark> if you're merging those together
[15:43:29 CDT(-0500)] <colinclark> git merge heidi/bar foo
[15:43:33 CDT(-0500)] <colinclark> you'd want the fast forward
[15:43:34 CDT(-0500)] <colinclark> ?
[15:44:13 CDT(-0500)] <michelled> I guess the question is why I'm merging them and whether I care about the black line in that case
[15:44:22 CDT(-0500)] <colinclark> THE BLACK LINE
[15:44:30 CDT(-0500)] <colinclark> ok
[15:44:33 CDT(-0500)] <michelled> for the most part I've only merged branches together because I want to see if something will fix what I'm working on
[15:44:37 CDT(-0500)] <colinclark> ok
[15:44:45 CDT(-0500)] <michelled> and I intend on backing out the merge afterward
[15:44:47 CDT(-0500)] <colinclark> So, maybe you can explain more about fast forwarding
[15:44:53 CDT(-0500)] <colinclark> why it's something you either want or don't want
[15:45:20 CDT(-0500)] <michelled> for me it's all about the pretty network graph which despite our repeated conversations I still love and use

[15:46:12 CDT(-0500)] <michelled> I want to keep the top line of that stable so I have an at a glance view of work packages that have been done in the repo
[15:46:49 CDT(-0500)] <michelled> so in order to keep it stable, I need a commit to go with a merging in a branch
[15:47:25 CDT(-0500)] <colinclark> so why does a fast forward affect the lineage of the line?
[15:47:27 CDT(-0500)] <michelled> when a merge is clean - which most of our are - and a fast forward happens then there is no commit
[15:48:09 CDT(-0500)] <michelled> by saying 'no-ff' I'm saying - please mark this spot where I'm merging in branch x with a commit
[15:48:15 CDT(-0500)] <michelled> at least that's what I think is happening
[15:48:17 CDT(-0500)] <colinclark> ok
[15:48:22 CDT(-0500)] <colinclark> that makes some sense
[15:51:57 CDT(-0500)] <colinclark> Bosmon2: Any guesses why only the crazy file queue view tests ended up receiving this "registeredFrom" property?
[15:54:20 CDT(-0500)] <huslage> gotta run
[16:03:38 CDT(-0500)] <Bosmon2> colinclark - I think it is the special case of having a demands block which specifies arguments directly, but then undergoes no further expansion
[16:04:27 CDT(-0500)] <colinclark> ah
[16:04:52 CDT(-0500)] <Bosmon2> I can't be completely sure... there are so many twisty turning passages in IoC expansion

[16:05:50 CDT(-0500)] <colinclark>

[16:06:43 CDT(-0500)] <Bosmon2> I remember in the old days, having books delivered would sometimes be too cold to touch after lying on the doorstep
[16:06:47 CDT(-0500)] <Bosmon2> This one is almost too hot to touch

Browser not supported