Options:
Previous day, Next day
- # Session Start: Thu Oct 15 00:00:00 2015
- # Session Ident: #css
- # [00:04] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
- # [00:35] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
- # [00:36] * Joins: estellevw (~estellevw@public.cloak)
- # [00:38] * Quits: estellevw (~estellevw@public.cloak) ("Snuggling with the puppies")
- # [00:42] * Joins: rego (~smuxi@public.cloak)
- # [00:54] * Quits: plh (plehegar@public.cloak) ("Leaving")
- # [01:07] * Quits: tgraham (~user@public.cloak) (Ping timeout: 180 seconds)
- # [01:17] * Quits: lajava (~javi@public.cloak) (Ping timeout: 180 seconds)
- # [01:20] * Joins: jdaggett (~jdaggett@public.cloak)
- # [01:23] <adenilson> Anyone around that could answer a few questions I have on http://www.w3.org/TR/css-round-display-1/
- # [01:23] <adenilson> ?
- # [01:36] * Joins: Florian (~Florian@public.cloak)
- # [02:11] * Quits: adenilson (~anonymous@public.cloak) (adenilson)
- # [02:32] * Joins: stakagi (~stakagi@public.cloak)
- # [03:37] * Quits: dbaron (~dbaron@public.cloak) ("8403864 bytes have been tenured, next gc will be global.")
- # [03:45] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
- # [03:49] * Joins: Florian (~Florian@public.cloak)
- # [05:08] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [05:09] * Joins: stakagi (~stakagi@public.cloak)
- # [05:12] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [05:13] * Joins: stakagi (~stakagi@public.cloak)
- # [05:45] * Joins: estellevw (~estellevw@public.cloak)
- # [05:53] * Quits: estellevw (~estellevw@public.cloak) (Ping timeout: 180 seconds)
- # [05:53] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [05:53] * Joins: stakagi (~stakagi@public.cloak)
- # [06:14] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [06:16] * Joins: stakagi (~stakagi@public.cloak)
- # [06:34] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [06:34] * Joins: stakagi (~stakagi@public.cloak)
- # [06:38] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
- # [06:38] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [06:38] * Joins: stakagi (~stakagi@public.cloak)
- # [06:39] * Joins: Florian (~Florian@public.cloak)
- # [07:05] * Quits: myles (~Adium@public.cloak) (Client closed connection)
- # [07:05] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [07:05] * Joins: myles (~Adium@public.cloak)
- # [07:07] * Joins: stakagi (~stakagi@public.cloak)
- # [07:15] * Joins: glazou (~glazou@public.cloak)
- # [07:15] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [07:15] * Quits: glazou (~glazou@public.cloak) (glazou)
- # [07:15] * Joins: stakagi (~stakagi@public.cloak)
- # [07:34] * Quits: myles (~Adium@public.cloak) ("Leaving.")
- # [07:34] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [07:35] * Joins: stakagi (~stakagi@public.cloak)
- # [07:44] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
- # [07:51] * Joins: rego (~smuxi@public.cloak)
- # [08:15] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [08:16] * Joins: stakagi (~stakagi@public.cloak)
- # [08:21] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
- # [08:21] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [08:22] * Joins: stakagi (~stakagi@public.cloak)
- # [08:28] * Joins: rego (~smuxi@public.cloak)
- # [08:28] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [08:29] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
- # [08:30] * Joins: Florian (~Florian@public.cloak)
- # [08:30] * Joins: stakagi (~stakagi@public.cloak)
- # [08:37] * Quits: Florian (~Florian@public.cloak) (Ping timeout: 180 seconds)
- # [08:42] * Joins: glazou (~glazou@public.cloak)
- # [09:10] * Joins: Robert (~Robert@public.cloak)
- # [09:11] * Joins: zcorpan (~zcorpan@public.cloak)
- # [09:17] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
- # [09:20] * Joins: tgraham (~user@public.cloak)
- # [09:20] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [09:21] * Joins: stakagi (~stakagi@public.cloak)
- # [09:26] * Joins: zcorpan (~zcorpan@public.cloak)
- # [09:35] * Quits: stakagi (~stakagi@public.cloak) (Client closed connection)
- # [09:42] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
- # [09:49] * Joins: rego (~smuxi@public.cloak)
- # [09:50] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
- # [09:59] * Joins: antonp (~Thunderbird@public.cloak)
- # [09:59] * Quits: jdaggett (~jdaggett@public.cloak) (jdaggett)
- # [10:01] * Joins: Ms2ger (~Ms2ger@public.cloak)
- # [10:08] * Quits: rego (~smuxi@public.cloak) (Ping timeout: 180 seconds)
- # [10:10] * Joins: rego (~smuxi@public.cloak)
- # [10:12] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
- # [10:19] * Joins: rego (~smuxi@public.cloak)
- # [10:21] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
- # [10:28] * Joins: rego (~smuxi@public.cloak)
- # [10:30] * Joins: Florian (~Florian@public.cloak)
- # [10:33] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
- # [10:42] * Joins: rego (~smuxi@public.cloak)
- # [10:46] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
- # [10:46] * Joins: Florian (~Florian@public.cloak)
- # [10:51] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
- # [10:53] * Quits: Florian (~Florian@public.cloak) (Ping timeout: 180 seconds)
- # [10:59] * Joins: rego (~smuxi@public.cloak)
- # [11:07] * Joins: Florian (~Florian@public.cloak)
- # [11:08] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
- # [11:09] * Joins: Florian_ (~Florian@public.cloak)
- # [11:09] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
- # [11:10] * Quits: Florian_ (~Florian@public.cloak) (Client closed connection)
- # [11:17] * Joins: lajava (~javi@public.cloak)
- # [11:32] * Joins: Florian (~Florian@public.cloak)
- # [11:43] * Quits: Robert (~Robert@public.cloak) ("Leaving")
- # [12:21] * Quits: Bert_ (bbos@public.cloak) (Client closed connection)
- # [12:25] * Joins: Bert (bbos@public.cloak)
- # [12:44] * Joins: Florian_ (~Florian@public.cloak)
- # [12:51] * Quits: Florian (~Florian@public.cloak) (Ping timeout: 180 seconds)
- # [12:52] * Joins: jdaggett (~jdaggett@public.cloak)
- # [12:56] * Quits: Florian_ (~Florian@public.cloak) (Client closed connection)
- # [12:56] * Joins: Florian (~Florian@public.cloak)
- # [12:58] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
- # [12:59] * Joins: Florian (~Florian@public.cloak)
- # [13:04] * Quits: Ms2ger (~Ms2ger@public.cloak) (Ping timeout: 180 seconds)
- # [13:06] * Quits: Florian (~Florian@public.cloak) (Ping timeout: 180 seconds)
- # [13:24] * Joins: Ms2ger (~Ms2ger@public.cloak)
- # [13:59] * Joins: Florian (~Florian@public.cloak)
- # [14:00] * Quits: jdaggett (~jdaggett@public.cloak) (jdaggett)
- # [14:01] * Joins: plh (plehegar@public.cloak)
- # [14:07] * Quits: Florian (~Florian@public.cloak) (Ping timeout: 180 seconds)
- # [14:17] * Quits: dauwhe (~dauwhe@public.cloak) (Client closed connection)
- # [14:29] * Quits: lajava (~javi@public.cloak) (Ping timeout: 180 seconds)
- # [14:51] * Joins: Florian (~Florian@public.cloak)
- # [15:23] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
- # [15:31] * Joins: Florian (~Florian@public.cloak)
- # [15:38] * Quits: plh (plehegar@public.cloak) ("Leaving")
- # [15:38] * Joins: plh (plehegar@public.cloak)
- # [15:59] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
- # [16:02] * Joins: anssik (~uid10742@public.cloak)
- # [16:19] * Quits: krit (~sid15081@public.cloak) (Client closed connection)
- # [16:20] * Joins: krit (~sid15081@public.cloak)
- # [16:38] * Joins: Florian (~Florian@public.cloak)
- # [16:52] * Quits: Ms2ger (~Ms2ger@public.cloak) (Ping timeout: 180 seconds)
- # [17:02] * Joins: Ms2ger (~Ms2ger@public.cloak)
- # [17:24] <cbiesinger> TabAtkins: fantasai: how are negative margins on flex items supposed to behave?
- # [17:27] <TabAtkins> cbiesinger: Like a negative-margin element in any other layout mode?
- # [17:28] <TabAtkins> cbiesinger: Like, the layout algo works on the margin box. So a negative margin just makes that box smaller.
- # [17:29] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
- # [17:30] <cbiesinger> TabAtkins: oh
- # [17:30] <cbiesinger> TabAtkins: it doesn't affect the positioning of the box?
- # [17:30] <cbiesinger> that was ambiguous
- # [17:30] <TabAtkins> Of course it does, the same way that a positive margin does.
- # [17:31] <cbiesinger> TabAtkins: ok so we shrink the margin box, we run the layout algorithm, and then we move the box left according to its negative margin-left?
- # [17:31] <TabAtkins> Yes. Exactly the same as a positive margin-left. ^_^
- # [17:32] <cbiesinger> fair enough :p
- # [17:32] <cbiesinger> of course we need to floor the size at 0
- # [17:34] <TabAtkins> Yeah, the size of a box is always floored at 0.
- # [17:36] * Joins: bkardell_ (~uid10373@public.cloak)
- # [17:37] * Quits: dholbert (~dholbert@public.cloak) (Ping timeout: 180 seconds)
- # [17:39] * Joins: dholbert (~dholbert@public.cloak)
- # [17:53] * Quits: glazou (~glazou@public.cloak) ("bbl")
- # [18:16] * Quits: Ms2ger (~Ms2ger@public.cloak) ("nn")
- # [18:37] <SimonSapin> I don’t remember finding a clear point in the spec that says floor at zero, except as a side effect of min-width
- # [18:38] <SimonSapin> weasyprint had negative widths before it implemented min-width
- # [18:53] <cbiesinger> TabAtkins: to confirm, width: 20px; margin-left: -20px; should show no content, right?
- # [18:53] <cbiesinger> wait, that's not right
- # [18:53] <cbiesinger> it should overflow if nothing else
- # [18:53] <cbiesinger> where did my content go?
- # [18:56] <cbiesinger> ok here we go
- # [19:12] * Quits: anssik (~uid10742@public.cloak) ("Connection closed for inactivity")
- # [19:13] <TabAtkins> SimonSapin: Yeah, I believe you that it's underspecified
- # [19:13] <TabAtkins> cbiesinger: Yes, overflow
- # [19:23] <cbiesinger> TabAtkins: but should the flex algorithm freeze the width to zero in this case? (note: it doesn't right now, in chrome/ff)
- # [19:28] * Joins: darktears (~darktears@public.cloak)
- # [19:33] * Quits: darktears (~darktears@public.cloak) ("Linkinus - http://linkinus.com")
- # [19:33] * Joins: darktears (~darktears@public.cloak)
- # [19:45] <TabAtkins> cbiesinger: Is it inflexible?
- # [19:45] <TabAtkins> (I'm wondering if you're asking because this is a special case, or if it's a "width + margin" vs "just width" thing.)
- # [19:56] * Joins: zcorpan (~zcorpan@public.cloak)
- # [19:59] <cbiesinger> TabAtkins: it is shrinkable but not growable; maybe I should try flex: none;. I consider it a special case because a negative margin is unusual...
- # [20:00] <TabAtkins> negative margin is no more unusual than positive margin.
- # [20:09] <cbiesinger> if you say so :p
- # [20:10] <cbiesinger> TabAtkins: anyway basically my question is: http://jsbin.com/cabaqemoza/edit?html,css,output
- # [20:10] <cbiesinger> TabAtkins: should there be any red visible?
- # [20:12] * Joins: myles (~Adium@public.cloak)
- # [20:12] <cbiesinger> TabAtkins: (I ask because you say the algorithm works on the margin box, and this sets the margin box to zero)
- # [20:13] <TabAtkins> Yeah, looking in now.
- # [20:15] <TabAtkins> Okay, it's correct that the flexbox is 20px wide - item2's margin box shrinks to 0, so the summed size of the contents is only 20px.
- # [20:15] <TabAtkins> Oh wait, no, the flexbox is only 10px in this example.
- # [20:15] <TabAtkins> Hm, I guess it's considering item2 as being -10px wide.
- # [20:16] <TabAtkins> fantasai: Refresh me - margin boxes *do* get clamped to zero size, right? They can't go negative and cause a container to be *smaller* than the rest of the contents?
- # [20:17] <TabAtkins> (I think she's driving right now.)
- # [20:17] <cbiesinger> oh, I was assuming that part was definitely wrong -- I was more wondering if item2 should end up at 0 width?
- # [20:17] <TabAtkins> Ah!
- # [20:17] <cbiesinger> but if you think that the container should not be 20px wide, let me know!
- # [20:18] <TabAtkins> No, it shouldn't. It's width is 20px, the negative margin just movies it.
- # [20:18] <cbiesinger> ok.
- # [20:18] <TabAtkins> I *think* the container should be 20px, but let me check some other examples real quick
- # [20:22] <TabAtkins> Okay, yeah, setting up the same situation with floats, the container doesn't get smaller than its other contents. (The super-negative-margin child gets clamped to zero for the purpose of sizing calculations, in other words.)
- # [20:22] <TabAtkins> So yeah, the flexbox should be 20px wide (one 20px child, one 0px child)
- # [20:23] <cbiesinger> well, one 20px child, and an "out of flow-ish" 20 px child, right? totaling 20px
- # [20:23] * Joins: dauwhe (~dauwhe@public.cloak)
- # [20:26] <cbiesinger> that was poor wording, but my point is that it is still 20px wide
- # [20:26] <cbiesinger> just doesn't affect the container width
- # [20:28] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
- # [20:29] * Quits: bkardell_ (~uid10373@public.cloak) ("Connection closed for inactivity")
- # [20:31] * Joins: Florian (~Florian@public.cloak)
- # [20:34] <TabAtkins> Yes. One 0px margin-box child. ^_^
- # [20:36] <cbiesinger> true. right. :)
- # [20:37] * cbiesinger goes to write a full reftest for this
- # [20:39] * Quits: Florian (~Florian@public.cloak) (Ping timeout: 180 seconds)
- # [21:08] * Joins: dbaron (~dbaron@public.cloak)
- # [21:15] * Joins: adenilson (~anonymous@public.cloak)
- # [21:18] <gsnedders> So, uh, yeah. I'm going to be trying to get almost all of the 2.1 testsuite moved over to reftests in the months after TPAC.
- # [21:18] <gsnedders> FWIW.
- # [21:54] * Quits: antonp (~Thunderbird@public.cloak) (antonp)
- # [21:56] <TabAtkins> gsnedders: +1
- # [21:57] <cbiesinger> TabAtkins: that reviewable thing is something I'll have to get used to
- # [21:57] <TabAtkins> I've never used it before either.
- # [21:57] <TabAtkins> I'm fine with sticking to GH reviewing if you want. ^_^
- # [21:58] <cbiesinger> eh, either way
- # [22:00] <cbiesinger> TabAtkins: so it looks like this travis thing builds everything twice, why is that?
- # [22:00] <TabAtkins> Not a clue, I've had nothing to do with the testing setup.
- # [22:00] <cbiesinger> heh, fair enough
- # [22:03] <gsnedders> cbiesinger: I should build every commit in the PR
- # [22:03] <cbiesinger> gsnedders: there's only one
- # [22:03] <cbiesinger> gsnedders: does it build the base revision separately...?
- # [22:04] <gsnedders> cbiesinger: it shouldn't. weird if it actually is.
- # [22:05] <cbiesinger> gsnedders: https://travis-ci.org/w3c/csswg-test/builds/85603877 - two sections of "Building" messages
- # [22:05] <cbiesinger> TabAtkins: looks good now?
- # [22:05] <cbiesinger> ah you must be kidding me, our <video> controls depend on the broken handling of negative margins
- # [22:06] <TabAtkins> cbiesinger: Yaaaaay, at least those can b efixed.
- # [22:07] <cbiesinger> yeah
- # [22:24] * heycam|away is now known as heycam
- # [23:34] <fantasai> TabAtkins: Just got online
- # [23:35] <fantasai> TabAtkins: 20 mi @ 20mph sucks
- # [23:35] * fantasai got herself tea and a monitor and is all set to read scrollback ^o^
- # [23:36] <fantasai> TabAtkins, cbiesinger: Margin boxes do not get floored at zero
- # [23:36] <TabAtkins> fantasai: For the purpose of sizing container.
- # [23:37] <fantasai> TabAtkins, cbiesinger however boxes and containing blocks get floored at zero
- # [23:37] <TabAtkins> Yeah, that. ^_^
- # [23:37] <cbiesinger> ?
- # [23:37] <cbiesinger> what does it mean that containing blocks get floored at zero?
- # [23:38] <fantasai> There is no such thing as a negative-width containing block
- # [23:38] <fantasai> Ordinarily this is because containing blocks are derived from content boxes, which are floored at zero
- # [23:38] <fantasai> (or from padding boxes, which are likewise floored at zero)
- # [23:39] <fantasai> but containing blocks derived from line-broken inline boxes could theoretically be negative
- # [23:39] <fantasai> but they're not
- # [23:39] <fantasai> I don't remember how they're not, but they're not
- # [23:39] <cbiesinger> ah I see
- # [23:40] <fantasai> however margin boxes can be negative :)
- # [23:40] <cbiesinger> but you agree that <div style="display: flex"><div style="width: 20px"></div><div style="width: 20px; margin-left: -40px;"> <-- flexbox should have a width of 20px not 0?
- # [23:41] <fantasai> if you have -20px margins and a 10px content box, then your effective margin-box size is -30px
- # [23:42] <TabAtkins> fantasai: Point is that if a container is content-sized, and has one 20px child and one child with {width: 20px; margin-left: -30px;}, the container should be 20px wide, not 10px.
- # [23:43] * Quits: shepazu (schepers@public.cloak) (Ping timeout: 180 seconds)
- # [23:43] <TabAtkins> This is how floats work, and it's how Flexbox should work too.
- # [23:43] <fantasai> Floats are different
- # [23:44] <fantasai> The model you want is heights in block containers
- # [23:44] <cbiesinger> you two duke it out, I gotta go :p
- # [23:44] * cbiesinger will read scrollback later
- # [23:44] <fantasai> If you have one 20px height child, and one child with height: 20px; margin-bottom: -30px
- # [23:44] <fantasai> Then you have a 10px height container
- # [23:44] <fantasai> Floats are different because your main axis is still downward
- # [23:45] <fantasai> You place the first float, and then you're like "hey, I still got room for this other float, because it doesn't take up any room"
- # [23:45] <fantasai> It's like placing a negative-width inline box next to it
- # [23:46] <TabAtkins> Ugh, I didn't realize that block heights worked differently, that's dumb.
- # [23:46] <TabAtkins> I tried it with inlines, but they went screwy in a way I couldn't figure out. ^_^
- # [23:47] <fantasai> :)
- # [23:47] <fantasai> Block heights make sense
- # [23:48] <fantasai> your height is the bottom edge of your last child
- # [23:48] <fantasai> that's it
- # [23:48] <fantasai> Ending up with the height of the previous child because you have a negative margin that covers it is weird
- # [23:48] <TabAtkins> A float containing floats should be in the same situation - its right edge should be the right edge of its last floating child. (Pretend that there are no other hcildren.) But it's not.
- # [23:49] <TabAtkins> It just sums up the widths of its floating children, flooring at zero, and becomes that width.
- # [23:52] <fantasai> If it did that, then that would make sense
- # [23:52] <fantasai> But that's not the answer you gave me above
- # [23:52] <fantasai> So I'm confused
- # [23:52] <fantasai> If you ran that forumla, you'd get 10px above.
- # [23:52] <fantasai> er, zero
- # [23:52] <fantasai> depending whether we were talkign about -30 or -40 px margins..
- # [23:54] <TabAtkins> ...no, it would be 20px, like I said up above.
- # [23:54] <TabAtkins> *flooring the widths of its floating children at 0*
- # [23:55] <fantasai> oh
- # [23:55] <TabAtkins> that phrase might not have been clear on its scope, sorry
- # [23:55] <fantasai> I read that as flooring the width of the containing block :)
- # [23:55] <fantasai> which is what actually makes sense imho
- # [23:55] <TabAtkins> Not what happens, tho, and I dispute you on it necessarily making sense. ^_^
- # [23:55] <fantasai> flooring children at zero would be really weird if you had a bunch of them
- # [23:55] <TabAtkins> How?
- # [23:56] <fantasai> Put some stuff that's not negative after the negative child ^_^
- # [23:56] <fantasai> If you floor each child at zero when calculating the width
- # [23:56] <fantasai> then the width ends up being a lot wider than the content
- # [23:57] <TabAtkins> That's not how floats work, at least in Blink.
- # [23:57] * TabAtkins tries it in FF.
- # [23:57] * Joins: jdaggett (~jdaggett@public.cloak)
- # [23:57] <TabAtkins> Or FF.
- # [23:57] * fantasai doesn't understand what you're saying
- # [23:58] <TabAtkins> http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0A%3Ccontainer%3E%0A%20%3Cchild%3Efoo%3C%2Fchild%3E%0A%20%3Cchild%3Ebar%3C%2Fchild%3E%0A%20%3Cchild%3Ebaz%3C%2Fchild%3E%0A%3C%2Fcontainer%3E%0A%3Cstyle%3E%0Abody%20*%20{%20margin%3A%202px%3B%20padding%3A%202px%3B%20background%3A%20rgba%280%2C0%2C0%2C.2%29%3B%20}%0Acontainer%2C%20chi
- # [23:58] <TabAtkins> ld%20{%20display%3A%20block%3B%20float%3A%20left%3B}%0Achild%20{%20height%3A%2020px%3B%20}%0Achild%3Anth-child%282%29%20{%20margin-left%3A%20-40px%3B%20}
- # Session Close: Fri Oct 16 00:00:00 2015
Previous day, Next day
Think these logs are useful? Then please donate to show your gratitude (and keep them up, of course). Thanks! — Krijn