/irc-logs / w3c / #css / 2015-10-08 / end

Options:

# Session Start: Thu Oct 08 00:00:00 2015
# Session Ident: #css
# [00:02] * Quits: gregdavis (~udavig4@public.cloak) (gregdavis)
# [00:07] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [00:10] * Joins: jcraig (~jcraig@public.cloak)
# [00:14] * Joins: rego (~smuxi@public.cloak)
# [00:15] * Quits: SteveZ (~SteveZ@public.cloak) (Ping timeout: 180 seconds)
# [00:17] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
# [00:19] * Joins: jcraig_ (~jcraig@public.cloak)
# [00:20] * Joins: antonp (~Thunderbird@public.cloak)
# [00:20] * Quits: jcraig_ (~jcraig@public.cloak) (jcraig_)
# [00:21] * Joins: jcraig_ (~jcraig@public.cloak)
# [00:23] * Quits: jcraig (~jcraig@public.cloak) (Ping timeout: 180 seconds)
# [00:23] * jcraig_ is now known as jcraig
# [00:27] * Quits: jcraig (~jcraig@public.cloak) (jcraig)
# [00:37] <fantasai> TabAtkins: The entirety of the terms index in the snapshot is linking to the EDs
# [00:37] <fantasai> I'm pretty sure at least some of them have corresponding TR anchors :)
# [00:37] <TabAtkins> kk, will check it out in a bit
# [00:37] <fantasai> I tossed a copy of the Snapshot to Bert, but commented out the indices
# [00:38] <fantasai> Goal being to have *something* up with the new text, so we can fix the boilerplate
# [00:38] <fantasai> while we work on getting all the indices to generate correctly
# [00:38] <TabAtkins> Yup, seems fine to me.
# [00:40] <fantasai> plinss: Can we get permission to republish Display by, say, Friday?
# [00:40] <fantasai> plinss: Would like to put it in the pipeline for Tuesday
# [00:41] * fantasai thinks we could do this over email, there's really nothing substantive
# [00:44] <plinss> fantasai: sure, will send cfc email, is the changes section up to date?
# [00:45] <fantasai> I'll check real quick, but https://lists.w3.org/Archives/Member/w3c-css-wg/2015OctDec/0021.html is up to date
# [00:50] <fantasai> plinss: Thanks!
# [00:50] <plinss> np
# [00:52] * Joins: Florian (~Florian@public.cloak)
# [00:57] <fantasai> TabAtkins: Is having <a>fragment</a> throw warnings because it's a term defined in the URL spec Works As Intended?
# [00:57] <TabAtkins> Yes.
# [00:58] <TabAtkins> Link warnings happen more often as we add more specs to the db and get more collisions.
# [00:58] <fantasai> TabAtkins: Would it be terrible to let it default to thinks within a WG without warning?
# [00:58] <TabAtkins> Unfortunately.
# [00:58] <fantasai> s/thinks/things/
# [00:58] <TabAtkins> Yes - we've had this discussion in the issue tracker before. ^_^
# [00:59] <fantasai> TabAtkins: If "start" "end" or "box" end up on the warning-trigger-list, I think we need to rethink the strategy
# [00:59] <fantasai> TabAtkins: but I'll fix fragments manually for now
# [00:59] <TabAtkins> Sure.
# [00:59] <TabAtkins> Just do a link default, obvs.
# [00:59] * Quits: Florian (~Florian@public.cloak) (Ping timeout: 180 seconds)
# [00:59] * fantasai can never remember the syntax
# [01:00] <TabAtkins> Teh error message gives it to you!
# [01:00] <fantasai> Oh,that's a different one than the other one
# [01:00] <fantasai> There's ... two, right?
# [01:01] <TabAtkins> Yeah, one in the metadata, and one in a spearate block.
# [01:01] <fantasai> kk
# [01:01] <TabAtkins> I messed up with the metadata one, and designed a bad syntax.
# [01:02] * fantasai isn't sure the new one is much better
# [01:02] <TabAtkins> The separate block uses my standard InfoTree syntax, which is extensible.
# [01:02] <TabAtkins> Now it's just key/values. You have to remember the keys, but not the arbitrary syntax ordering.
# [01:02] <TabAtkins> And it looks like CSS ^_^
# [01:02] <fantasai> nah
# [01:03] <fantasai> css-backgrounds-3 { values: fixed, local; terms: background paintin garea; } would look like CSS
# [01:03] <fantasai> ^_^
# [01:04] * fantasai would probably find that easier to remember, she isn't sure how to add more values now other than duplicating everything
# [01:04] * Quits: lajava (~javi@public.cloak) (Ping timeout: 180 seconds)
# [01:04] <TabAtkins> That's the part that's not like CSS. It's instead like Python.
# [01:04] <TabAtkins> Just nest and indent.
# [01:05] <TabAtkins> Stuff higher in the "block" is shared by everything nested underneath.
# [01:05] * fantasai prefers the CSS syntax, more DRY
# [01:06] <TabAtkins> It's not as extensible - now, for each link text, figure out hwo to specify a "for" value for it.
# [01:06] <TabAtkins> Triplets aren't enough, which is what I messed up earlier.
# [01:06] <fantasai> same way as '' '' then
# [01:07] <fantasai> But yeah
# [01:08] <TabAtkins> WebKit has the "normal" mandatory behavior.
# [01:08] <TabAtkins> You can flick past things, it just forces you to stop on one of them.
# [01:09] <TabAtkins> http://software.hixie.ch/utilities/js/live-dom-viewer/saved/3669
# [01:09] <fantasai> okay
# [01:09] * Joins: jcraig (~jcraig@public.cloak)
# [01:09] * fantasai will reboot and check IE for you
# [01:09] <TabAtkins> danke
# [01:10] <fantasai> TabAtkins: Does it support proximity?
# [01:11] <fantasai> TabAtkins: also need to check 1D vs. 2D snapping behavior (put two snapped boxes far away on a diagonal)
# [01:11] * Joins: estellevw (~estellevw@public.cloak)
# [01:11] <fantasai> (bottom right one off-screen to the right)
# [01:12] * fantasai starts shutting down
# [01:13] <TabAtkins> Nope, they only do mandatory.
# [01:16] * Joins: adenilson (~anonymous@public.cloak)
# [01:17] * Joins: zcorpan (~zcorpan@public.cloak)
# [01:18] <fantasai> oh
# [01:18] <fantasai> I don't have a touchscreen
# [01:18] * Joins: jdaggett (~jdaggett@public.cloak)
# [01:18] * fantasai doesn't know how to test this
# [01:23] <fantasai> TabAtkins: Apparently I'd need to upgrade the OS to test :|
# [01:24] * Quits: antonp (~Thunderbird@public.cloak) (antonp)
# [01:24] <fantasai> TabAtkins: and, there really isn't any space left on my windows partition for that >_<
# [01:24] <TabAtkins> np, we'll deal with it later.
# [01:25] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
# [01:25] * fantasai takes the opportunity to install windows updates and reboot again...
# [01:31] <TabAtkins> Hmmmmm, their implementation of -coordinate/-destination doesn't seem to work?
# [01:32] <TabAtkins> At least, I couldn't get it to work, and also I just beachballed the browser while typing.
# [01:32] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [01:33] <fantasai> They don't have one
# [01:33] <fantasai> They have a weird other syntax
# [01:33] <TabAtkins> They... claim to?
# [01:33] <TabAtkins> https://www.webkit.org/blog/4017/scroll-snapping-with-css-snap-points/
# [01:33] <fantasai> oh, webkit you mean
# [01:33] * fantasai thought you meant msft
# [01:36] <fantasai> No idea! Ask smfr?
# [01:36] <TabAtkins> Ah, found a test page with working examlpes.
# [01:36] <TabAtkins> They do 2d snapping correctly.
# [01:36] <TabAtkins> (all at once)
# [01:37] <TabAtkins> That said, their implementation is super bad, and somehow leaks some scroll momentum out of the container into the rest of the page.
# [01:37] <TabAtkins> So if you try to flick-scroll in a container element that uses snap points, it'll scroll the page a little bit, too.
# [01:37] <fantasai> weird
# [01:38] <TabAtkins> Le sigh.
# [01:38] * Quits: jcraig (~jcraig@public.cloak) (jcraig)
# [01:39] * fantasai ran out of space trying to windows update and is now trying to solve that problem
# [01:40] * Joins: jcraig (~jcraig@public.cloak)
# [01:41] <TabAtkins> Oh, cool, here's a detail we should capture:
# [01:42] <TabAtkins> if mandatory snapping, and the most appropriate snap point is such that snapping to it would overscroll you, snap to it *as much as possible in each axis*.
# [01:42] <fantasai> good call
# [01:42] <TabAtkins> They've got some centered images here, and the ones on the edges can'ta ctually center, as that would overscroll the container.
# [01:42] <fantasai> right
# [01:42] <fantasai> They also mention clamping the scroll-snap-points-x/y properties to the scrollable area
# [01:42] <TabAtkins> I'll put it in the mechanics section.
# [01:43] <fantasai> that also needs to be mentioned
# [01:43] <TabAtkins> Ah yeah.
# [01:44] <TabAtkins> Yo, I fixed the GitHub issues thing.
# [01:44] <fantasai> Be worth mentioning in the issue about snapping to the edges of the scrollable area, that safari implements that automatically
# [01:44] <fantasai> Cool, thanks :)
# [01:46] <TabAtkins> Oh, do they?
# [01:47] <fantasai> That's what it says in their post
# [01:47] <fantasai> See Standards Compliance and Future Work
# [01:48] <fantasai> last paragraph
# [01:48] <fantasai> er, second to last
# [01:48] <fantasai> second half
# [01:49] <TabAtkins> Ah, indeed.
# [01:49] <TabAtkins> Oh, didn't notice that they explicitly called out the "unreachable snap point" thing.
# [01:49] <TabAtkins> I just noticed it in their behavior. ^_^
# [01:51] <fantasai> Based on the description, it sounds like they generate snaplines across the whole canvas
# [01:52] <fantasai> Which is a reasonably simple behavior, but not particularly useful.
# [01:52] <TabAtkins> Right.
# [01:52] <fantasai> We should specify that 1D snap positions are scoped to the viewport
# [01:52] <TabAtkins> Already handled.
# [01:52] <TabAtkins> 2d are, too.
# [01:53] <fantasai> So if none of the scroll snap area is in the viewport, it's not considered a snap
# [01:53] * Quits: tantek (~tantek@public.cloak) (tantek)
# [01:53] <fantasai> No, I think we're not super clear on that... It's gotta be in the definition of valid snap position
# [01:53] <fantasai> It's not a valid snap position to snap to a thing that's off-screen
# [01:53] <TabAtkins> It's one of the #choosing bullet points.
# [01:54] <fantasai> Yeah, I know what you're referring to :)
# [01:54] <fantasai> Choosing is about which snap position you choose. I'm saying we didn't say clearly that a scroll position that's aligned to a 1D snap area isn't a valid snap position.
# [01:55] <fantasai> if that snap area is offscreen
# [01:55] <fantasai> Okay, updates installed, time to restart. I'll come up with an edit :)
# [01:55] <TabAtkins> But... that's exactly what it says already.
# [01:56] <TabAtkins> "should be ignored if their elements are sufficiently far outside..."
# [02:00] <fantasai> Yeah, that's a different issue.
# [02:00] <fantasai> also important to note
# [02:00] <fantasai> but different issue.
# [02:07] <TabAtkins> Okay, I don't understand the distinction.
# [02:09] * fantasai will help you in a sec
# [02:10] <fantasai> TabAtkins: https://hg.csswg.org/drafts/rev/f590c73d4ef9
# [02:11] <fantasai> TabAtkins: That's hopefully clear, but if not, I'll try to explain more the distinction.
# [02:11] <TabAtkins> That looks good, but yeah, I don't understand the distinction.
# [02:12] * Quits: jcraig (~jcraig@public.cloak) (jcraig)
# [02:13] <fantasai> The distinction is between what's a valid snap position, and which snap position we should choose.
# [02:13] <fantasai> The text I added specifies that a thing that's off screen doesn't cause a snap
# [02:13] * Joins: jcraig (~jcraig@public.cloak)
# [02:13] <fantasai> The text you have specifies that a thing that's far off screen and not along the "corridor" is not a snap we should try to reach
# [02:14] <TabAtkins> ...yes?
# [02:14] <TabAtkins> Still not understanding what's different about that.
# [02:14] <TabAtkins> It's literally "if the element generating the snap point is offscreen, don't snap to that snap point".
# [02:14] <TabAtkins> That seems to be exactly what your section is saying, too.
# [02:14] <fantasai> You're saying "don't snap to that position", I'm saying "it's not a snap position"
# [02:15] * Quits: dbaron (~dbaron@public.cloak) (Ping timeout: 180 seconds)
# [02:15] <fantasai> And yours is a should, not a must, with good reason
# [02:15] <fantasai> It's up to the UA which snap position they choose
# [02:15] <fantasai> But it's up to us to define what's a valid snap position
# [02:15] <fantasai> and aligning an off-screen element is not a valid snap position
# [02:16] <fantasai> If you want to choose that off-screen element to snap to
# [02:16] <fantasai> you have to scroll over until it's at least somewhat on-screen in order to snap to it
# [02:17] * Quits: jcraig (~jcraig@public.cloak) (jcraig)
# [02:22] * Quits: adenilson (~anonymous@public.cloak) (adenilson)
# [02:24] <TabAtkins> I fail to see the essential difference between "don't" and "invalid", but ok, whatever.
# [02:24] <TabAtkins> Note its affect on mandatory, tho.
# [02:26] <fantasai> Good point
# [02:26] <fantasai> the difference is, without the text I added
# [02:26] <fantasai> mandatory would create a mandatory snap point even if the element is off screen
# [02:26] <fantasai> so if there were no other closer elements
# [02:27] <fantasai> you'd snap to that off-screen element's top edge
# [02:27] <fantasai> despite that it's off-screen
# [02:27] <fantasai> With the text I added, that's not a valid place to snap
# [02:27] <fantasai> because there's nothing to see
# [02:27] <fantasai> so either you'll keep looking down the corridor much much further away
# [02:28] <fantasai> or go backwards
# [02:28] <fantasai> or scroll in both directions
# [02:28] <TabAtkins> My text would *also* do that?
# [02:28] <fantasai> no
# [02:28] <fantasai> it wouldn't require scrolling in both directions in order to snap to the element
# [02:28] <fantasai> it says "don't choose this element, go find another one that's better"
# [02:31] <fantasai> without my text, if there are only off-screen elements and you're mandatory, you could snap to any of them while they're off-screen
# [02:31] <fantasai> my text says there's no snap position available if all the elements are off-screen
# [02:31] <fantasai> This is the scroll-snap-scope issue we discussed
# [02:31] <TabAtkins> No, my text does that too.
# [02:31] <TabAtkins> I'm seriously thinking you've misunderstood what my text says.
# [02:32] <TabAtkins> (It does it with a "should", but that's the only diff.)
# [02:35] <fantasai> also says "far offscreen"
# [02:35] <TabAtkins> Sure. Lining up with something just a bit off seems fine.
# [02:36] <fantasai> that you can't see at all?
# [02:36] <TabAtkins> You'll see it when you align to it.
# [02:36] <fantasai> even once it's snapped in?
# [02:36] <fantasai> *right*
# [02:36] <fantasai> That's my point
# [02:36] <fantasai> We don't have text that says it's only snapped when it's on the screen
# [02:36] <fantasai> if it's top-aligned, it could be 50000px off-screen and still top-aligned
# [02:36] <fantasai> without the paragraph I added, it's still considered snapped
# [02:37] <fantasai> even though it's off screen
# [02:37] <fantasai> because it is off-screen but *nonetheless* top-aligned to the viewport
# [02:37] <TabAtkins> ...but with my text, it would *also* be ignored if 50000px off-screen.
# [02:38] <fantasai> okay, fine, 2px off-screen
# [02:38] <fantasai> can't see even 1px of it
# [02:38] <fantasai> your text doesn't say to necessarily ignore that -- it's close off-screen
# [02:39] <fantasai> but it also doesn't say that in order to snap to that, you need to bring it into view
# [02:39] <TabAtkins> Sure, but if you scroll a bit more, it'll be visible. I don't wanna lock down UA strategies too much.
# [02:39] <TabAtkins> I only MUSTed when something was clearly wrong.
# [02:39] <fantasai> no, if it's to the right of the viewport, scrolling down will not bring it into view
# [02:40] <fantasai> but if it's off to the right and 10px down, scrolling down 10px will align its top to the viewport's top
# [02:42] <TabAtkins> Okay, so if that's the sole difference you see ("outside" vs "far outside"), then okay.
# [02:43] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [02:48] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [02:49] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [02:53] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [03:05] * Joins: Florian (~Florian@public.cloak)
# [03:14] * Joins: dbaron (~dbaron@public.cloak)
# [03:17] * fantasai waves to dbaron
# [03:18] * dbaron waves back
# [03:20] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
# [03:26] <fantasai> TabAtkins: Thoughts on the issue 9?
# [03:27] <TabAtkins> ...what are "space travel" effects?
# [03:28] <fantasai> Like imagine a giant black canvas (kinda like you do for slides sometimes)
# [03:29] <fantasai> and there is some content at fairly far distances apart scattered throughout
# [03:29] <fantasai> you want to snap to the next content thing, traversing all the blackness
# [03:30] <fantasai> at high speed and without much fuss, and not having the UA let the wind out of your sails because it thought maybe you want to look at the blackness
# [03:30] <fantasai> you could even use the geocities space background, for extra greatness
# [03:31] <TabAtkins> That seems like a use-case that we shouldn't care about.
# [03:31] <TabAtkins> As opposed to "mandatory, but snap-points were spaced out for a desktop screen, and work badly on mobile"
# [03:32] <fantasai> That's why we have the overly large elements thing
# [03:32] * Joins: tantek (~tantek@public.cloak)
# [03:33] <TabAtkins> Regardless, "space travel" seems like a bizarre and ridiculous use-case to optimize for.
# [03:33] <fantasai> I'm just re-imagining your slide decks, which kinda do that ^_^
# [03:34] <TabAtkins> I'm willing to bet that bad cases of "overly spaced-out mandatory points that dont' hit the overly-large condition" are, like, 10000% more common than "space travel".
# [03:34] <TabAtkins> My slide decks don't use snap-points. ^^;
# [03:34] <fantasai> Because they weren't invented yet :D
# [03:36] <fantasai> But if that's something you're concerned about, I think it goes in the definition of mandatory, not here
# [03:36] <fantasai> It's "mandatory, but not if we think you're wrong"
# [03:37] <fantasai> i.e. It's not so much about choosing snap points (which this section is about) as about how seriously we're taking "mandatory"
# [04:06] * Quits: myles (~Adium@public.cloak) ("Leaving.")
# [04:17] * Joins: zcorpan (~zcorpan@public.cloak)
# [04:25] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
# [04:50] * Quits: tantek (~tantek@public.cloak) (tantek)
# [05:09] * Joins: myles (~Adium@public.cloak)
# [05:19] * Joins: zcorpan (~zcorpan@public.cloak)
# [05:24] * Quits: jdaggett (~jdaggett@public.cloak) (jdaggett)
# [05:26] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
# [05:31] * Quits: decadance (~decadance@public.cloak) (Ping timeout: 180 seconds)
# [05:39] * Quits: estellevw (~estellevw@public.cloak) ("Snuggling with the puppies")
# [05:40] * Joins: decadance (~decadance@public.cloak)
# [06:09] * Joins: jdaggett (~jdaggett@public.cloak)
# [06:24] * fantasai waves to jdaggett
# [06:24] <fantasai> jdaggett: Will you be at TPAC?
# [06:24] <jdaggett> greetings!
# [06:24] <jdaggett> arrive sun eve, leave late tues aft
# [06:25] <jdaggett> so only for mon/tues csswg meetings
# [06:25] <fantasai> jdaggett: fair enough
# [06:25] <jdaggett> having run right now with system fonts under OSX 10.11
# [06:25] <liam> aww you'll miss halloween! (or, yay, you won't miss halloween :-) )
# [06:26] <jdaggett> turns out the ‘system’ generic
# [06:26] <fantasai> jdaggett: Any plans to draft up the Fonts 4 diffspec?
# [06:26] <jdaggett> can’t be implemented without accessing private API’s
# [06:26] <jdaggett> weee…
# [06:26] <fantasai> :/
# [06:26] <jdaggett> fun fun
# [06:26] <fantasai> File a bug on dino to make them public :)
# [06:27] <jdaggett> heh
# [06:28] <jdaggett> well, need to get it to work first…
# [06:28] <jdaggett> turns out for japanese, the OS effectively implements font-family: sys-font-en, sys-font-ja
# [06:28] <jdaggett> :)
# [06:29] <jdaggett> intending to do editing next week, including Fonts4 diffspec
# [06:29] <fantasai> Do we have the ability to combine fonts like that with @font-face?
# [06:29] <fantasai> cool!
# [06:29] * fantasai forgets
# [06:29] <jdaggett> combine means what?
# [06:29] <fantasai> I remember combining them with unicode-range
# [06:30] <fantasai> e.g. you can @font-face system { unicode-range: <english stuff>; } @font-face system { unicode-range: <japanese stuff>; } to combine two different fonts into one "family", right?
# [06:30] <jdaggett> <div lang=“ja” style=“font-family: myfont, system”>?
# [06:30] <jdaggett> hmmm
# [06:31] <jdaggett> fantasai’s new @font-face syntax…?
# [06:31] <jdaggett> go for it!!
# [06:31] * fantasai can't remember it at all
# [06:31] <jdaggett> no, can’t “override” generics via @font-face
# [06:31] <fantasai> right, right
# [06:31] <fantasai> but
# [06:31] <fantasai> s/system/foo/
# [06:31] <fantasai> But, is what Apple does sometihng like that?
# [06:32] <fantasai> or is it "if the system-font-en is missing glyphs, then use system-font-ja"/
# [06:32] <fantasai> ?
# [06:32] <jdaggett> keep in mind you don’t really *need* unicode-range to make fontlists that are a composite set of fonts
# [06:33] <jdaggett> e.g. font-family: latin-font, ja-font
# [06:33] <fantasai> Yes
# [06:33] <jdaggett> will use Latin glyphs from latin-font, japanese glyphs from ja-font
# [06:33] <fantasai> do you need them to make a composite set of fonts referenced by a single name?
# [06:33] <fantasai> or can we also combine by fallback
# [06:33] <jdaggett> hmmm
# [06:34] <jdaggett> the problem under OSX is that there’s no simple API call to say “tell me the system font for ja”
# [06:35] <jdaggett> you can only ask for the “font cascade” for the system font
# [06:35] <jdaggett> and the CJK prioritization isn’t done, J is always favored
# [06:36] <jdaggett> that’s the part that requires a private API
# [06:36] <jdaggett> :(
# [06:37] <jdaggett> so under the covers, lang=zh-Hans, font-family: system, is kinda complicated :D
# [06:39] <jdaggett> weeee...
# [06:39] <fantasai> File a bug on myles to add an API to make it straightforward? :)
# [06:39] <fantasai> I mean, he wanted it in the spec, he should make it easy to implement ;)
# [06:39] <jdaggett> yeah
# [06:40] <myles> hrm
# [06:40] <jdaggett> oh, hi myles
# [06:40] <myles> jdaggett: it will be nice to meet you in person at TPAC
# [06:40] <jdaggett> yes! i’ll be there…
# [06:41] <myles> jdaggett: i saw when you said earlier :)
# [06:41] <myles> jdaggett: please let me know if you have trouble with system fonts on OS X! :D
# [06:42] <jdaggett> sure, will do
# [06:42] <jdaggett> long and short of it is that I think CTFontCopyDefaultCascadeListForLanguages needs to be public
# [06:42] <myles> mmmm
# [06:43] <jdaggett> otherwise no way to get the correct behavior with CTFontCreateForString
# [06:43] <myles> jdaggett: what are you trying to do?
# [06:43] <jdaggett> when lang=zh-Hans
# [06:43] <jdaggett> implement correct fallback for system fonts
# [06:44] <myles> i haven't done that yet in webkit x.x
# [06:44] <jdaggett> not too hard but need that magic CoreText private call I think…
# [06:45] <myles> jdaggett: i can't speak for CoreText, but i think you should file that request as a radar and i'll make sure that team knows about it
# [06:46] <jdaggett> okay, I’ll see if I can get it to work for us, then file a radar bug
# [06:46] <myles> jdaggett: thanks :) please let me know the id of the bug you file
# [06:46] <myles> you have my email
# [06:46] <jdaggett> sure
# [06:46] <jdaggett> yup
# [06:46] <jdaggett> the semantics of ‘system’ are going to be a bit hard to define I think
# [06:46] <myles> you were on the call when we discussed them c.c
# [06:47] <jdaggett> it’s not simply “use the system font for lang”
# [06:47] <jdaggett> under OSX it’s really “use the font cascade for lang” which is subtlely different
# [06:47] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [06:48] <myles> my idea, which i don't know if it will work or not, is to essentially treat "system" as a macro for the full list of fonts
# [06:48] <myles> and to use the SPI you mentioned earlier
# [06:48] <myles> but i haven't done it so i don't have any information
# [06:48] <jdaggett> right, that will work
# [06:49] <myles> the only weird thing about that is
# [06:49] <myles> there are 20-some odd fonts in any CoreText fallback chain
# [06:49] <jdaggett> yep
# [06:49] <myles> and CoreText fallback chains in general are expected to cover pretty much all unicode code points
# [06:49] <myles> which means if you say "system, foo" the foo will pretty much never be used
# [06:49] <myles> which …. i guess is expected
# [06:50] * Joins: estellevw (~estellevw@public.cloak)
# [06:50] <myles> anyway, i have a bug to implement this stuff, so we'll see when i can carve out time to work on it
# [06:51] <jdaggett> yes, but effectively OSX *always* uses the default Latin font for other scripts, so digits in Japanese UI appear in San Francisco, not in .Hiragino Kaku Gothic UI
# [06:52] <jdaggett> so i guess that ‘system’ is just defined loosely to mean “use the platform-defined set of fonts used for a given locale”
# [06:54] <myles> so, i've always wanted to ask, does mozilla have an office in japan?
# [06:55] <jdaggett> yep!
# [06:55] * Joins: gregdavis (~udavig4@public.cloak)
# [06:55] <jdaggett> only 15 or so of us here but it’s cozy
# [06:56] * Joins: rego (~smuxi@public.cloak)
# [06:57] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [06:58] <myles> so cool :)
# [06:59] <myles> jdaggett: are all 15 of you engineers?
# [06:59] <jdaggett> complete with a drill press, always handy for browser development…
# [06:59] <myles> lolol really
# [06:59] <jdaggett> about half are engineers, rest marketing and non-profit related activities
# [07:00] <jdaggett> the drill press is part of the “maker” effort here
# [07:00] <myles> jdaggett: oh i really enjoyed https://vimeo.com/132748238
# [07:01] <myles> good job 👍
# [07:01] <jdaggett> oh cool!
# [07:02] <jdaggett> yeah, i was happy how it went
# [07:02] <myles> :)
# [07:02] <myles> jdaggett: bahaha i just saw this now https://twitter.com/nattokirai/status/646858484785418240
# [07:03] <myles> jdaggett: i just did some work on that today, actually
# [07:03] <jdaggett> great!!
# [07:03] <myles> i committed a TrueType font for testing
# [07:03] <jdaggett> yeah, saw that bug
# [07:03] <myles> you have CC'ed yourself on all my bugs ;-)
# [07:04] <jdaggett> i was surprised you were writing fonts in c++
# [07:04] <jdaggett> yup!
# [07:04] <myles> how else would you write them?
# [07:04] <myles> surely not by hand
# [07:04] * Zakim myles, you typed too many words without commas; I suspect you forgot to start with 'to ...'
# [07:04] <myles> Zakim: what
# [07:04] <jdaggett> fonttools ttx
# [07:04] <jdaggett> wonderful python package
# [07:04] <myles> yeah, i could have used a tool
# [07:04] <myles> but … i dunno
# [07:04] <myles> that's not really how i fly
# [07:04] * Joins: rego (~smuxi@public.cloak)
# [07:05] <myles> :/
# [07:05] <jdaggett> heh
# [07:05] <myles> i actually find it valuable to understand each byte in the file
# [07:05] <myles> maybe i am misguided
# [07:05] <myles> :/
# [07:05] <jdaggett> well, hard to get everything just right
# [07:06] <myles> anyway, without commenting on future releases, i really hope i can finish font-feature-settings and font-variant-* …. soon
# [07:06] <jdaggett> ttx is really useful for taking an existing font and adding or tweaking features
# [07:06] <jdaggett> yup, looking forward to it
# [07:06] <myles> (having the source to CoreText makes it much easier to make fonts my way haha......)
# [07:06] <jdaggett> heh
# [07:07] <myles> though honestly i should read harfbuzz
# [07:07] <myles> that link you gave me to the coverage API in harfbuzz was very helpful
# [07:07] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [07:07] <jdaggett> well, knowing both is handy for comparisons
# [07:07] <myles> yes
# [07:08] <myles> and you are right, CoreText currently does not expose anything relating to coverage tables as API
# [07:08] <myles> so this is my battle
# [07:09] <jdaggett> yeah, seems like the NSFont/CoreText API’s for system fonts is a work-in-progress
# [07:09] <myles> no comment
# [07:09] <jdaggett> heh
# [07:09] <jdaggett> there was one API that was discussed at WWDC that’s not public yet… :P
# [07:09] <myles> which one?
# [07:10] <jdaggett> [NSFont systemFontOfSize:weight:]
# [07:10] <myles> that's not public?
# [07:10] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [07:10] <jdaggett> not in any header I’ve found
# [07:10] <myles> :(
# [07:11] <myles> well, yet again
# [07:11] <myles> i cannot speak for CoreText
# [07:11] <jdaggett> you can seen it in the symbol dump...
# [07:11] <myles> "symbol dump" is probably my queue to change the topic of conversation
# [07:11] <jdaggett> yep :D
# [07:12] <myles> anyway, i'm going to head off
# [07:12] <jdaggett> ok, cool
# [07:12] <jdaggett> have a good one
# [07:12] <myles> jdaggett: please do not hesitate to file radars on me, i really will read them i promise
# [07:12] <myles> (i know many people don't think we do)
# [07:12] <jdaggett> great!!
# [07:12] <myles> also: buzilla bugs mean you can follow along with progress, so you might perfer those hahaha
# [07:13] <myles> *prefer
# [07:13] <jdaggett> right, understood
# [07:14] <jdaggett> my experience with radar bugs is that someone in india does the triaging…
# [07:14] <jdaggett> but i’ll try again
# [07:14] <myles> no comment c.c
# [07:14] <myles> but !!
# [07:14] <jdaggett> heh
# [07:14] <myles> if you tell me about them
# [07:14] <myles> i can make sure …. good things happen
# [07:15] <myles> :X
# [07:15] <jdaggett> excellent!!
# [07:15] * Joins: gregdavis_ (~udavig4@public.cloak)
# [07:15] <myles> have a good night (day?) jdagget
# [07:15] <myles> byebe
# [07:15] <myles> *bye
# [07:16] * Joins: rego (~smuxi@public.cloak)
# [07:16] <myles> ****jdaggett
# [07:17] <jdaggett> later
# [07:19] * Quits: gregdavis (~udavig4@public.cloak) (Ping timeout: 180 seconds)
# [07:19] * gregdavis_ is now known as gregdavis
# [07:23] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [07:29] * Joins: johanneswilm (~johannes@public.cloak)
# [07:39] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [07:47] * Joins: rego (~smuxi@public.cloak)
# [08:02] * Quits: dbaron (~dbaron@public.cloak) ("8403864 bytes have been tenured, next gc will be global.")
# [08:17] * Joins: zcorpan (~zcorpan@public.cloak)
# [08:25] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
# [08:27] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [08:33] * Joins: zcorpan (~zcorpan@public.cloak)
# [08:33] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [08:40] * Quits: estellevw (~estellevw@public.cloak) ("Snuggling with the puppies")
# [08:47] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [08:47] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [08:47] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [08:51] * Quits: gregdavis (~udavig4@public.cloak) (Client closed connection)
# [09:03] * Quits: johanneswilm (~johannes@public.cloak) (Ping timeout: 180 seconds)
# [09:06] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [09:11] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [09:13] * Joins: antonp (~Thunderbird@public.cloak)
# [09:18] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [09:19] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [09:32] * Joins: Ms2ger (~Ms2ger@public.cloak)
# [09:58] <zcorpan> fantasai: how do you explain https://github.com/w3c/csswg-drafts/commit/7fdebce23887f5e8e9ef2964611aa169e8315476#diff-dddabab940a9a268f9d0c18889cc48ebR76 ? it makes the alt text unreadable :-)
# [10:00] * Quits: jdaggett (~jdaggett@public.cloak) (jdaggett)
# [10:11] * Joins: lajava (~javi@public.cloak)
# [10:16] * Joins: Florian (~Florian@public.cloak)
# [10:25] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [10:52] <zcorpan> https://github.com/w3c/csswg-drafts/pull/57
# [10:59] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
# [11:07] * Quits: myakura (~myakura@public.cloak) ("Leaving...")
# [11:42] * Joins: antonp1 (~Thunderbird@public.cloak)
# [11:45] * Quits: antonp (~Thunderbird@public.cloak) (Ping timeout: 180 seconds)
# [11:45] * antonp1 is now known as antonp
# [13:37] * Joins: jdaggett (~jdaggett@public.cloak)
# [14:07] * Joins: plh (plehegar@public.cloak)
# [14:13] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [14:18] * Joins: tommyjtl_ (~tommyjtl@public.cloak)
# [14:18] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [14:45] * Quits: tommyjtl_ (~tommyjtl@public.cloak) (Client closed connection)
# [14:56] * Joins: dauwhe (~dauwhe@public.cloak)
# [15:13] * Joins: shepazu_ (schepers@public.cloak)
# [15:16] * Quits: shepazu (schepers@public.cloak) (Ping timeout: 180 seconds)
# [15:21] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [15:21] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [15:21] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [15:42] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [15:42] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [15:45] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [15:45] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [16:13] * shepazu_ is now known as shepazu
# [16:17] * Joins: Florian (~Florian@public.cloak)
# [16:48] * Joins: bkardell_ (~uid10373@public.cloak)
# [16:50] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
# [17:03] * Joins: adenilson (~anonymous@public.cloak)
# [17:05] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [17:06] * Joins: Florian (~Florian@public.cloak)
# [17:07] * Quits: jdaggett (~jdaggett@public.cloak) (Ping timeout: 180 seconds)
# [17:13] * Joins: rego (~smuxi@public.cloak)
# [17:14] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
# [17:15] * Joins: jdaggett (~jdaggett@public.cloak)
# [17:42] * Quits: Ms2ger (~Ms2ger@public.cloak) (Ping timeout: 180 seconds)
# [17:54] * Quits: lajava (~javi@public.cloak) (Ping timeout: 180 seconds)
# [18:03] * Quits: dauwhe (~dauwhe@public.cloak) (Client closed connection)
# [18:04] * Joins: dauwhe (~dauwhe@public.cloak)
# [18:06] * Joins: shepazu_ (schepers@public.cloak)
# [18:08] * Joins: shepazu__ (schepers@public.cloak)
# [18:12] * Quits: shepazu (schepers@public.cloak) (Ping timeout: 180 seconds)
# [18:14] * Quits: shepazu_ (schepers@public.cloak) (Ping timeout: 180 seconds)
# [18:15] * shepazu__ is now known as shepazu
# [18:30] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
# [18:44] * Joins: dbaron (~dbaron@public.cloak)
# [19:19] * Joins: estellevw (~estellevw@public.cloak)
# [19:21] * Quits: estellevw (~estellevw@public.cloak) ("Snuggling with the puppies")
# [19:22] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [19:35] * Joins: rego (~smuxi@public.cloak)
# [19:39] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [19:46] * Joins: rego (~smuxi@public.cloak)
# [19:46] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [19:55] * Joins: rego (~smuxi@public.cloak)
# [19:59] * Quits: bkardell_ (~uid10373@public.cloak) ("Connection closed for inactivity")
# [19:59] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [20:02] <TabAtkins> Merged.
# [20:06] * Joins: rego (~smuxi@public.cloak)
# [20:08] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [20:11] * Quits: dbaron (~dbaron@public.cloak) ("8403864 bytes have been tenured, next gc will be global.")
# [20:18] * Joins: rego (~smuxi@public.cloak)
# [20:18] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [20:27] * Joins: rego (~smuxi@public.cloak)
# [20:37] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [20:46] * Joins: rego (~smuxi@public.cloak)
# [20:48] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [20:58] * Joins: rego (~smuxi@public.cloak)
# [20:59] * Joins: zcorpan (~zcorpan@public.cloak)
# [21:02] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [21:07] <zcorpan> TabAtkins: is css's default.css automatically mirrored to bikeshed's stylesheet.include, or is that a manual step?
# [21:09] * Joins: rego (~smuxi@public.cloak)
# [21:13] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [21:20] * Joins: rego (~smuxi@public.cloak)
# [21:22] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [21:29] * Joins: rego (~smuxi@public.cloak)
# [21:31] * Quits: tommyjtl (~tommyjtl@public.cloak) (Client closed connection)
# [21:31] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [21:31] * Joins: tommyjtl (~tommyjtl@public.cloak)
# [21:33] <TabAtkins> zcorpan: It's a manual step. I was gonna make the fix later today.
# [21:34] <zcorpan> TabAtkins: ok, great
# [21:35] <TabAtkins> And done.
# [21:38] * Joins: rego (~smuxi@public.cloak)
# [21:38] * Quits: tommyjtl (~tommyjtl@public.cloak) (Ping timeout: 180 seconds)
# [21:40] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [21:48] * Joins: rego (~smuxi@public.cloak)
# [21:49] * Joins: AndroUser (~androirc@public.cloak)
# [21:49] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [21:57] * Joins: rego (~smuxi@public.cloak)
# [22:02] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [22:08] * Joins: rego (~smuxi@public.cloak)
# [22:08] <fantasai> zcorpan: No idea. That's clearly a bug
# [22:08] <zcorpan> fantasai: fixed now :-)
# [22:08] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [22:09] <zcorpan> TabAtkins: thx
# [22:10] * AndroUser is now known as JonathanC
# [22:17] * Joins: rego (~smuxi@public.cloak)
# [22:18] * heycam|away is now known as heycam
# [22:18] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [22:26] * Joins: rego (~smuxi@public.cloak)
# [22:34] * Quits: antonp (~Thunderbird@public.cloak) (antonp)
# [22:38] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [22:45] * Joins: rego (~smuxi@public.cloak)
# [22:51] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [22:54] * Joins: lajava (~javi@public.cloak)
# [22:58] * Joins: rego (~smuxi@public.cloak)
# [22:58] * Joins: estellevw (~estellevw@public.cloak)
# [22:58] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [23:05] * Quits: dauwhe (~dauwhe@public.cloak) (Client closed connection)
# [23:09] * Joins: rego (~smuxi@public.cloak)
# [23:11] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
# [23:18] * Joins: rego (~smuxi@public.cloak)
# [23:18] * Quits: estellevw (~estellevw@public.cloak) ("Snuggling with the puppies")
# [23:21] * Joins: dbaron (~dbaron@public.cloak)
# [23:40] * Joins: estellevw (~estellevw@public.cloak)
# [23:40] * heycam is now known as heycam|away
# [23:40] * Quits: jdaggett (~jdaggett@public.cloak) (jdaggett)
# [23:44] * Quits: estellevw (~estellevw@public.cloak) ("Snuggling with the puppies")
# [23:46] * Quits: plh (plehegar@public.cloak) ("Leaving")
# [23:53] * Joins: estellevw (~estellevw@public.cloak)
# Session Close: Fri Oct 09 00:00:00 2015

Think these logs are useful? Then please donate to show your gratitude (and keep them up, of course). Thanks! — Krijn