Options:
- # Session Start: Tue Mar 04 00:00:00 2014
- # Session Ident: #css
- # [00:18] * Quits: sgalineau (~sgalineau@public.cloak) ("Textual IRC Client: www.textualapp.com")
- # [00:33] * Joins: dauwhe_ (~dauwhe@public.cloak)
- # [00:33] * Quits: dauwhe (~dauwhe@public.cloak) (Client closed connection)
- # [00:58] * Quits: dauwhe_ (~dauwhe@public.cloak) (Client closed connection)
- # [00:58] * Joins: dauwhe (~dauwhe@public.cloak)
- # [01:09] * Joins: Dwight_Stegall (~YIM@public.cloak)
- # [01:09] * Quits: Dwight_Stegall (~YIM@public.cloak) ("I'm bored, time to go.")
- # [01:10] * Joins: Dwight_Stegall (~YIM@public.cloak)
- # [01:26] * Quits: lmclister (~lmclister@public.cloak) ("")
- # [01:35] <SimonSapin> TabAtkins: Tring to grok the <grid-line> syntax … I think <custom-ident> | [ <integer> && <custom-ident>? ] is identical to <integer> || <custom-ident> in terms of grammar.
- # [01:35] <SimonSapin> Is there a reason not to write it that way?
- # [01:37] <TabAtkins> Yes, it's identical. Writing it that way makes it clearer that they are different syntax forms, and gives a better connection to the definition of each form down the in <dl>.
- # [01:39] <SimonSapin> Eh. This doesn’t fit at all in Gecko’s nsCSSValue :)
- # [01:40] <TabAtkins> Hey, black-box compatible. Do what you want.
- # [01:40] <SimonSapin> sure
- # [01:40] <SimonSapin> I’m complaining about Gecko more than about the spec
- # [01:41] <TabAtkins> Ah, kk.
- # [01:44] <SimonSapin> In Servo each property has its own pair of data types for specified and computed values
- # [01:44] <SimonSapin> <3 Rust enum
- # [01:53] * Quits: rhauck (~Adium@public.cloak) ("Leaving.")
- # [01:58] * Joins: jet (~junglecode@public.cloak)
- # [02:42] * Quits: jet (~junglecode@public.cloak) (jet)
- # [03:02] * Quits: Dwight_Stegall (~YIM@public.cloak) ("I'm bored, time to go.")
- # [04:28] * Joins: dauwhe_ (~dauwhe@public.cloak)
- # [04:34] * Quits: dauwhe (~dauwhe@public.cloak) (Ping timeout: 180 seconds)
- # [05:52] * Joins: jet (~junglecode@public.cloak)
- # [05:52] * Parts: jet (~junglecode@public.cloak) (jet)
- # [07:20] <fantasai> TabAtkins: btw, I'm free to work on Thursday if you want
- # [07:28] * Joins: Dwight_Stegall (~YIM@public.cloak)
- # [08:34] * Quits: Dwight_Stegall (~YIM@public.cloak) ("I'm bored, time to go.")
- # [08:35] * Joins: Dwight_Stegall (~YIM@public.cloak)
- # [08:36] * Quits: Dwight_Stegall (~YIM@public.cloak) ("I'm bored, time to go.")
- # [08:41] * Joins: Dwight_Stegall (~YIM@public.cloak)
- # [08:41] * Quits: Dwight_Stegall (~YIM@public.cloak) ("I'm bored, time to go.")
- # [08:54] * Joins: Dwight_Stegall (~YIM@public.cloak)
- # [09:07] * Quits: Dwight_Stegall (~YIM@public.cloak) ("I'm bored, time to go.")
- # [09:10] * Joins: Dwight_Stegall (~YIM@public.cloak)
- # [09:16] * Quits: Dwight_Stegall (~YIM@public.cloak) ("I'm bored, time to go.")
- # [09:16] * Joins: Dwight_Stegall (~YIM@public.cloak)
- # [09:19] * Quits: Dwight_Stegall (~YIM@public.cloak) ("I'm bored, time to go.")
- # [09:22] * Joins: Dwight_Stegall (~YIM@public.cloak)
- # [09:33] * Quits: dbaron (~dbaron@public.cloak) (Ping timeout: 180 seconds)
- # [09:36] * Joins: Ms2ger (~Ms2ger@public.cloak)
- # [09:41] * Joins: jet (~junglecode@public.cloak)
- # [10:12] * Joins: zcorpan (~zcorpan@public.cloak)
- # [10:19] * Joins: plh (plehegar@public.cloak)
- # [10:19] * Joins: plh3 (plehegar@public.cloak)
- # [10:26] * Quits: plh (plehegar@public.cloak) (Ping timeout: 180 seconds)
- # [10:26] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
- # [10:27] * Joins: zcorpan (~zcorpan@public.cloak)
- # [10:28] * Joins: zcorpan_ (~zcorpan@public.cloak)
- # [10:28] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
- # [10:30] * Quits: zcorpan_ (~zcorpan@public.cloak) (Client closed connection)
- # [10:30] * Joins: zcorpan (~zcorpan@public.cloak)
- # [10:37] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
- # [10:44] * Quits: jet (~junglecode@public.cloak) (jet)
- # [11:10] * Quits: Dwight_Stegall (~YIM@public.cloak) ("I'm bored, time to go.")
- # [11:31] * Joins: zcorpan (~zcorpan@public.cloak)
- # [11:36] * plh3 is now known as plh
- # [11:38] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
- # [12:08] * Quits: birtles (~uid16523@public.cloak) ("Connection closed for inactivity")
- # [12:32] * Joins: zcorpan (~zcorpan@public.cloak)
- # [12:35] * Quits: plh (plehegar@public.cloak) ("Leaving")
- # [12:39] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
- # [13:14] * Joins: zcorpan (~zcorpan@public.cloak)
- # [13:21] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
- # [14:14] * Joins: zcorpan (~zcorpan@public.cloak)
- # [14:18] <SimonSapin> TabAtkins, fantasai: a note in the Grid ED would be helpful to explain what’s going on with the three versions of the layout algorithm
- # [14:22] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
- # [14:38] * Quits: Ms2ger (~Ms2ger@public.cloak) (Ping timeout: 180 seconds)
- # [15:01] * Joins: Dwight_Stegall (~YIM@public.cloak)
- # [15:14] * Quits: dauwhe_ (~dauwhe@public.cloak) (Client closed connection)
- # [15:15] * Joins: dauwhe (~dauwhe@public.cloak)
- # [15:15] * Joins: zcorpan (~zcorpan@public.cloak)
- # [15:22] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
- # [16:16] * Joins: zcorpan (~zcorpan@public.cloak)
- # [16:17] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
- # [16:18] * Joins: zcorpan (~zcorpan@public.cloak)
- # [16:25] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
- # [16:38] * Joins: glenn (~gadams@public.cloak)
- # [17:05] * Joins: rhauck (~Adium@public.cloak)
- # [17:07] * Joins: zcorpan (~zcorpan@public.cloak)
- # [17:07] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
- # [17:07] * Joins: zcorpan (~zcorpan@public.cloak)
- # [17:20] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
- # [17:20] * Joins: zcorpan (~zcorpan@public.cloak)
- # [17:28] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
- # [17:37] * Joins: jet (~junglecode@public.cloak)
- # [17:44] * Joins: lmclister (~lmclister@public.cloak)
- # [17:45] * Joins: Ms2ger (~Ms2ger@public.cloak)
- # [17:46] * Joins: shepazutu (schepers@public.cloak)
- # [17:52] * Quits: shepazu (schepers@public.cloak) (Ping timeout: 180 seconds)
- # [17:52] * shepazutu is now known as shepazu
- # [18:12] * Joins: dbaron (~dbaron@public.cloak)
- # [18:13] * Quits: rhauck (~Adium@public.cloak) ("Leaving.")
- # [18:13] * Joins: adenilson (~anonymous@public.cloak)
- # [18:19] * Joins: tantek (~tantek@public.cloak)
- # [18:25] * Quits: Dwight_Stegall (~YIM@public.cloak) ("I'm bored, time to go.")
- # [18:29] * Quits: Ms2ger (~Ms2ger@public.cloak) (Ping timeout: 180 seconds)
- # [18:29] * Joins: Dwight_Stegall (~YIM@public.cloak)
- # [18:41] * Quits: jet (~junglecode@public.cloak) (jet)
- # [18:46] * Joins: rhauck (~Adium@public.cloak)
- # [19:00] * Joins: Ms2ger (~Ms2ger@public.cloak)
- # [19:11] * Joins: jet (~junglecode@public.cloak)
- # [19:17] * Quits: Dwight_Stegall (~YIM@public.cloak) ("I'm bored, time to go.")
- # [19:22] * Quits: Ms2ger (~Ms2ger@public.cloak) ("bbl")
- # [19:31] <TabAtkins> fantasai: I have a phone interview from 3:30 to 4:15.
- # [19:31] <TabAtkins> on Thursday.
- # [19:33] * Quits: rhauck (~Adium@public.cloak) ("Leaving.")
- # [19:33] * Joins: rhauck (~Adium@public.cloak)
- # [19:48] * Joins: Dwight_Stegall (~YIM@public.cloak)
- # [19:50] * liam hopes the telephone passes the interview
- # [19:51] <TabAtkins> liam: They're usually pretty bad at it.
- # [19:54] * Joins: tantek_ (~tantek@public.cloak)
- # [19:54] <liam> :D
- # [20:21] * Quits: tantek_ (~tantek@public.cloak) (tantek_)
- # [20:22] * Joins: Ms2ger (~Ms2ger@public.cloak)
- # [20:38] * Quits: jet (~junglecode@public.cloak) (jet)
- # [20:55] * Quits: rhauck (~Adium@public.cloak) ("Leaving.")
- # [21:03] * Joins: plh (plehegar@public.cloak)
- # [21:03] * Joins: plh3 (plehegar@public.cloak)
- # [21:03] * Quits: plh (plehegar@public.cloak) (Client closed connection)
- # [21:04] * plh3 is now known as plh
- # [21:09] <shepazu> does text-decoration trigger reflow, or just repaint?
- # [21:11] <SimonSapin> shepazu: that’s up to implementations, but IIRC it doesn’t affect layout
- # [21:13] <shepazu> SimonSapin, thanks :)
- # [21:14] <shepazu> SimonSapin, how can you tell if any given property affects reflow?
- # [21:14] <Ms2ger> If you can, there's an implementation bug :)
- # [21:15] <SimonSapin> shepazu: if in any of the specs’ layout algorithms, the size or position of any fragment depends (possibly indirectly) on that property
- # [21:16] * Quits: dbaron (~dbaron@public.cloak) (Ping timeout: 180 seconds)
- # [21:17] <SimonSapin> but it’s actually a fuzzy concept
- # [21:18] <SimonSapin> since many things can affect the "overflow area" of some element
- # [21:18] <SimonSapin> which is how much you can scroll if that element has scrollbars
- # [21:18] <SimonSapin> so… I’m not sure
- # [21:19] <shepazu> ah, well, seemed like something that affects performance like that might have been easy to pinpoint
- # [21:20] * Quits: Dwight_Stegall (~YIM@public.cloak) (Client closed connection)
- # [21:55] * Joins: dbaron (~dbaron@public.cloak)
- # [21:58] * Quits: tantek (~tantek@public.cloak) (tantek)
- # [22:28] * Joins: rhauck (~Adium@public.cloak)
- # [22:29] <TabAtkins> SimonSapin: I added a note to all three of the algo sections.
- # [22:31] <SimonSapin> TabAtkins: cool
- # [22:33] <SimonSapin> TabAtkins: any more idea what to do about <custom-ident> vs. shorthands?
- # [22:33] * Joins: jet (~junglecode@public.cloak)
- # [22:33] <TabAtkins> The best I've got is:
- # [22:33] <TabAtkins> 1) Only make the automatic restriction apply to keywords in the same simple group/function (or top-level, if it lives in the top level).
- # [22:33] <TabAtkins> (This is what I said in the email.)
- # [22:34] <TabAtkins> 2) Explicitly require spec authors to define what types of keywords are disallowed, with an easy way to say "whatever ones are used by the grammar". Not doing so is a spec error.
- # [22:34] <SimonSapin> simple group is what Syntax calls a block?
- # [22:34] <TabAtkins> Sorry yeah, simple block.
- # [22:35] <SimonSapin> so CSS-wide keywords are only disallowed at property value top-level?
- # [22:35] <TabAtkins> Yeah, I think so.
- # [22:36] <SimonSapin> that simplified Grid’s <line-names>, good
- # [22:36] <SimonSapin> simplifies*
- # [22:37] <TabAtkins> Yeah, it makes the <line-names> completely unrestricted.
- # [22:40] * Joins: dholbert (~dholbert@public.cloak)
- # [22:40] <SimonSapin> yeah, same block/functions sounds like a good boundary. It may still restrict cases that are not actually ambiguous, but it’s easy to define
- # [22:40] <TabAtkins> Yup yup.
- # [22:40] <TabAtkins> So I think we shouldn't do my #2 above, but probably make sure it's *possible* to specify a particular list of disallowed keywords.
- # [22:41] <SimonSapin> why?
- # [22:41] <TabAtkins> It's probably fine to just default to "disallow anything else in the same context".
- # [22:44] <SimonSapin> where do we have <custom-ident> yet? counter names, counter styles, Grid. More?
- # [22:44] * Joins: tantek (~tantek@public.cloak)
- # [22:45] <astearns> I should probably change named flows to <custom-ident>
- # [22:45] <TabAtkins> Animation names and font names *should* be using <custom-ident>, but they dont'. (And they'd need to customize the disallowed stuff.)
- # [22:48] <TabAtkins> SimonSapin: How did you and I miss these bugs in the parse algos? >_<
- # [22:48] <SimonSapin> yeah :/
- # [22:49] <SimonSapin> typical case of "I know what this is supposed to say"
- # [22:50] <SimonSapin> I’ve heard the argument that implementers should not also be editors of the spec they implement, for this reason
- # [22:50] <TabAtkins> Heh, yeah.
- # [22:51] <dauwhe> GCPM uses <custom-ident> for the names of named strings, etc. But pages spec uses <identifier> for named pages.
- # [22:51] <TabAtkins> Hm, the consume algorithms are somewhat inconsistent on whether they start with the current or next input token. I think I need to make them consistent.
- # [22:52] <SimonSapin> "CR means we found all the bugs already, right?" :)
- # [22:54] <TabAtkins> It means we're done with design work, and hope there's only minor bugs left.
- # [22:54] <SimonSapin> hope sounds right
- # [22:55] <SimonSapin> TabAtkins: re <custom-ident>, with the same group/function thing, Grid doesn’t have a problem with shorthands anymore
- # [22:55] <TabAtkins> Yeah.
- # [22:55] <SimonSapin> but I wonder if non-ambiguous things could be come ambiguous when stuck in the same shorthand
- # [22:56] <TabAtkins> Yes.
- # [22:56] <TabAtkins> That is, yes they can.
- # [22:56] <SimonSapin> so maybe we shouldn’t make shorthands special
- # [22:56] <TabAtkins> What do you mean by "special"?
- # [22:57] <SimonSapin> http://lists.w3.org/Archives/Public/www-style/2014Mar/0011.html
- # [22:58] <SimonSapin> if <line-names> didn’t have parens
- # [22:58] <TabAtkins> Still don't know which behavior you intend to mean is "special".
- # [22:59] <SimonSapin> if <line-names> didn’t have parens, should "dense" be excluded in <line-names> in the 'grid' shorthand?
- # [22:59] <SimonSapin> current spec says yes
- # [23:00] <SimonSapin> which means we can’t always serialize the shorthand (in a way that’s legal to parse), as you pointed out
- # [23:00] <TabAtkins> Yes. So which behavior do you think is "special"? ^_^
- # [23:01] <TabAtkins> I completely understand what the possible behaviors are, just not which you consider "normal" and which you consider "special".
- # [23:02] <TabAtkins> Or rather, which one you'd prefer. That's the actual information I'd like to know.
- # [23:02] <SimonSapin> one option is to change the rules to make shorthands an exception: when parsing <custom-ident> in a shorthand, only exclude other keywords from the same longhands
- # [23:02] <SimonSapin> but that’s also problematic, as it could introduce ambiguities in parsing
- # [23:02] <TabAtkins> What do you mean by "the same longhands"?
- # [23:03] <TabAtkins> Case in point - what keywords are disallowed for the <counter-style> when writing 'list-style'?
- # [23:03] <TabAtkins> Just the keywords from 'list-style-type'? Or all the keywords from all the list-style longhands?
- # [23:03] * TabAtkins is sorry, but you keep using words in ambiguous ways, so I don't know which options you're referring to. :/
- # [23:04] <SimonSapin> 'subgrid' is in the same longhand (grid-template-{columns,rows}) as <line-names>), 'dense' isn’t
- # [23:04] <SimonSapin> although they can both appear in the 'grid' shorthand
- # [23:04] <TabAtkins> Okay, "from the longhand the <custom-ident> appears in".
- # [23:04] <SimonSapin> yes
- # [23:04] <TabAtkins> The ambiguity clearly breaks that.
- # [23:05] <SimonSapin> yes, so the exception for shorthands that I described above is a bad idea
- # [23:05] <TabAtkins> Though the ambiguity is only troublesome if there's a meaningful ordering constraint mixed in as well.
- # [23:05] <TabAtkins> But still, that's too subtle to base a distinction on.
- # [23:05] <SimonSapin> but not having it is also a bad idea, for serialization
- # [23:05] <TabAtkins> Sometimes you can't serialize shorthands.
- # [23:05] <TabAtkins> That's already something that happens.
- # [23:05] <SimonSapin> oh, ok
- # [23:06] <TabAtkins> Can't think of an example off the top of my head, but they exist.
- # [23:06] <SimonSapin> right
- # [23:06] <SimonSapin> so, the proposed change is to only exclude keywords (including css-wide) that appear at the same "level"
- # [23:07] <TabAtkins> Ugh, I have current/next token bugs littering these algorithms.
- # [23:07] <TabAtkins> Yes.
- # [23:07] * Quits: Ms2ger (~Ms2ger@public.cloak) ("nn")
- # [23:07] <TabAtkins> Switching everything to consistently start by looking at the next input token, ignoring the current one.
- # [23:07] <SimonSapin> can we try to resolve on it tomorrow?
- # [23:07] <TabAtkins> Yeah. I won't be in the call, but go for it.
- # [23:08] <SimonSapin> (I have patches in review blocked on this :))
- # [23:08] <SimonSapin> (well, not just on this but still)
- # [23:13] * Quits: renoirb (renoirb@public.cloak) ("Having some rest")
- # [23:24] * Joins: renoirb (renoirb@public.cloak)
- # [23:24] <SimonSapin> TabAtkins: Can I leave today’s Syntax issues to you?
- # [23:24] <TabAtkins> Please do, since I'm in the mdidle of them right now.
- # [23:25] <SimonSapin> cool
- # [23:37] * Quits: adenilson (~anonymous@public.cloak) (adenilson)
- # [23:38] * Quits: plh (plehegar@public.cloak) ("Leaving")
- # [23:40] <TabAtkins> All right, Syntax should be done now.
- # Session Close: Wed Mar 05 00:00:00 2014
The end :)