Options:
Previous day, Next day
- # Session Start: Fri Oct 02 00:00:00 2015
- # Session Ident: #css
- # [00:30] * Quits: lajava (~javi@public.cloak) (Ping timeout: 180 seconds)
- # [00:35] * Joins: Florian (~Florian@public.cloak)
- # [00:42] * Quits: Florian (~Florian@public.cloak) (Ping timeout: 180 seconds)
- # [01:18] * Quits: adenilson (~anonymous@public.cloak) (adenilson)
- # [01:31] * Joins: MikeJoel_real (~null@public.cloak)
- # [01:35] * Quits: MikeJoel (~null@public.cloak) (Ping timeout: 180 seconds)
- # [01:58] * Joins: Florian (~Florian@public.cloak)
- # [02:25] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
- # [02:32] * MikeJoel_real is now known as MikeJoel
- # [02:34] * Parts: MikeJoel (~null@public.cloak)
- # [02:39] * Joins: Florian (~Florian@public.cloak)
- # [02:39] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
- # [02:39] * Joins: Florian (~Florian@public.cloak)
- # [02:57] * Quits: myles (~Adium@public.cloak) ("Leaving.")
- # [03:07] * Quits: Florian (~Florian@public.cloak) ("Leaving...")
- # [03:20] * Joins: estellevw (~estellevw@public.cloak)
- # [03:26] * Quits: dbaron (~dbaron@public.cloak) ("8403864 bytes have been tenured, next gc will be global.")
- # [04:42] <dauwhe> fantasai: maybe sometime later in week of Oct 12, or early in the next week.
- # [05:09] * liam books a flight :)
- # [06:33] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
- # [06:37] * Joins: dbaron (~dbaron@public.cloak)
- # [06:40] * Joins: rego (~smuxi@public.cloak)
- # [06:54] * Quits: rego (~smuxi@public.cloak) (Client closed connection)
- # [07:01] * Joins: rego (~smuxi@public.cloak)
- # [07:44] * Quits: estellevw (~estellevw@public.cloak) ("Snuggling with the puppies")
- # [08:28] * Joins: antonp (~Thunderbird@public.cloak)
- # [08:34] * Joins: zcorpan (~zcorpan@public.cloak)
- # [09:15] * Joins: Ms2ger (~Ms2ger@public.cloak)
- # [09:26] * Joins: Florian (~Florian@public.cloak)
- # [09:49] * Joins: lajava (~javi@public.cloak)
- # [09:53] * Quits: Florian (~Florian@public.cloak) (Client closed connection)
- # [10:14] * Quits: dbaron (~dbaron@public.cloak) (Ping timeout: 180 seconds)
- # [11:41] * Quits: dholbert (~dholbert@public.cloak) (Ping timeout: 180 seconds)
- # [11:43] * Joins: dholbert (~dholbert@public.cloak)
- # [13:00] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
- # [13:02] * Joins: zcorpan (~zcorpan@public.cloak)
- # [14:21] * Quits: lajava (~javi@public.cloak) (Ping timeout: 180 seconds)
- # [14:36] * Joins: plh (plehegar@public.cloak)
- # [14:39] * Quits: dauwhe (~dauwhe@public.cloak) (Client closed connection)
- # [15:53] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
- # [15:55] * Joins: lajava (~javi@public.cloak)
- # [16:04] * Joins: dauwhe (~dauwhe@public.cloak)
- # [16:20] * Quits: dauwhe (~dauwhe@public.cloak) (Client closed connection)
- # [16:44] * Joins: darktears (~darktears@public.cloak)
- # [16:49] * Joins: estellevw (~estellevw@public.cloak)
- # [16:51] * Joins: dauwhe (~dauwhe@public.cloak)
- # [16:57] * Quits: estellevw (~estellevw@public.cloak) ("Snuggling with the puppies")
- # [17:05] * Joins: zcorpan (~zcorpan@public.cloak)
- # [17:12] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
- # [17:26] * Quits: tgraham (~user@public.cloak) (Ping timeout: 180 seconds)
- # [17:27] * Joins: estellevw (~estellevw@public.cloak)
- # [17:27] * Quits: estellevw (~estellevw@public.cloak) ("Snuggling with the puppies")
- # [17:30] * Quits: Ms2ger (~Ms2ger@public.cloak) (Ping timeout: 180 seconds)
- # [18:15] * Joins: dbaron (~dbaron@public.cloak)
- # [18:49] * Quits: plh (plehegar@public.cloak) ("Leaving")
- # [18:54] * Joins: plh (plehegar@public.cloak)
- # [18:55] * Joins: adenilson (~anonymous@public.cloak)
- # [19:06] * Joins: myles (~Adium@public.cloak)
- # [19:32] <TabAtkins> plinss: Yo, I'd like to chat with you about exposing and scraping propdef table info, so I can compile propdef indexes across specs.
- # [19:53] * Joins: zcorpan (~zcorpan@public.cloak)
- # [20:01] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
- # [20:05] * Joins: estellevw (~estellevw@public.cloak)
- # [20:05] * Quits: estellevw (~estellevw@public.cloak) ("Snuggling with the puppies")
- # [20:13] * Quits: dbaron (~dbaron@public.cloak) ("8403864 bytes have been tenured, next gc will be global.")
- # [20:23] * Quits: myles (~Adium@public.cloak) ("Leaving.")
- # [20:30] <fantasai> dauwhe: I can do the week of the 12th, but not that Friday; or I can do the Monday after if necessary. Flying out Tuesday.
- # [20:34] <dauwhe> fantasai: How about Wednesday the 14th? It seems traditional to meet on CSS day :)
- # [20:35] <fantasai> I think I can swing that
- # [20:35] <fantasai> Let me know if schedule changes, etc. I'm fairly flexible, since I've got no scheduled commitments except being in RI by noon that Friday
- # [20:36] <fantasai> (and, of course, flying out on Tuesday)
- # [20:38] <dauwhe> It's exciting to think of making progress on CSS Content :)
- # [20:51] <plinss> TabAtkins: hit me
- # [20:56] * Joins: dbaron (~dbaron@public.cloak)
- # [21:02] * Joins: myles (~Adium@public.cloak)
- # [21:03] <TabAtkins> plinss: So, um, what I said. Can we start scraping propdef/descdef/elementdef tables for their information, so I can retrieve it and generate cross-spec propdef indexes?
- # [21:04] <TabAtkins> We can either do this with the existing markup, or I can start outputting stuff in special attributes for you.
- # [21:06] <plinss> TabAtkins: yeah, I can add it, looking at the current markup now…
- # [21:14] <TabAtkins> plinss: Cool. All I really need is to be able to tell what spec it came from, so I can tell which propdefs to put into a given index.
- # [21:15] * fantasai back
- # [21:15] <plinss> do you want a different api call to get propdef data or should it be bundled into the anchor data?
- # [21:16] <plinss> TabAtkins: scraping the tables looks straightforward without extra markup
- # [21:16] <TabAtkins> That's what I thought.
- # [21:16] <fantasai> probably should be a different call, or at least different options; we already schlepping around a lot of data no?
- # [21:16] <TabAtkins> I don't care whether it's in the anchor data or another call.
- # [21:16] <TabAtkins> fantasai: It's only relevant when running `bikeshed update`
- # [21:17] <TabAtkins> And I'll want to normally pull all of that data anywya.
- # [21:17] <fantasai> fair
- # [21:17] <plinss> should I just store the data in the <td> as a string or parse it (like the value)
- # [21:17] <fantasai> that does take awhile though :)
- # [21:17] <TabAtkins> plinss: what do you mean by "parse"?
- # [21:17] <plinss> take https://drafts.csswg.org/css-grid/#track-sizing as an example
- # [21:18] <plinss> for the ‘value’ do you want: “none | <track-list> | <auto-track-list> | subgrid <line-name-list>?”
- # [21:18] <plinss> or a structure with the values broken out somehow
- # [21:18] <fantasai> string should be fine
- # [21:18] <TabAtkins> Ideally I'd like something I can still autolink with
- # [21:19] <plinss> (note that simply storing it as a string loses the links)
- # [21:19] <fantasai> the links will break if they're local anyway
- # [21:19] <fantasai> it needs to be reparsed by Bikeshed
- # [21:19] <TabAtkins> Yeah, this is why I was wondering about handing you the data manually in an attr, so I could preserve the markup appropraitely for autolinking.
- # [21:20] <fantasai> At the moment, we do need to pull data from old specs, though
- # [21:20] <fantasai> ones that aren't on bikeshed
- # [21:21] <plinss> fantasai: I can always fall back to the string value
- # [21:21] <TabAtkins> That's fine/irrelevant here.
- # [21:21] <plinss> so, yeah, putting structured data like that into an attr would help a lot, avoid parsing errors
- # [21:21] <plinss> s/attr/attrs
- # [21:22] <TabAtkins> Okay, so how about, if there's a data-original-markup on the <td>, use that, otherwise just use the string?
- # [21:22] <TabAtkins> That way I can control which fields get the treatment based on whether they need it or not.
- # [21:22] <plinss> TabAtkins: fine by me if that’s all you need
- # [21:23] <TabAtkins> I *think* so. I'll make the changes on my side, and we'll see how well it works.
- # [21:23] <plinss> do you need to know if I used the data-original-markup value?
- # [21:23] <TabAtkins> No.
- # [21:23] <plinss> ok
- # [21:23] <TabAtkins> I do need escaping to be consistent, though.
- # [21:23] <TabAtkins> So I guess if you take the string, html-escape it?
- # [21:23] <plinss> how about no?
- # [21:24] <plinss> :-)
- # [21:24] <TabAtkins> I mean, I obviously can't have both "<a ...>foo</a>" and "<track-list>" coexisting in the dataset.
- # [21:24] <plinss> I was thinking if I take the string I’d be stripping all the markup
- # [21:25] <TabAtkins> Yes.
- # [21:25] <TabAtkins> But that doesn't help "<track-list>"
- # [21:25] <plinss> you’d get that as literally: “<track-list>”
- # [21:25] <plinss> so why do you need it escaped?
- # [21:25] <TabAtkins> And that's the problem I'm pointing to. That's not okay, because I'll *also* have raw HTML (when you take a data-original-markup value)
- # [21:26] <TabAtkins> So I can't tell whether a < is supposed to be a less-than sign or a tag start.
- # [21:26] <fantasai> why not pull the data and relink it?
- # [21:26] <fantasai> plaintext
- # [21:26] <fantasai> You know it's a value
- # [21:26] <fantasai> you know what property it's for
- # [21:26] <fantasai> That's all the information you usually have to generate a link
- # [21:26] <TabAtkins> Because people sometimes have to specify autolinking details to disambiguate things.
- # [21:27] <fantasai> sure, but the thing you're specifying is what property it's for
- # [21:27] <fantasai> and that's information you have already
- # [21:27] <TabAtkins> Not always, no.
- # [21:27] <TabAtkins> Autolinks can exist in more than just the Values: line!
- # [21:27] <fantasai> no?
- # [21:27] <fantasai> oh, you mean in the rest of the propdef table
- # [21:27] <fantasai> That's a lot less important
- # [21:28] <fantasai> If you don't link anything in the rest of the propdef table for the index
- # [21:28] <fantasai> that's fine
- # [21:28] <TabAtkins> Doesn't matter, it's still important imo.
- # [21:28] <TabAtkins> No reason to fall down here.
- # [21:28] <fantasai> It's not. It's an index table
- # [21:28] <fantasai> just click on the property to get the full definition
- # [21:28] <fantasai> Most people will be looking to click on either the property name or a value name
- # [21:28] <fantasai> they don't care about Computed Value or other things like that
- # [21:29] <fantasai> not on a propindex table
- # [21:29] <TabAtkins> That's cool, but again, there's no reason for me to fall down on this. It's not more difficult to handle than not.
- # [21:29] <fantasai> We don't even include those in the table
- # [21:29] <fantasai> http://www.w3.org/TR/CSS/#properties
- # [21:30] <fantasai> It's the name, value, applies to, inherited, percentages, and media
- # [21:30] <TabAtkins> And anyway, I'm not super-enthused about having to reparse possibly-ambiguous naked text with the assumption that it'll reverse in a simple way.
- # [21:30] <TabAtkins> "Applies to" matters too
- # [21:31] <plinss> So, back to escaping, not sure why you need html escaping on my end
- # [21:32] <TabAtkins> plinss: Because someone can write, in their markup, "<foo>", which when you grab the plain text, will look like "<foo>", which when fed back to Bikeshed, will look like markup.
- # [21:32] <TabAtkins> And Bikeshed will have other fields which *do* contain markup (via the data-original-markup attribute), so it needs to be able to tell consistently whether something is markup or not.
- # [21:32] <TabAtkins> Either with an indicator in the data, or with the data just being self-consistent.
- # [21:34] <plinss> ok, got it, so when I don’t have a data-original-markup you just need anything that looks like markup to be entities instead
- # [21:34] <plinss> (after I strop all the actual markup)
- # [21:34] <TabAtkins> Yes.
- # [21:34] <plinss> s/strop/strip/
- # [21:34] <plinss> I’m good with that
- # [21:34] <TabAtkins> Which is just the &< conversion.
- # [21:34] <plinss> right
- # [21:39] * Joins: estellevw (~estellevw@public.cloak)
- # [22:34] * Quits: estellevw (~estellevw@public.cloak) ("Snuggling with the puppies")
- # [22:36] <fantasai> plinss: pushing seems to hang a lot atm, fwiw
- # [22:43] <TabAtkins> *pushing to hg
- # [22:43] <fantasai> yeah, it's totally hung up now
- # [22:43] <fantasai> and github mirror is behind >_<;
- # [22:56] * Quits: dauwhe (~dauwhe@public.cloak) ("")
- # [23:16] <fantasai> Discussing scroll-snapping
- # [23:16] <fantasai> Map case is clear use case for 2D
- # [23:16] <fantasai> but most cases will want 1D snapping
- # [23:16] <fantasai> should make that easy to get, easiest to get
- # [23:17] <fantasai> because don't want authors to get 2D snapping unless they think about it and want it in
- # [23:17] <fantasai> should not be dependent on how the alignment is specified
- # [23:17] <fantasai> because 1D start snapping and center snapping ar both legit uses
- # [23:17] <fantasai> so need some other way to distinguish....
- # [23:17] * fantasai is dumping notes into IRC from discussion with Tab
- # [23:19] * Quits: darktears (~darktears@public.cloak) ("Leaving...")
- # [23:27] <TabAtkins> align: <position-1d> | [ block <position-1d> || inline <position-1d> ] | align <position-2d>
- # [23:30] <fantasai> <p> | [ x <p> || y <p> ] | xy <p>
- # [23:33] * Joins: estellevw (~estellevw@public.cloak)
- # [23:34] <fantasai> <position-1d> = start | end | center | <percentage> | <length> | edges
- # [23:35] <fantasai> <position-1d> = start | end | center | edges | <percentage>
- # [23:37] <fantasai> <position-2d> = <position-1d>{1,2}
- # [23:37] <fantasai> ?
- # [23:41] * Quits: plh (plehegar@public.cloak) ("Leaving")
- # [23:41] <fantasai> Discussiong bgposition syntax vs alignment syntax
- # [23:41] <fantasai> Allowing <length>s here is harmful, because we have the scroll-snap margin for adjusting the scroll snap area box
- # [23:41] * Quits: adenilson (~anonymous@public.cloak) (adenilson)
- # [23:42] <fantasai> <length> here would duplicate the functionality
- # [23:42] <fantasai> We resolve the conflict on the UA side by adding, that's easy, but it creates a cascading problem
- # [23:42] <fantasai> complicates the model for authors because 2 ways to do the same thing
- # [23:42] <fantasai> that interact in unexpected ways if different bits of code try to do it with different techniques
- # [23:43] <fantasai> scroll-snap margin gives better behavior because it affects the scrollable area
- # [23:43] <fantasai> the common thing (show more space) is positive
- # [23:43] <fantasai> and it handles overflow better, which helps the user
- # [23:43] <fantasai> So it's best to not provide <length>, and to guid the author down the path of scroll-snap-margin
- # [23:43] * plinss fantasai: server/pushing should be sorted now
- # [23:44] <fantasai> <position-1d> aligns with the alignment properties
- # [23:44] <fantasai> <percentage> is new, but would totally make sense there inasmuch as it makes sense here
- # [23:44] <fantasai> the cases where you'd use it are similar
- # [23:44] <fantasai> it represents a percentage between start and end
- # [23:45] * Quits: estellevw (~estellevw@public.cloak) ("Snuggling with the puppies")
- # [23:45] * fantasai plinss thanks
- # [23:51] * Joins: estellevw (~estellevw@public.cloak)
- # [23:55] <fantasai> +longhands for padding: trbl
- # [23:56] <fantasai> +longhands for area; trbl
- # [23:57] <fantasai> drop scroll-snap-group
- # [23:57] * Quits: estellevw (~estellevw@public.cloak) (estellevw)
- # Session Close: Sat Oct 03 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