IRC log of css on 2008-12-03

Timestamps are in UTC.

16:56:40 [RRSAgent]: RRSAgent has joined #css
16:56:40 [RRSAgent]: logging to http://www.w3.org/2008/12/03-css-irc
16:56:51 [glazou]: Zakim, this is Style
16:56:51 [Zakim]: ok, glazou; that matches Style_CSS FP()12:00PM
16:57:13 [Dsinger]: We had an agenda etc so I assume so...
16:57:26 [glazou]: anne: we do have a call
16:57:50 [Zakim]: + +1.858.354.aaaa
16:58:08 [plinss_]: zakim, +1.858.354 is me
16:58:08 [Zakim]: +plinss_; got it
16:58:09 [glazou]: Zakim, who is on the phone?
16:58:10 [Zakim]: On the phone I see dsinger, glazou, plinss_
16:58:17 [Dsinger]: Glazou: on bus on iPhone on mute to say 9:25...sorry
16:58:19 [Zakim]: -plinss_
16:58:22 [glazou]: Dsinger: ok
16:58:37 [Zakim]: +plinss_
17:00:30 [Dsinger]: Old. Sorry to hear that
17:00:40 [Dsinger]: Oof
17:00:45 [glazou]: sh** happens
17:00:45 [Zakim]: -dsinger
17:01:29 [Zakim]: +Bert
17:01:31 [Zakim]: +[Microsoft]
17:01:55 [sylvaing]: Zakim, [Microsoft] has sylvaing
17:01:55 [Zakim]: +sylvaing; got it
17:02:04 [Zakim]: +dsinger
17:02:08 [Zakim]: +saloni
17:02:26 [Dsinger]: Dsinger has joined #css
17:02:48 [Dsinger]: Zakin, mute me
17:03:11 [Dsinger]: Zakim, mute me
17:03:11 [Zakim]: dsinger should now be muted
17:03:55 [dbaron]: dbaron has joined #css
17:03:56 [Dsinger]: A big tyou to AT&T for dropping me...
17:04:05 [Zakim]: +[Mozilla]
17:04:22 [dbaron]: Zakim, [Mozilla] has David_Baron
17:04:22 [Zakim]: +David_Baron; got it
17:04:34 [anne]: Zakim, passcode?
17:04:34 [Zakim]: the conference code is 78953 (tel:+1.617.761.6200 tel:+33.4.89.06.34.99 tel:+44.117.370.6152), anne
17:05:25 [Zakim]: +[IPcaller]
17:05:35 [anne]: Zakim, [IPCaller] is me
17:05:35 [Zakim]: +anne; got it
17:05:55 [dbaron]: Zakim, who is on the phone?
17:05:55 [Zakim]: On the phone I see glazou, plinss_, Bert, [Microsoft], dsinger (muted), saloni, [Mozilla], anne
17:05:57 [Zakim]: [Mozilla] has David_Baron
17:05:57 [Zakim]: [Microsoft] has sylvaing
17:07:29 [Dsinger]: Not on an iPhone, sorry
17:07:33 [glazou]: ScribeNick: glazou
17:07:56 [glazou]: plinss: extra agenda items ?
17:08:07 [glazou]: plinss: apparently not
17:08:15 [glazou]: plinss: first item on the agenda
17:08:31 [plinss_]: http://wiki.csswg.org/spec/css2.1#issue-86
17:08:41 [glazou]: plinss: issue 86
17:08:46 [Zakim]: +SteveZ
17:08:53 [glazou]: plinss: list item marker placement follow-up
17:09:08 [alexmog]: alexmog has joined #css
17:09:11 [glazou]: plinss: we had discussion about this on www-style and twitter
17:09:20 [dbaron]: Zakim, who is noisy?
17:09:25 [glazou]: plinss: text-align vs. list-style-position inside/outside
17:09:30 [Zakim]: dbaron, listening for 10 seconds I heard sound from the following: plinss_ (5%), glazou (24%), SteveZ (53%)
17:10:03 [glazou]: glazou: I can summarize
17:10:33 [glazou]: glazou: inside makes the marker glue to the text
17:10:49 [glazou]: dbaron: this change affects page that use no cssat all
17:11:05 [glazou]: dbaron: we're worried about the larger set that don't use list-style-position
17:11:12 [glazou]: s/set/set of users
17:11:23 [glazou]: plinss: I am missing something here
17:11:30 [glazou]: dbaron: because ouside is the default
17:11:54 [Zakim]: +??P7
17:12:03 [glazou]: dbaron: in a css-based renderer, we probably need a new value of list-style-position and make it the new default value
17:12:15 [dbaron]: Zakim, ??P7 is fantasai
17:12:15 [Zakim]: +fantasai; got it
17:12:42 [glazou]: plinss: I see the problem about changing the behavior about that
17:12:51 [glazou]: dbaron: I am not even sure changing is the right word
17:13:02 [dbaron]: since what everybody does disagrees with the spec
17:13:28 [glazou]: saloni: I also pinged www-style about floats and it seems to be the same issue
17:13:41 [glazou]: saloni: there is divergence between specs and implems
17:14:13 [glazou]: saloni: align with the start of the first line box
17:14:29 [fantasai]: fantasai has joined #css
17:14:34 [glazou]: saloni: then we can deal with center
17:14:45 [glazou]: saloni: put marker at some distance of line box is ok to me
17:14:46 [fantasai]: RRSAgent: pointer
17:14:46 [RRSAgent]: See http://www.w3.org/2008/12/03-css-irc#T17-14-46
17:15:23 [fantasai]: RRSAgent: make logs public
17:15:40 [glazou]: stevez: there is a convention here in publishing
17:16:06 [glazou]: stevez: the bullets don't overlap the marker
17:16:27 [glazou]: glazou: a desirable behavior is marker above the list item if centered
17:16:50 [Dsinger]: Sure if it's numbered
17:17:15 [glazou]: stevez: yes that's a common case
17:17:20 [Dsinger]: Sub-lists?
17:17:53 [glazou]: glazou: I was just quoting the comments I got from web designers
17:18:06 [glazou]: all: we can do that in css3 in different ways
17:18:15 [glazou]: fantasai: use :before
17:18:27 [Dsinger]: I cannot imagine where centered sublists go
17:18:48 [glazou]: plinss: do we have a consensus on initial issue here ?
17:19:11 [glazou]: fantasai: previous resolution on overflow
17:19:15 [glazou]: dbaron: saying what ?
17:19:35 [glazou]: saloni: marker that does not screw with div that has overflow
17:20:35 [glazou]: saloni: more sites using floats that text-align
17:21:17 [glazou]: fantasai: text-align shifts line box
17:21:31 [Zakim]: -SteveZ
17:21:54 [Zakim]: +SteveZ
17:21:54 [fantasai]: fantasai questioned whether we can get the designers' expected behavior for text-align if we define its position as an offset
17:21:57 [glazou]: saloni: what if we took the max of the list item box or the first line box ?
17:22:09 [arronei]: arronei has joined #CSS
17:22:46 [sylvaing]: Zakim, [Microsoft] has alexmog
17:22:46 [Zakim]: +alexmog; got it
17:22:56 [glazou]: dbaron: I am not sure I understand
17:23:23 [glazou]: saloni: the max of width of the entire list item and the width of 1st line box
17:23:47 [glazou]: saloni:if you pick the wider box, and offset from that one
17:23:59 [glazou]: dbaron: the list item itself is always wider than the line box
17:24:43 [fantasai]: fantasai: need to be very careful in defining this because text-indent also needs to work properly here
17:25:12 [dbaron]: neither text-indent nor text-align affect the size of the line box
17:25:27 [dsinger_]: dsinger_ has joined #css
17:25:28 [glazou]: plinss: no resolution here, we don't agree on behaviour
17:25:31 [Zakim]: +dsinger.a
17:25:40 [Zakim]: -dsinger
17:26:14 [glazou]: dbaron: neither text-align nor text-indent affect the size of the line box
17:26:29 [glazou]: plinss: affected by a float ?
17:26:31 [glazou]: dbaron: yes
17:26:44 [glazou]: dbaron: I believe
17:27:23 [glazou]: can I recommend another minute taker, my phone connection is really bad
17:27:34 [sylvaing]: should we have one or more testcases and agree on what the current specified behavior is for them ? since it seems current implementations are not useful in figuring what that is...
17:27:54 [fantasai]: ScribeNick: fantasai
17:28:26 [fantasai]: fantasai: Defining the marker position as an offset from the start of the first line box will not solve the float problem in a reasonable way
17:28:34 [fantasai]: fantasai: It will place the marker on top of the float
17:28:41 [fantasai]: fantasai: that is not what designers want
17:29:22 [dbaron]: dbaron: we need the float displacement properties in css3-box for that
17:30:11 [fantasai]: fantasai: Is the line box as wide as the containing block except in the presence of floats?
17:30:15 [fantasai]: dbaron: yes, I think that's the case
17:31:02 [fantasai]: stevez: It would seem to me that if you have a float for it to push the line box and the float
17:31:22 [fantasai]: stevez: since the marker is outside the line box, you would have to push it in further in order for the bullet to have room to appear
17:31:45 [fantasai]: fantasai: then subsequent lines would have to be pushed that extra amount, too
17:32:04 [fantasai]: dbaron: There are some proposals in css3-box that deal with this problem, about how much to push things in
17:32:17 [fantasai]: dbaron: they need a little more work, but they mostly solve the problem
17:32:30 [fantasai]: bert: They do solve the problem. They're maybe a little hard to understand.
17:33:00 [fantasai]: ?: but we're discussing css2.1
17:33:19 [dbaron]: s/?:/SteveZ:/
17:33:29 [fantasai]: dbaron: Well, if you define that they're positioned relative to the line box now, then you can define how much they're pushed in later
17:34:05 [glazou]: CSS Charter was just approved
17:34:29 [sylvaing]: time is 9:34. Should we summarize the requirements for a write-up ?
17:34:30 [fantasai]: fantasai: I don't mind what we define here, as long as it doesn't conflict with our previous resolutions or what designers expect for text-align
17:35:16 [fantasai]: saloni: I don't think we asked designers about float behavior
17:35:39 [fantasai]: fantasai: No, we didn't. But if you ask them about that, I think you'll get responses asking for float-displace with reasonable default behavior for lists
17:37:06 [anne]: I have to go, sorry
17:37:13 [anne]: (not that I could hear very well anyway)
17:37:19 [Zakim]: -anne
17:38:52 [fantasai]: ?: but that's not css2.1
17:39:13 [fantasai]: saloni: We want IE8 to be able to follow the spec and be consistent with other browsers
17:39:38 [fantasai]: saloni: Other browsers have slight differences in how much offset, but they're consistent about interpretation of how the float affects the marker's placement
17:40:27 [fantasai]: peter: and the marker moves with text align?
17:40:36 [fantasai]: saloni: yes
17:40:50 [fantasai]: dbaron: I'm seeing that the marker moves with the text in FF and Opera, but not in Webkit
17:41:20 [fantasai]: dbaron: or at least an older webkit
17:41:36 [fantasai]: dbaron: We got feedback from a small number of web designers
17:42:23 [fantasai]: dbaron: And when Mozilla behaved the other way we got bugs filed
17:42:55 [dbaron]: On http://dbaron.org/css/test/2008/text-align-list-item , Mozilla, Opera, and WinIE7 put the bullets next to the text, and WebKit puts the bullets along the left
17:43:09 [fantasai]: peter: I'm not sure if the use cases that Steve is citing are outside markers or inside markers. For inside markers we want the bullet to stay with the text, but for outside we want them to stay outside.
17:43:23 [fantasai]: stevez: IIRC table of contents with centered items keep the number with the text
17:44:12 [fantasai]: stevez: It seems to me that the default value of list-item-position could be a different value that matches what the browser does today
17:44:18 [Zakim]: + +1.408.454.aabb
17:44:23 [fantasai]: fantasai doesn't think that's a good idea
17:45:01 [dino]: dino has joined #css
17:45:26 [fantasai]: The only thing that would be different between outside and this new value would be the marker's position
17:45:47 [fantasai]: You don't want to affect the default text-wrapping behavior
17:46:50 [fantasai]: (how the text wraps for the second line wrt bullet position)
17:47:14 [fantasai]: SteveZ: I think what Saloni is asking for is to codify existing bheavior
17:47:52 [sylvaing]: ie8 has to choose between interop or following the spec. we'd love to do both :)
17:48:05 [fantasai]: Saloni: I don't think our proposal changes the meaning of outside or inside
17:50:03 [fantasai]: SteveZ: Does anyone expect the marker to be on top of the float?
17:50:27 [fantasai]: fantasai: No, but because of the way the css model works we can't fix that without css3
17:50:53 [fantasai]: SteveZ: So it sounds like the best thing we can do is leave it undefined, because that's not desireable behavior
17:51:20 [fantasai]: Saloni: that's not the behavior in the spec
17:52:02 [fantasai]: Saloni: The spec implies that the marker doesn't move to make room for the float
17:52:45 [fantasai]: SteveZ: So sounds like we should make it undefined
17:53:06 [fantasai]: Peter: Do we have consensus to make behavior for floating clearly undefined for css2.1?
17:53:20 [fantasai]: Peter: have an opportunity to fix in css3
17:54:02 [fantasai]: fantasai: If we leave the behavior undefined here
17:54:20 [fantasai]: fantasai: In CSS3 we can have a float-displace value that computes differently depending on whether the box is display: list-item or not
17:55:28 [fantasai]: dbaron: float-displace requires the author to set something on an ancestor of the list-item.
17:55:33 [fantasai]: dbaron: It's hard to do that automaticallky
17:55:40 [dbaron]: (something == indent-edge-reset, I think)
17:56:22 [fantasai]: SteveZ: If we choose the undefined route, we have a chance of defining a better default in css3
17:56:40 [fantasai]: dbaron: So, that would fix the problem because we need the position of the /text/ undefined, not just the position of the marker
17:57:07 [fantasai]: SteveZ: So maybe we leave that undefined. Allow the marker to cause some suitable displacement of the text to avoid overlaps
17:57:18 [fantasai]: dbaron: Maybe. Depends on the actual text.
17:57:30 [fantasai]: ACTION: SteveZ write a proposal that allows for reasonable behavior
17:57:30 [trackbot]: Created ACTION-120 - Write a proposal that allows for reasonable behavior [on Steve Zilles - due 2008-12-10].
17:58:39 [fantasai]: Steve: I agree with the point that without going to css3 we can't really solve the problem
17:58:59 [fantasai]: Steve: Leaving it undefined allows existing implementations to be compliant, and it allows us to make a reasonable solution in css3
17:59:31 [fantasai]: Peter: we're out of tim
17:59:32 [fantasai]: e
17:59:51 [fantasai]: Peter: Make sure you rejoin the group.
17:59:52 [Zakim]: -dino
17:59:54 [Zakim]: -[Mozilla]
17:59:54 [Zakim]: -saloni
17:59:55 [Zakim]: -SteveZ
17:59:56 [Zakim]: -glazou
18:00:05 [Zakim]: -plinss_
18:00:08 [Zakim]: -fantasai
18:00:15 [Zakim]: -[Microsoft]
18:00:17 [Zakim]: -Bert
18:00:31 [melinda]: melinda has joined #CSS
18:00:38 [Zakim]: -dsinger.a
18:00:40 [Zakim]: Style_CSS FP()12:00PM has ended
18:00:44 [Zakim]: Attendees were dsinger, glazou, +1.858.354.aaaa, plinss_, Bert, sylvaing, saloni, David_Baron, anne, SteveZ, fantasai, alexmog, +1.408.454.aabb, dino
19:39:22 [Lachy]: Lachy has joined #css
19:45:57 [Zakim]: Zakim has left #css