IRC log of css on 2008-03-26
Timestamps are in UTC.
- 16:13:49 [RRSAgent]
- RRSAgent has joined #css
- 16:13:49 [RRSAgent]
- logging to http://www.w3.org/2008/03/26-css-irc
- 16:13:53 [dbaron]
- RRSAgent, make logs member
- 16:14:15 [dbaron]
- Meeting: CSS Working Group face-to-face, San Diego, CA
- 16:14:24 [dbaron]
- Chair: Daniel Glazman, Peter Linss
- 16:14:25 [Bert]
- Added Web site redesign as a topic to the list of topics that don't have a time slot yet. (Forgot to do that earlier.)
- 16:14:28 [dbaron]
- Topic: Introductions
- 16:15:03 [dbaron]
- Scribe: Elika Etemad
- 16:15:07 [dbaron]
- Scribe: David Baron
- 16:15:10 [dbaron]
- ScribeNick: dbaron
- 16:15:31 [fantasai]
- ScribeNick: fantasai
- 16:15:33 [fantasai]
- introductions
- 16:17:40 [dbaron]
- Present: Daniel Glazman, Tantek Çelik, Chris Lilley, Jason Cranford Teague, Steve Zilles, Bert Bos, Anne van Kesteren, Alex Mogilevsky, Aaron Eicholtz, Molly Holzschlag, Ming, Elika Etemad, Peter Linss
- 16:18:29 [fantasai]
- Daniel: Main topic for the day is the Charter.
- 16:18:37 [fantasai]
- Daniel: Current charter expires July 1st
- 16:18:39 [dbaron]
- i/Daniel:/Topic: Charter
- 16:18:42 [plinss]
- plinss has joined #css
- 16:18:52 [alexmog]
- alexmog has joined #css
- 16:18:59 [fantasai]
- Daniel: This document needs to be ready by the end of April
- 16:19:06 [fantasai]
- Daniel: we have a lot to discuss
- 16:19:22 [fantasai]
- Daniel: given recent discussions
- 16:19:29 [chris]
- chris has joined #css
- 16:19:42 [fantasai]
- Daniel: - openness in wg, Apples proposals for animations etc.,
- 16:19:49 [fantasai]
- Daniel: how we work.. lots of topics on our plate
- 16:20:00 [chris]
- rrsagent, here
- 16:20:00 [RRSAgent]
- See http://www.w3.org/2008/03/26-css-irc#T16-20-00
- 16:20:10 [fantasai]
- Daniel: We should start with the easiest thing to solve: communication channels for wg
- 16:20:23 [fantasai]
- Daniel: For time being we have private mailing list, which serves for w3c-process-related issues only
- 16:20:35 [fantasai]
- Daniel: public mailing list that is used only for technical discussions, open to everyone to contribute
- 16:20:38 [fantasai]
- Daniel: we have IRC channel
- 16:20:45 [fantasai]
- Daniel: And we have issue tracking tools
- 16:21:07 [fantasai]
- Daniel: One proposal to get more openness in the wg is to make the wg list publicly readable but not publicly writeable
- 16:21:20 [fantasai]
- Daniel: so everyone can see what we're doing, but we don't have a lot of noise
- 16:21:57 [fantasai]
- Alex: if it's publicly readable, people can discuss it on the public writeable list
- 16:22:12 [tantek]
- tantek has joined #css
- 16:22:24 [tantek]
- greetings
- 16:22:33 [tantek]
- I like IRC+wiki
- 16:22:44 [Arron]
- Arron has joined #CSS
- 16:23:16 [fantasai]
- Elika: I don't understand what exactly you're proposing
- 16:23:46 [fantasai]
- Elika: Are you saying to move the technical discussion onto this new publicly-archived private list?
- 16:24:23 [fantasai]
- That seems to be the idea
- 16:24:33 [fantasai]
- Tantek: I propose the IRC channel be publicly-logged
- 16:24:50 [fantasai]
- Daniel: I object to that because it is another communication channel..
- 16:24:59 [fantasai]
- Daniel: costs too much bandwidth
- 16:25:30 [fantasai]
- Tantek: I'm not saying everyone should read it, just that it should be logged
- 16:25:33 [fantasai]
- Daniel: That's fine
- 16:26:26 [fantasai]
- Chris shows rrsagent's permalink feature
- 16:27:12 [SteveZ2]
- SteveZ2 has joined #css
- 16:27:50 [jason_cranfordtea]
- jason_cranfordtea has joined #css
- 16:27:54 [fantasai]
- Bert: We don't have a public IRC channel.. #css has always been a group-only channel
- 16:28:07 [Ming]
- Ming has joined #css
- 16:28:56 [fantasai]
- Daniel: So... we can take that step if/when we make the group list public
- 16:29:51 [fantasai]
- Bert: I want a private group channel
- 16:30:05 [jason_cranfordtea]
- jason_cranfordtea has joined #css
- 16:30:28 [fantasai]
- Steve: Tantek, if the channel is logged I'm not convinced I don't have to read it
- 16:30:53 [fantasai]
- Tantek: This makes it easier to not read it, because someone can always point you to the logs
- 16:30:59 [chris]
- Makes it easier to point to prior discussion
- 16:31:43 [fantasai]
- Steve: I can live with that *if* people post a pointer to important discussions on the mailing list
- 16:33:00 [fantasai]
- Bert: If I have a discussion with someone on IRC, then I will post the snippet from my log to the mailing list. That's fine.
- 16:33:15 [fantasai]
- Bert: It's different from having the whole world read everything I write -- it's edited
- 16:33:35 [fantasai]
- Bert: There are several ...
- 16:33:47 [fantasai]
- Bert: When I write an email I take my time to tailor it to my audience.
- 16:33:58 [fantasai]
- Bert: When I type in IRC, I know who is there, so I can speak appropriately
- 16:34:17 [fantasai]
- Bert: If the whole world can see, then I have to explain everything I say
- 16:34:56 [fantasai]
- Bert: I don't mind having a world-readable channel, but I want a channel to talk only with you
- 16:35:27 [fantasai]
- Molly: I think this is becoming a discussion of philosophy.
- 16:35:37 [fantasai]
- Molly: There ar going to be different levels of privacy and how work gets done
- 16:35:52 [fantasai]
- Molly: In the HTMLWG there's a lot going on and it creates a pandemonium
- 16:36:10 [fantasai]
- Molly: An all or nothing solution isn't going to get us anywhere today.
- 16:36:16 [fantasai]
- Molly: We need a balance.
- 16:36:35 [fantasai]
- Bert: I'm looking for efficiency. I spend all my time communicating, not doing anything.
- 16:36:49 [fantasai]
- Bert: If I want to say something, I write an email. IRC is chatting in the hallway
- 16:37:20 [fantasai]
- Peter: I don't think a log of the IRC is meant to make proclamations to the world or to say "this is the way its going to be"
- 16:37:27 [fantasai]
- Peter: The logs are for us.
- 16:38:08 [fantasai]
- Peter: They're public to give transparency to the group, so the public can see what we're working on, that we're working.
- 16:38:23 [fantasai]
- Peter: We have other channels for official communication.
- 16:38:43 [fantasai]
- Peter: These logs need a big disclaimer "These are not official statements even by the people who made the statements", etc
- 16:39:21 [fantasai]
- Molly: Who's going to be using those logs as news?
- 16:39:53 [fantasai]
- Jason: My concern is that a lot of times you say things simply to hear your own arguments, not to take a stand for that position.
- 16:40:18 [fantasai]
- Jason: People hold you accountable for what you say, even if you're just thinking out loud
- 16:40:23 [fantasai]
- Jason: Even if you say so
- 16:40:52 [fantasai]
- Jason: You don't expect to be held accountable for everything you say at the water-cooler
- 16:41:32 [fantasai]
- ..
- 16:41:47 [fantasai]
- Daniel: I hope that everyone agrees that having the logs available for members is useful
- 16:42:59 [fantasai]
- Daniel: If our technical discussion is public, then we will be pointing to these logs. We can't make them Member-only if the public wants to see.
- 16:43:06 [fantasai]
- Elika: You can copy and paste the relevant parts.
- 16:43:20 [fantasai]
- Chris: Minuts
- 16:43:31 [fantasai]
- Chris: I assume minutes will be public once the grouup goes public.
- 16:44:00 [fantasai]
- Chris: You can make the minutes private first, and then have everyone check the minutes, then publish the minutes later.
- 16:44:21 [fantasai]
- Chris: Then next step is to take the minutes private, and unless someone objects the minutes go out 24 hours later.
- 16:44:46 [fantasai]
- chris: the third option is to make the minutes public as soon as they're written
- 16:45:13 [fantasai]
- Chris: If we're not using that third option for minutes, then we can't have public IRC logs
- 16:45:17 [fantasai]
- Chris: I recommend the second option
- 16:46:16 [fantasai]
- Bert: I find the edited minutes, the ones that have headings and spliits things into paragraphs, those are much easier to read than rrsagent minutes
- 16:46:35 [fantasai]
- Bert: The minute taker should spend an hour after the meeting to clean up the minutes
- 16:47:41 [fantasai]
- Daniel: That's a point I made when we started using IRC..
- 16:48:01 [fantasai]
- Daniel: We have a lot of work to do, though, but I think the WG needs to focus on our work not the minutes
- 16:48:26 [dbaron]
- I'd note that minutes generated by scribe.perl are a lot more readable than raw IRC logs, which some people have posted as minutes occasionally.
- 16:48:58 [fantasai]
- Daniel, Chris: If minutes are taken on IRC then people can correct, add missing statements, etc.
- 16:49:00 [anne]
- I agree that IRC should be publicly logged. That's how it works for the WHATWG, HTML WG and XHTML 2 WG already.
- 16:49:56 [fantasai]
- Steve: The standard problem with W3C, the problem is getting IPR that's not covered under the royalty-free policy into the specs...
- 16:49:57 [glazou]
- anne, a ftf is for TALKS please :-)
- 16:50:26 [fantasai]
- Alex: If most technical discussion happens on public lists ...
- 16:50:35 [fantasai]
- Alex: If I don't want something public, I just don't say it
- 16:50:41 [glazou]
- anne, in other words, you're opinion is interesting so bring it using voice :-)
- 16:50:53 [fantasai]
- Alex: A part of motivation of publishing everything is to prevent things like web fonts discussion from happening again
- 16:51:13 [fantasai]
- Anne: I think it should be publicly logged.. it works that way for WHATWG, HTMLWG, and XHTML2 WG
- 16:51:15 [Zakim]
- Zakim has joined #css
- 16:51:20 [chris]
- I agree with Anne
- 16:51:21 [fantasai]
- Molly: How are people using the logs in those groups?
- 16:51:26 [dbaron]
- q+ to question read-only assumption
- 16:52:15 [fantasai]
- Molly: The problem Bert and others are concerned about is that the logs become a source of news and information
- 16:52:29 [chris]
- q+
- 16:52:31 [fantasai]
- Bert: People spend a lot of time trying to interpret the IRC logs to figure out who meant what
- 16:52:56 [fantasai]
- Anne: If people have questions about the logs, someone can ask.
- 16:53:12 [tantek]
- q+ to note that the microformats community has also been successfully using a publicly logged IRC channel.
- 16:53:19 [fantasai]
- Molly: That's great for the WG, but how does that work for people who aren't in the WG and post about things on their blog?
- 16:53:42 [fantasai]
- dbaron: I think a lot of people were making the assumption that there are things we want to be member-writeable and publicly readable.
- 16:53:50 [fantasai]
- dbaron: But I think that's discouraging for a lot of contributors.
- 16:54:08 [fantasai]
- dbaron: One of the reasons we want to make things more public, is that we want people to help us
- 16:54:16 [tantek]
- dbaron, that's not my assumption. I'd like more things to be publicly read/writeable.
- 16:54:16 [fantasai]
- dbaron: and to channel their energy into something that's useful.
- 16:54:25 [Zakim]
- dbaron, you wanted to question read-only assumption
- 16:54:31 [fantasai]
- dbaron: If all these things we do are not publicly-writeable, then they will be discouraged and go away.
- 16:54:43 [fantasai]
- Bert: I agree with that. People see lots of things going on that they can't influence.
- 16:54:56 [fantasai]
- Bert: I like the mailing list best, it's 2-way communication rather than blogs or wiki
- 16:55:01 [dbaron]
- want people to help us and to channel their energy
- 16:55:09 [fantasai]
- Daniel: Currently technical discussions are all writeable, being on www-style.
- 16:55:20 [fantasai]
- Tantek: CSS IRC channel is also writeable.
- 16:55:27 [fantasai]
- Bert: Then I won't join anymore.
- 16:55:35 [fantasai]
- Daniel: What else do you want to be writeable?
- 16:56:16 [chris]
- ack chris
- 16:56:16 [fantasai]
- Daniel: property discussions? patent policy discussions?
- 16:56:26 [fantasai]
- dbaron: public comment on patent policy was very important
- 16:56:33 [glazou]
- SteveZ2: join the queue ?
- 16:56:34 [SteveZ2]
- Q+
- 16:56:43 [anne]
- Zakim, q+ to say that the Forms WG, HTML WG, and XHTML2 WG allow the public to join their group after agreeing to the PP
- 16:56:45 [Zakim]
- I see tantek, SteveZ, anne on the speaker queue
- 16:57:10 [fantasai]
- Chris: When Molly was talking about interpreting IRC logs, maybe it's a problem that there's no summary of the discussions?
- 16:57:40 [fantasai]
- Chris: If a summary of what happened in the F2F is available, how many people would want to dig though the IRC logs?
- 16:57:59 [fantasai]
- Molly: I agree that everything should be open, I'm just concerned about the practicality of it
- 16:58:10 [chris]
- ack tan
- 16:58:10 [Zakim]
- tantek, you wanted to note that the microformats community has also been successfully using a publicly logged IRC channel.
- 16:58:24 [fantasai]
- Tantek: In microformats, we're using the IRC logs a lot
- 16:58:32 [fantasai]
- Tantek: It's more accessible to more people.
- 16:58:45 [fantasai]
- Tantek: I found that it's very hard to keep up with the WHATWG mailing list etc.
- 16:58:54 [fantasai]
- Tantek: It's much harder to out-talk people on IRC
- 16:59:10 [fantasai]
- Tantek: You're right that if there's a distilled version, that's what people will read.
- 16:59:38 [fantasai]
- Tantek: As soon as we set up IRC logging on the microformats community, the effects were 99 percent positive.
- 16:59:54 [fantasai]
- Tantek: It shows the process of debate
- 16:59:58 [fantasai]
- Tantek: which is really healthy
- 17:00:25 [glazou]
- ack SteveZ2
- 17:00:35 [fantasai]
- Steve: I see 2 conflicting positions, and I'm on both of them.
- 17:00:36 [chris]
- ack ste
- 17:00:55 [fantasai]
- Steve: First one is, there's a goal for transparency and public participation. I think that's a good goal to have, so I can't be against it.
- 17:01:08 [fantasai]
- Steve: On the other hand, there's the level of work to participate, and that's the one that strongly concerns me.
- 17:01:27 [chris]
- rrsagent, here
- 17:01:27 [RRSAgent]
- See http://www.w3.org/2008/03/26-css-irc#T17-01-27
- 17:01:36 [fantasai]
- Steve: I only work part=time, and I'm hoping things progess on a weekly rather than hourly
- 17:02:14 [fantasai]
- Steve: I just feel that there's somewhat difference in certain aspects of CSS, and i'll pick on margin collapsing as one, where you need more knowledge than I have to make useful contributions to the discussion
- 17:02:26 [fantasai]
- Steve: in other areas, like the name of a property, where public input is very useful.
- 17:02:41 [fantasai]
- Steve: My concern is that this isn't a level playing field in which one particular solution works really where everywhere.
- 17:02:56 [fantasai]
- Steve: For margin collapsing, I think it's important that someone write a complete coherent proposal.
- 17:03:10 [fantasai]
- Steve: My understanding is that these were discussed at meetings, and then written up as a coherent proposal later.
- 17:03:20 [fantasai]
- Steve: I dont' think this discussion is taking into account all the different things we have to do.
- 17:03:31 [fantasai]
- Steve: And my fear is that I'll have to follow more work and more discussions
- 17:04:02 [fantasai]
- Steve: I'm heartened by Tantek's positive experience, but I'm not convicned that those discussions have the same complexity as issues we have in CSS
- 17:04:12 [molly]
- Q+
- 17:04:25 [tantek]
- Steve, for complex proposals and discussions, we have found that using a wiki works best currently.
- 17:04:48 [fantasai]
- Daniel ask for straw poll
- 17:06:11 [fantasai]
- Molly: if the logs are public, these *need* to have that disclaimer Peter was talking about
- 17:07:42 [fantasai]
- Steve: The consequence of HTML's method of working is that a lot of people have given up following the HTMLWG
- 17:09:12 [fantasai]
- Daniel: Most of the people able to solve the very complex issues we have are inside this room
- 17:09:21 [fantasai]
- dbaron: Maybe two of them are in this room, but one of them is out there.
- 17:09:30 [fantasai]
- dbaron: We want our discussions public so that we can find out about that person.
- 17:09:47 [fantasai]
- Tantek: I'd say there are more people with expertise outside this WG than inside it
- 17:10:14 [fantasai]
- Elika: I think there are not more people with expertise, but more people with potential
- 17:10:42 [fantasai]
- Daniel: QuestionA: log IRC channel?
- 17:10:52 [fantasai]
- Daniel: QuestionB: make logs public?
- 17:11:03 [fantasai]
- Daniel: Yes, yes
- 17:11:06 [fantasai]
- Tantek: yes yes
- 17:11:09 [fantasai]
- dbaron: yes yes
- 17:11:13 [fantasai]
- chris: yes yes
- 17:11:17 [fantasai]
- jason: yes yes
- 17:11:21 [fantasai]
- Steve: yes abstain
- 17:11:30 [fantasai]
- Bert: no opinion, no
- 17:11:34 [fantasai]
- Anne: yes yes
- 17:11:37 [fantasai]
- Alex: yes yes
- 17:11:40 [fantasai]
- Arron: yes yes
- 17:11:47 [dsinger_]
- dsinger_ has joined #css
- 17:11:47 [fantasai]
- Elika: yes abstain
- 17:11:51 [fantasai]
- Peter: yes yes
- 17:12:20 [dbaron]
- Molly: yes yes
- 17:12:34 [fantasai]
- Daniel: Question is openness of member-only mailing list..
- 17:12:37 [fantasai]
- Steve: take a straw poll
- 17:12:59 [fantasai]
- Steve: Question is, should we create a separate mailing list that is publicly-archived, but member-only, in which the work of the work of the wg can be carried out
- 17:13:14 [fantasai]
- Anne: I want to make this a multipe choice question
- 17:13:53 [fantasai]
- Anne: I want to make the wg anyone can agree to the policy and join
- 17:13:55 [fantasai]
- the discussions
- 17:14:05 [fantasai]
- Anne: that's how Forms and .. work
- 17:14:08 [fantasai]
- Chris: That's not true
- 17:14:12 [fantasai]
- Chris: That's how it works for HTML
- 17:14:17 [fantasai]
- Chris: but not the others
- 17:15:39 [chris]
- For the others, everyone has agreed to the patent policy but its w3c members and invited experts (in the real sense ir invited and expert) who can post to the public WG list
- 17:15:48 [fantasai]
- Daniel clarifies proposal: Leave technical discussion on www-style
- 17:16:27 [fantasai]
- Make new closed, publicly-archived mailing list for other discussion
- 17:16:39 [fantasai]
- Peter: we should keep the private list for things like "here's my cell phone number, meet for dinner"
- 17:16:50 [fantasai]
- Chris: Make a member-only web page, there's too much over head for that.
- 17:17:02 [chris]
- Suggest you use a Member-only web page for that sort of stuff 9cell phone numbers etc)
- 17:17:06 [fantasai]
- Molly: So the new list is readable to all, writeable only to WG members
- 17:17:12 [fantasai]
- Anne: restricted to W3C business
- 17:17:13 [chris]
- s/9cell/(cell/
- 17:17:25 [fantasai]
- Molly asks for examples
- 17:17:38 [hyatt]
- hyatt has joined #css
- 17:17:56 [fantasai]
- Daniel: Organizing meetings, agendas, charter, etc.
- 17:18:04 [fantasai]
- RRSAgent: pointer
- 17:18:04 [RRSAgent]
- See http://www.w3.org/2008/03/26-css-irc#T17-18-04
- 17:18:09 [fantasai]
- hyatt, ^
- 17:18:26 [fantasai]
- Daniel: With the public mailing list I have to filter all the time.
- 17:18:45 [fantasai]
- Daniel: There are a lot of people who only contribut noise
- 17:18:58 [fantasai]
- Daniel: For the daily life of the WG I don't want any noise
- 17:19:08 [fantasai]
- Alex: I think we have a good separation right now between technical, and process
- 17:19:18 [fantasai]
- Alex: I'm in favor of not changing anything
- 17:20:07 [fantasai]
- Daniel: Question A: Should we have a new publicly-readable closed mailing list that obsoletes the current mailing list
- 17:20:24 [fantasai]
- Daniel: Question B: Should this mailing list be publicly writeable?
- 17:20:32 [dbaron]
- Daniel: The topic of this list is non-technical (process), just like w3c-css-wg today.
- 17:20:40 [fantasai]
- s/current mailing list/w3c-css-wg mailing list/
- 17:20:51 [fantasai]
- Daniel: no no
- 17:21:03 [fantasai]
- Tantek: yes abstain
- 17:21:12 [fantasai]
- dbaron: yes yes
- 17:21:16 [fantasai]
- Chris: yes no
- 17:21:22 [fantasai]
- Jason: no no
- 17:21:25 [fantasai]
- Steve: no no
- 17:21:29 [fantasai]
- Bert: no no
- 17:21:36 [fantasai]
- Anne: yes yes
- 17:21:39 [fantasai]
- Alex: no no
- 17:21:43 [fantasai]
- Arron: no no
- 17:21:47 [fantasai]
- Molly: yes no
- 17:21:56 [fantasai]
- Elika: abstain no
- 17:22:01 [fantasai]
- Peter: abstain no
- 17:22:26 [fantasai]
- Seem to have consensus on second question
- 17:22:28 [fantasai]
- not on first
- 17:23:08 [fantasai]
- RESOLVED: IRC channel will be logged constantly
- 17:23:15 [fantasai]
- RESOLVED: IRC logs will have a disclaimer
- 17:23:22 [fantasai]
- RESOLVED: IRC logs will be public
- 17:23:33 [fantasai]
- Topic: Telecon time
- 17:23:57 [fantasai]
- Daniel: I would like to change the weekly telecon, I have a personal constraint on Tuesdays -- I have my children with me
- 17:24:04 [fantasai]
- Daniel: I would like to change the date of the telecon
- 17:24:31 [fantasai]
- Daniel: The time is fine
- 17:24:43 [fantasai]
- Peter: Proposed to move to Wednesday, same time
- 17:24:53 [fantasai]
- dbaron: Wednesdays are sort of worse for me..
- 17:25:05 [fantasai]
- dbaron: but that may change every 6 months anyway
- 17:25:13 [fantasai]
- Monday mornings are terrible for several people
- 17:25:48 [fantasai]
- noon in Boston
- 17:26:16 [fantasai]
- RESOLVED: Telecons move to Wednesday at noon in Boston
- 17:26:24 [fantasai]
- BREAK
- 17:45:06 [hyatt]
- hyatt has joined #css
- 17:47:47 [glazou]
- resume
- 17:47:49 [glazou]
- hello hyatt
- 17:47:56 [hyatt]
- hi
- 17:47:58 [glazou]
- hyatt, we're going to discuss goals of CSS
- 17:48:06 [glazou]
- goals of next charter sorry
- 17:48:19 [fantasai]
- Topic: Goals and Deliverables
- 17:48:31 [molly]
- molly has joined #css
- 17:48:37 [fantasai]
- Peter: We'd like to get CSS2.1 done as quickly as possible.
- 17:48:48 [fantasai]
- Peter: we have 10 years not publishing a REC
- 17:48:52 [fantasai]
- Peter: We have IPR issues..
- 17:49:06 [fantasai]
- Peter: Need to move things forward to REC, and if we have issues deal with them in errata.
- 17:49:20 [fantasai]
- Anne: What are the IPR issues?
- 17:49:41 [fantasai]
- Chris: When Members join a wg they make commitments wrt their patents
- 17:49:57 [fantasai]
- Chris: When something becomes a REC, the royalty-free license starts to apply
- 17:50:27 [fantasai]
- Chris: Another risk is if you have something in CR, there's an assumption that it's stable.
- 17:50:45 [fantasai]
- Chris: There are other implementors (not in this room) that need a stable document
- 17:51:24 [fantasai]
- Elika: At the same time, I don't want us to push things through just to push them through. We've done it already, and it's not had good results.
- 17:52:18 [chris]
- http://www.w3.org/Consortium/Patent-Policy-20040205/Overview.html
- 17:52:33 [fantasai]
- Daniel: One question about pushing CSS2.1 to REC. Where do we stop?
- 17:52:53 [fantasai]
- Daniel: We will always find deep technical issues. At some point we have to stop, publish the document, and use the errata system.
- 17:53:40 [fantasai]
- Elika: We are using errata for CSS2.1
- 17:54:10 [fantasai]
- Anne: The main problem is test coverage.
- 17:54:23 [fantasai]
- Anne: Once you have test coverage, you can see if the implementations match the spec and if not there's a problem.
- 17:54:58 [chris]
- q+ to say There needs to be a test coverage report, and an implementation report on those parts that have tests
- 17:55:01 [fantasai]
- Anne: THe main problem is that we keep bringing up new issues and discussing the issues, instead of making the test suite first and then trying to resolve those.
- 17:55:03 [chris]
- ack anne
- 17:55:03 [Zakim]
- anne, you wanted to say that the Forms WG, HTML WG, and XHTML2 WG allow the public to join their group after agreeing to the PP
- 17:55:45 [mjs]
- mjs has joined #css
- 17:57:33 [fantasai]
- ...
- 17:57:48 [fantasai]
- dbaron: We've already gone back to Last Call to drop features. I think we're pretty good there.
- 17:58:00 [fantasai]
- dbaron: we can still improve the test suite once we're at REC, and that will continue to improve interoperabiliyt
- 17:58:04 [glazou]
- ack molly
- 17:58:08 [fantasai]
- Molly: What is the status of test suite discussion?
- 17:58:18 [fantasai]
- Molly: Are we discussing that today? Are we discussing it later?
- 17:59:12 [fantasai]
- ...
- 17:59:28 [fantasai]
- Molly: We have a test suite that needs to be completed or near completed and available
- 17:59:36 [fantasai]
- Molly: And we need two implementations that pass the test suite.
- 17:59:47 [chris]
- ack molly
- 17:59:54 [chris]
- ack chris
- 17:59:54 [Zakim]
- chris, you wanted to say There needs to be a test coverage report, and an implementation report on those parts that have tests
- 18:00:12 [fantasai]
- Chris: There needs to be a test coverage report, that's a new requirement and mostly ignored by wg.
- 18:00:21 [fantasai]
- Elika: we have that..
- 18:00:31 [fantasai]
- Chris: And an implementation report that say which implementations pass or fail.
- 18:00:40 [fantasai]
- Chris: There are two notions of completeness.
- 18:00:53 [SteveZ]
- SteveZ has joined #css
- 18:01:02 [fantasai]
- Chris: One is implementability.. can two implementations be interoperable by relying only on the spec.
- 18:01:14 [fantasai]
- Chris: Another type of test shows 'can web designers use these specification'
- 18:01:20 [fantasai]
- s/technology/
- 18:02:04 [fantasai]
- In the first case you only need two implemenations for each test.. doesn't need two implementations that completely pass
- 18:02:07 [fantasai]
- http://www.w3.org/Style/CSS/Test/CSS2.1/current/xhtml1/by-section.xht
- 18:02:48 [fantasai]
- Chris: I would like to say that for each spec there is ongoing test suite work.
- 18:02:50 [SteveZ]
- Q+
- 18:02:53 [fantasai]
- Chris: in the charter
- 18:03:02 [chris]
- ack ste
- 18:03:08 [fantasai]
- Steve: First comment: I think Chris said it, but it's important
- 18:03:39 [fantasai]
- Steve: Every feature needs to have two implementations. Doesn't have to be the same implementation. Collectively you need 2 implementations.
- 18:03:53 [fantasai]
- Steve: I thought I detected that the test suite is the next most important thing to do.
- 18:04:08 [fantasai]
- Steve: Processing errata stuff is done mostly become the testcases weren't appearing..
- 18:04:17 [glazou]
- q+
- 18:04:19 [fantasai]
- Steve: That we're processing errata because we don't have test cases to process
- 18:04:37 [fantasai]
- Steve: We should shift our work to reviewing those test cases.
- 18:05:19 [fantasai]
- Daniel: We can move the problem from stopping the spec at some point to stopping the test suite at some point
- 18:05:39 [fantasai]
- Steve: You need a point to say "I've got an acceptable level of coverage".
- 18:05:59 [fantasai]
- dbaron: I would disagree.
- 18:06:05 [fantasai]
- dbaron: I think we should keep adding tests.
- 18:06:06 [glazou]
- ack glazou
- 18:06:25 [fantasai]
- dbaron: I think we can get to a point that we say "This test suite is good enough to base our decision to enter PR on"
- 18:07:57 [dbaron]
- we should decide the test suite is good enough for entering PR, use that version of the test suite to enter PR, and *continue* work on the test suite and errata
- 18:08:43 [fantasai]
- RESOLVED: dbaron's statement accepted
- 18:09:15 [fantasai]
- http://www.w3.org/Style/CSS/Test/CSS2.1/current/xhtml1/by-section.xht
- 18:11:33 [glazou]
- q+
- 18:11:47 [fantasai]
- Molly wants concrete goals
- 18:12:52 [glazou]
- ack glazou
- 18:13:57 [molly]
- scribe molly
- 18:14:22 [alexmog]
- scribenick: molly
- 18:15:05 [molly]
- Peter: Sounds like Anne is talking about changing policy within the charter
- 18:16:06 [molly]
- Anne: What we're currently doing will probably take another decade or so. Every time you add new tests, they have to be fixed
- 18:16:15 [molly]
- Anne: There's a lot of precedence for that
- 18:16:25 [fantasai]
- Anne: Adding tests for a section uncovers errors in that section
- 18:16:47 [fantasai]
- Alex: I think the spec is pretty good, and very usable for creating implementations.
- 18:16:54 [molly]
- Alex: We seem to agree usually that the spec is in pretty good shape, and it is usable to create implementations. We're talking about how to define rec in order to create rec
- 18:16:57 [fantasai]
- Alex: We're talking about how to make the sepc a recommendation
- 18:17:21 [fantasai]
- Alex: When we get the quality of test suite same as quality of spec, then we can talk about making that a recommendation
- 18:17:42 [fantasai]
- Alex: In my opinion, the spec is very high quality as far as a spec of this complexity goes.
- 18:17:52 [fantasai]
- scribenick: fantasai
- 18:18:01 [dbaron]
- dbaron has joined #css
- 18:18:43 [fantasai]
- Molly: I suggest we put some solid criteria down about when we'd feel comfortable moving forward with the test suite
- 18:18:50 [fantasai]
- Molly: Define the baseline we want for moving to PR
- 18:19:37 [mjs]
- mjs has joined #css
- 18:20:02 [fantasai]
- Elika: I think once we have all of Microsoft's tests in, and hixie's tests in, then we'll have a much better idea of how we're doing and how much coverage we have and how far we have to go
- 18:20:09 [fantasai]
- Chris: How much work is that?
- 18:20:27 [fantasai]
- Anne: Hixie has more than a thousand tests, they haven't been updated, and they need review and reformatting
- 18:20:54 [fantasai]
- Arron: Microsoft's tests aren't completely added in
- 18:21:15 [fantasai]
- Dbaron: Some of these tests are blocked by licensing issues
- 18:21:32 [fantasai]
- Anne: Hixie won't contribute his tests until we have a reasonable licensing policy for the test suite
- 18:21:37 [fantasai]
- Steve: There's discussion going on
- 18:22:25 [fantasai]
- Steve: One person said "I don't understand what the problem is. You can use the test without modifying it."
- 18:23:01 [fantasai]
- Chris: The tests are usually released under W3C Document license. The harness is under the software license.
- 18:23:37 [fantasai]
- Chris: Third parties can't take the test suite and tweak it.
- 18:23:52 [fantasai]
- Chris: But the Working Group can release a derived version that is under a different license.
- 18:24:26 [fantasai]
- dbaron: Would it be acceptable to W3C to move CSS2.1 to PR under a test suite managed by a different organization?
- 18:24:50 [fantasai]
- ACTION Chris: If CSSWG can use test suites provided by an external organization to move from CR to PR?
- 18:25:00 [fantasai]
- Steve: I'd be totally opposed to that.
- 18:25:42 [tantek]
- dbaron, I would trust the group to make the right decision there.
- 18:25:50 [fantasai]
- Steve: If W3C can get a copy of the tests under W3C Document License and use that, that should be fine
- 18:26:15 [fantasai]
- Anne: W3C Software License has some problems.. we need either MIT or BSD
- 18:26:41 [fantasai]
- Alex: So if I hear Chris correctly, we can republish every test under whatever license that we want.
- 18:26:59 [fantasai]
- Chris: yes
- 18:27:16 [Bert]
- http://www.w3.org/Consortium/Legal/2002/copyright-software-20021231 is the W3C Software License. It's not viral.
- 18:27:31 [Bert]
- And doesn't require feedback.
- 18:28:03 [fantasai]
- Tantek: As an implementor, making changes to the tests is very useful and important.
- 18:28:11 [fantasai]
- Tantek: I don't see any problem with making changes. It's just claims of conformance.
- 18:28:46 [dbaron]
- Tantek made the point about using tests at intermediate states in the progress of an implementation.
- 18:28:49 [fantasai]
- Elika: That's why I said we should license under BSD and use a trademark policy that prevents people from claiming a derivative is a conformance test suite
- 18:29:06 [dbaron]
- in other words, working around some implementation bugs by changing the tests, so that the tests can be used to test the things that work.
- 18:29:18 [fantasai]
- Bert: We don't have legal resources to enforce trademark law
- 18:30:03 [fantasai]
- dbaron: Having worked on large open source projects, licenses that require notice of changes per file are problematic
- 18:30:24 [fantasai]
- Chris: No, it only requires to say where you derived it from and the fact that it's been changed.
- 18:30:35 [chris]
- http://www.w3.org/Consortium/Legal/2002/copyright-software-20021231
- 18:30:36 [fantasai]
- dbaron: It requires *what* changes have been made and when
- 18:30:54 [chris]
- yes, agreed the date is an issue
- 18:31:10 [fantasai]
- Chris: Ok, I now agree this is problem.
- 18:32:22 [molly]
- Fantasai: With the exception of existing tests, and Microsoft's tests, with very minor exceptions we will be able to license under BSD without W3C permission
- 18:33:35 [fantasai]
- Alex: From what Chris said, we don't have to be blocked by W3C lawyers. We can publish two identical sets of tests, one under W3C Document License that is the formal compliance test
- 18:34:16 [fantasai]
- Alex: and one identical set of tests under a similar license.
- 18:34:17 [chris]
- http://www.w3.org/2004/10/27-testcases.html
- 18:34:20 [fantasai]
- s/similar/different/
- 18:37:15 [fantasai]
- Elika: If what Chris says is true, then I would like to propose that we publish two test suites, one under W3C Document License labelled CSS Conformance Test Suite, and one under BSD 3-clause labelled something else.
- 18:37:34 [chris]
- ========= quote ==========
- 18:37:42 [chris]
- 2. Grant of License for Contributed Test Cases Published Outside a W3C Recommendation
- 18:37:42 [chris]
- The Contributor hereby grants to the W3C, a perpetual, non-exclusive, royalty-free, world-wide right and license under any Contributor copyrights in this contribution to copy, publish, use, modify, and to distribute the contribution under the W3C Document License, as well as a right and license of the same scope to any derivative works prepared by the W3C and based on, or incorporating all or part of the contribution. The Contributor further agrees that any
- 18:37:42 [chris]
- derivative works of this contribution prepared by the W3C shall be solely owned by the W3C.
- 18:37:42 [chris]
- The Contributor states, to the best of her/his knowledge, that she/he, or the company she/he represents, has all rights necessary to contribute the Materials.
- 18:37:45 [chris]
- W3C will retain attribution of authorship to the Contributor. Whenever modifications are made to the Materials, this fact, and the nature of the modifications, will be clearly signalled in the distributed version thereof. The W3C makes no a-priori commitment to support or distribute contributions.
- 18:37:51 [chris]
- ======== end quote ========
- 18:41:14 [molly]
- Under discussion: Understanding that we can place a baseline goal on our test suite that will provide good enough coverage to get CSS 2.1 to PR
- 18:41:17 [mjs_]
- mjs_ has joined #css
- 18:41:41 [molly]
- then, we can continue testing, but we move the spec forward
- 18:43:04 [fantasai]
- Topic: Scope of Charter
- 18:43:43 [fantasai]
- Daniel: Hyatt discussed their proposal with Bert and Elika and Bert felt that their proposals were outside the scope of the current charter
- 18:43:55 [fantasai]
- Daniel: (animations, transitions, filters)
- 18:43:56 [hyatt]
- specifically that animations were.
- 18:44:01 [hyatt]
- i don't think transforms were.
- 18:44:12 [fantasai]
- Anne, Elika: There was a discussion at a telecon
- 18:44:18 [hyatt]
- (we have never proposed anything related to filters, so i'd prefer we just not worry about filters for now)
- 18:44:40 [fantasai]
- Daniel: My personal opinion:
- 18:44:55 [fantasai]
- Daniel: Filters were first proposed in 1998 from microsoft. I think it's normal for them to be in the scope of this WG
- 18:45:18 [fantasai]
- Daniel: Transitions are not about how a property works, but .... I think it's in scope
- 18:45:37 [fantasai]
- Daniel: Animations would be outside the scope, but folding and unfolding elements in a tree would be in the scope of this charter.
- 18:45:44 [fantasai]
- Daniel: I think the three of them are in the scope of the charter.
- 18:45:50 [fantasai]
- Daniel: But this is for the current charter.
- 18:46:00 [fantasai]
- Daniel: I think all the problems/issues raised by the proposals from Apple
- 18:46:19 [fantasai]
- Daniel: We need strong collaboration with SVG and SMIL, which already have similar technologies
- 18:46:29 [fantasai]
- Daniel: Besides these things that we can solve, that we should solve the proposal
- 18:46:39 [fantasai]
- Daniel: I think it falls into the scope of the charter.
- 18:46:44 [fantasai]
- Steve: I'd like to ask a clarification question
- 18:46:52 [dbaron]
- Agree that it should be within charter.
- 18:47:13 [fantasai]
- Steve: The question we discussed before was whether it was in the previous charter. I think the discussion should now be whether we should put these in the scope of our next charter
- 18:47:28 [fantasai]
- Alex: I don't see anything wrong with bringing something that has to do with Rendering and making that a CSS3 Module
- 18:47:48 [fantasai]
- Alex: We resist this because we have a lot of work already, but I don't think that's enough grounds for rejecting it
- 18:48:05 [fantasai]
- Alex: If someone is coming with the offer of a specification, test suites, etc.
- 18:48:32 [fantasai]
- Daniel: Apple has a plan to submit extended animations to this CSSWG.
- 18:49:05 [fantasai]
- Chris: If you have a specification that specifies how properties evolve over time, then the group that defines properties has that in scope
- 18:49:06 [glazou]
- (said in an email IIRC)
- 18:49:11 [fantasai]
- Chris: but the group that does timing is also invovled.
- 18:49:42 [fantasai]
- Peter: I don't think we need to discuss a new animation technology, but application of existing technologies to CSS is something we should sicuss
- 18:50:04 [fantasai]
- Daniel: We already discussed e.g. the 'binding' property, a CSS property that allows you to do something different
- 18:50:22 [fantasai]
- Jason: You can call it behavioral, but I consider it adding a temporal element to your styles.
- 18:50:41 [fantasai]
- dbaron: I disagree that we should avoid creating new animation technology
- 18:51:16 [fantasai]
- dbaron: If someone brings a technology that is significantly simpler and solves the use cases, we should consider that.
- 18:51:56 [dbaron]
- dbaron: We shouldn't ignore them, but we also shouldn't assume that they were designed for the same types of implementations just because they're W3C RECs.
- 18:52:02 [fantasai]
- Peter: I don't disagree that we can't investigate a new technology, but we should not ignore existing technologies.
- 18:52:32 [fantasai]
- Steve: The key thing to add is, you need to investigate the interoperability of the new technology and the existing technology.
- 18:52:53 [fantasai]
- Steve: For example, those of us that are investing vertical text agree that the existing glyph-orientation property in SVG and XSL is not solving the problem well
- 18:53:07 [fantasai]
- Steve: We came up with text-orientation, but we took the time to define what would happen if both get used.
- 18:53:23 [dbaron]
- (note that my comment was a response to Peter's comment minuted afterwards)
- 18:53:39 [fantasai]
- Peter: I don't think anyone disagrees that we should be working on this
- 18:54:00 [fantasai]
- Daniel: I brought this up to discuss scope. Other things we should consider are..
- 18:54:16 [fantasai]
- Daniel: if anyone else has other ideas to contribute.
- 18:54:36 [fantasai]
- Steve: I'm concerned about the approach you're taking.
- 18:54:52 [fantasai]
- Steve: The potential space of scope of this WG then becomes the display of content on the screen.
- 18:55:19 [fantasai]
- Steve: I'm concerned about coming up with a work program that pushes things forward.
- 18:55:36 [fantasai]
- Daniel: The intention is to have a better wording of the goals section that determines what is in scope and what's not.
- 18:55:58 [fantasai]
- Steve: I'd like the charter to lay out explicitly what is in scope, because of getting patent clearance.
- 18:56:23 [fantasai]
- Steve: If you make it totally open-ended you'll get pushback on IPR issues.
- 18:56:39 [fantasai]
- Steve: Chris Wilson was very careful about what went into scope sections in previous charters.
- 18:56:55 [fantasai]
- Steve: Maybe there's a new category of work item which is not a deliverable per se, but a whitepaper or discussion
- 18:56:56 [chris]
- Initial AC review does need for the scope to be reasonably well defined.
- 18:57:10 [fantasai]
- Steve: So that we can carry on the discussions and document the results, but not create a specification yet
- 18:57:41 [fantasai]
- Steve, Chris: If you want then to take on the area as a deliverable, then you have to amend the charter
- 18:57:42 [mjs]
- mjs has joined #css
- 18:57:53 [fantasai]
- dbaron: how often is that done?
- 18:57:56 [fantasai]
- Bert, Chris: It happens
- 18:58:26 [fantasai]
- Bert: Advanced layout and grid were considered split off from existing modules
- 18:58:52 [fantasai]
- Chris: Over time the charters are becoming more tight about scope
- 18:59:46 [fantasai]
- Steve: For advanced layout and grid, nobody in the WG thought they were out of scope. If someone in the WG felt they were out of scope, we'd have to amend the charter.
- 19:00:25 [fantasai]
- Anne: So, there seems to be nothing else. Can we agree that Apple's proposals are in scope?
- 19:00:35 [fantasai]
- Steve: I'm concerned about resources.
- 19:00:48 [fantasai]
- Anne: Since Apple is offering resources, I think it's fine.
- 19:01:37 [fantasai]
- Anne: Given that there was previously some confusion about whether these are in scope or not, we should make it explicit
- 19:01:50 [mjs_]
- mjs_ has joined #css
- 19:03:32 [fantasai]
- Steve feels these are not in scope in the current charter.
- 19:04:15 [fantasai]
- Chris: Also, since there's another group that has these things in scope there will be a lot of argument over it.
- 19:04:24 [fantasai]
- Chris: I think we should just put it in scope for the next charter.
- 19:05:22 [anne]
- anne has joined #css
- 19:05:32 [fantasai]
- Molly: we don't have a good working structure with SVG
- 19:05:54 [molly]
- Clarification: We don't have a good working structure
- 19:06:12 [fantasai]
- Steve: Putting it in the charter is the official notice to other WGs that we're going to work on it and consider it in scope.
- 19:06:23 [molly]
- Concerned that will impact related WG such as SVG
- 19:06:38 [chris]
- s/we should just put it in scope/if the group decides to work on this, it should be explicitly listed/
- 19:06:38 [fantasai]
- Steve: The other reason to put it in scope is to be clear on IPR
- 19:07:04 [fantasai]
- Steve: It's not that Adobe would reject the charter, but we would need to consult a different part of the company about it
- 19:07:42 [fantasai]
- Alex: So we'd have to do a charter update to bring animations etc. in scope
- 19:08:03 [fantasai]
- Molly: There's something that concerns me about what Hyatt is proposing.
- 19:08:13 [fantasai]
- Molly: It will divert us from a lot of important things.
- 19:08:34 [fantasai]
- dbaron: That's a prioritization consideration.
- 19:08:56 [fantasai]
- dbaron: We have to prioritize our existing things, and we can prioritize those as well.
- 19:08:58 [hyatt]
- grid will divert us from a lot of important things also
- 19:09:13 [hyatt]
- any new proposal has to be prioritized
- 19:09:26 [fantasai]
- dbaron: We should include them, and then prioritize the whole group
- 19:09:36 [fantasai]
- dbaron: I don't think they'll sort at the bottom of the list
- 19:09:54 [fantasai]
- Jason: Apple didn't propose these because they wanted something new. They saw important use cases.
- 19:10:17 [chris]
- The proposed it because it ships in iPhone,in fact ....
- 19:10:30 [fantasai]
- RESOLVED: Agreed that the areas proposed are areas that could go in the charter.
- 19:10:54 [fantasai]
- STeve: We need to do a realistic cut of what we'll work on in the charter, and the second part determines what goes in
- 19:11:42 [fantasai]
- Topic: Prioritization
- 19:12:18 [fantasai]
- Daniel: I think we need the list of modules we have for CSS3, and look at them carefully to know
- 19:12:40 [fantasai]
- Daniel: how important they are, how interested implementors are to implement them.
- 19:12:50 [anne]
- anne has left #css
- 19:13:02 [anne]
- anne has joined #css
- 19:13:18 [fantasai]
- Daniel: Peter and I are willing to have confidential, NDA discussions with individual implementors about these
- 19:13:39 [fantasai]
- Peter: It doesnt' have to be confidential, but we're willing to do that
- 19:13:53 [anne]
- (aside: http://my.opera.com/desktopteam/blog/2008/03/26/opera-and-the-acid3-test 100/100, small rendering glitch left)
- 19:13:53 [fantasai]
- Peter: We don't want modules in our charter that only one company cares about.
- 19:14:16 [fantasai]
- Tantek: I think there are two things you could measure.
- 19:14:34 [fantasai]
- Tantek: You could ask for an advocate within the working group to drive the module
- 19:14:55 [fantasai]
- Tantek: The other is, ask implementors "no interest, some interest, strong interest". Doesn't disclose when you'll ship
- 19:15:13 [fantasai]
- Tantek: I think having that feedback from browser vendors is key
- 19:15:36 [fantasai]
- Tantek: With those 2 pieces of information, then you could say .. the minimum for a module to get in is two strong plus one advocate
- 19:16:27 [fantasai]
- Tantek: I think we should have that instead of a priorities discussion.
- 19:16:39 [fantasai]
- Tantek: decide by demand rather than personal bias
- 19:17:18 [fantasai]
- Jason: Another factor to consider is web designer demands
- 19:17:57 [fantasai]
- Tantek: designers can influence this by either becoming advocats or by convincing implementors to go for it
- 19:18:54 [glazou]
- q+
- 19:19:17 [fantasai]
- Molly: I think Jason's point is that the public should also be able to give feedback aside from these two things
- 19:20:22 [fantasai]
- Molly: WaSP and Microsoft worked together to come up with feature priorities.
- 19:20:25 [glazou]
- ack glazou
- 19:21:11 [fantasai]
- Tantek: If we post to www-style with our priorities, what's been cut, and our rationale and ask for comments, people will speak up
- 19:22:53 [fantasai]
- RESOLVED: Gather information for modules: whether it has an advocate, which browser implementors have no interest, some interest, strong interest.
- 19:23:10 [Bert]
- A quick analysis of comments on the WaSP blog: http://www.w3.org/Style/Group/2008/feedback
- 19:25:15 [fantasai]
- ACTION Daniel and Peter: Collect information on: advocate? #implementors no/some/strong interest and prepare anonymized report
- 19:26:14 [fantasai]
- RESOLVED: post to www-style for feedback
- 19:26:49 [glazou]
- LUNCH
- 19:55:40 [alexmog]
- alexmog has joined #css
- 20:05:00 [mjs]
- mjs has joined #css
- 20:10:45 [jason_cranfordtea]
- jason_cranfordtea has joined #css
- 20:28:19 [anne]
- anne has joined #css
- 20:31:47 [MG]
- MG has joined #css
- 20:36:51 [plinss]
- plinss has joined #css
- 20:39:06 [Ming]
- Ming has joined #css
- 20:49:51 [dbaron]
- dbaron has joined #css
- 20:50:20 [dbaron]
- ScribeNick: molly
- 20:51:54 [glazou]
- hyatt: ping
- 20:52:02 [hyatt]
- glazou: pong
- 20:52:02 [fantasai]
- Steve: We should discuss the telecons and how we use them, because one of the problems going into this meeting was that our current methods are blocking participation
- 20:52:22 [fantasai]
- Steve: on a related note, I'm hesitant to include Apple's proposals in our work if they're not participating in the WG
- 20:52:43 [fantasai]
- ...
- 20:53:03 [fantasai]
- Steve: For things that aren't reaching closer, are we reaching agreement on what problem we're trying to solve?
- 20:53:32 [glazou]
- -- quote from David Hyatt --
- 20:53:34 [glazou]
- I can bring that up with my management. The face to face CSS WG meetings I have attended have (with the notable exception of a fantastic margin collapsing conversation between Ian, Tantek, David Baron and myself) largely a waste of time. This is something I'd like to go into in more detail, but it would take a lot longer email. Maybe you can turn this around.
- 20:53:42 [fantasai]
- s/.../Steve: We need to minute when an agreement has been reached/
- 20:53:43 [glazou]
- The big problem the CSS WG has right now is that there is a big pile of CSS3 modules, and a given WG member may be interested in only some of them. You really need to break up meetings and tasks by module I think. For example, I could not care less about CSS 3 paged media. We may implement it, but frankly printing is a pretty low priority for us. I don't want to sit through a bunch of conference calls or face to face meetings where a topic I have no interest in
- 20:53:46 [chris]
- trackbot-ng, status
- 20:53:57 [glazou]
- I find it hard to attend teleconferences for which I am only interested in a small portion of the topics and very hard to justify traveling halfway across the world if the meeting is not going to be extremely productive.
- 20:54:01 [glazou]
- -- end of quote --
- 20:54:52 [tantek]
- I sympathize and agree with many of David Hyatt's concerns.
- 20:55:16 [fantasai]
- Summary: Need to get closure on issues we discuss and minuting that closure. Also where such closure is not there, neet to clarify what exactly is the problem being discussed.
- 20:55:38 [fantasai]
- Steve: David Baron made the point that telecons aren't a good place for brainstorming solutions.
- 20:56:07 [SteveZ]
- SteveZ has joined #css
- 20:56:11 [chris]
- s/quote from/quote (with his permission) from/
- 20:57:30 [fantasai]
- Daniel reads Hyatt's comments
- 20:58:07 [molly]
- Chris: another possibility is having more telecon with topic specifics
- 20:58:17 [glazou]
- mail sent
- 20:58:17 [molly]
- Chris: don't have to go if not interested
- 20:58:49 [fantasai]
- Chris: first possibility to limit telecons to previously-stated topics.. might result in long time between discussion of a topic
- 20:58:50 [molly]
- Steve: The default rule is that you have to make 2 out of 3 meetings
- 20:59:07 [molly]
- Chris: You can make this formal and do it by task force
- 20:59:20 [hyatt]
- There are many issues in CSS that are complex enough that solutions cannot be determined to be "right" at a face to face meeting. Often you need to go off and think about it for a while, to write test cases, perhaps even to implement a bit in order to really understand whether such a solution will work. I think telecons and face to face meetings encourage rushed/immediate resolution of issues that may require deeper investigation;conversation (for which email ma
- 20:59:21 [fantasai]
- Molly: You could also organize by module
- 20:59:48 [molly]
- Peter: My fear of going down that path is that time will go by where there's 3 or 4 meetings where people who don't have interest in those topics will disconnect
- 20:59:58 [molly]
- Peter: from the group, it falls off their radar, we've lost them
- 21:00:06 [dbaron]
- hyatt, end of line was cut off after "(for which email ma"
- 21:00:08 [molly]
- Chris: That's why you have Task Forces
- 21:00:27 [chris]
- and then a plenary call that everyone attends
- 21:00:27 [molly]
- Peter: My other concern is that sometimes people have insights that people are too close to the problem do not see
- 21:00:57 [molly]
- Peter: For example, it might be good to have Dave's eyes on paged media /because/ he doesn't have a primary interest in it
- 21:01:26 [molly]
- Chris: If thrashing things out in painstaking detail is going to prevent people from participating . . .
- 21:01:39 [mjs]
- mjs has joined #css
- 21:01:47 [molly]
- Jason: Rather than one check-in, periodic check-ins
- 21:02:06 [SteveZ]
- q+
- 21:02:31 [fantasai]
- Peter: I don't want to see someone check out while the wg thrashes on a module for 6 months, then look at it the first time and say "we can't implement that" and throw it back to square one
- 21:02:34 [molly]
- Peter: I understand the concept of people not involving themselves in things they don't care about
- 21:02:46 [chris]
- q+ to agree with Jason that specs need to be updated frequently so others can read the 'latest' at any time
- 21:02:51 [molly]
- Peter: I do want people to stay involved and interested, open to suggestions
- 21:02:53 [dbaron]
- ack SteveZ
- 21:03:17 [molly]
- SteveZ: We often spend an entire meeting on a given topic, and we often don't know what that topic will be
- 21:03:56 [Hixie]
- fantasai: any editor should be working closely with the browser vendors right from the start, before even writing one line of the spec, so that should never be able to happen
- 21:04:05 [molly]
- SteveZ: I find it hard to believe that David (Hyatt) isn't interested in 2.1 - I know I've tuned out things I don't care about, I'm doing other things online, if I hear something relevant
- 21:04:39 [molly]
- SteveZ: It's a valuable contribution - you can't say you're only going to attend the part you're interested in - commitment to the working group to commit to the entire body of the working group
- 21:06:09 [chris]
- ack chris
- 21:06:09 [Zakim]
- chris, you wanted to agree with Jason that specs need to be updated frequently so others can read the 'latest' at any time
- 21:06:36 [molly]
- Chris: picking a point up from Jason - things should be checked in regularly, things shouldn't be a surprise
- 21:06:50 [molly]
- Fantasai: A lot of times people won't read the spec until their interested in it
- 21:06:56 [molly]
- their/they're
- 21:07:20 [Hixie]
- fantasai: if no browser vendor is involved, then the work is likely pointless
- 21:07:53 [molly]
- Jason: What are those weekly meetings meant to be? Let's step back and define it
- 21:09:00 [molly]
- Glazou: My perception is that it's mostly to resolve issues that require more detail
- 21:09:07 [fantasai]
- conversation
- 21:11:29 [dbaron]
- q+ on telecons vs. email and on resolution granularity
- 21:11:42 [dbaron]
- q+ to talk about telecons vs. email and on resolution granularity
- 21:11:57 [dsinger_]
- q+
- 21:12:41 [dbaron]
- ack dbaron
- 21:12:41 [Zakim]
- dbaron, you wanted to talk about telecons vs. email and on resolution granularity
- 21:13:02 [fantasai]
- dbaron: Somebody mentioned using telecons where things would take longer on email. I think that makes sense.
- 21:13:03 [dsinger_]
- Hi I am happy to attend, but I cannot claim to be a CSS expert overall
- 21:13:16 [fantasai]
- dbaron: I think we've also used telecons to discuss things that would be shorter on email.
- 21:13:26 [dsinger_]
- I am a media person, partic. the audio/video elements, and some support for the transitions, animations etc.
- 21:13:37 [fantasai]
- dbaron: In many cases the editor would be able to come up with the solution that everyone would agree with
- 21:13:37 [glazou]
- dsinger_: wait
- 21:13:45 [dsinger_]
- I'd probably get up to speed on CSS, but it's not true I'd have intelligent things to dsay about printing
- 21:13:50 [fantasai]
- dbaron: and not use up group time discussing the different options
- 21:13:50 [glazou]
- dsinger_: will give you the floor as soon as dbaron has ended
- 21:14:18 [fantasai]
- dbaron: Second point, Bert made the point that we need to have resolutions on all these things
- 21:14:31 [fantasai]
- dbaron: we don't need to have very granular
- 21:14:45 [fantasai]
- dbaron: Someone can go do work, and we can accept the list of changes, or accept them minus these two
- 21:14:50 [fantasai]
- Bert: we do that a lot
- 21:16:20 [fantasai]
- dbaron: we often don't have a proposed resolution at the point when we start discussing it.
- 21:17:22 [molly]
- LOLspec by Eric Meyer: http://www.flickr.com/photos/mollyeh11/2364990122/
- 21:17:44 [fantasai]
- Elika: I think in most cases we should have proposed text before discussing an issue
- 21:18:00 [fantasai]
- Elika: But sometimes I'll have an issue and I'm not sure what to do, and it's helpful to have a few minutes of group time to get direction and ideas
- 21:18:36 [fantasai]
- Arron: we'd need to make sure we cut discussion after say 5 min on such issues, then go back and discuss it again later
- 21:18:49 [fantasai]
- Molly: ...
- 21:19:08 [fantasai]
- Molly: In the i18n working group, we'd have an agenda.. and then we'd throw out the agenda depending on who's there or not there.
- 21:19:47 [fantasai]
- Molly: I'm not hearing these problems just in this wg, btw, a lot of w3c groups have these problems
- 21:19:55 [fantasai]
- Steve: three things
- 21:20:01 [fantasai]
- Steve: We need to agree that we made a decision
- 21:20:17 [fantasai]
- Steve: If something has come up, trying to decide who it belongs to but also clarifying what is the issue
- 21:20:39 [hyatt_]
- hyatt_ has joined #css
- 21:20:44 [fantasai]
- Steve: Thirdly there is sometimes useful when there is not closure on the mailing list to have a verbal discussion that may identify some of the hidden pieces of the problem that need to come out on the table for people to start communicating
- 21:20:54 [fantasai]
- Steve: Or the third one is trying to get rid of communication barriers
- 21:21:09 [fantasai]
- Steve: my second category is aimed at what Elika was talking about getting hints
- 21:21:21 [fantasai]
- Steve: That could probably be limited to a 5-10 minute discussion
- 21:21:33 [fantasai]
- Steve: And of course there's administrative discussions like where's the next meeting
- 21:22:20 [fantasai]
- Steve: The goal of outlining that ..
- 21:22:29 [fantasai]
- Steve: We have agendas in the four broad areas.
- 21:22:45 [fantasai]
- Steve: The first part would be making decisions. People should participate in the first part.
- 21:23:08 [fantasai]
- Steve: Then we should put more discussive items towards the end of the agenda, so people can tune out or turn off if they're not interested in the topic
- 21:23:21 [fantasai]
- Steve: But regular participation would be expected for the beginning of the meeting
- 21:24:04 [fantasai]
- Daniel: We don't predict how long a topic will take: we might predict 15 min, and it takes the hour
- 21:24:10 [fantasai]
- Steve: The chair can postpone the discussion
- 21:24:34 [fantasai]
- Peter: We want to set the expectation that everyone participates in the first 20min 30min of the call and then they cna drop off
- 21:24:42 [glazou]
- dsinger_: go ahead here please
- 21:24:45 [glazou]
- write here
- 21:24:50 [fantasai]
- Peter: we have so many telecons where only 4 people participate, and we have 14 ppl here
- 21:25:14 [dsinger_]
- Hi . (Sorry 'bout that, I forgot how zakim q'd and unq'd speakers.) I am happy to attend, but I cannot claim to be a CSS expert overall. I am a media person, partic. the audio/video elements, and some support for the transitions, animations etc. I'd probably get up to speed on CSS, but it's not true I'd have intelligent things to say about e.g. printing (at least initially). so, given a large spec., you will get (and want) 'area experts'. agendas help...as
- 21:25:24 [dsinger_]
- ack
- 21:25:29 [glazou]
- we read
- 21:25:53 [glazou]
- dsinger_: text cut at "agendas help...as"
- 21:25:57 [chris]
- david if you said anything after "help ..as" we missed it
- 21:26:17 [dsinger_]
- as do plenary sessions plus area-specific breakout groups
- 21:26:31 [dsinger_]
- q-
- 21:27:23 [fantasai]
- Molly: Going back to issue of communication and best ways to ensure baseline participation
- 21:27:32 [fantasai]
- Molly: Peter was talking about e.g. first 15 minutes
- 21:27:40 [fantasai]
- Molly: Maybe there could be one set meeting that is an all-hands call
- 21:28:21 [fantasai]
- Daniel: I have mixed feelings.
- 21:28:45 [fantasai]
- Daniel: If one meeting is mandatory, then the others are seen as optional
- 21:29:09 [fantasai]
- Daniel: When you use the first 15 min, then its easier to stay longer for the full hour
- 21:30:00 [fantasai]
- Chris: You need to send out agendas at least 24 hours in advance
- 21:30:06 [fantasai]
- Chris: Planning agendas in advance
- 21:30:13 [fantasai]
- Chris: Identify who needs to be there for the topic
- 21:30:27 [glazou]
- we don't want to say the first 15/20 minutes, each agenda sent will determine what's mandatory or not
- 21:30:41 [dsinger_]
- hm. I've seen this done with periodic 'full' mandatory (maintain status by attending) meeting with much more frequent area-specific breakouts that are self-select. real decisions are made at 'full' meetings, only recommendations from breakouts...
- 21:31:08 [fantasai]
- Peter: We should send out agendas Friday or monday before, so people know what's coming up
- 21:31:11 [chris]
- lists of planned topics and dates are useful, allow planning. Regrets should also be sent in advance
- 21:31:19 [fantasai]
- Peter: And if someone can't make it we know ahead of time and can reschedule topic
- 21:31:38 [fantasai]
- Peter: Another pet peeve: a lot of our telecons start late
- 21:31:47 [fantasai]
- Peter: We schedule to start 5 min before the hour
- 21:31:58 [fantasai]
- Peter: So 5 min to call in and we start on the hour.
- 21:32:53 [fantasai]
- RESOLVED: Telecons scheduled for 5 min before noon Boston time on Wednesday, start at noon
- 21:33:17 [fantasai]
- Peter: We could reinstate the dollar-a-minute rule from Gecko meetings
- 21:33:26 [fantasai]
- Peter: If you're late, it's a dollar a minute up to $20
- 21:33:56 [dsinger_]
- (10 cents) * (number of minutes late) * (number of people waiting when you join) !!
- 21:35:21 [fantasai]
- RESOLVED: Agenda due Monday before agenda-writer goes to sleep or sun comes up next morning, whichever comes first
- 21:36:09 [glazou]
- dsinger_: so you're suggestins Apple is going to sponsorize all our social events in this WG ? ;-)
- 21:36:22 [dsinger_]
- tee hee
- 21:36:48 [dsinger_]
- if I collect those fines, it should be easy...
- 21:37:00 [fantasai]
- RESOLVED: If you can't make the meeting, or going to be late send regrets -- due the day before (should be after agenda is sent)
- 21:37:37 [fantasai]
- Peter: Nonattendance without sending regrets is bad, risks putting you in bad standing
- 21:37:58 [fantasai]
- Steve: standing is more lenient if you send regrets than if you don't
- 21:38:09 [glazou]
- we use to work like that, let's go back to that
- 21:38:13 [glazou]
- s/use/used
- 21:41:15 [chris]
- Anne is willing to give telcons a go if they are made better
- 21:41:52 [fantasai]
- Peter: If we're minuting in IRC, and you're participating back, then attendance via IRC is acceptable.
- 21:43:12 [fantasai]
- Peter: Participating on IRC doesn't mean merely being online.
- 21:44:03 [glazou]
- s/emergency/exceptional
- 21:44:22 [fantasai]
- RESOLVED: Participation over IRC is acceptable for telecons in exceptional circumstances.
- 21:45:27 [fantasai]
- RESOLVED: First topics on agenda must be resolving issues that require consensus, later topics can be more exploratory discussions
- 21:45:56 [fantasai]
- RESOLVED: Attendance at first part of telecon is mandatory, first part defined by agenda and probably ~20min
- 21:46:33 [fantasai]
- Topic: Advocates for Modules
- 21:48:36 [fantasai]
- CSS2.1: advocated by entire CSSWG..
- 21:49:16 [chris]
- link to what is being projected?
- 21:49:20 [fantasai]
- dbaron: We don't necessarily need the CSS3 Module for something if we don't have urgent things to add to them
- 21:49:36 [fantasai]
- dbaron: for example if we're not adding features to Inheritance and Cascading, then that module is not urgent
- 21:49:47 [fantasai]
- dbaron: because it's already defined in CSS2.1
- 21:50:18 [fantasai]
- chris, http://www.w3.org/Style/CSS/current-work
- 21:51:48 [fantasai]
- discuss dependencies, css2.1 css3
- 21:52:58 [fantasai]
- Elika suggests reading http://www.w3.org/TR/css-beijing since these dependencies are clarified there
- 21:53:55 [molly]
- All: Determining advocate roles
- 21:54:17 [molly]
- Daniel: Advocate for selectors module
- 21:54:44 [fantasai]
- Tantek: can add a fourth level of interest: partially implemented
- 21:55:02 [molly]
- Daniel: Yes but you have partial implementations for years
- 21:55:18 [molly]
- Tantek: It's actually sufficient to succeed with two partial implementation
- 21:55:23 [fantasai]
- Scribenick: Molly
- 21:56:00 [molly]
- Tantek: Adds fourth level
- 21:56:16 [molly]
- Mobile Profile ?
- 21:56:34 [chris]
- Svante Schubert
- 21:56:49 [molly]
- Svante (for MP? - to be decided)
- 21:57:22 [molly]
- CSS Snapshot: Fantasai
- 21:57:57 [molly]
- Namespaces: Fantasai and AVK
- 21:58:40 [anne]
- s/AVK/AvK/
- 21:59:04 [molly]
- action: fantasai add language regarding advocates to roadmap
- 21:59:04 [trackbot-ng]
- Created ACTION-12 - Add language regarding advocates to roadmap [on Elika Etemad - due 2008-04-02].
- 21:59:19 [molly]
- Paged Media: Fantasai
- 21:59:28 [chris]
- I would like to see the current-work document also linking to test suites (if they exist)
- 21:59:44 [molly]
- Print Profile: Fantasai
- 22:00:17 [fantasai]
- (both under HP identity)
- 22:00:38 [molly]
- values and units: Hakon, David
- 22:01:18 [molly]
- David: Let me put it this way: Of the sort of three drafts that I think I want to work on implementing in the six months, this is the only one that lacks an advocate right now
- 22:01:31 [molly]
- Tantek: Says you have strong interest as a browser vendor, that's what I hear
- 22:04:18 [molly]
- Cascading and Inheritance: Hakon?
- 22:05:13 [molly]
- CSS text/text layout: Paul, Steve, Fantasai
- 22:05:39 [molly]
- CSS Ruby: Paul, Steve
- 22:06:01 [molly]
- Generated Content for Paged Media: Hakon?
- 22:06:07 [molly]
- SteveZ: That needs to be split
- 22:06:41 [molly]
- SteveZ: This is not a trivial task. The whole area of grids/advanced layout/generated content for paged media intersects in ways that need to be sorted out
- 22:06:59 [molly]
- Fantasai: Not so much Generated Content
- 22:07:40 [molly]
- Alex: These are inter-related
- 22:07:56 [molly]
- Backgrounds and Borders: Bert and Fantasai
- 22:09:09 [molly]
- Fonts: Jason and John D?
- 22:09:38 [molly]
- Bert: Paul is interested
- 22:09:57 [molly]
- Daniel: Interest from Apple
- 22:11:15 [molly]
- Discussion: fonts, text effects Jason, Bert, Chris
- 22:11:16 [anne]
- http://webkit.org/blog/85/introducing-text-stroke/
- 22:11:45 [molly]
- Basic Box Model: Bert
- 22:11:55 [molly]
- Multicolumn Layout: Hakon
- 22:13:01 [molly]
- Advanced Layout: Bert
- 22:13:13 [molly]
- media queries: Hakon, Anne probably taking over
- 22:13:22 [molly]
- speech "you gotta be kidding"
- 22:13:51 [molly]
- speech: No advocate currently, public call
- 22:14:18 [molly]
- Color: David
- 22:14:32 [molly]
- Chris: Problem with color module, for much of it you can conform without doing anything
- 22:14:46 [molly]
- Chris: it's very difficult to test
- 22:15:26 [molly]
- Basic UI: Jason
- 22:15:28 [chris]
- I'd like to discuss that later in the meeting
- 22:16:05 [molly]
- CSS Scoping: Daniel, but suggest we drop it, I might write a note about it
- 22:16:19 [molly]
- Grid positioning: Alex
- 22:16:52 [molly]
- CSSOM View Module: Anne
- 22:16:59 [molly]
- Extended Box Model: Bert
- 22:17:14 [molly]
- Object Model: Anne
- 22:18:43 [molly]
- Syntax: noone until it becomes an issue
- 22:18:55 [molly]
- lists: Arron
- 22:19:55 [molly]
- Chris: web fonts was removed from CSS Charter - would be done by SVG, doesn't need a group
- 22:20:55 [molly]
- Discussion: advocate? part of charter? Determination is no
- 22:21:09 [molly]
- Jason is interested
- 22:21:17 [molly]
- David: Ought to be
- 22:21:22 [molly]
- SteveZ: um no
- 22:21:37 [molly]
- Peter: Let's table this
- 22:21:48 [molly]
- Peter: Will note as separate topic for discussion
- 22:21:56 [molly]
- tables: "deprecate!"
- 22:22:04 [molly]
- tables: Saloni
- 22:22:10 [molly]
- tables: David
- 22:22:38 [molly]
- Reader Media Type: no advocate
- 22:22:53 [chris]
- s/SVG, doesn't need a group/SVG, out of scope for this group/
- 22:22:54 [molly]
- positioning (David "deprecate!")
- 22:23:35 [molly]
- No specific advocate at this point
- 22:23:47 [molly]
- generated and replaced content: Fantasai
- 22:24:11 [molly]
- line layout: Steve
- 22:25:18 [chris]
- q+ to ask How does the Math module fit with "A MathML for CSS Profile"
- 22:25:49 [molly]
- Hyperlink Presentation: Jason, Bert
- 22:26:59 [molly]
- style attribute syntax: no advocate
- 22:27:05 [chris]
- ack chris
- 22:27:05 [Zakim]
- chris, you wanted to ask How does the Math module fit with "A MathML for CSS Profile"
- 22:27:11 [chris]
- http://www.w3.org/TR/2007/WD-mathml-for-css-20071214/
- 22:28:26 [molly]
- Math: Bert
- 22:29:22 [molly]
- CSS Presentation Levels: no advocate
- 22:29:31 [molly]
- Aural Style Sheets: no advocate
- 22:30:12 [molly]
- CSS TV Profile: no advocate
- 22:30:18 [molly]
- Behavioral Extensions: Ian
- 22:33:06 [molly]
- ****break*****
- 22:33:08 [glazou]
- fantasai: and we'll need a Selectors Level 4
- 23:07:26 [molly]
- Daniel: We have between 1-1.5 hours, we listed five things we wanted to discuss or state
- 23:08:13 [molly]
- Daniel: New modules to see (ie Apple); what is an implementation wrt test suites; possible division of complex specs of primer/language; features * that we do not use in this working group at all; potential new members
- 23:08:22 [molly]
- Daniel: Any preference for the order?
- 23:08:31 [molly]
- Peter: New modules first
- 23:09:09 [molly]
- Peter: focus remainder of day on charter / group issues
- 23:09:34 [molly]
- Steve: Questioning whether the group should decide deprecated modules
- 23:09:52 [molly]
- Steve: I don't want to do it now but it seems somewhere in the agenda we talk about this
- 23:10:05 [molly]
- Tantek: I think that can be postponted 'til after new charter
- 23:10:34 [molly]
- Tantek: How long do we give for advocates?
- 23:10:41 [molly]
- Peter: Need to get charter finished by end of April
- 23:10:56 [molly]
- Daniel: Next conference call is on Wednesday
- 23:11:22 [molly]
- Daniel: I suggest no more than two weeks, publish the list with no advocates to www-style
- 23:11:36 [molly]
- Peter: We had some proposed new modules we have advocates for
- 23:11:56 [molly]
- Fantasai: Level 4 selectors - from webstandards.org
- 23:12:13 [molly]
- Daniel: says yes, no dissent
- 23:12:38 [molly]
- Anne: Interested in flexible box model and Apple's proposal
- 23:12:56 [molly]
- Peter: Apple was proposing filters, transitions, animations, transformations
- 23:13:08 [molly]
- Anne: Extension to media queries as well
- 23:13:20 [Bert]
- http://www.w3.org/mid/15590069-5956-459A-99AA-86872198C940@apple.com
- 23:13:32 [glazou]
- http://lists.w3.org/Archives/Public/www-style/2008Mar/0313.html
- 23:14:26 [molly]
- Steve: One comment on the transforms - the 2d piece of it is already work in progress within the working group of which I am the advocate. The separate piece 3d piece is separate
- 23:15:25 [molly]
- Steve: I think that defines a viewpoint, but it gets very much more involved in graphics than the 2d piece
- 23:16:02 [molly]
- Peter: We can call for new modules but we ask for advocate with it
- 23:16:10 [molly]
- Daniel: Must be in the scope
- 23:16:39 [molly]
- Fantasai: If you open that up, you get people with no technical ability - you have to evaluate their abilities
- 23:17:45 [molly]
- Fantasai: module idea+advocate+two implementers willing to show strong interest within the charter period
- 23:17:55 [anne]
- +test suite
- 23:18:22 [molly]
- Daniel: one change re: CR tests. For the time being an implementation is a shipped product
- 23:18:47 [molly]
- Peter: We have an opportunity to open this up in the new charter
- 23:19:04 [molly]
- Chris: This is where the distinction between test suites I was describing earlier comes in
- 23:19:58 [molly]
- David: One of the motivations for the way it is now, maybe someone will implement something that isn't interoperable, they release it, they find the problem
- 23:20:14 [molly]
- Peter: The test is to test the specification
- 23:20:30 [molly]
- David: Is the spec compatible with the web is probably something we should be testing
- 23:21:41 [molly]
- Daniel: Implementations are supposed to demonstrate implementability, not that the spec is good for the web
- 23:21:46 [fantasai]
- David: If a browser isn't going to ship it even after implementing it, we probably don't want it
- 23:22:59 [molly]
- Peter: If the feature shows up in a nightly build over two implementations out of CR > PR
- 23:23:16 [molly]
- Peter: If one of the implementations pulls out, we can then pull it out of the spec.
- 23:23:28 [molly]
- Anne: Which specs can be moved forward on nightly builds
- 23:24:16 [molly]
- Tantek: I'd like to open that up to at least one
- 23:24:49 [molly]
- Chris: It's very damaging to have things sitting their waiting
- 23:25:05 [molly]
- Anne: You also have to demonstrate web compatibility
- 23:25:23 [molly]
- Peter: By having two implementations that pass a test suite we have enough to start finding compatibility problems
- 23:26:01 [molly]
- Alex: I would like to ship new features without having to use the prefix
- 23:26:45 [molly]
- Chris: If it's in CR, you can take away the prefix, it's risky
- 23:26:57 [molly]
- Anne: I don't think it's that risky
- 23:26:58 [chris]
- s/their waiting/there waiting, especially if there are multiple implementations/
- 23:27:45 [molly]
- Steve: The original specs for vertical text went to CR and stayed in CR for a number of years and had one implementation (prior to CR), then went out, went back into WD, has been split up and moderately drastic changes
- 23:28:36 [molly]
- Steve: I wanted to cite an example where additional input during CR made change
- 23:28:56 [molly]
- Peter: We're talking about what happens /after/ CR
- 23:29:16 [molly]
- Discussion: What is CR?
- 23:29:24 [molly]
- Peter: We're calling for testing of the spec
- 23:30:34 [molly]
- general discussion re process
- 23:30:51 [molly]
- Tantek: Changed my opinion, I'd be fine with nightly builds
- 23:31:09 [molly]
- David: Publicaly available
- 23:31:28 [molly]
- David: Wait a few weeks, makes sure it sticks
- 23:32:07 [molly]
- Peter: In order to exit CR, it has to pass the test suite, your buffer zone there is your criteria to pass the test suite - it's going to have to be out in the public, automatic buffer
- 23:32:16 [molly]
- Fantasai: I don't think there's a problem of writing in a minimum time
- 23:32:49 [molly]
- Chris: Another thing that's commonly excluded: Mobile - not delivered to users w/out buying phone
- 23:33:33 [dbaron]
- Tantek: Public point-of-sale is allowed.
- 23:34:15 [molly]
- Anne: I really still don't see why waiting is bad
- 23:34:22 [molly]
- Tantek: Because we've waited too long already
- 23:34:41 [molly]
- Peter: Do you have a very good reason for NOT making this change
- 23:34:59 [molly]
- Anne: First of all, I'd like to agree w/ Bert - if the implementations stay there for a year, then ship, it means that it's more stable
- 23:35:07 [tantek]
- e.g. read this: http://www.w3.org/TR/2003/CR-css3-color-20030514/#status
- 23:35:10 [molly]
- Anne: Can be used by authors
- 23:35:20 [molly]
- Alex: For this purpose a beta is still useful
- 23:35:26 [tantek]
- per what is a "feature", what does "interoperable" mean, and what is an "implementation"
- 23:35:37 [molly]
- Daniel: We just saw that Opera is passing Acid3
- 23:37:39 [molly]
- Daniel: We are not putting a stamp that this is going to be conformant to everything, it's a snapshot
- 23:38:00 [molly]
- Daniel: The nightly is enough to show implementability
- 23:38:17 [molly]
- Bert: The requirement is also that it matches the web's architecture
- 23:38:33 [molly]
- Daniel: that's why having some weeks, we'll see those issues
- 23:39:42 [fantasai]
- Bert: Are we changing this just so we can get a REC faster?
- 23:39:57 [fantasai]
- Bert: The reason we have this requirement is so that once we get to REC we don't have to change it
- 23:40:08 [fantasai]
- Molly: It can help expedite some of this
- 23:40:09 [chris]
- +1 to Peter
- 23:40:12 [fantasai]
- Peter: I want us to be less timid
- 23:40:27 [fantasai]
- Peter: I don't want us to be reckless, but less perfectionist
- 23:40:49 [fantasai]
- Straw poll, who's in favor of including publicly-available builds, nightly or shipped or otherwise
- 23:41:03 [fantasai]
- Anne against, others in favor or abstain
- 23:41:18 [glazou]
- hyatt_, dsinger_: ping
- 23:41:57 [fantasai]
- Molly: We don't have to move the spec forward based on this, this just opens up the option of doing so
- 23:42:04 [hyatt_]
- glazou: yes?
- 23:42:05 [glazou]
- "are you in favor of including publicly-available builds, nightly or shipped or otherwise to the list of valid implems for moving specs outside of CR ?"
- 23:42:06 [dsinger_]
- hi. I think if you want to know "is this spec. consistently implementable?" then a nightly build that passes the test suite answers the question. One rather doubts a vendor will want to go backwards on their cmpaibility, surely?
- 23:42:13 [dsinger_]
- that's a yes
- 23:42:18 [hyatt_]
- yes also
- 23:42:20 [glazou]
- thanks
- 23:42:35 [anne]
- dsinger_, if it "breaks the Web" -- sure
- 23:42:38 [dbaron]
- My one other comment was: why have this discussion now. Instead, let's see if we can finish the css3-selectors PR draft *before* we have two shipping implementations and discuss this if we manage to finish the spec first?
- 23:42:50 [dsinger_]
- cmpaibility -> compatibility, lfgdjty -> dyslexia
- 23:42:54 [fantasai]
- RESOLVED: two publicly-available builds (shipped,nightly, or otherwise) can qualify a spec for PR minimum of one month after the build has shipped
- 23:45:11 [fantasai]
- ACTION Peter: write updated wording for CR drafts exit criteria
- 23:45:29 [dsinger_]
- anne: if the vendor discovers that correctly implementing the spec. breaks the web or their users' compatibility, I would hope they come right back to the group and wail (and resist promotion of the spec. until the issue is resolved)
- 23:46:18 [fantasai]
- Topic: Primer
- 23:46:54 [dbaron]
- q+
- 23:46:55 [fantasai]
- Chris: split spec into intro and examples vs. conformance bit
- 23:47:54 [dbaron]
- ack dbaron
- 23:47:59 [tantek]
- suggestion re: action 3 above:
- 23:48:00 [anne]
- dsinger_, true, but it might happen after spec promotion depending on how long it takes for it to be discovered
- 23:48:01 [tantek]
- change "is shipping (i.e. development, private or unofficial versions are insufficient)." to "is shipping (i.e. development, private or unofficial versions are insufficient, however nightly builds are sufficient)."
- 23:48:11 [anne]
- dsinger_, but maybe it's too theoretical and everyone else is right :)
- 23:48:20 [fantasai]
- dbaron: Given the resource constraint problem in this group, I think some of the material that would be in the primer is stuff that could be written outside the group by other people
- 23:48:30 [fantasai]
- dbaron: It might not be quite as effective, but given resource constraints..
- 23:48:56 [fantasai]
- Molly: WaSP should have been rewriting the specs in normal language for over 10 years. It just doesn't happen
- 23:49:07 [fantasai]
- dbaron: It's a real choice between us doing the primer or us doing other specs
- 23:49:28 [fantasai]
- dbaron: I would hope there are people in the community who can write good documentation
- 23:49:52 [fantasai]
- Molly: and a lot of the documentation out there is crap
- 23:50:17 [fantasai]
- Molly: one criticism of w3c is that it's not very friendly to outsiders
- 23:50:23 [alexmog]
- q+
- 23:50:32 [fantasai]
- Molly: I think it's important to W3C to find people from outside and find those resources and pull them in to address these issues
- 23:50:50 [chris]
- s/split spec into/one option is to split specs into/
- 23:50:55 [fantasai]
- Molly: There's a big gap between where specs get made and how they're used
- 23:50:57 [dbaron]
- q+
- 23:50:58 [jason_cranfordtea]
- I'd like to speak
- 23:51:05 [fantasai]
- Alex: Is there a middle ground?
- 23:51:18 [jason_cranfordtea]
- q+
- 23:51:23 [fantasai]
- Alex: If you read something about margin collapsing, there's the details but also some motivation of why we want this.
- 23:51:31 [fantasai]
- Alex: In some cases there's more explanation, some less
- 23:51:48 [fantasai]
- Alex: When I'm reading the spec and I find a discrepency between the motivation and the normative text, I raise it as an issue
- 23:52:21 [fantasai]
- Chris: What you're picking up on is informative vs. normative and chunking it up
- 23:52:37 [fantasai]
- Tantek: As an implementor I want it integrative
- 23:52:44 [dbaron]
- ack alexmog
- 23:53:46 [tantek]
- q+
- 23:54:02 [fantasai]
- Elika
- 23:54:19 [fantasai]
- dbaron: I think we should look at this as the same way of finding advocates as other stuff
- 23:54:29 [tantek]
- I think this is a *really* bad idea.
- 23:54:37 [fantasai]
- dbaron: If the person who is editing CSS3 XX whatever at the technical level is willing to advocate the introductory matieral, fine
- 23:54:53 [tantek]
- IMHO the specs that require primer vs. actual (or whatever) are specs that basically say, we're too complicated, too wordy to bother with.
- 23:54:59 [fantasai]
- dbaron: If not, if someone else isn't willing to be the advocate for it then I'd rather drop that and at least get the spec done so that we can have implementations, make progress
- 23:55:06 [fantasai]
- dbaron: than pull somebody away from something else
- 23:55:07 [dbaron]
- ack dbaron
- 23:55:22 [fantasai]
- Steve: So some questions
- 23:55:36 [fantasai]
- Steve: Is it a good idea to have some material that is non-normative that helps people understand the spec?
- 23:55:53 [fantasai]
- Steve: Is it a good idea to integrate that material into the spec or create it as a separate document or in a wiki etc?
- 23:55:59 [fantasai]
- Steve: How do we find people to do that?
- 23:56:21 [fantasai]
- Steve: ...
- 23:56:29 [fantasai]
- Steve: I'd cite the example of the XML Schema Primer
- 23:56:33 [chris]
- @tantek - actually no, its a mechanism to make the normative part terse and clear
- 23:56:44 [fantasai]
- Steve: which took a document that was unreadable and made it understandable
- 23:56:56 [glazou]
- ack jason_cranfordtea
- 23:57:28 [fantasai]
- Jason: Some of us make our living out of explaining this stuff to different audiences. We know that this stuff is dense, and we need to target the audience.
- 23:57:33 [tantek]
- every spec that requires back and forth between two URLs havs been A LOT harder to read/grok IMHO
- 23:57:40 [fantasai]
- Jason: Have we really addressed who is the audience for what we are producing?
- 23:57:40 [tantek]
- e.g. WAI
- 23:57:50 [fantasai]
- Jason: Do we need to address audiences outside our primary audience?
- 23:57:56 [tantek]
- also
- 23:58:03 [fantasai]
- Jason: I'd suggest our primary audience is implementors
- 23:58:13 [fantasai]
- Jason: Secondary is developers
- 23:58:21 [fantasai]
- Jason: And tertiary is the web designers
- 23:58:26 [tantek]
- the more you separate two pieces of related text (whether in the same document, or across documents), the more you risk them becoming inconsistent
- 23:58:27 [glazou]
- tantek: please, wait until you can use _voice_, you know that old protocol ;-)
- 23:58:40 [tantek]
- glazou, just writing down my thoughts before I forget ;)
- 23:58:46 [glazou]
- bwahaha :-)
- 23:58:47 [fantasai]
- Jason: Writing is a very specific skill, and editing is a very specific skill
- 23:59:08 [fantasai]
- Jason: We have people who are employed full-time by W3C, but I've never heard of writers employed full-time by W3C..?