Warning:
This wiki has been archived and is now read-only.
Teleconference.2007.10.24/Minutes
These minutes have been approved by the Working Group and are now protected from editing. (See IRC log of approval discussion.)
See also: IRC log
Contents
Attendees
- Present
- Achille Fokoue, Alan Ruttenberg, Bernardo Cuenca Grau, Carsten Lutz, Deborah McGuinness, Elisa Kendall, Evan Wallace, Fabien Gandon, Giorgos Stoilos, Ian Horrocks, Jeff Pan, Jeremy Carroll, Michael Smith, Ratnesh Sahay, Rinke Hoekstra, Sandro Hawke, Tommie Meyer, Vipul Kashyap, Zhe Wu, Bijan Parsia, Boris Motik, Peter Patel-Schneider, Uli Sattler, Vit Novacek
- Regrets
- Chair
- Alan Ruttenberg
- Scribe
- Elisa Kendall
Agenda
See http://www.w3.org/2007/OWL/wiki/Teleconference.2007.10.24/Agenda.
Approval of Minutes
Alan Ruttenberg: NOT RESOLVED: Our first working drafts, to be published before the 3-month heartbeat, will be one: (1) Structural Specification, (2) Semantics. We may include (3) RDF Mapping in this list. These are based on the text for each of these at http://www.webont.org/owl/1.1/.
RESOLVED: based on email discussions, this issue is reopened at this meeting.
Wiki
RESOLVED: Accept minutes of meeting 10/24/2007.
Peter Patel-Schneider: with regard to ACTION 1 from last week, will need this resolved in the near future. How can we transition this to the wiki if this isn't resolved?
Sandro Hawke: Action 1 is the log-ins?
Peter Patel-Schneider: Yes. While we can redirect links, we can't merge at this point. One strategy may be to rename accounts, use open id.
Sandro Hawke: If the history is attributed to your old login name rather than new, that may not be a big issue.
Peter Patel-Schneider: It probably isn't too much of a problem if my inputs are split over two logins.
Rinke Hoekstra: Everyone can create logins with new names, since nothing has been edited yet.
Sandro Hawke: Maybe W3C logins are the proper logins to use.
Alan Ruttenberg: Let's take this up this week and try to resolve it.
Ian Horrocks: There are more people on the call than have accounts on the wiki; everyone needs to create a wiki account.
Alan Ruttenberg: The front page should represent consensus of what we all think it should be and what W3C would like it to be. If someone wants to change the front page, they should send a request to the chairs and they will address it accordingly.
Vipul Kashyap: Why is it a bad idea for anyone to add what they want (not disagreeing with guidelines, just asking). When something is approved or disapproved, some sort of reasons should be given, and an alternative proposed.
Alan Ruttenberg: We should add these to the agenda and discuss. I'm in favor of taking a wiki approach, that said, what does it mean to be as permissive as possible?
Alan Ruttenberg: We should develop a set of policies that everyone can live with, but the front page may be special - the front page may need to look a certain way depending on W3C policies, etc.
Vipul Kashyap: I agree with what you're saying but this should evolve over time.
Alan Ruttenberg: Addition to agenda for next meeting to discuss further.
Bijan pointed out Michelle's table from use cases.
Alan Ruttenberg: Any other comments on this proposal?
Alan Ruttenberg: RESOLVED: Edits to Front Page should be vetted by WG chairs.
Vipul will work on functional requirements based on this for next time.
Publication Schedule
Alan Ruttenberg: There was a fair amount of discussion on this topic last week, that we would effectively put the current set of documents as a public working draft, sometime before the heartbeat requirement for a public working draft, around the 6th or 8th of January. Subsequently, there was discussion in email regarding both content and process issues.
Alan Ruttenberg: In light of this, we decided to slow down a bit and reconsider how we should proceed. See http://lists.w3.org/Archives/Public/public-owl-wg/2007Oct/0170.html
Alan Ruttenberg: What we're proposing is that the editing of the 3 documents, whatever comes out of this, will be the first draft, but we will migrate these to the wiki. Effectively, everyone who has an interest in contributing would be able to edit them in the wiki and make them our own.
Alan Ruttenberg: The procedure would be to post an issue to the tracker. Use pages on the wiki to draft changes, issues will be discussed during meetings and changes made based on resolutions. Shortly before, or during f2f, we will make a decision about publication as first WD. Then sandro will take wiki contents and format them back out as draft documents depending on group consensus.
Peter Patel-Schneider: Not that it's likely to come to pass, but I'm a bit uneasy about a decision to not publish for some period of time.
Alan Ruttenberg: It would be self-imposed...
Ian Horrocks: I wouldn't describe this as a publication black-out - we'll decide before a certain date.
Peter Patel-Schneider: Well, shortly before...
Bijan Parsia: One thing that caused a lot of difficulty - the proposal was to publish this week, then we might target the F2F instead, and say that's what we're going for, when we feel ready.
Deborah McGuinness: I was reading some of the email, are we now proposing that we're going to have three documents come out ...
I want to make sure we don't have something that only theoreticians can read. Is the structural specification something we expect to be readable by a broad audience?
Alan Ruttenberg: We were trying to triage to see which documents were available for publication first.
Alan Ruttenberg: The reasoning wasn't to avoid publication of the more readable documents. We should take the same approach to the other docs, post issues, and publish them when they are ready.
Deborah McGuinness: Now matter how hard we work, we're not going to be able to make the semantics document, for example, readable for a large audience. We need another document to come out at the same time that is readable by the broader audience.
Alan Ruttenberg: ...What will the users go to as the ultimate set of documents we produce, and what are the steps in the process. I think what you're saying is that you want some kind of overview document to be in the first set we publish.
Deborah McGuinness: Yes, or something that can take its place.
Alan Ruttenberg: It might be confusing to have some version of the user docs if there is flux in the set of features discussed.
Deborah McGuinness: But if there isn't something available that is readable, people won't be able to follow what's going on.
Bijan Parsia: I'm strongly against this - reasons include that we don't have such a document available for OWL 1.1 yet, although quite a few people have been able to comment on what we do have. There is a limit even with the OWL Guide on who can read it and comment effectively. Most working groups I have participated in recently publish tech docs first, long before publishing a guide, until the tech docs are farther along. I don't think that other WGs would have a problem with publishing the tech docs well in advance of the guide. We can deal with some questions in email along the way.
Vipul Kashyap: I agree with Deborah and disagree with Bijan. I believe that this depends on who we believe our audience is.
Bijan Parsia: I work with end users all the time on various lists, and my primary customer is the end user (people building and using ontologies).
Vipul Kashyap: This reinforces my claims, because he provides these things to implementors and technical teams.
Alan Ruttenberg: He has clarified that this isn't his primary audience.
Vipul Kashyap: If you think your end user is someone using OWL, then that's not the same as an end user who is a business user presenting this technology to a CIO or business user.
I'm in favor of Deborah's position - the first thing in the SW dev lifecycle is to go out and develop use cases (end user use cases rather than technical use cases).
We should bring this issue of use cases forward in the development cycle
The table Michelle developed is fantastic - we need to do the same thing in [other domains] - we need to give primacy to the end user.
Ian Horrocks: It seems to me that alot of this argument in favor of guides, etc. is predicated on the belief that end users will look at OWL; in a way they care less about OWL and more about the editors they use that are based on the standards. I believe that the goal is to bring all of these documents into the domain so that we can work on them as soon as possible. There is an overview document that isn't great but provides some idea of the features which we can point people to.
This whole discussion has drifted quite away from the agenda - if people believe that we should start working on a guide or other doc we should propose an agenda item for that for next week.
Jeremy Carroll: If the best way to move forward is to propose to create these end user facing documents ... I believe that a requirements document would be useful, but agree with Ian ...
Vipul Kashyap: I was attempting to differentiate my view of an end user ...
Alan Ruttenberg: There is broad agreement that we should have good user facing docs. The current procedure doesn't preclude that, but talks about focus on the tech docs and working on them via the wiki.
Alan Ruttenberg: Do you believe that we have adequate support for these user docs (to Deb and Vipul)?
Vipul Kashyap: The issue may be sequencing - we should add this to a future agenda.
Alan Ruttenberg: Do you agree that we can migrate the three proposed technical documents to the wiki as soon as possible, that people can raise issues, etc., and that we can move forward to work on these between now and the F2F?
Vipul Kashyap: Maybe we should consider trying to publish a use cases and requirements document at the same time, which we can take up in the next agenda.
Deborah McGuinness: I was pushing much more for the overview than the guide - updating the guide is a massive amount of work.
Deborah McGuinness: My worry is that if the focus is just on the three docs, that work on an overview would be relegated to the end of telecons and ultimately wont get addressed.
Deborah McGuinness: I'm trying to support a broad user base of scientists and business people who are trying to work with this. Something like an overview is really critical for them.
Alan Ruttenberg: To address the issue of these getting short shrift on the agenda - Ian and Alan are responsible for the agenda, and will make sure it gets appropriate time. I will take this seriously to make sure that all of the concerns of people in the group are considered.
Alan Ruttenberg: Are there any objections currently to adopting this proposal to the group?
Alan Ruttenberg: Objections?
Alan Ruttenberg: This does not preclude an additional resolution that we would publish another document on a similar schedule.
Alan Ruttenberg: Not hearing any objections, and with one abstention from Jim ...
He indicated support for the chairs proposal.
Document Migration
Ok, we're getting close to the end of an hour ...
Ian Horrocks: We should take a quick stab at the next agenda item ... what form these documents are going to take in the wiki or otherwise.
Alan Ruttenberg: There seems to be productive discussion on this in the mailing list at this point
Ian Horrocks: I would like to get this moved forward as quickly as possible so that people can start working on the documents.
Peter Patel-Schneider: I would prefer that we keep the meetings to 90 minutes
Alan Ruttenberg: There are a number of people that can only attend for 60 minutes, and this is an issue for the entire working group, so we should address it at the next meeting.
Bijan Parsia: I'd like to talk about the task forces - and have some guidance about how we are going to proceed with reviews.
Alan Ruttenberg: So in between the wiki docs and task forces is the issues list, tracker, and so forth.
Alan Ruttenberg: Should we address those first or skip to the task forces?
Peter Patel-Schneider: I'd like to be able to have something happen before the next telecon - like moving docs to the wiki, or getting the issues list set up, or something.
Alan Ruttenberg: The last discussion was on how we might mark up the documents.
Peter Patel-Schneider: We have documents in html, let's just move them over to the wiki and figure out how to deal with the math tags later.
Alan Ruttenberg: When you say html, does that mean wiki vs. html?
Sandro Hawke: Right now the docs are pure html, for math will they use images or what...
Ian Horrocks: The structural specification uses images, so it would be possible to move the other two over that don't use images, and deal with the images later.
Alan Ruttenberg: I think use of wiki mark-up is alot cleaner, I'm concerned with broad participation in editing.
Alan Ruttenberg: There is a converter that seems to do reasonably well, but we should be sure that we can use html if that's not the case.
Sandro Hawke: We don't need full html to do the equations ...
Peter Patel-Schneider: I would be very surprised if there were scripts.
Sandro Hawke: We would convert to wiki markup but not wiki math - math equations stay in html.
Alan Ruttenberg: The complicated stuff is in html, but the uncomplicated stuff goes to wiki markup. Sandro and I will do this before next week.
Boris Motik: The important bits of markup in the documents are always in style sheets, so it may be possible to automate the process of migration.
Alan Ruttenberg: That's our thought as well; the OWL WG front page was done in a similar way. Any objections?
Task Forces / Tracker
Alan Ruttenberg: If we can get one or two of the activities identified for task forces we would be in good shape. If people can start working on issues related to backwards compatibility audits, for example, they can submit issues, comments.
Bijan Parsia: We may want to create wiki pages for this.
Alan Ruttenberg: There are two people that should be involved in this - I'll poke at Jim and Jeremy. We should define what this means. As soon as we can start working on test cases against the specs the better as well. If we can start to create test cases, then a harness that can be run right out of the wiki. If we can identify people who want to start working on this, that would be good.
Bijan Parsia: I can start to send email out to people who might want to work on this.
Alan Ruttenberg: We should create a wiki page, and start adding these issues to the wiki.
Sandro Hawke: The tracker doesn't support this yet.
Alan Ruttenberg: On the tracker page - the last thing i noticed is that the tracker didn't let me enter an issue due to permissions.
Alan Ruttenberg: This is just an introduction to get the topic started.
Ian Horrocks: We skipped over this issue list in order to get to the task forces, but getting the issues and tracker sorted out is important to supporting the task forces. Now we're in a position where we may end up with two parallel issues lists.
Alan Ruttenberg: My understanding was that we were going to migrate the OWL 1.1 issues to the tracker once it was working.
Ian Horrocks: So what is the issue with doing that?
Alan Ruttenberg: I think it means someone has to manually do this, so perhaps we should manually move them over once the tracker is up and running.
Ian Horrocks: Qk, we can do this offline, but we need to get to it.
Alan Ruttenberg: Agreed.
We wanted to have a publicly open place (the google site) for people to enter issues, but that the working group should use the tracker.
Ian Horrocks: By moving issues to the tracker now, what happens to new issues added to the google site, and what is the status of those that are moved?
Alan Ruttenberg: They are all reported, and Alan and Ian have to decide how to address those.
It should be a lower priority to deal with new publicly added issues, but we should decide what to do with them going forward.
Ian Horrocks: It would be a big commitment to continue monitoring the other list and moving issues on an ongoing basis.
Alan Ruttenberg: Maybe what we can do is have people monitor this and move them accordingly.
Bijan Parsia: Maybe we can set it up so that they can be monitored automatically and moved.
Alan Ruttenberg: We might be taking on more work that we are chartered to do, which I'm reluctant to do right now.
Ian Horrocks: Just to conclude, the initial action is that we will divvy up the work to manually move these issues to the tracker once the tracker is working sufficiently to do so.
Alan Ruttenberg: Anything else we want to cover at this point?
Alan Ruttenberg: Move to adjourn
Ian Horrocks: second