[13:39:25] --- reschke has joined
[13:39:58] --- reschke has left
[15:21:18] --- LOGGING STARTED
[15:25:26] --- reschke has joined
[15:25:40] --- reschke has left
[21:46:44] --- reschke has joined
[21:50:43] --- SAH has joined
[21:54:09] --- hardie has joined
[21:55:55] --- Lisa D has joined
[21:56:08] <Lisa D> Presentation: <http://www.sharemation.com/~milele/public/dav/presentations/webdav-wg-ietf59.ppt>
[21:56:36] <hardie> I'll be your jabber scribe for the session; please let me know if there are issues you want raised in the meeting.
[21:57:05] <hardie> Agenda bashing, ACL and ordered collection status, wg charter, progress, redirect draft, interop report, RFC2518bis issues, PATCH proposal
[21:57:14] <hardie> Agenda issues welcome--plenty of time.
[21:57:34] <hardie> Here to declare "partial success", and go over why we didn't finish what we didn't finish.
[21:57:43] <hardie> WG extant for 7 years; recharter once.
[21:57:53] --- paf has joined
[21:58:19] <hardie> Completed goals RFC2518, ACL, ordered collections (RFC3648)
[21:58:44] <reschke> disagree with status of BIND -- there are no unresolved *technical* issues with locks
[21:58:51] <hardie> Official goals, not complete: property registry, binding, redirect references, ACL goals document, DRAFT standard status(2518bis)
[21:59:09] <hardie> The property registry just has been as necessary as thought, no demand, so no action.
[22:00:10] <hardie> Binding did produce a draft, but it is stalled. Lisa's personal opinion is that the draft cannot be used, since it breaks locking. The authors disagree, and the number of participants is not large enough to get past with more community input.
[22:00:55] <hardie> Redirect references--might go to standard, but not clear that there will be implementations. There has been no common way of authoring these resources, but still not clear that there would be implmentations.
[22:02:02] <reschke> Last time issues were raised with BIND vs LOCK, there has been a WG dicussion, and IMHO all open questions have been answered. If somebody feels that there's a problem, the issue should be re-raised on the mailing list. Until then, the author(s) feel that there is no issue.
[22:02:06] <hardie> ACL goals document, author not pushing it (Lisa), not clear it's needed if ACL is done
[22:03:11] <hardie> Drive to finish 2581bis great during the interoperability phase, but now energy stalled over discussions of recycle at proposed with moves of locking info to a separate draft.
[22:03:43] <hardie> Lisa saying that Webdav may need to be in interop mode for a while before going to Draft standard.
[22:03:56] <hardie> Lisa repeats to Larry that our result code is 207 "Partial Success"
[22:04:07] <hardie> Other work: DASL, property data types, quota, http patch
[22:04:42] <hardie> There are interoperable implementations of DASL, but it has stalled. Not clear
[22:04:59] <hardie> if this should go to informational, experimental, or proposed standard.
[22:05:01] <hardie> Comments?
[22:05:05] <reschke> I feel the sheer number of interoperable servers and clients show the success of the group.
[22:05:24] <hardie> Is that a comment for the mic?
[22:05:44] <hardie> If so, please preface it with a "mic", so I know to do it.
[22:06:37] <reschke> mic: DASL: I feel we have one major thing to do, which is to define datatype-based queries. Right now this only works because servers have some hardwired logic about which live properties have what type. Need to make that more flexible.
[22:07:15] <hardie> Larry says experimental is used when there is doubt on whether it is the right approach. It's a way of gauging buy-in
[22:07:25] <reschke> mic: I feel the sheer number of interoperable servers and clients show the success of the group. mic: I feel the sheer number of interoperable servers and clients show the success of the group.
[22:07:53] <hardie> Lisa replies to Julian: this seems like a rat hole, but since it is an individual submission, he can do so if he feels it is needed
[22:07:58] --- Glenn Parsons has joined
[22:08:27] <reschke> mic: DASL - It think we have a lot of buy-in. However we need to encourage the current implementes to help finishing the draft (instead of steadily coming up with new feature requests :-)
[22:08:41] --- michael has joined
[22:09:01] <hardie> Lisa agrees with Julian
[22:09:36] <hardie> Larry says that one way to push things forward is to publish as experimental, to get a stable document, maybe encourage further implmentation
[22:10:04] <reschke> mic: DASL - if there's a broad consensus that leaving this (data typing) out of the spec and publishing it is preferrable, we can certainly do it. But again, I'd like to seee some WG discussion about that before.
[22:10:08] <hardie> Lisa: What is there is a small group Jim, Julian, Lisa, that feels that it is ready, can it go forward on the standards track.
[22:11:23] <reschke> log: DASL issues list is at <http://greenbytes.de/tech/webdav/draft-reschke-webdav-search-issues.html>
[22:11:24] <hardie> Lisa agrees with Julian, but is concerned that broad consensus not possible with current state of the group
[22:11:51] <hardie> Ted suggests that we take the existing DASL work to proposed and put the data typing stuff to experiemental
[22:12:06] --- Lisa D has left: Disconnected
[22:12:10] --- Lisa D has joined
[22:12:12] <hardie> Lisa agrees that might works.
[22:12:38] <hardie> Lisa is typing the issues list into a browser
[22:12:43] <reschke> mic: what's holding us is basically limited time -- we have many drafts to work on, and recently ACL, BIND and REDIRECT got more attention. Once the first two are done, DASL may get more attention again.
[22:13:37] <hardie> Lisa disagrees; it is not limited time, but it is limited participation. When there is two people on one side and two people on the other, there is no way to determine consensus.
[22:14:01] <reschke> mic: DASL - I partly agree, yet I'd recommend to delay that decision until the WG can focus on DASL (that is when ACL is published and BIND is submitted)
[22:14:38] <Lisa D> Can you send the URL again to DASL issues?
[22:15:06] --- galvinjamesm has joined
[22:15:08] <hardie> Larry suggest that publishing the common set might be possible; Lisa responds not possible with BIS
[22:15:09] <Lisa D> I don't feel bind can be submitted. IMO, it breaks locks.
[22:15:26] <SAH> <http://greenbytes.de/tech/webdav/draft-reschke-webdav-search-issues.html>
[22:15:34] <reschke> mic: DASL - I don't think we have that situation. The DASL/datatyping question was discussed by the WG in January 2003 during the interim WG meeting, and the consensus of the participants was that we want the typing stuff in.
[22:16:27] <reschke> mic: Lisa, if you really feel that there's an open issue with BINDs vs LOCKs, please raise it on the mailing list.
[22:16:36] <hardie> Ted interprets this as saying DASL datatyping is not as contentious as BIND would be; Lisa says that is possibly true
[22:16:46] <Lisa D> I have raised it on the list.
[22:16:57] <hardie> I've seen that raised multiple times.
[22:17:09] <hardie> (the last was really my comment, not channeling the room)
[22:17:23] <hardie> is there something on the issues list in particular for us to look at?
[22:17:30] <reschke> mic: BIND/LOCK - yes, it was raised *and* answered. I'll provide a URL.
[22:17:49] <reschke> DASL issue: no -- just informational.
[22:18:32] <hardie> Lisa answers: it was responded to, but her concerns have not been addressed
[22:19:14] <reschke> If her concerns weren't addressed, she should have said so. It's impossible to resolve issues unless we finish discussions.
[22:19:23] --- jlcjohn has joined
[22:20:13] <Lisa D> Ted is explaining his belief that clients need deterministic behavior from servers doing WebDAV locking.
[22:20:25] <Lisa D> Surprising the client by having lock status change -- is a bad thing.
[22:20:50] <Lisa D> If you started to experiment with this you'd be affecting existing WebDAV clients too.
[22:21:00] <reschke> I don't think anybody disagrees with that.
[22:21:32] <Lisa D> Ted: if you deployed this with new methods to access things that are hard links then clients would know when they're not dealing with deterministic behavior.
[22:21:59] --- jlcjohn has left: Disconnected
[22:22:35] <reschke> Lisa: There *is* deterministic behaviour. If you feel otherwise, please describe that case in detail.
[22:23:22] <hardie> Lisa suggests that Ted's word "deterministic" should be "apparent" or "predictable".
[22:23:36] <hardie> The client has always expected the rug to be there; now it can not be there.
[22:23:53] <hardie> Whether the client knows the rug is being pulled out or not, the client is still upset
[22:24:04] <hardie> (That was Lisa)
[22:24:38] <hardie> Lisa suggests that a large group might be possible in a more focused effort on this topic, and then she and Larry decide that is wishful thinking.
[22:24:49] <reschke> Folks, I'm not sure what we're dicussing here. The interaction of BIND vs LOCK is clearly defined by GULP (5.6), all that is open is where to put this information (which isn't new, just a clarification of RFC2518).
[22:25:54] <hardie> Lisa: the disagreement is more fundamental.
[22:26:19] <hardie> When you lock a binding, old clients believe that this blocks the e-tag changing
[22:26:28] <hardie> (ie. blocks resource changes)
[22:26:43] <Lisa D> Yes -- but even though my client has locked one binding, another client can change the underlying resource via another binding.
[22:26:50] <reschke> log: URL for GULP: <http://lists.w3.org/Archives/Public/w3c-dist-auth/2004JanMar/0001.html>
[22:27:42] <hardie> Ted asks whether Jim agrees to the proposal to close the working group; Lisa responds that he is more reluctant, but does agree
[22:27:44] <reschke> mic: Lisa: yet, the other client can do that but only if it supplies the lock token. Where's the issue here???
[22:27:54] <Lisa D> I claim that opening the possibility for the underlying resource to change while a lock is held is an incompatible change.
[22:28:01] --- Glenn Parsons has left: Disconnected
[22:28:17] <reschke> mic: Lisa, yes that would be an incompatible change, but it is not the case.
[22:28:31] <hardie> her understanding is that the lock is on the binding, not the underlying resource
[22:28:42] <hardie> if the lock is on the underlying resource, that is not how the draft reads to Lisa
[22:28:54] <hardie> or her understanding of the explanations given heretofor
[22:29:03] <reschke> mic: WG group closure - I'd like to understand how this would affect the ongoing work of the active working group members.
[22:29:04] <Lisa D> That's not how I've read the draft, nor the explanations Julian and Geoff have given.
[22:29:30] <Lisa D> Work can proceed as individual submissions
[22:29:53] <reschke> mic: In which case Lisa's understanding is partially wrong. LOCKs affect resources (directly) and the namespace (URI mappings) indirectly.
[22:30:37] <reschke> Lisa: then please point to the sections/comments that made you get that opinion, so that we can fix them.
[22:30:42] <hardie> People may be happier if there is no working group
[22:30:47] <hardie> (Lisa)
[22:31:31] <reschke> mic: I'd like to see a WG dicussion of the pros/cons of closing the WG.
[22:32:20] <hardie> Patrik: when you are going to close the working group you look at the documents that are the working group items. You don't close the mailing list, and you decide on each of these how to proceed
[22:32:21] <reschke> mic: right now we have a small but steady flow of document submissions, so the WG provably is doing it's work (although slowly)
[22:32:21] <Lisa D> Of course. That's what we're having.
[22:32:31] <Lisa D> (a WG discussion of closing the WG)
[22:32:44] <hardie> Lisa responds that they are getting very narrow review
[22:33:08] <reschke> Lisa: the WG is the set of readers of the mailing list.
[22:33:41] <hardie> The WG is the active participants on the mailing list (Ted responds)
[22:35:15] <hardie> Patrik: it is the WG which decides whether a document is a work item
[22:36:44] <hardie> Patrik: update the charter to reflect the things that actually will be done, get the working group to agree whether it can resolve something that is a milestone.
[22:37:14] --- jlcjohn has joined
[22:38:18] <hardie> Ted: digression: a new charter requires new review
[22:38:34] <hardie> by the IETF community
[22:40:13] <hardie> Lisa: not much more we can accomplish here; we have to go back to the list to complete the discussion.
[22:40:25] <hardie> Larry says "let's go through the WG status on goals one more time"
[22:40:34] <hardie> Abandond property registry
[22:40:58] <hardie> Lisa: now hearing something new on Binding; republish bindings draft, get review
[22:41:37] <hardie> query: is the bindings draft expired?
[22:41:51] <hardie> Larry is there a chance to get some of these things out?
[22:41:56] <reschke> mic: No, the binding draft is up-to-date with one open issue.
[22:42:25] <reschke> URL: <http://greenbytes.de/tech/webdav/#draft-ietf-webdav-bind>
[22:42:38] --- galvinjamesm has left
[22:42:56] <hardie> Lisa: possible,
[22:43:01] <reschke> (last version dated December)
[22:43:09] <hardie> Bindings issues cover move and copy as well
[22:43:18] <hardie> may not be as critical
[22:43:38] <reschke> mic: there are no open issues regarding MOVE/COPY.
[22:43:44] <hardie> redirect references could go to standard
[22:44:03] <hardie> Lisa: not raised in the face of belief of blocking lock issues
[22:44:26] <hardie> If the other is not an issue, may need to go through those as well.
[22:45:01] <hardie> blue sheets interrupt
[22:45:04] <reschke> mic: I'd like to ask everybody to actually raise issues with published drafts. The authors maintain official issues lists, and as long as it's not on the issues list, nobody will take care of it.
[22:46:29] <reschke> mic: re BIND -- I think the right way to proceed is to close the remaining open issue and then to last-call the document, so that it gets the attention it deserves.
[22:47:47] <hardie> draft standard status discussion: we've collectively put a lot of work into that draft, feeding the results of those interop events into the draft
[22:48:00] <hardie> now discussion of doing something completely different, though.
[22:48:22] <hardie> Maybe people feel that more playing around with Webdav is necessary
[22:49:00] <hardie> Larry asks why abandon the path of going to draft?
[22:49:25] <hardie> Lisa: personal opinion is that it is very hard to get consensus
[22:49:57] <reschke> I don't think we should abandon that plan. We just need to be careful not to put in new stuff that is controversial.
[22:49:59] <hardie> Interop folks not being heard here.
[22:50:26] --- paf has left
[22:51:20] <reschke> (for the log: latest redirect ref draft & issues at: <http://greenbytes.de/tech/webdav/#draft-ietf-webdav-redirectref-protocol>)
[22:51:23] <hardie> issue raised of Lisa needing a light touch on the bis draft because she is an author
[22:51:57] <hardie> Larry suggests forcing the issues into the light
[22:52:48] <hardie> Larry says bugging people maybe necessary--not waiting for people to respond, but going out and getting them involved.
[22:53:43] <hardie> Anyone want to volunteer to do reviews of bis and bind?
[22:55:59] <Lisa D> Ted's recommending to do last call on some of our remaining docs, but to require N independent reviews to be posted in order to successfully move on.
[22:56:29] <Lisa D> Without those N reviews -- e.g. 3 independent reviews -- Ted says it would no longer be considered a WG item.
[22:56:57] <Lisa D> Ted: it's up to the chairs to determine if the participation of the WG is sufficient to call consensus and come to completion.
[22:57:23] <Lisa D> Ted: but before coming to that state, Ted suggests the chairs call for a minimum # of reviews.
[22:57:34] <Lisa D> Ted: then the reviews also indicate whether another cycle is needed on the document.
[22:57:44] <reschke> mic: note that BIND already went through WG last call three years ago, the current spec is a major rewrite in order to simpliy and to address LC comments (similar with REDIRECT)
[22:57:58] <hardie> Lisa: thanks for attending; if you want to talk to her later, she's here all week.
[22:58:02] <hardie> Thanks!
[22:58:10] --- SAH has left
[22:58:19] --- hardie has left
[22:59:17] --- reschke has left
[22:59:55] --- jlcjohn has left
[23:06:22] --- Lisa D has left: Disconnected
[23:13:57] --- Lisa D has joined
[23:19:33] --- anewton has joined
[23:22:59] --- anewton has left