See also: IRC log
<trackbot-ng> Date: 02 October 2007
<jo> Meeting: Content Transformation Task Force
<jo> scribenick: SeanPatterson
<jo> http://www.w3.org/2005/MWI/BPWG/Group/track/products/12
ACTION-550
Feedback from several people
<jo> Jo: Comments on how to take proposed text forward
<jo> SeanP: Have a response prepared to your comment (Jo) and am preparing responses to others
<jo> ... how do we handle this usually
<jo> jo: not unusual to have divergent views and not unhealthy at this stage
<jo> sean: will post responses with a view to driving forward on some consensus
<jo> ... will try to get stuff out over this week
Jo: Aaron has made a contribution
on 9/22
... It was an outline of section 2.3
... No comments yet so we can't do anything yet.
Andrew: Apologize for tardy response. Has some notes.
Jo: Go ahead and send the notes.
Andrew: Aware of amout of
discussion of topic.
... Will get contribution in by next conf. call.
Jo: Aware of discussion as well.
Jo: On me. Can close that.
Jo: On Andrew.
Andrew: Sometimes don't get
actions sent to me.
... Sometimes doesn't get each email to mailing list.
Jo: Looks like there is a problem on Andrew's end.
Andrew: Will check archives for
email messages.
... Has raised issue; no contributions yet.
Jo: Can't close yet--mark it as pending.
Andrew: Sent an email on this action on 9/25.
Jo: Not showing up on
Trackbot.
... It did show up on Trackbot.
Jo: Have done half of this.
Jo: On Rhys, he did that
<jo> ACTION: Jo to find out what has happened to the zakim slot [recorded in http://www.w3.org/2007/10/02-bpwg-minutes.html#action01]
<trackbot-ng> Created ACTION-572 - Find out what has happened to the zakim slot [on Jo Rabin - due 2007-10-09].
<jo> (Jo notes that we are not quorate for a decision but useful to discuss)
<jo> (Andrew notes debate and noted that there is a body of opinion in favour of preserving original user agent, asks Sean for comment on the problems with that)
<jo> Sean: main problem is that lots of web servers that return nothing or "please upgrade browser if you send the original user agent
<jo> Andrew: any idea of how many sites do this
<jo> ... i.e. either we have a problem with blocking sites or with mobile sites whichever way we do it so it would be useful to know some numbers
<jo> Sean: ... the long tail of the Web won't change, whereas mobile sites are assumed to be still under maintenance and so greater possibility of change
<jo> ... I will ask for whatever data on this
<jo> Andrew: It would be useful if other WG colleagues could get a fix on this
<jo> ... I note that dotMobi is doing "find" has dotMobi got any data on this
<jo> Jo: yes, lots, so has Google I expect, I'll see what I can do, may be in confidence if I can get anything
<jo> s/I'll see what/
<jo> Andrew: wonder if we should use non-standard headers at all
<jo> ... should we use the Via instead of X-Mobile-Gateway
<jo> Sean: the problem is that the comment field can be stripped from Via
<jo> ... I know that Jo proposed a way of using Via, but there could be quite a lot of info we want to add
<jo> ... which would be hard to do using Jo's proposal
<jo> ... e.g. multiple transformations servers - etc.
<jo> jo: I know we have been talking specifically at the HTTP header level, necessarily because that's what a lot of the discussion is about, but I think we should actually focus our efforts on developing Magnus's work - i.e. first figure out who is saying what to whom, and then figure how that maps to HTTP headers, existing or otherwise
<jo> sean: i agree with that approach
<jo> jo: I think we can state in principle we prefer not to invent new HTTP headers but can't decide for now whether that is achievable or not
<jo> ... I suggested 8 different use cases for analysis of who needs to say what to whom
<jo> ... but better to have that discussion by email
<jo> Sean: Would prefer Magnus to be present
Andrew: Jo, could you explain your use of Vary header--thought that i could only be used in the context of cache
Jo: Can use Vary header for
CT
... Says that server has variance of a particular URI
... is useful for saying that a server is an adapting
server
... certainly does no harm of including it
Andrew: I think you are correct
Jo: At dotMobi we have suggested
that people use it.
... If you masquerade the user agent and got back a Vary
header, you know that you are in trouble
... We've held off (at dotMobi) in making recommendations
because of this group
<jo> Sean: I agree that the Vary header can be used for that purpose
<jo> ... it's a reasonable use of it to signal that there are different representations
<jo> [jo: at dotMobi we wanted the actions of the trancoding proxy to be figured out in this group]
<jo> [none]
<jo> [call closed at 15:52 BST]
This is scribe.perl Revision: 1.128 of Date: 2007/02/23 21:38:13 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/rrrsagent, make logs public// Succeeded: s/27/22/ Succeeded: s/It/... It/ Succeeded: s/556/566/ Succeeded: s/can't talk about this./he did that/ Succeeded: s/iwll/will/ FAILED: s/I'll see what// Succeeded: s/I'll what/I'll see what/ Succeeded: s/gigure/figure/ Succeeded: s/google/Google/ Found ScribeNick: SeanPatterson Inferring Scribes: SeanPatterson WARNING: Replacing list of attendees. Old list: jo SeanP Andrew New list: Magnus_Lonnroth Default Present: Magnus_Lonnroth Present: Jo SeanP Andrew Regrets: Rhys Agenda: http://lists.w3.org/Archives/Public/public-bpwg-ct/2007Oct/0002.html Found Date: 2 Oct 2007 Guessing minutes URL: http://www.w3.org/2007/10/02-bpwg-minutes.html People with action items: jo[End of scribe.perl diagnostic output]