See also: IRC log
francois: [review of what
happened at the F2F and pointer to summary]
... jo proposed an editor's meeting ... I can host at Sofia (SE
France)
... who would be able to come?
... dates would be next week or the week after
<andrews> No, sorry
<SeanP> I don't think I'll be able to make it
jo: needs to be before Easter
francois: perhaps we can wait an extra week?
heiko: could we do it in the UK?
jo: need a host in the UK
... if noone else wants to attend then Francois and I can sort
this out between us
PROPOSED RESOLUTION: Francois and Jo to roganise work as they see fit and present group with final pre-FPWD
<andrews> +1 (and thank you)
<rob> +1
<francois> +1
RESOLUTION: Francois and Jo to roganise work as they see fit and present group with final pre-FPWD
jo: the only bits that open to
discuss in that context are the bits after 3.1.4
... following resolution at F2F
<francois> Aaron's contribution
francois: this is content for 2.5
control of the behavior of the proxy
... goes in the same direction as we agreed to go
... maybe inconsistent with other things we say
... question is how should proxy indicate it could
transform?
... if the content has not been modified then the proxy should
indicate that a transformed version is available
... how should this happen?
... I don't see how this can happen without transforming the
page
magnus: there could be an insterstitial page
francois: yes, I suggested that but it would be a lousy user experience
heiko: i think it should be only done once
seanP: you could insert a little link
francois: yes, I noted that in response to aaron, but that would imply inconsistent behavior
seanP: one possiblity is some user agreement that it you wanted it you could get it
francois: the list of options
should contain this as an option that the user can set
... sometimes you are looking at a page and you wish you could
transform it or not
... if the server issued a cache-control no-transform the link
would be missing
... the proxy MAY indicate that a transformed version is
available only where it decided not to transform
seanP: further on your point about having a link sometimes and sometimes not might be confusing, I am not really sure that would be a bad user experience
rob: there is the issue about
confusion and there is also the question about wanting to split
the page and in those circumstances the user might want to
transform in this case
... so its worth saying you can switch it off and on but saying
how is probably out of scope
heiko: can we differentiate three things - leave things completley untouched, second, just headers and footers
francois: we are trying to allow
the server to have a "switch off" and if we allow adding a link
that is something to be avoided
... we could complete Aaron's text to say that this should not
be done to pages that can not be transformed
<francois> PROPOSED RESOLUTION: keep aaron's text as it is for 2.5.1
<francois> current editor's text
jo: Don't think we should discuss adding a link to the unmodified content to say that it could be modified
<Magnus> Transforming proxies SHOULD provide to their users an indication that the content being viewed has been adapted for
<Magnus> mobile presentation. If the content has not been modified, because of
<Magnus> server or user preferences, but transformation is possible and
<Magnus> potentially useful, the proxy MAY indicate that a transformed version
<Magnus> is available.
jo: that could appear later in the document - this section is an overview
francois: take jo's point that
whatever we resolve on the second bit it could go later in the
document
... magnus, would you leave out the bit on saying that the
content can be modified
magnus: my main problem is that
the resolution can't be left as is, as it has a number of
comments
... I'd be happy with just taking the first sentence and then
saying that the original content can be retireved without
saying how exactly
... could also have the bit on sticky sessions
... I'm trying to look at this in the context of the document
which is an overview
... [discusses how it can be worded]
<francois> Transforming proxies MAY provide to their users:
<francois> - An indication that the content being viewed has been adapted for
<francois> mobile presentation.
<francois> - An option to view the original, unmodified content.
<francois> - some "sticky" preferences to their users.
francois: this is to emphasize the point that this is just a discussion
seanp: we have talked about providing a link to unmodified content - what to do if the unmodified content crashes the phone?
francois: good point but does not
remove the need for a link in the generic case, the proxy could
offer a warning
... in that case
<francois> Transforming proxies SHOULD provide to their users:
<francois> - An indication that the content being viewed has been adapted for
<francois> mobile presentation.
<francois> - An option to view the original, unmodified content.
<francois> - some session settings to their users.
jo: I'd prefer to leave the
normative wording even though this is less of a prescriptive
section, otherwise there's a danger that the document may
appear to contradict itself when it goes on to discuss it
later
... I think it's SHOULD on the first two and a MAY on the
third
<francois> Transforming proxies SHOULD provide to their users:
<francois> - An indication that the content being viewed has been adapted for
<francois> mobile presentation.
<francois> - An option to view the original, unmodified content.
<francois> They MAY also provide session settings to their users.
+1
<andrews> +1
<Magnus> +1
<scribe> ACTION: Jo to reword 2.5.1 along the lines proposed by Francois [recorded in http://www.w3.org/2008/03/11-bpwg-minutes.html#action01]
<trackbot-ng> Created ACTION-706 - Reword 2.5.1 along the lines proposed by Francois [on Jo Rabin - due 2008-03-18].
<SeanP> +1
francois: additional qeustion, should we mention examples of session settings
magnus: that would be OK
francois: relates to the list of
directives we removed from the document
... seems to me to be useful
magnus: examples being: never transform no matter what, only transform if absolutely necessary, transform as much as possible and so on
jo: that is similar to the list we will try to codify in POWDER as server preferences
<scribe> ACTION: Jo to include examples in 2.5.1 bullet 3 per the dicussion above [recorded in http://www.w3.org/2008/03/11-bpwg-minutes.html#action02]
<trackbot-ng> Created ACTION-707 - Include examples in 2.5.1 bullet 3 per the dicussion above [on Jo Rabin - due 2008-03-18].
heiko: I expect POWDER is for the
device capability database
... it's a follow on to WURFL?
jo: no
... we have been looking at it as a way of describing server
preferences
heiko: we need to capture the dependences on other projects, we need to understand device capabilities
francois: we don't want to depend
on POWDER
... and don't want to be delayed by it
... or indeed the DDR
... on 2.5.2
<francois> Transforming proxies MUST provide support for control over the content
<francois> transformation process by origin servers.
<francois> These control mechanisms are detailed in section 3 (Behavior of Components).
<hgerlach> sorry quys, I have to leave for the next call:-(, cheers Heiko
<andrews> +q
andrew: are we only referring to no-transform here?
francois: the controls are in section 3, at the moment it is true that no-transform is the only control we have but maybe there will be more control by POWDER later
jo: points out that vary and warning transformation applied are both means of control
francois: I don't think we should list warning transformation applied as a means of control
jo: agreed
<scribe> ACTION: Jo to update 2.5.2 in accordance with discussion and Seoul resolution on preferences [recorded in http://www.w3.org/2008/03/11-bpwg-minutes.html#action03]
<trackbot-ng> Created ACTION-708 - Update 2.5.2 in accordance with discussion and Seoul resolution on preferences [on Jo Rabin - due 2008-03-18].
francois: aaron leaves 2.5.3 untouched, do we need to say more
<SeanP> This is another place where some examples may help.
jo: not sure
francois: it's a bit useless as is so examples would help
<francois> ACTION: daoust to write some examples for 2.5.3 [recorded in http://www.w3.org/2008/03/11-bpwg-minutes.html#action04]
<trackbot-ng> Created ACTION-709 - Write some examples for 2.5.3 [on François Daoust - due 2008-03-18].
jo: I'll try to update the doc in the next couple of days
<francois> Close ACTION-666
<trackbot-ng> ACTION-666 Draft section 2.6 listing user control options that SHOULD be supported closed
[adjourned]