RE: CT problem statement

Hello Nigel, 
 
Thanks for your response.
 
I guess I'm a bit more optimistic about the participants in the task
force. I think we have people who represent users, content developers and
mobile operators, and we each have views on what we need to be able to do
to support the constituencies we represent. For example, 
my experience is that the adaptation server providers do provide a
reasonable proxy for content providers. Content providers are the main
users of such technology. One of the difficulties I've noted in nearly 7
years working within W3C is that content providers tend to feel that it is
not appropriate to participate directly in discussions that they perceive
as highly technical. That is a shame, but is understandable.
 
In terms of outputs from the task force and discussion, while the results
will necessarily be a set of technically oriented guidelines, there is no
reason why subsequent, public discussion might not raise non-technical
matters in criticising the materials that we produce.
 
As to the particular question you raise about masquerading, my personal
view is that there are perfectly legitimate use cases in which this
behaviour is needed. The issue seems to me to be in making it the default,
and in not having some intelligence in the communication so that other
equally legitimate use cases, such as delivery of the original desktop
site or delivery of an optimised mobile experience, are prevented.
 
That's the area we're concentrating on in the TF. We are currently
collecting material associated with the guidelines that we will produce.
We hope that by appropriate use of existing HTTP headers, the various
players in the delivery chain, including the browser, proxies and origin
servers can announce their capabilities in such a way that each knows when
to provide additional processing and when to step aside.
 
Taking my TF leader's hat off, from a personal perspective, I'm with you.
My day job is all about helping people create optimised mobile content and
the last thing I want is a proxy attempting to transform materials already
appropriate for the device.
 
Best wishes
Rhys 

  _____  

From: Nigel Choi [mailto:nigelchoi@yahoo.com] 
Sent: 26 September 2007 02:49
To: Rhys Lewis
Cc: public-bpwg-ct@w3.org
Subject: Re: CT problem statement


Hi Rhys,

Thank you for your detailed reply, and I appreciate the attempts in
incorporating feedbacks from me.

Perhaps what I said about manipulation is hyperbole (the conspiracy
theorist in me is getting the the best of me :) ). But it appears that
those who engage in this Task Force initially are mostly Operators and
Content Transformation Server vendors. Whether intentional or not, one
would think from the standpoint that they are most familiar with, while
neglecting the needs of others that are affected by the technology. In
this case, it's easy for Content Transformation Server providers to
justify their actions for the sake of "providing desktop information",
while neglecting the consequences of their decisions and actions on Mobile
Content Providers and Users. I can just see that spelled out from the
latest draft sent by Sean Patterson. Your list of interested parties are
"browser providers, transforming proxy providers, adaptation server
providers," but it needs to be much wider. Everyone, from Content
Providers, Network Operators and Users are affected by Content
Transformation. Following this thought, I can understand why Jo and others
are calling for participation here.

I don't deny the value of Transformation in advancing and encouraging the
use of the mobile web. The issues of Content Transformation has been in
the back of my mind for a while. Back when I was working for a Mobile
Content Provider, we were receiving lots of traffic from the Google Web
Transcoder to our already mobile-optimized servers.  I've  always wondered
if there are ways to get around it (there is now). On the other hand, I
use various Transcoders to read desktop-only sites when I am on the go.

It's true that Transcoders has been masquerading the User-Agent header
since this technology is invented. But is that the right way to do it?
That all existing transcoders are doing it doesn't necessarily mean it is
right, and the recent Vodafone controversy just puts the issue up front
and center. The context in which this is done needs to be carefully
examined, and I think it may involve more than a pure technical
discussion.

Regards,
Nigel.


----- Original Message ----
From: Rhys Lewis <rhys@volantis.com>
To: Nigel Choi <nigelchoi@yahoo.com>
Cc: public-bpwg-ct@w3.org
Sent: Monday, September 24, 2007 12:13:36 AM
Subject: RE: CT problem statement


Hello Nigel,
 
As chair of the Content Transformation Task Force within the Mobile Web
Initiative Best Practices (MWI BP) group at W3C, I'm naturally very
interested in your comments. Can I just point out that Dan Appelquist is
one of the joint chairs of the MWI BP, along with Jo Rabin. He is not
actually chair of the task force.
 
I should say at the outset that I don't recognise any manipulation of the
task force by anyone, including Vodafone. I would be interested in knowing
why you feel that this is, or may be happening. This may simply be a
failure of communication on my part. The task force is relatively new.
 
You are of course correct in that standards activities take time. However,
I'm very encouraged by the level and breadth of participation in the CT
task force. The problem of ensuring that the various software entities,
that may be involved in the delivery of material from servers to clients,
know about each other and are able to behave appropriately, has been known
about for some time. Only recently, however, have the technologies
actually been deployed, leading to the kinds of difficulty that we have
seen this summer on the Web.
 
The solution to the problem involves not only origin servers that can
perform adaptation and transforming proxies, but mobile browsers too.
Ideally, I would like to see a situation where users can choose whether to
receive an optimised mobile view of a particular site, or the original
desktop version. The transformations necessary to receive the desktop
version might happen in a proxy, or, as is increasingly the case with
modern, mobile browsers, in the device itself.
 
There is a real tension between ensuring that the needs of authors who
create optimised, mobile sites are met while also meeting the needs of
users who want to be able to see desktop materials if they choose. The
content transformation task force is trying to address this tension. The
good news is that all of the interested communitites (browser providers,
transforming proxy providers, adaptation server providers) are engaged.
The even better news is that there is a feeling that by simply using
existing HTTP faciltites, we may be able to provide a solution which could
be quickly implemented. It is early days, and only time will tell.
 
The W3C operates by consensus. Only when agreement is reached within the
CT task force, and within the entire MWI BP,  will any work be published.
At that stage, the entire membership of the W3C and the public have the
right to challenge the work formally. Not until all comments have been
properly addressed can the work proceed to the point of being a
recommendation. This is part of the reason why standardisation takes time,
but it is a vital part of a process that ensures that all interested
parties have the right to participate and to influence the outcome of the
work. My personal experience is that this is a very open and transparent
process. My expectation is that we will be able to provide a solution that
is acceptable to all. 
 
Can I just add my thanks to you for providing the specific, substantive
comments that Jo requested in a separate e-mail exchange. My personal
feeling is that the guidelines will be primarily a technology document, as
that is generally the purpose of W3C work. The problem statement may well
include discussion of the rationale for the technology, and specific use
cases. 
 
Very best wishes
 
Rhys Lewis, Chair MWI BP Content Transformation Task Force