W3C home > Mailing lists > Public > public-bpwg@w3.org > March 2009

RE: Raising a new issue on the need for a statement on the roles and mutual respect of author/user/transformer.

From: Rotan Hanrahan <rotan.hanrahan@mobileaware.com>
Date: Thu, 12 Mar 2009 15:00:58 -0000
Message-ID: <D5306DC72D165F488F56A9E43F2045D301E9FE3C@FTO.mobileaware.com>
To: "MWI BPWG Public" <public-bpwg@w3.org>
Hello Luca,

Very interesting to be engaging in a substantive dialogue with you after
so long a time. Let me respond a little to the points you raise. I
regret I cannot give them the full elaboration that they may require, as
unfortunately I am pressed for time.

> To which extent does the transcoding craziness jeopardize the concept
of 
> the web as a whole?

By "transcoding craziness" I assume you mean the presence of transcoding
solutions that may appear to be operating without clear and respectful
guidelines regarding how they treat the content of Web authors and the
browsing consumers of content. It is a matter of public record that the
introduction of transcoding proxies in some cases has resulted in some
less-than-optimal experiences, both for content producers and for
consumers. It is certainly possible for transcoders to provide their
services in a manner that will not result in such an experience, and
thus could achieve their stated objective of improving the general Web
experience (for mobile users). It is my understanding that this is what
the CT guidelines intend to facilitate, just like other efforts in the
mobile community, which are intended to reduce/remove the "craziness".
One could argue that the introduction of these "guideline" initiatives
was in response to a perceived threat to the concept of the Web,
ensuring that the clever technology within transcoders would only be
used to improve the Web experience (for everyone, not just for the end
user). The guidelines are not complete, and not every transcoding
provider has agreed to abide by the emerging guidelines. Bodies such as
W3C cannot demand compliance, but a strong enough argument that shows
how particular guidelines can benefit everyone should encourage
compliance. There may still remain some who will shun these guidelines,
preferring instead to persist with the "craziness", interfering with the
author-user relationship in ways that are probably motivated by
commercial concerns rather than any concern for the Web itself. The
"craziness" could jeopardize the Web, but (properly managed) transcoding
does not have to be seen as a major risk.

> The foundation of the web is that individuals and companies invest
time 
> and money to create content because that content will be accessed by 
> end-users.

Indeed. Creating content without having to worry too much about how your
audience will access or render that content is one of the big benefits
of Web publishing, assuming that you are confident that your content
will be treated with respect. The Web experience is so much better
because you know that the time (and money) you invest will have
beneficial results.

> In the case of transcoders, we are talking about network elements
that, 
> from one day to the next, become alive and start playing tricks with
the 
> content that travels in the network!!!

Transcoders that suddenly and unexpectedly appear and start undoing the
effort you have invested in creating your content are unwelcome.
However, if you have not put in effort to facilitate a good user
experience in a particular delivery context, it should be comforting to
authors to know that it may yet be possible for the user to view your
content and get a good experience. And this would be at no expense to
the author. However, providing this "gap filling" service may involve
costs, and today these costs are primarily covered by the users. Indeed,
the provision of the service may be profitable. But when the technology
is pushed further purely for commercial reasons (such as the
injection/removal of advertising), it begins to look like the proxy is
behaving like a content author, and now the author/user relationship has
been affected. The proxy is not respecting the intentions of the
original content author, but instead is encroaching on the author's
role. When the relationship between the original author and the end
consumer has been affected in this way, the Web suffers.

> I mean, are you all blind? or crazy? can't you see how much confusion 
> and disruption this "technology" can bring to the way the Internet and

> the Web have worked all of these years?

Are we blind to the issues? Are we crazy? I cannot speak for others, but
I hope nobody thinks I am blind to the issues, and I hope you are not
questioning my sanity. Unmanaged, without guidelines, the transcoding
proxy technology has the capacity to create confusion and disruption.
Properly managed and with well-reasoned guidelines, the technology has a
role to play, and the Web can benefit. Hopefully the guidelines will be
well-received and adopted by the technology providers, and hopefully the
guidelines will mandate the respect and fairness that I suggested when I
raised the issue.

> Anyway, there seem to be agreement that transcoders are a "temporary" 
> and "transitional" technology. In this case, you don't need a 
> recommendation or a spec. Just guidelines which may even happen
outside 
> of W3C. If you want W3C to make these guidelines happen, there is only

> one logical thing to do: demand that transcoders do the maximum effort

> possible to preserve authors' intentions, and err on the side of not 
> transcoding.

The current batch of transcoding solutions may be
temporary/transitional, but I think the Web will always have layers of
advancement, so there will always be some subset of the Web community
that will struggle to gain access to certain content, and won't have the
benefit of a content author who has taken, or can/will take, their needs
into consideration. For this reason, I expect that there will always be
a role for some intermediate adaptation, and this makes it all the more
important for the Web to have proper guidelines on how proxies can be
"good citizens" of the Web. Among those guidelines, as I have previously
said, should be a requirement to keep improving the analysis of content
sources, and not just rely on a few known heuristics. I agree with you
that transcoders should do the maximum effort in this regard to discover
authors' intentions. However, in the absence of any evidence to the
contrary, I still believe that publication via the Web is done under the
assumption that "Web-like things" will happen, and so a certain amount
of adaptation may take place anyway. That does not give anyone free
license to edit/elide portions of content at will, as this would not be
respectful of the publication process per se. This goes beyond the Web;
it is at the heart of how society relates to intellectual property.

>> Authors, users and transformers should cooperate to help the Web
reach
>> its full potential, specifically: all can publish, all can consume.
> this can only happen by "enabling" authors, not by bastardizing their 
> content behind their back (with a good injection of operator-managed 
> banner ads in the someone else's content, please)

Enabling authors can mean giving them the means to adapt (like
MobileAware does today with our various products) or providing them with
the means to influence what other can do to their content (e.g. the HTTP
No-Transform/Vary headers, POWDER, etc.), which would only work if the
authors are assured that the guidelines are respected.

Speaking personally, I would be concerned about someone altering my
authored content, making it look like I was the one providing that
content that in fact had been inserted by a third party. That includes
inserting advertising into my content. I like the idea mentioned by
Bryan recently: "editrights" (viz. copyrights), which might give me a
way to advise current and future proxies what I consider to be
reasonable with respect to subsequent adaptations of my content. It
might even be possible for me to delimit subsets of my content to ensure
that future adaptation takes my wishes into consideration, which is
possibly one use case for the W3C's XHTML Role technology. If I had my
own adaptation technology (which, interestingly, I do!) and a means for
adding metadata (Headers/Role/RDF/etc.) to influence 3rd-party
adaptation, then I would feel "enabled", to use your word. We are not
yet at that position because we haven't figured out the metadata, nor
have we agreed the guidelines for proxies to follow in the
presence/absence of such metadata, even though a lot of progress is
being made.

---Rotan.

-----Original Message-----
From: public-bpwg-request@w3.org [mailto:public-bpwg-request@w3.org] On
Behalf Of Luca Passani
Sent: 12 March 2009 12:54
To: MWI BPWG Public
Subject: Re: Raising a new issue on the need for a statement on the
roles and mutual respect of author/user/transformer.


Rotan, I think the issue you raised fails to consider one basic aspect:

To which extent does the transcoding craziness jeopardize the concept of

the web as a whole?

The foundation of the web is that individuals and companies invest time 
and money to create content because that content will be accessed by 
end-users.

In the case of transcoders, we are talking about network elements that, 
from one day to the next, become alive and start playing tricks with the

content that travels in the network!!!

I mean, are you all blind? or crazy? can't you see how much confusion 
and disruption this "technology" can bring to the way the Internet and 
the Web have worked all of these years?

Anyway, there seem to be agreement that transcoders are a "temporary" 
and "transitional" technology. In this case, you don't need a 
recommendation or a spec. Just guidelines which may even happen outside 
of W3C. If you want W3C to make these guidelines happen, there is only 
one logical thing to do: demand that transcoders do the maximum effort 
possible to preserve authors' intentions, and err on the side of not 
transcoding.

 > Authors, users and transformers should cooperate to help the Web
reach
>  its full potential, specifically: all can publish, all can consume.

this can only happen by "enabling" authors, not by bastardizing their 
content behind their back (with a good injection of operator-managed 
banner ads in the someone else's content, please)

Luca


Rotan Hanrahan wrote:
>
> Jo has invited me to raise an ISSUE, so that something more specific 
> relating to the CT document can be considered. I have already 
> presented the background to my concerns in [1], and I now add this 
> email to raise the ISSUE, which I hope will be added to the Issue
Tracker.
>
> = Summary =
>
> Absent some indication from the author of the original content, it is 
> fair to presume that an author publishes content via the Web knowing 
> that "Web-like things" may happen to that content, but also having an 
> expectation that such Web-like things would be "reasonable". Where the

> author has provided some indication of how the content should be 
> treated, this should be respected. Nevertheless, it is also fair to 
> assume that publication via the Web has an implicit intent of 
> publication to the widest possible audience. Where there are users 
> within that Web audience who, for whatever reason, have trouble 
> consuming the content, the architecture of the Web provides for 
> remedial action via a transforming proxy. That proxy must take into 
> account the gap between what the author provided (and intended) versus

> what the user can consume (and prefers). Authors, users and 
> transformers should cooperate to help the Web reach its full 
> potential, specifically: all can publish, all can consume.
>
> = Issue =
>
> = = =
>
> ISSUE: The CT document is missing some statement recognising the role 
> and expectations of the main parties (author, consumer, transformer) 
> and the need for mutual understanding and respect of the others' 
> needs/expectations. Perhaps also some suggestion from the BPWG on how 
> to prioritize the different needs/expectations would be useful, as a 
> general principal, especially given that there will be conflicts to 
> resolve.
>
> = = =
>
> = Illustration =
>
> I think this issue could be addressed by a sentence or two in the 
> introduction, to which we can refer when consider the technical 
> matters elsewhere. For example, with the principle of mutual fair 
> respect, one could examine the technical suggestion (4.1.5.3) for 
> "User Selection of Restructured Experience" and ask questions like:
>
> - Does this respect the efforts of an author to provide alternative 
> representations?
>
> - Does this respect the need of a user to obtain alternative 
> representations?
>
> - Does the proxy fairly take into account the gap between author and
user?
>
> - Is it fair to allow the user to select the proxy's representation 
> over that of the author?
>
> In the above example, the first three points appear to be in keeping 
> with a principle of mutual respect. The last point shows the potential

> tension between author and user, but the CT guidance suggests that it 
> is the user who makes the decision (not the proxy) and we should also 
> consider the assumption of "Web-like things" happening to what you 
> publish, in which case the user's choice prevails. Thus 4.1.5.3 
> appears to be reasonable.
>
> However, an explicit indication from the author that transformation 
> not permitted would show that the author was not making an assumption 
> regarding Web-like things happening. Perhaps in this case the author's

> representation should prevail. Personally I would find such an 
> attitude by the author to be narrow and short-sighted because the 
> author would be intentionally limiting the Web audience, ignoring what

> is technically possible today and what might be possible in the 
> future. Unfortunately, the existing mechanisms (e.g. HTTP 
> No-Transform) are not expressive enough. I'd like there to be a HTTP 
> Avoid-Transform, so that there's flexibility to do the right thing 
> under circumstances we can't predict, but I know that creating new 
> technology is out of scope.
>
> = Closing =
>
> We in MobileAware have considered what this means for our own 
> products/services. We aim to give an excellent end-user experience 
> regardless of the browsing device, and we hope that any intermediary 
> would respect our efforts by not transforming our content unless the 
> user's circumstances strongly demand it. At some time in the future 
> there may be some unanticipated delivery context (e.g. the spherical 
> display of a holographic browser) and rather than deny people access 
> to our content via such an amazing device we would accept an 
> intermediary stepping in to do additional adaptation. Indeed, we might

> even work directly with an intermediary to help them adapt our content

> to the niche market of spherical holographic displays. It's all about 
> mutual respect.
>
> ---Rotan.
>
> [1]
_http://lists.w3.org/Archives/Public/public-bpwg/2009Mar/0073.html_
>
Received on Thursday, 12 March 2009 15:01:41 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:43:00 UTC