Re: [CTG]: Novarra going for mobile sites too

Charles McCathieNevile wrote:
>
> Luca, which features exactly do you consider the abusive ones? 
> Inserting advertising,

obviously abusing, if done without the consent of the content owner.

> collecting some user data, 

abusive. In italy it's illegal. You can legally only collect the data 
for that session, unless the user has approved explicitly that you 
collect their data.

> compressing code? 
this I can be OK with, as long as it happens at a low enough level, does 
not break HTTPS and does not make any difference at the higher levels.


> Is there something wrong with their privacy policy, or is it the fact 
> that they provide a service that reduces image resolution?

reducing image resolution is a potentially abusive move, particularly 
with mobile-optimised sites which have already made an effort to optimize.

>
> As the article you reference says, the interesting question is whether 
> they can avoid pissing off users. The rest of what they do is simply a 
> service they make available, and people can take it up or not.

as I have argued in the past, this is not the case. People buy mobile 
devices and choose operators for different reasons. Internet 
connectivity is a deciding factor only for a tiny fraction of users. 
They'll use data services because they discover them on their devices, 
not because they have been proactive in acquiring them. So, no, people 
cannot really decide (and that's already assuming a user knows what is 
going on). Too many other services, binding contracts and voice-call 
rates are attached to their device/operator.


> As I understand it nobody is being forced to use this service under 
> some unreasonable arrangement.

the problem here is that the content owner is the one who gets the 
damage. I have said multiple times that supporters of transcoders are 
ignoring the rights of content owners and keep repeating "user, user, 
user". This is wrong. Content owners have rights too and must be 
respected. This point I raised several times and nobody ever challenged 
it (they just ignored it and kept arguing about users)


>
> I have not seen any convincing rationale to explain a moral difference 
> (i.e. whether or not it something is "abusive", or "a hack") between 
> me zooming to and reflowing a paragraph or choosing to use "mobile 
> view" with Opera mini, choosing to set a minimum font size in my 
> browser, using it to apply a client-side user style sheet or 
> userJS/Greasemonkey script to a specific site, opting to use someone 
> else's configuration that does the same thing, paying for a service 
> that does the transformation for me, using a service that selectively 
> chooses what content and presentation to send to me on the server 
> side, and simply accepting the default rendering provided by my choice 
> of browser for a single version of content authored and served to all 
> comers.
>

As I have written multiple times, the fact that certain features are 
technically possible does not mean that they are legitimate. I am sure 
that content owners are not happy with users changing the fonts with 
which they read their (content owners') website, but, as long as this 
remains strictly in the "user domain", there is little you can do about 
it. Quite a different thing is when a third party company places a 
component in the network and this component intercepts traffic and 
produces derivative work. This is NOT ok.

> Each of these scenarios has been a part of the web since there was 
> more than one browser, one server, and zero proxies, a time when few 
> people were even aware of the Web. It is true that such an environment 
> presents challenges to people who think that they should control 
> exactly what the user sees, but such people also seem not to have 
> noticed that some people wear glasses in order to change the way the 
> world looks, and other people don't see it at all and therefore 
> interact with it through hearing, touch, smell etc. The web merely 
> expands the range of capabilities for both provider and consumer to 
> adapt the communication in an attempt to better meet their desires.

again, technical feasibility does not equal legitimacy.


>
> The "One Web" vision, which you seem to have misunderstood

For the record, I did not misunderstand it. W3C failed to say what it 
means in practice.


> as some attack on the right to free expression, merely suggests that 
> it is sensible 

well, "sensible" is an interesting adjective. Some one-web supporters 
say "one web = must", some say "one-web= should, but you don't have to". 
This was never clarified W3C because a compromise was found in keeping 
its real meaning ambiguous


> for both users and authors to treat the Web as a single communication 
> system, in which all of these adaptation possibilities are present.

yeah, right, the Opera vision.

> Following this approach, attempts to somehow divide the Web into 
> several "sub-Webs" which are assumed to be fundamentally incompatible 
> and *requiring* completely seperate treatment is to misunderstand both 
> the capacity of the Web and the apparent patterns of use, which 
> demonstrate both clear variation in people's preferences and a 
> demonstrated desire on the part of many users to treat "Web sites" as 
> a single consistent service accessible in different ways.

I disagree with this viewpoint because I think mobile is such that 
better applications are built by accepting the view that mobile is 
fundamentally different from the big web. But this is not really the 
reason I am here. The reason I am here is that I don't want that, in the 
name of one-web or whatever else, someone jeopardizes those achievements 
which still allow developers to treat the mobile web as a platform (UA 
string and HTTPS, for example).

>
> In this context, transcoding per se is morally neutral, 

yes, in your wildest dreams.

> providing one part of this adaptability. It is also done for a variety 
> of reasons, including a desire to transform the presentation of 
> something to a form that a user finds more convenient (for example 
> applying mobile mode on Opera mini), making a perceived "default 
> desktop-style presentation" available on different devices (as with 
> Opera mini in its current default mode), providing operator branding 
> for an access service (as done by Vodafone), and so on. Whether these 
> various uses of the technology succeed or fail in the market is for 
> their intended users to determine, by choosing them or not. 

again, the fact that users may like certain services is not a 
justification that these "services" are legitimized per se, any more 
than P2P clients are (and believe me, they are popular among "users")


> The fact that *I* don't like to receive my content with added 
> advertising is *one* factor in what makes me choose one method of 
> accessing content through the Web over another. It isn't reasonable 
> for me to insist that nobody provide advertising as a way of mentising 
> a service I want to use - that is a restriction of a legitimate 
> business model, and ultimately of service providers' right of 
> expression. It is reasonable for me to choose a service that doesn't 
> do that, if I want.

If I create a website, the site is popular, and I make $1000 a month 
with AdSense, I get damaged when a  transcoder sucks my pages, strips my 
adsense ads (or avoids injecting them, which is the same). Full stop. 
If, on top of that, they even inject someone else's ads, I get even more 
crazy. Is this so hard to understand? I have seen Opera being born. You 
guys know how hard it is to make a living with advertisement, why are 
you doing this to other companies? why are you supporting the moral 
legitimacy of disrupting other companies' business model?


>
> In the same way, various possibilities (such as the good work of you 
> and others on WURFL) exist to allow content providers to offer various 
> adapted versions of their service to provide them with a perceived 
> advantage in delivering what they think will be most helpful to their 
> intended users, and various possibilities exist in most browsers for 
> the user to adapt content received (beyond the various default 
> renderings, which are known not to be consistent across different 
> browsers) which are intended to make the browser more useful to its 
> intended users.

So:
- do not change the UA,
- respect HTTPS.

this will allow developers to customize content.

Users which will go out of their way to transcode web content by 
opting-in will be outside of scope of CTG.
Companies which disrupt the model above won't be able to do it in W3C's 
name. So simple.

>
>
> PS: In the unlikely event that this fails to convince you that code 
> transformation as a service is a legitimate part of the web, you may 
> like to tilt at a new windmill: 
> http://www.opera.com/press/releases/2009/02/12/ (incorporating Opera 
> Mini functionality into the Desktop brower) which includes the quote
>
> [[[
> ...Opera Turbo supports Opera's belief in "One Web", providing equal 
> access to the Internet regardless of the device or network quality. 
> ... To Opera, One Web means being able to take the one true Web and 
> make it available to people on their own terms.
> ]]]
>
> ("one true Web"! Gotta love press releases :) )
>
> Or an antecedent: 
> http://my.opera.com/chaals/blog/2009/01/22/opera-mini-on-mac-os (about 
> people using Opera mini on desktop computers).
>

I am not really sure about what you want to prove with this. To me, it 
proves that Opera has been pushing and promoting One-Web within W3C 
because it served its own specific marketing needs. Which is legitimate, 
but please don't try to present them as Opera doing anyone a service. 
This is only serving the requirements of Opera and other transcoder 
vendors. The web is based on the foundation that content providers can 
control the way their content is presented to end-users. Legitimizing 
the removal of this pillar is "suicidal".

Luca

Received on Thursday, 12 February 2009 13:42:49 UTC