W3C home > Mailing lists > Public > public-coremob@w3.org > April 2012

Re: How many browsers matter Re: Ringmark is now open source

From: Charles McCathieNevile <chaals@opera.com>
Date: Fri, 13 Apr 2012 09:05:47 +0200
To: "Robin Berjon" <robin@berjon.com>
Cc: public-coremob@w3.org
Message-ID: <op.wcozbxfdwxe0ny@widsith-3.local>
On Thu, 12 Apr 2012 15:06:11 +0200, Robin Berjon <robin@berjon.com> wrote:

> On Apr 11, 2012, at 20:39 , Charles McCathieNevile wrote:
>> I hope this clarifies why I think the current assumption is only  
>> relevant if this group is looking at a "US/Europe Wide Web".
>
> For values of US/Europe that include Japan, Korea, Australia, New  
> Zealand, Singapore but exclude Poland, Russia, the Ukraine, or Belarus.

Sure. And there are a few million more people to add and subtract if we  
want to get pedantic. But since my goal is not to be inflammatory, but to  
better understand the exercise (and how valuable it really is to me), I  
think it's a reasonably fair approximation to what we are working with at  
the moment.

> Inflammatory implications aside, I don't think that you can usefully  
> make a geographic characterisation here.

Ah. Here we are getting to the core of the disagreement. For a developer  
wanting to make stuff work in Russian, or China, or BRIC countries, or  
Latin America, or a number of other large markets, this group currently  
seems to say "forget it. Most people can't use ring 0, so do your stuff in  
english or german or in Japan". I had hoped that a descriptive starting  
point would be "here is what you can use in the world".

I think the way forward to do that would be to break down apps a bit more  
into different sets of requirements. I realise it doesn't fit the nice  
neat ring model so well, and there is a trademark or two covering global  
activities symbolised by five rings...

> As described before, this group aims to improve interoperability
> in a given part of the ecosystem (where we're seeing specific
> problems),

As I said before, that is a legitimate aim. But it isn't quite what the  
spec claims it did.

> with the natural consequence that interoperability will be
> improved across the spectrum by making it easier to target content
> throughout the board.

Not necessarily. By ignoring enormous markets and segments, we end up  
assuming that they will develop the same way. This is not true, as a  
glance at the varied global histories of e-commerce shows.

If a major system develops that doesn't work the way we thought it would,  
then we can get major interoperability breakage.

> Yes, that part of the ecosystem is more present in some parts of the  
> world than others. That is true of all efforts that attempt to improve  
> things in any segment. This isn't about the EU/US Web any more than MWBP  
> is about the Saharan Web.

Actually, I disagree. This *is* about the EU/US (as a rough approximation)  
web. It is about mirroring a couple of systems that hold a large share of  
the market there (but are apparently far less relevant outside that part  
of the Web) into the Web. MWBP specifically focused at a level that made  
it work in both the Sahara and the US (one of which had expensive phones,  
while the other had functional mobile commerce - both things that would  
fit into subsequent rings...)

As I said, that's a legitimate choice. But if that is what the group is  
doing, it should be far more explicit in the specification about the  
methodology, and probably mention the potential limitations of that  
methodology.

>> I would personally prefer the group to take on the more substantial  
>> challenge of being relevant to the world, rather than withdraw to being  
>> a de facto documentation team for a couple of companies who could  
>> easily do that for themselves.
>
> It's a good thing that's not what we're doing then!

Indeed. I hope we don't end up there accidentally.

Cheers

Chaals

-- 
Charles 'chaals' McCathieNevile  Opera Software, Standards Group
     je parle français -- hablo español -- jeg kan noen norsk
http://my.opera.com/chaals       Try Opera: http://www.opera.com
Received on Friday, 13 April 2012 07:06:23 UTC

This archive was generated by hypermail 2.3.1 : Friday, 19 April 2013 17:36:46 UTC