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

Re: [minutes] 12 May 2009 Teleconference

From: Luca Passani <passani@eunet.no>
Date: Thu, 14 May 2009 11:30:17 +0200
Message-ID: <4A0BE4A9.3070906@eunet.no>
To: Mobile Web Best Practices Working Group WG <public-bpwg@w3.org>

After seeing this message and another one with an Opera employee 
offlist, I went back to the thread to figure out where the disconnect 
was, since it wasn't making sense to me until that point.

So, there was an implicit assumption I was making that transcoders can 
spoof the UA and, when not restrained, they'll go ahead and happily do 
it. This applies to the transcoders by Novarra, Infogin, Openwave and 
ByteMobile (the major four). As an aside, ByteMobile spoofs as OperaMini.
This is what I was referring to when I said that transcoders replace 
"the user-agent string to fool web sites". This has been the core of the 
discussion all the way.

I understand that Bruce's answer referred to the fact that the Opera 
browser and the Opera Mini browser/proxy combination do not spoof the UA 
string. Which is correct.

I am still perplexed by the fact that user-agents (i.e. browsers) have 
been pulled inside a discussion that was focused on transcoders (i.e. 
proxies), but I think there is enough to take back my statement about 
Bruce not knowing what he is talking about.

Apologies for the confusion.

Of course, this does not change my viewpoint that, by voting against the 
introduction of the MobileOptimised heuristic in CTG, Opera has taken 
the part of transcoder vendors and disrespected developers, content 
owners and the common goal of a healthy mobile ecosystem.

Luca


Sangwhan Moon wrote:
> <snip>
>
> I would like to sincerely apologize to other subscribers of the list 
> about this vendor specific spam, please feel free to mark it as read 
> if it is not of your interest. There was a wrong point that I had to 
> clarify before discussion proceeds further.
>
>> Bruce, can you look at yourself in the mirror after saying this? do 
>> you realise that here you are talking about a "technology" 
>> (transcoders, albeit the term "technology" gives them way more honor 
>> than they deserve) which replaces the user-agent string to fool web 
>> sites?
>
> I have to disagree with your point here, I checked on the 
> implementations that utilizes transcoding and the user-agent string 
> seems to be intact.
>
> Opera 10.0:
> User-Agent: Opera/10.00 (Macintosh; Intel Mac OS X; U; ko) Presto/2.2.1
>
> Opera 10.0 with Turbo Mode on:
> User-Agent: Opera/10.00 (Macintosh; Intel Mac OS X; U; ko) Presto/2.2.1
>
> As you can see above, both have the exact same user agent. If you 
> would like to see for yourself, please try the feature enabled test 
> build with Opera Turbo at: http://labs.opera.com/downloads/
>
> It would be greatly appreciated if some pre-research was done before 
> making such claims on a public list from now on.
>
> Kind regards,
>
Received on Thursday, 14 May 2009 09:31:04 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 14 May 2009 09:31:04 GMT