Re: [wmlprogramming] Verizon, guidelines

This thread ends here for me. I think I understand CTG enough to be able 
to represent it for what it is when it goes out of "work in progress" mode.

Only note is about default configuration: the concept of default 
configuration does exist for transcoders. Vendors who have signed the 
Manifesto have had to make changes to their default configuration in 
order to comply. For some, the work was minor. For others it has implied 
a large amount of work, so they asked if they could sign and claim 
compliance starting with version X.Y (which they got away with doing by 
linking to a statement on their website which explained these details).

Complying to the Manifesto is possible, it's fair and it's an effective 
way to install transcoders responsibly.

Luca

Tom Hume wrote:
>
>
> On 23 Dec 2008, at 14:45, Luca Passani wrote:
>
>>> Indeed, but UA-spoofing is a red herring, as it won't happen for 
>>> made-for-mobile services under CTG anyhow.
>> It's not a red herring. It's the linchpin of the whole discussion. By 
>> spoofing the UA, mobile sites no longer know the final recipient of 
>> their content is a user with a mobile phone.
>
> But it won't happen.
>
>>>> This is a really shakey argument Luca: W3C are to blame when an 
>>>> unfinished document is misquoted, yet get no credit for sorting the 
>>>> problem out.
>> you see how irritating you can  be? W3C is to blame because they have 
>> created a working draft which was easily misrepresentable. I warned 
>> about this before it happened and it has happened exactly as I had 
>> described. Where did I get credits or acknowledgement for that?
>
> If you're going to quote a document that says "don't quote me" and 
> then misquote it anyway... any document can be misrepresented.
>
>> Secondly, W3C has fixed nothing. Novarra/Verizon are still spoofing.
>
> m. and .mobi addresses are now whitelisted, which gave developers a 
> way to quickly ensure their services worked.
>
> As someone who benefitted from this, I'm quite grateful.
>
>>> At the same time, when Manifesto signatories abusively transcode 
>>> then you take no responsibility personally and praise them for 
>>> reacting.
>> Mystification has obviously become your favorite sport. First, the 
>> Manifesto requests that transcoders are compliant for their default 
>> configuration. This allows developers to turn to abusive operators 
>> for complaints. This is to prevent that the disgusting theater that 
>> VodaUK and Novarra set up last year, when Voda was blaming Novarra 
>> for the abuse, and Novarra was blaming Voda.
>> With the Manifesto responsibilities are clear.
>
> Without knowing the details of how transcoders are configured and how 
> they deviate from the default configuration (whatever that is), I 
> don't see how that makes things clearer.
>
> If a transcoder ships with 3 example config files in a directory, 
> which one is default? Sounds v woolly to me.
>
>> But I am not done. In spite of the fact that I am not required to 
>> react to anything, I did took up the task of contacting transcoder 
>> vendors when developers were reporting a problem. I could only 
>> observe that they were very responsive and fixed the problems 
>> effectively. Again, there is no evidence of this with any Novarra 
>> installation. Whitelisting m.* and *.mobi was already part of the 
>> platform. Fixing the reference to W3C to acknowledge that CTG is work 
>> in progress does not look like much of an intervention to me.
>
> This illustrates my point: when you go out of your way to fix things, 
> it's good. When others do, it's bad.
>
>>> Well, there are reasons, like the use case I've outlined for where 
>>> users might prefer to get a transcoded version.
>> yes, but you don't need CTG to authorize you to do it. Just create a 
>> transcoder that spoofs.  Creating CTG establishes a dangerous 
>> precedent: you change the rules single handedly, W3C will come and 
>> clean up for you later.
>
> I don't understand how this sentence follows on from what I've said.
>
> I've said that UA spoofing is acceptable in certain limited situations 
> (like the one I've outlined). You've said it's acceptable in other 
> limited situations.
>
>>>>> I don't see how this goes in favour of Novarra, can you explain 
>>>>> that assertion? Content tasting causes problems, lots of 
>>>>> developers have pointed this out.
>> Novarra does not support content tasting. If CTG were to require 
>> content tasting to avoid UA-Spoofing on mobile sites, this would mean 
>> extra work for Novarra in order to make their platform compliant.
>
> Is that a good thing? That CTG mandates content tasting in order to 
> create work?
>
>> Also, content tasting causes problems to mobile sites mainly. With 
>> the Manifesto, mobile sites are not subject to content tasting.
>
> Gotcha - OK, I see that now.
>
>>> So we're in agreement that double-hits are bad; yet the Manifesto 
>>> promotes them and CTG prohibits them.
>> The Manifesto does not promote them. The Manifesto allows them if a 
>> transcoder really has to. Don't try to play tricks please.
>> This exception was added in the process of finding a good compromise 
>> between mobile developers and transcoder vendors.
>
> :) Can you not see that this is exactly the same reason for having the 
> limited circumstances where transcoding is permitted in CTG?
>
>>> This is enough. You keep doing CTG. I keep doing the Manifesto. 
>>> Please refrain from saying that the Manifesto and CTG are similar. 
>>> They are fundamentally different.
>
>
> The heuristics in each, and the spirit of protecting mobile-ready 
> content, would suggest otherwise.
>

Received on Tuesday, 23 December 2008 18:42:02 UTC