Share your know-how
Dear Brian et. al,
I know this has been repeated and talked about passionately in other threads. But all discussions inadvertantly gets deflected to a discussion about W3C standards, which is non-existent and beside the point. As a developer who has years of experience developing mobile content, I've seen a lot of interference (or screw-ups) by carriers. But none is so bad as to changing the mobile phone's User-Agent. So I would like to ask a question, point blank:
What is the rationale behind changing the User-Agent header?
Please do not mention W3C because the standard does not exist. Vodafone is the one doing it, so please answer as Vodafone. Don't act as if a standard already exists.
I do understand and actually somewhat appreciate your efforts to automatically transcode desktop websites, if what you say about increasing mobile web usage is true. The more people that are used to browsing on the phone, the better for the ecosystem as a whole.
You already provide publishers a way to tell the WTE not to transcode, some are based on standards (mobile web sites should send the proper Content-Type and DOCTYPE anyways). I do applaud the effort.
I am trying to think from your standpoint here. The one thing that I do not understand, is why you have to change the User-Agent. I would like you to also think from the standpoint of mobile website publishers.
Sites that are desktop-only will serve largely the same desktop content, no matter what the incoming User-Agent is. These sites can be transcoded by the WTE. These site owners may have no problem having their site transcoded, because by definition, they don't care to adapt their content to mobile devices. Even if they do detect the User-Agent for desktop browser capabilities, they probably don't care too much about browsers that are not PC or mainstream. In this case, what's the point of faking as a different User-Agent, as sending the phone's User-Agent will largely yield the same result?
Sites that adapt their content to mobile phones rely on the User-Agent to identify the incoming device. Adapting their content to specific devices are their bread and butter. By faking as a desktop PC, Vodafone is effectively fooling these sites, telling them to return desktop/non-adapted content, and have WTE ruin their carefully engineered and crafted markup and user interface specifically targeted to various mobile phones. Some of these pages may contain carefully placed ads, so money and revenue for these sites are at stake. Is there any wonder that you are incurring the wrath of all these mobile publishers? In this case, it is of course better off for everyone involved if you just leave the User-Agent alone.
So it seems to me that changing the User-Agent is pointless and harmful, from a technical standpoint. Telling us to read from a different header is completely beside the point. (Besides it is not a standard!) It just creates unnecessary inconveniences for mobile content publishers that should not be there in the first place.
Unless, well, you create the inconvenience for reasons other than a technical one. I won't go down that path just yet, and I want to hear what you have to say.
So, tell us, what is the rationale behind changing the User-Agent header?
