Re: What's wrong with UA sniffing and server side processing?.

On Friday, June 28, 2013 at 4:10 PM, Andy Davies wrote:

> On 28 June 2013 12:04, Marcos Caceres <w3c@marcosc.com (mailto:w3c@marcosc.com)> wrote:
> >  
> >  
> > On Thursday, June 27, 2013 at 2:34 PM, Darrel O'Pry wrote:
> >  
> > > <unlurk />
> > >  
> > > There seems a be a general dislike for UA sniffing, but I don't see much reasoning behind it. Why are people opposed to UA sniffing? What's wrong with UA sniffing to do device detection and determine device characteristics?
> > See:
> > https://developer.mozilla.org/en-US/docs/Browser_detection_using_the_user_agent
> >  
> > For example:
> > http://blogs.msdn.com/b/ieinternals/archive/2011/06/30/perils-of-user-agent-sniffing-browser-mode-document-mode-compatibility-view.aspx
>  
>  
> I think UA sniffing is never going away but I'd prefer to see it only used for limited cases e.g. where we want different markup for different UAs and feature detection doesn't deliver.

Agreed.   

> The examples I can think of are replacing async/defer on script elements and the proposed lazyload attribute on images with JavaScript loaders for older UAs rather than always using JS.  
>  
> That said where feature detection can be used I'd prefer to use it but there areas where it may not be the most appropriate solution.
  
Agreed. There are even case where feature detection is really just "iPhone detection" in disguise… I've been guilty of that, and I'm sorry :)   

> > > I understand reservations about server side processing, particularly when you don't have access or control over the server?
> > Yes, this affects a significant number of developers (at least the research we - RIGC- has done seems to suggest this… speaking of which, I don't know where the link to that research went. Would be good to have that in the Wiki and if there is supporting research, would be good to add to the wiki too).
>  
>  
> Without some understanding as to who completed the survey I think we have to be careful what conclusions can be drawn from the results. Do we have any ideas what sort of organisations the respondents came from and what size sites they run?
Yes, was just a non-probablistic sample (survey we ran on twitter). We didn't get the data you mention above - just "do you have access to your server" or some such. We would like to run more scientific surveys, but alas we engineers make crappy sociologists. Again, apologies that I don't have the link! I'm hopeful Mat can provide it.     
>  
> I can understand the need to avoid a server-side solution for things like epub but I wonder how much the desire to avoid a server-side approach is limiting the possibilities for a solution (not that I've got an magic answers there).

We are certainly not avoiding it. There are actually a few people who run CDNs in the group and we've tried to make sure we accommodate their needs (e.g., http://www.cdnconnect.com/). Adam Bradley can probably speak to this (hopefully in positive terms! :) ).  

Kind regards,
Marcos

Received on Friday, 28 June 2013 15:50:42 UTC