Re: The "polyglot" approach to fedsocweb

On 10/09/2012 04:10 PM, Sandeep Shetty wrote:
> On Wed, Oct 10, 2012 at 12:03 AM, Melvin Carvalho
> <melvincarvalho@gmail.com> wrote:
>> I'm thinking more and more this is a great idea.
>>
>> But why stop at identifiers?  Maybe we can polyglot the whole social web to
>> include data, posts and APIs!
>
> The polyglot approach is expensive to build and even more expensive to
> maintain. Does anyone have an example where this approach has worked
> in the past in some other context?
>

I think there's a number of examples out there already. For example:

* Most feed readers speak both RSS (various versions) and Atom.

* Lots of services support login with Facebook, Twitter and possibly 
Google, etc, even though it requires custom work for each. A couple 
years ago this set also often included OpenID.

I think in practice this is just how "standards" (in the de-facto sense) 
happen: you start off with lots of competing things, people grumble and 
implement some subset of them, and the ones that end up implemented in 
the most places end up winning. It sucks in the mean time, of course.

Received on Tuesday, 9 October 2012 23:25:52 UTC