Shipping Atom products prematurely

[Mark Nottingham <mnot@xxxxxxxx>]

I've just changed the old, pre-WG drafts that I host on mnot.net to 
clearly state that they're not to be implemented, and are superseded by 
the work here.

One of the things that drove me to do this is the increasing incidence 
of vendors shipping products and services that "support" Atom, even 
though it's clearly marked as a pre-draft (on my site) and only for 
experimental implementation (in the WG's drafts).

Working on experimental Atom implementations is great because it helps 
us get experience and find problems. However, shipping it in 
mass-market products before it's done is likely to lead to a lot of 
interoperability problems, end-user support issues and even barriers to 
adoption as the spec matures.

As such, I think the WG should proactively contact vendors of consumer 
products and services and ask them to refrain from integrating Atom 
support until it's done. They have the best of intentions by supporting 
Atom, but may do more harm than good by accelerating the adoption of it 
prematurely.

I'm not naming names here, but I have two vendors in particular in 
mind. I *don't* mean the smaller aggregator-only vendors or personal 
blogs who have made a choice to experimentally support Atom; I mean 
those that are pushing Atom to an audience who doesn't understand what 
the issues are.

Thoughts?

--
Mark Nottingham     http://www.mnot.net/