[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Shipping Atom products prematurely



* Mark Nottingham <mnot@xxxxxxxx> [2004-08-16 20:39-0700]
> 
> 
> On Aug 16, 2004, at 8:51 AM, Antone Roundy wrote:
> 
> >On Monday, August 16, 2004, at 09:22  AM, Mark Nottingham wrote:
> >>Ah; I misunderstood your question. My suggestion would be to not ship 
> >>Atom in mass-market products until it's done (or at least close to 
> >>being done). RSS is already out there.
> >>
> >I would think that we'd like to recommend not shipping Atom in 
> >mass-market PUBLISHING tools till it's done, but off the top of my 
> >head, I don't see any problem with shipping it in CONSUMING tools.  Am 
> >I missing something?
> 
> Vendor A ships 10 million units of their software that includes support 
> for reading Atom feeds. Now, every time we suggest that something 
> change in the spec, we run up against the "but it's already on 10 
> million desktops" argument.
> 
> This is especially worrisome, considering that we haven't yet made 
> versioning and extensibility bulletproof.

+1, please let's not have to leave extensibility and versioning for v2.0
because of this.

> I don't want this to be a huge issue; however, the expectations need to 
> be managed.

Yup. For sites *producing* Atom feeds, perhaps encourage a bit of
boilerplate at the top of the file in XML comments, perhaps something
like:

<!--  Early Adoption Warning
This document uses an in-progress draft of the Atom publishing
format. Atom is undergoing active development and standardization 
within the IETF, see http://www.ietf.org/html.charters/atompub-charter.html
Please note that the AtomPub WG will not allow deployment of early 
implementation of draft proposals to hinder improvements to the Atom
design. Implementor feedback is welcomed by the WG, however implementors 
should be cautious when deploying prototype Atom software to end-users. -->

...or words along those lines? Less wordy might be wise, for big sites
with lots of feeds... (eg. a pointer to text elsewhere perhaps)

cheers,

Dan