On 5/21/05, A. Pagaltzis <[EMAIL PROTECTED]> wrote: > > * Robert Sayre <[EMAIL PROTECTED]> [2005-05-21 19:05]: > > At this stage, changing the spec to suit religious preferences > > would be extremely arrogant. > > Please stop talking to people about "process bullshit" at one > occasion and turning around to chide others for "at this stage" > at the next.
As I said before, it doesn't matter if the spec is perfect if it takes forever. Making up requirements after last call is out of order. Eric is making up requirements and adding category elements to author elements. It is beyond ridiculous. Pardon me if I'm getting impatient, but I don't think any of the ideas raised in the past few days are critical improvements to the format. In fact, all I see is people twiddling their beanies about semantics, rather than explaining why they need X for their products (Wyman excluded--note that he raised his atom:id issue in last call). Bob's bits-on-the-wire approach to multiple ids is fine. We can put the DOS issue in the security concerns, but it really comes down to "don't believe everything you read on the internet". The current suggested text makes the terrible mistake of conflating atom:id and timestamps. I remember Graham once tried to re-raise the atom:modified idea, and was told that he was very close to being out-of-order. Well, it certainly is now. Not only is it offensive, but there's no reason to expect it will work, and no reason it couldn't be added if it does happen to work. So, really, we have folks who want to delay this spec because they think they've solved Distributed Versioning On The Internet. Robert Sayre
