On Thu, 04 Nov 2004 11:40:09 -0800, Walter Underwood <[EMAIL PROTECTED]> wrote:
I know we are all really tired of busted HTML and XML, but this doesn't seem like it solves the problem. Getting rid of bad feeds requires publicising them, not telling the publisher (who might not care).
Telling the publisher would probably help in 50% of the cases. Maybe more. However, the actual number of cases it helps in isn't essential. What is essential is that it helps.
Having 100% wellformed and valid producing systems would of course be best, but until we have that, we need to take other measures to fix the situation. Maybe even Six Apart can implement the error mechanism into Movable Type so that ERR requests are routed to them after a local analysis, so they can improve their tool centrally.
None of this is possible without an ERR mechanism. We could of course just continue down the endless road of lobbying for better tools, but until we reach the end of that road (which we never will), all other paths are worth exploring.
It is important to be able to figure out the feed software and version, so we can say the AtomBrainCreature 13.1 is busted and get the word out.
Do you take the responsibility to publish this information centrally somewhere? Or do you want an automated mechanism that sends some sort of error report directly to the software vendor?
But that requires notifying the implementor, not the publisher.
We need to notify both, but first and foremost the publisher -- it's his feed that's broken.
-- Asbj�rn Ulsberg -=|=- [EMAIL PROTECTED] �He's a loathsome offensive brute, yet I can't look away�
