Re: Idea for a process experiment to reward running code...

[Jari Arkko <[e-mail]>]

Brian, Martin,

> On 03/12/2012 06:01, Martin J. Dürst wrote:
>> One of the advantages of a standards organization such as the IETF is
>> cross-concern review. For the IETF, one very strong cross-concern is
>> security. Another one (also for my personally) is internationalization.
>> Another, more vague one, is general architecture. Early running code is
>> very often (not always) characterized by the fact that such
>> cross-concerns are actively or passively ignored.
> An excellent point. The fact that a hack works, and can be implemented,
> does not alter the fact that it's a hack. This is the sort of thing that
> cross-area review is supposed to look for. As a gen-art reviewer, I am
> sometimes surprised by what gets through to Last Call in the regular
> process - if the whole review process is squeezed down to a couple
> of weeks, we will definitely miss cross-area issues.

True.

>
> Encouraging running code is a Good Thing. Publishing sloppy specifications
> is a Bad Thing.

Yes. Of course, it is difficult tradeoff. No hard and fast rules here. You have to publish a timely spec to make an impact, but you also have to get it right. But how timely? How right?

>
> The Interop show network used to be a Very Good Thing. We've lost that,
> though I was delighted to see some actual running code at Bits-n-Bytes
> in Atlanta. More please. Maybe a prize for Best Demo?

I've been thinking about something along those lines. Following up in a second e-mail...

jari