Re: [art] Unichars / Modern Network Unicode / PRECIS

[Marc Petit-Huguenin <[e-mail]>]

On 3/19/24 6:17 PM, Paul Kyzivat wrote:
> On 3/19/24 8:55 PM, Steffen Nurpmeso wrote:
> 
>> Postel's law is wonderful.
>> I just reread RFC 2441, i think on Friday or Saturday.  And
>> i truly believe all those old guys did not only do a tremendous
>> job, but they were and are indeed America's (ie: U.S.) torch,
>> which, in my opinion, stands in total opposition to its military
>> and even more so politics.  Very off-topic and personal, of
>> course.
> 
> Postel's law is often the wise thing to do.
> But it is not a panacea. It isn't an excuse for sloppy specifications.
> 
> In particular "be conservative in what you send, be liberal in what you accept" only works when there is only one liberal and plausible way to interpret what you expect. While that is often true it isn't always true.

Yes!  And I believe that a specification can be verified to ensure that.  Basically, there exists a list of different messages that, when applied to a receiver in a specific state, always results into a unique new state, and the "be liberal in what you accept" part is about verifying that the receiver is conforming to that.  Conversely, the "be conservative in what you send" is about verifying that the sender is only able to send a message that is on that list if the intent is to reach that unique new state from a specific state.

That's the correct use of Postel's Law, and RFC 9414 needed only to say that, instead of throwing the baby out with the bathwater.

-- 
Marc Petit-Huguenin
Email: [e-mail]
Blog: https://marc.petit-huguenin.org
Profile: https://www.linkedin.com/in/petithug