Re: [art] On BCP 190

[Leif Johansson <[e-mail]>]

Hi Mark,

I didn’t expect this to be this hard to parse. 

I understand you are hoping BCP190 will have an effect in the future.

My question is by what means would you measure or determine success? I ask because I (and clearly others on the list) see a pretty big impedance missmatch between reality and BCP190.

Hope this helps to make what l’m asking for clearer.

Cheers Leif


Skickat från min iPhone

> 24 juli 2019 kl. 08:59 skrev Mark Nottingham <[e-mail]>:
> 
> Indeed.
> 
>> On 23 Jul 2019, at 11:38 pm, Adam Roach <[e-mail]> wrote:
>> 
>> On 7/23/19 18:25, Leif Johansson wrote:
>>> 
>>> Skickat från min iPhone
>>> 
>>>> 23 juli 2019 kl. 17:49 skrev Adam Roach <[e-mail]>:
>>>> 
>>>> Leif --
>>>> 
>>>> It seems to me, watching this conversation, that you and Mark are talking past each other. While I understand the points he is making, I'm having a hard time following your line of logic given that BCP 190 does, in fact, codify practices that have largely been applied to URI usage in the past (albeit only in an ad-hoc fashion prior to its publication). It's done a pretty reasonable job of keeping protocols from co-opting parts of the URI space so far, and will ostensibly continue to do so into the future.
>>>> 
>>> I read Mark saying that there are 1000s of APIs that get it wrong. That made me ask if (and if so you you know) BCP190 is actually working.
>> 
>> 
>> Ah. I'm pretty sure -- and Mark can confirm -- that what he meant to convey by that statement is that he's concerned about the impacts of 1000s of APIs eroding the usable URI namespace in the future if we were to abandon or substantially weaken the constraints outlined in BCP 190.
>> 
>> /a
> 
> --
> Mark Nottingham   https://www.mnot.net/
>