Regarding #5, it should be specified which characters should remain escaped.
I don't quite understand the question? It says "to the maximum extent possible"? User:Http://mylid.net/jernst
- Unless there's an exception for %25 (which resolves to %), this could mean repeatedly unescape until unescaping is no longer possible. i.e. %25252525252525252525252541 = A User:Http://dag.myopenid.com/
What does the algorithm create? Is the result the "user token" that I store in my database but what do I send e.g. as an openid.identity? For this one can perform steps 1, 3 and 4 but for example not 2. --Http://lukasrosenstock.net/ 07:10, 22 April 2006 (PDT)
- As proposed, its only purpose is an internal representation that allows implementations to compare URLs. It may also make sense as the primary keys in the data store for lookup. I'm not proposing that the user ever sees it (although one could extend the proposal to also apply there in some fashion, but let's not try to do too many things at thesame time.) User:Http://mylid.net/jernst