Alan DeKok said:|
"I suggest racing against the design guidelines document. After all,
it's only a draft. If the IPv6 document can get published first, then
the design guidelines matter a lot less."
[BA] Either the Design Guidelines advice is appropriate or it isn't.
If it isn't right, then we should fix it, because eventually it will become an RFC and "racing against" it won't work any more.
If it is right, then we should try to adhere to the guidelines -- that's what they are there for.
In this particular case, my take is that the Design Guidelines document is saying "don't do precisely what RFC 2868 did". That is quite sensible. However, it might also be helpful to say "Here are the problems that were caused by RFC 2868's scheme, and here is some advice to avoid those issues". That way, until we finish the Extended Attributes document, we'll have some way to address these kind of situations.