Arnt Gulbrandsen wrote:
Cyrus Daboo writes:Hi Timo,--On February 13, 2009 5:50:53 PM -0500 Timo Sirainen <tss at iki.fi> wrote:The IMAP SEARCH=INTHREAD and THREAD=REFS Extensions <http://www.ietf.org/internet-drafts/draft-ietf-morg-inthread-00.txt>Technically OK, but I was left wondering why the heck I would want to implement the new SEARCH capabilities if I were a client.This document is meant to answer some questions. It doesn't tell you to ask those questions.In my not at all humble opinion, there's a four-step process: 1. You need to want to do something. 2. You look for ways to do it. 3. You find some RFC. 4. You read it.4 comes after 1. Telling people in step 4 that they should want something is no good.This implies that an RFC must be findable for the people whose questions it answers, then it must answer their questions (and do so clearly enough that they understand it).
Last sentence: I agree.
I added some text to the address-search document, answering the all too common question "how do I implement so-and-so if the server does and doesn't implement this extension"? I guess similar stuff would be useful here.
Sure.
Also, I'll dream up some google searches and see that the draft is found.I think the document needs to have more examples or concrete use cases to better explain why these changes are important and how clients would make use of them, and what the benefit to the server is to implement.What are some RFCs that do this and benefit from it?
I can't remember any RFC doing this from the top of my head, but I think I've done something like this in SEARCHRES document.
I actually agree with this comment from Cyrus and I am happy to suggest some text.