OK, well I looked around to understand how I was supposed to go about editing the specs, and it appeared that the Editor has to make changes (spec has "this is governed by COSS") . However it appears that they are actually stored in github and the idea is that I fork and then send a pull request? If that's the idea, then fine, I wasn't aware that that was the process. I see now that at the top of the main RFC page it does say that, so perhaps the individual specs ought to be more explicit about the process for modification.
Thanks. -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Pieter Hintjens Sent: 02 February 2016 09:56 To: ZeroMQ development list <[email protected]> Subject: Re: [zeromq-dev] XRAP clarification: Are parameters/metadata case sensitive? I think at some stage we'll change COSS to reflect the more modern approach we use in ZeroMQ with C4.1. The state flow is good and works, but the edit/contribution process could be simpler. On Tue, Feb 2, 2016 at 10:55 AM, Pieter Hintjens <[email protected]> wrote: > I've pushed small edits for this. > > On Tue, Feb 2, 2016 at 10:39 AM, Tom Quarendon > <[email protected]> wrote: >> I'm certainly happy with that definition, it's what I've implemented. >> Pieter, my understanding of the COSS process is that given that you are the >> Editor of the spec, you have to make the mods? Oh, actually you are credited >> as "contributor", with no Editor. What does COSS process say happens in that >> situation :-) I think you are probably de facto Editor. >> Thanks. >> >> -----Original Message----- >> From: [email protected] >> [mailto:[email protected]] On Behalf Of Pieter >> Hintjens >> Sent: 02 February 2016 09:15 >> To: ZeroMQ development list <[email protected]> >> Subject: Re: [zeromq-dev] XRAP clarification: Are parameters/metadata case >> sensitive? >> >> ZMTP uses case-insensitive metadata names. I suggest we use the same >> convention. >> >> >> >> On Tue, Feb 2, 2016 at 10:03 AM, Tom Quarendon <[email protected]> >> wrote: >>> It’s not specified in the XRAP specification whether or not the >>> names in the parameter hash, or the returned metadata hash are case >>> sensitive or not. >>> >>> When actually attempting to implement this, I think this is >>> information that is required. >>> >>> In HTTP, header fields are case insensitive. In HTTP/2 I believe it >>> even goes as far as to say that the header names have to be passed in lower >>> case. >>> >>> >>> >>> I don’t know what would be most appropriate in this circumstance, I >>> think that having them case insensitive, or equivalently always >>> passed in lower case, would be most sensible. >>> >>> >>> >>> Assuming concensus is reached, can the spec be edited to reflect? >>> >>> Thanks. >>> >>> >>> _______________________________________________ >>> zeromq-dev mailing list >>> [email protected] >>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev >>> >> _______________________________________________ >> zeromq-dev mailing list >> [email protected] >> http://lists.zeromq.org/mailman/listinfo/zeromq-dev >> _______________________________________________ >> zeromq-dev mailing list >> [email protected] >> http://lists.zeromq.org/mailman/listinfo/zeromq-dev _______________________________________________ zeromq-dev mailing list [email protected] http://lists.zeromq.org/mailman/listinfo/zeromq-dev _______________________________________________ zeromq-dev mailing list [email protected] http://lists.zeromq.org/mailman/listinfo/zeromq-dev
