On 16/12/16 17:07, Andrew Newton wrote: > On Fri, Dec 16, 2016 at 11:33 AM, Keith Gaughan <ke...@blacknight.com> wrote: >>> Many people complained, hence RFC 4992. >> >> Here's a question: was BEEP really the problem, or was the problem >> really that IRIS itself never got enough traction? >> >> Because if IRIS got any traction, we wouldn't have needed RDAP. > > A multiplexing protocol is not trivial to implement correctly.
That's true, but we don't need a multiplexing protocol in this case, just pipelining. It's a simpler problem. > The problem with IRIS isn't that the XML was hard or the XPC or DCHK > were hard, but the compounded layers simply added to complexity for a > problem that was easily solved with common tools. Agreed, which is the point is was trying to make. >> The nice thing about Nominet's protocol is that it's trivial to >> implement and you get pipelining for free if you want it. While it has >> its issues (a way to get a list of field names would be nice, and >> headers giving request and response lengths would be preferable to >> using CRLF as terminators), it fits its purposes very well. > > Until you need protocol versioning, extensibility because no two > registries have the same problem, Unicode support, and security... > etc. etc. etc... And pretty soon you've re-invented the wheel. Not all protocols need all those things. Layering a DAC on top of RDAP makes the same mistake as layering DCHK on top of IRIS. The use case here is to check the availability of a domain and do it as quickly as possible. The constraints of a protocol like that differ from those of a directory access protocol like RDAP. The biggest two omissions are that the DAC protocol lacks any kind of error response, and there needs to be a way for the client to discover the response field mapping, whether that be that the responses be self-describing (such as line-delimited JSON documents) or a greeting, which includes the version and a description of the response. Those are just two options. There are other things that need to be covered, such as how fee information is transmitted and how this interacts with draft-ietf-regext-epp-fees. I'm not saying there aren't tonnes of potential issues with Nominet's DAC protocol, but it's a closer match to the needs of registrars than RDAP is. -- Keith Gaughan, Development Lead; PGP/GPG key ID: D5FC9D23 Blacknight Internet Solutions Ltd. <http://blacknight.host/> 12A Barrowside Business Park, Carlow, R93 X265, Ireland Registered in Ireland, Company No.: 370845 _______________________________________________ regext mailing list regext@ietf.org https://www.ietf.org/mailman/listinfo/regext
