See inline for my responses <as individual> On Aug 27, 2012, at 1:21 PM, Don Joslyn <[email protected]<mailto:[email protected]>> wrote:
Brian, I think you’re misunderstanding my reason for suggesting something other than LoST for discovery. Let me explain… 1. I support using an existing solution for discovery, such as LoST, as it could save a great deal of time and effort. I have not seen any meaningful discussions to help the paws group understand the pros/cons for LoST, or the understand the logistics of using it, so I’m trying to promote a useful exchange of information. 2. I do find it confusing that we’re spending all this time/effort discussing XML versus JSON (and potentially being forced to pick just one), but you think it does not matter when it comes to LoST (“We can evolve LoST if it'd decided that it's useful and doesn't screw up existing systems”) If we were going to use an existing LoST server, I assumed that it would already be using XML, thus it would matter for the master device because if PAWS uses JSON only then a master device would need to support both XML and JSON. We have never really discussed implementing a LoST server before, so I’m trying to figure out the plan. It’s unfortunate that we did not have time to discuss discovery in Vancouver. A LoST server intending to support Whitespace devices would need to support whatever encoding we would specify. That could complicate deployment, because XML encoding is driving current LoST deployments, but I don't think that is a limiting factor. It's a factor, but not a limiting one. 3. I’m trying to demonstrate other reasons that we may want to support XML, but my example was based on using an existing LoST server implemented with XML, and my assumption that we would not be re-writing the LoST standard (because I didn’t realize that was an option). I think it's an option. We would probably want to consult with ecrit, the work group that promulgated LoST. I have a few questions of my own (for Brian): 1. You still don’t care if we use XML or JSON, right? Right 2. You want the paws group to use LoST for discovery, right? Yes 3. You will not support any other discovery proposal, it must be LoST, right? I wouldn't put it that hard. If there is some good reason LoST won't work, sure. Don't know of any, but I'm willing to listen. 4. If paws uses JSON only, we will re-write the LoST spec and then implement a LoST server using JSON, right? I'd phrase that as "add a JSON encoding to LoST", that's not a rewrite, it's an update. It would be a fairly short document that normatively updated RFC5222. Thanks, Don From: Rosen, Brian [mailto:[email protected]] Sent: Monday, August 27, 2012 11:40 AM To: Don Joslyn Cc: [email protected]<mailto:[email protected]>; [email protected]<mailto:[email protected]>; [email protected]<mailto:[email protected]>; [email protected]<mailto:[email protected]> Subject: Re: [paws] XML schema versus JSON, vCard & iCal <as individual> So, you think you understand how to write location sensitive discovery for a specific service? You think you thoroughly understand how it works, what the pitfalls are, where you need flexibility? Further, you think that every service that needs location sensitive discovery should invent its own mechanism and every jurisdiction (read country) should support who knows how many equivalent - but different, mechanisms? Or, maybe, you could allow that a group of very smart people spent a great deal of time working out how location sensitive services should be discovered, and tried to come up with one mechanism that would work for a very wide variety of services. I mean, why limit encoding to XML or JSON? Why don't we come up with our own encoding? See in line for responses to the questions you asked On Aug 27, 2012, at 11:17 AM, Don Joslyn <[email protected]<mailto:[email protected]>> wrote: So if you use JSON for LoST, are you really supporting the LoST standard as it’s currently written? Standards evolve. We can evolve LoST if it'd decided that it's useful and doesn't screw up existing systems. Is there any chance that PAWS could use an existing LoST server? Sure Does one exist? Yes. It's still pretty early though. If one does not exist, then I’m assuming that someone would need to implement a LoST server for use by PAWS devices, correct? Maybe. In most places, I'd say, probably Would it then make sense to implement it with JSON, even though the standard as written uses XML? Sure, why not, assuming we have a good reason to use JSON for whitespace Maybe somebody should take a look at the XML messages described in the LoST protocol, to determine how easy it will or will not be to convert them to JSON. I'm not a JSON expert, but I doubt it would be hard. I'll ask an expert. Maybe we should just forget about using LoST, and go with the only discovery proposal that has been formally submitted to PAWS? Maybe we should stop re-inventing something that has already been invented and concentrate on things that need new work. Also, remember vCard/xCard - same arguments.
_______________________________________________ paws mailing list [email protected] https://www.ietf.org/mailman/listinfo/paws
