I wrote some generally useful code recently which i figured I should share. http://windmill.osafoundation.org/trac/browser/windmill/trunk/ ...
Mikeal Rogers
mikeal@...
Jan 3, 2007 7:39 am
413
... I appreciate your sharing this code. Unfortunately I have no idea what to do with it or how to install it. Could you provide a little guidance? Is it...
... The dispatcher is the underlying object for running any python method over json-rpc. You send it methods to be added for dispatch, or you can send it a...
Mikeal Rogers
mikeal@...
Jan 3, 2007 9:08 pm
415
... I'm glad to read this. JSON-RPC (1.0) currently proves to be the best solution for the communication between our new and legacy embedded devices (which...
I have not seem much posting recently around the WD1.1. There are some of us that have issues with different sections, but I would like to see WD1.1 moved...
... I personally think it's about time to put json-rpc on the track to an IETF RFC. Next week I'll be seeing Lisa Dusseault, the current application area ...
Mikeal Rogers
mikeal@...
Jan 7, 2007 12:59 am
419
... I'd probably not be regarded as really being a part of the json-rpc community so my opinion may be seen as irrelevant, but I would disagree with trying to...
*It requires that the request package be an array of request objects with unique id. For single requests, per the wd1.1 the id would be optional To clarify, if...
I am in agreement with Graham, pushing to RFC is not the solution to issues in the WD 1.1 . I would rather not step on other's work, but I'm concerned that...
... Part of the RFC process is to change the language in the specification. Looking at the spec now, there is very little should and must or suggested type...
Mikeal Rogers
mikeal@...
Jan 8, 2007 5:22 pm
423
... One of the things I hoped would happen when suggesting this course of action would be to see what people thought of the current working draft and bring...
Mikeal Rogers
mikeal@...
Jan 8, 2007 5:54 pm
424
I think it's not as bad as it sounds. Looking back at the last year, the majority of the discussion in this group has centered around WD 1.1, from feedback...
The only real unresolved issue I see is the multiple request in a single incoming request package. Matt, I'm sorry if you feel this has not been addressed. I...
Enter your vote today! A new poll has been created for the json-rpc group: Is JSON-RPC 1.1 (Working Draft 7 August 2006) headed in the right direction? In...
json-rpc@yahoogroups....
Jan 9, 2007 12:58 am
427
I'm afraid that many important points and concerns are getting lost in threads and discussions. There is the danger that we'll get some good traffic for...
Although self description of a system is important, I really do see it as optional albeit socially important for public facing servers. Just a quick question...
Enter your vote today! A new poll has been created for the json-rpc group: Section 6.2.1 of JSON-RPC 1.1 (WD Aug 07, 2006) permits positional and named...
json-rpc@yahoogroups....
Jan 9, 2007 7:09 am
430
... - 7.4. Error Object as noted, the error-codes are missing (or: are all 000). maybe we could use the same error-codes as some other RPCs, i.e. as XMLRPC? <<...
Distinct from mixing named and positional in one call, I'd also question the requirement for a service implemented using positional parameters to still be able...
Hi I'd like to make some (increasingly intrusive) suggestions. Why not move the procedure name into the URI? ... There are several reasons for this suggestion....
... The given poll is a bit simple - there are too many issues to say 1.1 good or bad. What I'd really rather see is a wiki (as you mentioned) where people...
Sorry, but this complicates things rather than making them simpler! Yes, it help someone implementing the server, but think about the scenario where a...
... I've played around with this idea some time ago and I would love to move in this direction (though to a different effect than you seem to be suggesting in...
Just wondering how a wiki would be different for "hashing out all the issues and ramifications for each point of the spec" than a threaded-discussion group...
... A wiki isn't great for having conversations, but if you use the comments plugin on trac it's a good way to note issues within a spec and have short...
Mikeal Rogers
mikeal@...
Jan 10, 2007 11:28 pm
439
Using the trac wiki to edit the 1.1 spec is a good idea. Collaborators who want to make notes or comments on the spec directly can separate them from spec text...
... Actually it's even easier than that with the [AddComment] macro. http://trac-hacks.org/wiki/AddCommentMacro -Mikeal...
Mikeal Rogers
mikeal@...
Jan 11, 2007 9:06 pm
441
... In other words, on the client side you want to save yourself some work by not providing separate client side endpoints, one for positional parameters and...
Atif, Thanks for the comments. ... Well now I'm curious. What were the different effects you were after? ... Well I guess so. Or ... you could just move the...