Search the web
Sign In
New User? Sign Up
xml-rpc · XML-RPC Discussion
? Already a member? Sign in to Yahoo!

Yahoo! Groups Tips

Did you know...
Want to share photos of your group with the world? Add a group photo to Flickr.

Best of Y! Groups

   Check them out and nominate your group.
Having problems with message search? Fill out this form to ensure your group is one of the first to be migrated to the new message search system.

Messages

  Messages Help
Advanced
RFC: lcdXML-RPC specification, round 3   Message List  
Reply | Forward Message #3912 of 6839 |
I've changed the version number to 1.2 to imply continuity with the XML-RPC
spec document (assuming base64 was 1.1). I'll throw this up on a web page
shortly.

====

lcdXML-RPC 1.2 specification

Introduction (Non-normative):

The XML-RPC specification document contains contradictory requirements
for string handling. As a result, it is not possible to conform to
it, and designers of individual XML-RPC systems have implemented both
sides of the contradiction. As there are many fielded XML-RPC
systems, it would not be popular or useful towards the overall goal of
interoperation to declare one side or the other to be in error. The
lcdXML-RPC specification lists specific areas that can result in
undefined behavior between lcdXML-RPC systems.

It is intended that existing XML-RPC implementations are compliant
with lcdXML-RPC. Implementations legitimately may claim compliance
with lcdXML-RPC, and this should provide more useful labeling for
interoperability. New implementations may provide programmatic
enforcement of the additional restrictions of lcdXML-RPC to assist in
the construction of interoperable systems.

Authors of lcdXML-RPC systems are encouraged to document the the
actual and intended behavior of their implementations when
encountering situations undefined in lcdXML-RPC.

It is intended that other standards may standardize behavior that is
undefined in lcdXML-RPC; in particular, defining the issue of UTF-8
encoding.


Amendment (Normative):

lcdXML-RPC 1.2 amends the XML-RPC spec as follows:
===
<string>, string-valued <value>s, and <name>s are restricted to the
set of US-ASCII characters 0x20-0x7E, plus CR, LF, and HT.
Implementations may treat CR or the sequence CR LF as LF.

"<" and "&" must be encoded as "&lt;" and "&amp;" respectively. ">"
may be encoded as "&gt;" and must be encoded as "&gt;" when occuring
in the sequence "]]>".

The behavior of implementations is undefined when receiving the
following constructs:

a) characters outside the range 0x20-0x7E plus CR, LF, and HT,

b) XML CDATA sections,

c) XML character references ("&#55;"), and

d) XML character entities other than "&lt;", "&gt;", and "&amp;".

Any transmission of these constructs is outside this specification and
may be used only by private agreement between endpoints.

Update Question #5 ("What characters are allowed in strings?") is deleted.
===


Rationale (Non-normative):

The restrictions above are necessary to ensure interoperability
between lcdXML-RPC systems using ad-hoc and full XML processors. In
particular,

o Implementations with full XML processors reject characters in the
range 0x00-0x1F (except CR, LF, HT).

o Implementations with full XML processors may treat their input as
UTF-8. In those systems, characters in the range 0x80-0xFF will be
processed as UTF-8 multibyte encodings. Some ad-hoc implementations
process characters in that range directly rather than encoding them
as UTF-8.

o Many ad-hoc processors do not generate or process CDATA sections,
character references, or all the XML character entities.

--
Jay




Sat Dec 1, 2001 7:34 pm

nop@...
Send Email Send Email

Forward
Message #3912 of 6839 |
Expand Messages Author Sort by Date

I've changed the version number to 1.2 to imply continuity with the XML-RPC spec document (assuming base64 was 1.1). I'll throw this up on a web page shortly....
Jay Carlson
nop@...
Send Email
Dec 1, 2001
7:33 pm

Hi Jay -- I've been watching this thread with interest, and getting some private emails asking for me to comment. First, a caveat -- my team is in the endgame...
Dave Winer
dwiner
Offline Send Email
Dec 2, 2001
12:18 am

... Yeah, I figured this out after I realized you hadn't been on the list in a bit. This stuff can obviously wait for the product ship. ... very ... ...
Jay Carlson
nop@...
Send Email
Dec 2, 2001
6:48 pm

Jay, thanks for continuing the thoughtfulness of the thread -- it's much appreciated, I want you to know that. Now -- here's a bit of philosophy. Corner-turns...
Dave Winer
dwiner
Offline Send Email
Dec 2, 2001
7:07 pm

Sounds great to me also as I would enjoy seeing the spec tightened up a bit to support interop on a greater level. One of the things that seems to be kicking...
Codepunk
codepunk2000
Offline Send Email
Dec 2, 2001
8:30 pm

... Ok have been away the weekend flying rockets, but thats another story :) Reading through the posts it seems that everyone is saying more or less the same...
colledgi@...
iaincolledge
Offline Send Email
Dec 3, 2001
11:27 am

Hi Iain ... I think what is probably needed is something like "The Annotated XML-RPC Specification"... where all the words of wisdom that have appeared in this...
David Somers
moundsmere
Offline Send Email
Dec 4, 2001
2:21 pm

Monday, December 03, 2001, 12:27:33 PM, colledgi wrote: [great postings from dave, cliff, iain and jay cut] creating something like a RFC or an 'Implementors...
Andreas Bolka
lord_earl_wa...
Offline Send Email
Dec 4, 2001
10:21 pm

... Sounds like it to me, and I have to agree with Jay 100% in his last post. To be honest I think the realisation that the spec is incorrect and impossible to...
iaincolledge
Offline Send Email
Dec 5, 2001
12:58 pm

XML-RPC and SOAP both have trouble handling complex data types like XML, unicode, or binary. They also seem like an aweful lot of work for simple stuff like...
S. Alexander Jacobson
alex@...
Send Email
Dec 5, 2001
5:13 pm

... Interesting idea, related to the REST concept i guess. If this is simpler than XML-RPC is really a question of perspective. Seen from a HTTP/web...
olavjunker
Offline Send Email
Dec 6, 2001
1:09 pm

... Actually from a script centric perspective, MIME-RPC is really much easier to use than XML-RPC. The issue is whether you need type information at code time...
S. Alexander Jacobson
alex@...
Send Email
Dec 6, 2001
9:17 pm

PLEASE IGNORE LAST MAIL. IT WAS AN ACCIDENTAL SEND. -Alex- ___________________________________________________________________ S. Alexander Jacobson i2x...
S. Alexander Jacobson
alex@...
Send Email
Dec 6, 2001
9:18 pm

... so ... be ... Let me restate to see if I understand what you're saying. Changes in the use of existing real-world protocols happen though rough consensus...
Jay Carlson
nop@...
Send Email
Dec 4, 2001
8:14 pm
Advanced

Copyright © 2009 Yahoo! Inc. All rights reserved.
Privacy Policy - Terms of Service - Guidelines - Help