Next in thread →
Next in month →
RE: [wsrp-interop] draft change request
Definitely stronger than I would prefer and an errata candidate,
as the "" replacement value would then have a new special
meaning.
Is ""
then ever a valid value for wsrp-mode & wsrp-windowState or does "where the
Producer has no value" imply that the current mode & windowState MUST now be
sent back?
Unfortunately, I will not be able to call in at the start of the meeting
tomorrow, so would just like to briefly re-iterate my opinion on this from last
week. Such a restriction is likely to (needlessly) cause interoperability
problems (due to producers that omit to replace some token) and consumers will
still need to protect themselves (from such producers).
If we
accepted this change request, I would filter written URLs at the
producer so that any remaining "{wsrp-*}" gets replaced with "". If we
also require the current mode and window state to be written into the URL
then that would be even more conformance tax IMHO. I would prefer only to have
MUSTs for the tokens that are actually needed to be replaced to achieve
consumer/producer interoperability.
regards,
Andre
-----Original Message-----
From: Rich Thompson
[mailto:]
Sent: 19 November 2003
14:44
To:
Subject:
[wsrp-interop] draft change request
Here is a draft for the change request I was asked to draft. I know
this is stronger than some would prefer, but hopefully it gives us a good
strawman for the discussion.Document: Spec
Section: 10.2.2 (Producer URL Writing)
Requested by: Interop
SC
Old text: These tokens are enclosed in curly braces (i.e. "{" and
"}") and contain the name of the portlet URL parameter that should be
replaced.
New text: These tokens are enclosed in curly braces (i.e. "{"
and "}") and contain the name of the portlet URL parameter which the Producer
MUST replace (using "" for those parameters where the Producer has no
value).
Errata?
Reasoning: The current ambiguity will make for
a lot of additional Consumer logic to protect itself.
Next in thread →
Next in month →