WTP should be able to run standalone
(especially since we need to be able to demo in in conference situations
where connectivity is unreliable). There should be no delays or blocking
validation problems.
One variant we should test is disconnected
use after a period of connected use so the cache gets populated.
We should also test out-of-the-box disconnnected
use, and disconnected use on new workspaces.
Arthur Ryman,
IBM Software Group, Rational Division
Craig Salter/Toronto/IBM@IBMCA Sent by: wtp-dev-bounces@xxxxxxxxxxx
10/18/2005 01:41 PM
Please respond to
"General discussion of project-wide or architectural issues."
To
wtp-dev@xxxxxxxxxxx
cc
Subject
[wtp-dev] proposed disconnected
testing phase
I'd like to propose setting aside a day or two of our M9 test phase to
dedicate to testing WTP without an internet connection. I know in
the past we've suffered from bugs in this area. Some of us are working
on solving some of these issues for M9. I'd like to ensure we give
this scenario enough focused attention during the M9 test phase to validate
fixes and flush out remaining bugs before we wrap up WTP 1.0 development.
For those of you who'd like to get started testing 'disconnected' early
a tracing tool is available that will allow you to capture stack traces
where 'http' internet accesses are attempted [1].
BTW, in WTP 1.5 I'd like to propose a development mode [2] that discourages
the use of internet resource references (that aren't mapped to local files
via XML Catalog etc.) and will provide error or warning messages to flag
these. This will be to support developers who want to tightly control the
resources that are getting fetched from the internet and ensure that disconnected
and connected development modes behave consistently (or at least as consistently
as possible). See the bug for more info.