Hi Folks,
The httpclient4-based provider has been in Equinox/Eclipse
milestones for a little while now [1], and so far there has only
been one clear regression reported...and then fixed [2]. I think
this is good news...and bodes very well for using httpclient4 in
Kepler.
I would, however, like to suggest that some more p2 community
testing...and public reporting...be enlisted before
Kepler...particularly around the use of proxies and, in
particular, NTLMv2 proxies. I've have personally done some proxy
testing in the environments available to me (a local proxy
server), but this is not at all complete...given the number of
types of proxy servers, the varying impls/versions of them, the
interaction between the platform proxy API and proxy config/auth
information, and other idiosyncracies of many...particularly
corporate...networking environments.
Given how well things have been going, I'm very optimistic that
the issues will be minor. But if additional issues are found, I
will commit to help deal with them...but depending upon timing I
may need to ask p2 committers and community to assist...especially
if the issue is tied to a specific proxy or network environment.
Thanks,
Scott
[1]
https://bugs.eclipse.org/bugs/show_bug.cgi?id=337449
[2]
https://bugs.eclipse.org/bugs/show_bug.cgi?id=402740
On 2/22/2013 7:55 AM, Pascal Rapicault wrote:
