Hi Claude, thanks for your contribution, I just have a couple of questions for you:
How was this patch generated?
It appears to replace almost the entire class some bits of which appear to have minimal changes which suggests a mismatch in your SVN EOL settings, this is an issue I've run into in the past when developing patches on Windows - see http://mail-archives.apache.org/mod_mbox/incubator-jena-dev/201111.mbox/%3CC0B6979A3CA668458B697E4EA907CA9401825F@CFWEX01.americas.cray.com%3E for ways to solve this. If you are able to regenerate the patch so that it contains only the actual changes that would be preferred as it is much easier to review and integrate if the patch only contains your changes.
Also have you considered writing any tests for this?
It should be fairly trivial to write a test to a known non-existent server (maybe localhost on some random port) which demonstrates that timeouts do apply as expected after these changes. While lack of tests is not a barrier to integration it's another things that makes it easier to review and integrate