On Tue, Nov 17, 2009 at 11:56 AM, Kalle Korhonen <kalle.o.korho...@gmail.com> wrote: > On Tue, Nov 17, 2009 at 11:01 AM, David Rees <dree...@gmail.com> wrote: >> On Tue, Nov 17, 2009 at 2:00 AM, Ciaran Wood <ciar...@gamesys.co.uk> wrote: >>> The issue you were thinking of is >>> https://issues.apache.org/jira/browse/TAP5-714 - there is a patch w/tests >>> attached to the issue so a committer can probably fix this one! >> >> Thanks, voted. BTW, it affects all versions of IE that I tested. >> It's a show-stopper, not a minor issue - should raise the priority, >> especially since it has a patch. > > Agree, maybe not a showstopper, but not a minor either.
Seems pretty critical to me if it breaks the browser with the single largest market share - but either way - the bug has been known for 6 months now with a patch - yet the patch hasn't been applied anywhere as far as I can tell. >> How can we encourage the developers to get a release out with some of >> these issues? Are people building their own local version with >> patches applied? > > Starting to veer off-topic, but I think the issue there is that Howard > has traditionally done all the releases, but he hasn't been actively > involved in bug fixing lately (and probably shouldn't need to be). I agree. Aren't there a handful of Tapestry devs? > Howard's also been complaining about some difficulties with releasing > but I don't know the details. I know though that setting up a fully > automated release process with Maven can sometimes be taxing but the > efforts are well worth it. Especially for bug fix releases, > confidently being able to roll up yet another point release is great > (at times, I release my projects a few times a week since it's so > simple). I fully agree with you that more frequent point releases for > T5 are badly needed. I concur as well, here. Maven should make it easy to roll out point releases, no? Apply patch to 5.1.0 branch - bump version number, build and deploy. > I'd avoid building your own versions if at all possible. Not because > building would be too difficult, but with local versions, you need to > deal with everything else included - deploying, maintaining, > distributing and keeping track of versions. Use nightly snapshots from > Formos if you have to and lock down to specific unique snapshot > whenever possible. I would like to avoid it as well - but do you have an alternative? It's either apply a 1-line patch to Tapestry 5.1.0.5 and figure out how to build it (call it 5.1.0.5.1 or 5.1.0.6-SNAPSHOT internally or something), or go through hundreds of lines of my own code replacing single quotes with double quotes and more testing to work around this bug. -Dave --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@tapestry.apache.org For additional commands, e-mail: users-h...@tapestry.apache.org
