Hi Torsten, Torsten Werner wrote:
On Sat, Feb 27, 2010 at 9:33 PM, Eric Lavarde <Eric@lavar.de> wrote:I'm not a big Java specialist, just a packager, but we've seen quite a lot of FTBFS bugs due to the upload of ant 1.8. I'm personally hit by one for FreeMind, so I would support Matthias' suggestion until someone explains what has actually changed in ant.We have a bug in gij, but not in ant.
Sure, but...
Well, just after ant 1.8 has been uploaded, we've got at least 10 FTBFS bugs. Second, if I look at http://apache.mirror.iphh.net/ant/README.html, I see quite an impressive amount of "Changes that could break older environments".One more argument: ant 1.8 is obviously not backwards compatible with ant 1.7; this makes it a transition, which shouldn't be started that late and without preparation in the release process.Why do you think that it is not backwards compatible?
And, third, specifically for FreeMind, I suspect I've been hit by: * when nested into builds that have been invoked by <ant> tasks <subant> might set the wrong basedir on the called projects. Bugzilla Report 30569.because upstream seems to have worked around this bug, build doesn't work anymore. So, of course, you might argue that the bug is in the build procedure of FreeMind (and you would be right), but fact is that ant doesn't work the same as it used to do (for whichever good reason), which is the definition of a lack of backwards compatibility.
I'm not against upgrading to ant, but I would had prefered to not be obliged to work in urgency to fix a problem caused by somebody else's upload.
Thanks, Eric
Torsten