On Thu, Apr 3, 2008 at 8:39 AM, Archie Cobbs <[EMAIL PROTECTED]> wrote:
> I think I have enough information to go and work on a new resolver based
> on these ideas... will report back later...
>
Here's a very rough first cut at a new "builder" resolver.
Here's the basic idea: it works like URLRe
On Fri, 04 Apr 2008, Stefan Bodewig <[EMAIL PROTECTED]> wrote:
> The root of the problem probably is that FailureRecorder expects the
> Test instance passed into addFailure to be an instance of the class
> under test - something that isn't true for JUnit4 wrapped tests in
> Ant.
it also assumes t
On Fri, 04 Apr 2008, Stefan Bodewig <[EMAIL PROTECTED]> wrote:
> The strange thing is that even the oldest version of
> JUnit4TestCaseFacade[3] I can find contains such a constructor.
http://junit.cvs.sourceforge.net/junit/junit/junit/framework/JUnit4TestCaseFacade.java?hideattic=0&revision=1.2&v
Hi all,
we can finally see[1] why the tests fail in Gump (and this is the only
remaining failure on vmgump, BTW)
[junit] nested build's System.err:
/srv/gump/public/workspace/ant/src/etc/testcases/taskdefs/optional/out/FailedTests.java:9:
cannot find symbol
[junit] symbol : constructor
publishPermissions is fine for me, we can use the "chmod mode" terminology in
the documentation of this attribute.
Maarten
- Original Message
From: Xavier Hanin <[EMAIL PROTECTED]>
To: Ant Developers List
Sent: Friday, April 4, 2008 9:45:57 AM
Subject: Re: svn commit: r644541 - in /ant
On Fri, Apr 4, 2008 at 9:24 AM, Maarten Coene <[EMAIL PROTECTED]>
wrote:
> Btw, in the manual for chmod, they use the "mode" terminology for this
> kind of information.
Maybe it's only me then, but I think mode is too generic in this situation.
Moreover we have resolveMode option on the resolve t
Btw, in the manual for chmod, they use the "mode" terminology for this kind of
information.
Maarten
- Original Message
From: Maarten Coene <[EMAIL PROTECTED]>
To: Ant Developers List
Sent: Friday, April 4, 2008 9:19:57 AM
Subject: Re: svn commit: r644541 - in /ant/ivy/core/trunk: CHANG
On Fri, Apr 4, 2008 at 9:19 AM, Maarten Coene <[EMAIL PROTECTED]>
wrote:
> It was hardcoded to 0600 in Scp.java before I made this change. To keep it
> BC, I defaulted it to the same value.
OK, I didn't see that, I thought it was using umask according to the
discussions on the issue. Sorry about
It was hardcoded to 0600 in Scp.java before I made this change. To keep it BC,
I defaulted it to the same value. As far as I could see, the umask wasn't used.
Since this attribute is only used for publishing modules, not for module
retrieval, I think we should at least keep "publish" in the attr
