RE: svn commit: r27468 - in trunk/subversion: libsvn_client tests/cmdline

Paul,
I am not sure what is causing this, some *spurious linkage* issue. It seem to happen only with ra_local, when I ran against ra_dav it passed.

By "*only via testsuite*" I meant running the test via our python testsuite. If I put a 'return' just before the problematic merge and do a *problematic* merge with the latest installed command line I could see the installed binary meeting the expectation.

P.S: From yesterday onwards(before r27468) I could see merge_tests 48 failing over ra_dav!!

With regards
Kamesh Jayachandran

-----Original Message-----
From: Paul Burba
Sent: Tue 10/30/2007 9:03 PM
To: dev@subversion.tigris.org; Kamesh Jayachandran
Subject: RE: svn commit: r27468 - in trunk/subversion: libsvn_client tests/cmdline
 
In IRC today:

<Kamesh> pburba: I could see merge_tests-71 failing since r27468 over ra_local *only via testsuite* not via the installed binary
<Kamesh> If I just revert r27468 then test passes via testsuite
<pburba> Kamesh: Can you paste the log, it passes for me.
<Kamesh> Paul: http://www.pastebin.ca/755205

Hi Kamesh,

I'm at a bit of a loss here, this test passes for me and it isn't failing on any of the buildbots. Can anyone else replicate this failure?

Also, could you explain "*only via testsuite* not via the installed binary"? I'm not sure what you mean by that.

Thanks,

Paul
Received on Tue Oct 30 20:54:25 2007