Hi James,
full regression testing isn't possible.
What does full regression testing mean :
- you have to run all automated test scripts on such CWS (hopefully
there are more than one bug fix integrated)
- you have to run these tests on all supported platforms
(Windows, Linux, Mac, Solaris Sparc, Solaris X86, ...)
- you have to make manual testing based on test case specification
for such feature
- additionally you should check this in different languages
As you can see, this is a mass to do and it will take some days/weeks
to realize this. So you have to ask for description, how to reproduce
the bug, how to verify the fix and what side effects are possible.
Based on this information (written down in the issue and/or in EIS for
the CWS) you have to create the test scenario. If this informations
are missing it will be very difficult to make a good QA. So ask for
such information.
How you could verify the issues and approve a CWS you can find
information in OOo Wiki under 'Approve a CWS' :
http://wiki.services.openoffice.org/wiki/Approve_a_CWS
Regards,
Thorsten
James Mckenzie wrote:
Full regression. One should NEVER depend on the developer knowing the effects
of their changes.
James
-----Original Message-----
From: Zhu Lihua <[EMAIL PROTECTED]>
Sent: Dec 3, 2007 8:22 PM
To: [email protected]
Subject: [qa-dev] A question about testing progress.
Hi,
If a developer modified some core code to fix a bug, it might influence several parts of the software. Will developer tell us tester which parts of the software need a focused testing? or should we start a complete regression testing to find out the new bugs?
Thank you!
Best regards,
zhulihua
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
