On 1/10/21 4:09 PM, Andrea Pescetti wrote:
On 05/01/2021 Arrigo Marchiori wrote:
I tried to ask if there is a procedure to follow for testing [1].
Do we have any? I can follow it for the Italian builds.

Things are very different depending on whether one is testing a dev build or a release candidate.

A list of manual tests close to the "checklist" you were looking for can be found here (and these apply to all builds):

https://wiki.openoffice.org/wiki/QA/TestCases/#AOO_3.4_General_Testing_Test_Cases

and, still on the wiki, there is also a more complete dump of the former Testlink cases that I cannot find right now.
Maybe these?
https://www.openoffice.org/qa/testcase/


A RC (Release Candidate), instead, is very different. We ship both the source code and convenience binaries, but the Apache release process focuses mostly on legal checks to be done on the source code.

Details about voting on a RC:
http://www.apache.org/legal/release-policy.html

An example of OpenOffice RC testing (this is what I usually do; but other people run different tests):
https://s.apache.org/openoffice-416-tests

There are also automated tests you can run against a build.
https://wiki.openoffice.org/wiki/QA/test_automation_guide

I don't use the eclipse instructions and the Ant command line way is at the end.

I've been trying to clean these up and it's work in progress.
The BVT test for trunk are in pretty good shape. All 80 pass for me now.
Some in 4.1 are still flaky though. pass/fail intermittently.
Threading/Processor speeds are my theory.
I'm working on cherry picking a lot of test fixes from trunk to 4.1 but they're not ready yet. There was a lot of API changes included that I don't think we should backport.
Anyway it won't matter after 4.2 anyway :)

Best regards,
Carl

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Reply via email to