Hi James,
James Mckenzie schrieb:
Clytie:
I will add that I agree the use of automated testing is paramount to the
completion of QA on OpenOffice.org.
We do not have enough people to do manual testing, even for the American
English version as we would
have to go through each menu, dialog, wizard, and help file (just to begin
with) in addition to testing all
functionality of the program.
Pleas don't mix up the diffent tests. The "go through all menus and
dialogs" test is not in Release Sanity scenario. So this test is not
mandatory for an approval.
And if we miss one item, our users will find them and maybe/maybe not repoort
it. More likely, they will tell their friends about the poor quality control and ask them to not use it.
*sigh* and why do you think, that testtool would help there? Someone
needs to write the scripts and might miss some functionality and / or
dialogs. Users might find the bugs and you are exactly in the same
situation.
There is *always* the risk to deliver a version that might have bugs.
And thats why the QA representatives are so important: they need to
judge the risk.
Furthermore I see more and more complaints about manual testing (some
people seemingly like to skip manual tests at all and do only automated
tests instead). But manual tests have something that you have not in
automated tests: a person who executes the tests and can use his / her
brain for the current test case. Using brain instead of a pre-written
test script is much more like the normal user's behaviour.
So after all .. I don't get your point: do we to much testing? Do we to
little testing? Pleas don't tell that we do the wrong testing .. rather
tell how to do the right testing.
André
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
