[Libreoffice-qa] Thread on QA done by l10n teams

[Nino Novak]

On Friday 02 March 2012, 18:26:32 Sophie Gautier wrote in libo-l10n:

> http://listarchives.libreoffice.org/global/l10n/msg04226.html

Hi Sophie, all, 

I'm not subscribed to the l10n list, so I'm trying to answer here. Also, I 
don't know about others, I just can tell what I'm doing myself:

> - Are you doing manual tests on the builds?

yes, if time allows on most of Beta and RC builds

> . using Litmus?

No, just for lack of time.

> - Are you running other types of tests

yes

> .which one?

"en passant silent testing" :-)

Means: I just install the release in parallel and try to use it as much as 
possible for my daily tasks. With critical documents, I work on a sandbox copy 
and replace the original later (only if all went smooth)

That way, testing does not take much time (just installing the version and 
creating a testing copy each time I want to work on a critical document) but 
is very effective as I perform "real tasks" (and not artificial test cases) 
with real data (i.e. sometimes large old documents).  

BTW, nobody will notice this kind of testing (of course unless a problem or 
bug comes up). That's why I've suggested to implement some kind of simple 
logging ability into test builds but unfortunately, this seems to be too 
complicated for the devs. 

BTW, I'm doing "functional testing" only but as I use the localized (German) 
Version, I hope I will notice l10n problems, too (and have so in past).

(I believe, that many many people do such "personal silent testing" without 
anybody knowing how many tests were performed on what platform and with which 
kind of documents)

Nino