Frank Schönheit - Sun Microsystems Germany wrote:
Hi Nikolai,
Hi Frank,
The Valgrind ID is indeed a needed information. It is used to identify
issues during the Valgrind run when IssueZilla ids are not yet known. It
is not needed for the developer fixing it, but helps to relate the tasks
easily to the valgrind-result-data.
Ah, so we have some database tracking the valgrind issues during the
test runs, and the IZ issues are submitted from this DB? This explains
the need for an ID, thanks for the clarification.
Well except from that it is basically just a file and not a database
where we keep track of what valgrind already found in previous runs and
for which errors tasks have already been submitted you are right there
is additional storage involved. With each valgrind run that file gets
updated. If an error already contained there would be found by valgrind
no new entry would be generated in that file. Only new things get a new
entry with a new Valgrind ID there. We than have a program submitting
issuzilla tasks for new issues contained in this file after a valgrind
run and which in turn updates state information in that file by looking
at the already submitted issuezilla tasks. As later on eventually
issuezilla tasks may be set to duplicate it´s quite possible that two
generated Valgrind ID´s might be fixed with one issuezilla task (when
you resolve the issuezilla issues to the last one in the list formed by
set duplicate-to fields of issues).
So I'd suggest:
- Summary looks like: Valgrind ID: <ID>, <ErrorText>
Valgrind <ID>: <ErrorText>
?
- The issues get the keyword "valgrind".
+1 :)
Thanks & Ciao
Frank
Kind regards,
Bernd
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
