On 29/01/14 19:00, Michael Brunnbauer wrote:
Hello Andy,
On Wed, Jan 29, 2014 at 06:40:05PM +0000, Andy Seaborne wrote:
[https://issues.apache.org/jira/browse/JENA-627]
The key issue here is resourcing.
It would take me a lot longer than one hour as I don't know your data or
indeed your query and it's history. You probably want me to spend my
time on the bits I can help with
So you *want* a complete minimal example.
Yes - and as much information as possible. I'm sure that hour spent
tried various things e.g. branch order, specific FILTERs.
I'm sure there is more information than the reduced test case. What
happens if the UNION branches were swapped? Why the LIMIT?
This may seem trivial but someone coming cold to a test case has to get
to grips with what's going on.
You think my minimal example may not be complete regarding my problem?
Why is that problematic? I provided a minimal example for a real Jena bug.
If you fix that bug and my problem persists, I have no problem coming up with
another minimal example.
There is a spectrum from making contributions/working on the code base
directly, through people testing development, people testing only
releases, to people who ask questions.
I can test release candidates - but not with all applications.
Sure - I would not expect anyone to run dev builds in production except
to explicitly pick up bug fixes. Do you have test or dev environments?
Because we run "clean trunk", releases are just the code in the dev tree
at the time of release so it gives you advance warning.
Imagesnippets -
the application that uncovered JENA-627 - is beta but in production mode.
IMO, a test case constructed from it would diverge from the application
soon (too many changes to the SPARQL queries).
Regards,
Michael Brunnbauer
