Re: Proposed tags for cataloging

[William Cohen <wcohen at redhat dot com>]

Frank Ch. Eigler wrote:
> wcohen wrote:
>
> > [...]  following is a list of tags with quick outline what each
> > does. [...]
>
> Tags that are not forseeably used by programs need not be such formal
> tags.  Do you see some e.g. test suite script behaving differently
> based on "version", "author", "subsystem", ...?  Might a user need to
> filter scripts by whether or not they do "reporting on-exit"?

Having a tag summarizing when and kind of output has value for testing; it gives 
the testing frame work some indication when or whether output should be 
expected. Users may also want to know whether a script prints data while running 
or on exit. People may be surprised that nothing comes out when running the 
script until control c is hit.

"exclusivearch" useful for scripts that are designed to only run on particular 
systems. Hopefully, this will be rarely used.

The "version" is allow people to know that the version in the catalog has changed.

"author" probably isn't interesting for most purposes

Maybe just lump the "subsystem" into the "keyword".

> IMO, let's keep it absolutely simple and get on with actually
> documenting/collecting all the samples.  keywords (which can be
> grepped, and indexed in the catalogue), and a summary sentence may
> well be enough.  The test system might like a
>
>    test-invocation: stap script ARG1 ARG2


The tst-invocation is good suggestion.

> to trigger its automatic execution during "make installcheck".

"status" also factor into the "make installcheck" only run ""production" as pass 
5 others run as pass 4 to check that they build.

-Will