On Wed, 2005-11-02 at 07:57 -0800, Wes Hardaker wrote:
Dave> - does the default configuration compile successfully?
Wes> I've always argued that that's not a fair test. We should be
Wes> determining if the most common code fails that test, not just
Wes> the default build.
Two comments:
a) I'd argue that the most common code should (normally)
be *part* of the default build.
b) There may well be particular exceptions to a), in which
case I'm quite happy to accept these as nominally part
of the default build for this purpose.
But that should definitely be an unusual case - the
expectation should be that common code would be included
by default, unless there are clear and good reasons.
Not being bothered to adjust the configure script doesn't
qualify, IMO :-)
(And such exceptions should be clearly identified in advance)
> Also, if SNMP GETs were broken for a certain major
> handler (say iterator code) in the core code
Then there should be a test case to check this!
We really ought to be trying to bring the quality of our practise
up to the level of our procedures - not weakening the procedures
to pander to our poor practises :-)
> It's impossible to define a concrete set of rules that will solve
> every possible problem we'll face at release time.
Of course. I've said time and time again that I regard such rules
as a framework to support our work, rather than a straighjacket
to limit our flexibility. But if the rules are generally sensible,
then we should try to avoid having to break them wherever possible -
and address the underlying deficiencies when we do.
In this case, I'd suggest that this means:
a) Looking at the possibility of including embedded perl
support within the default build (with suitable checks
to ensure the environment is suitable, naturally).
Or at the very least, clearly documenting that this
should be regarded as part of the core functionality
(and explaining why it's not part of the default config)
*and*
b) Writing a "make test" case to check that this is
functioning correctly. If that had been done before,
we'd presumably have picked up this problem earlier.
> That being said, I'm sort of on the fence about whether the embedded
> perl support falls on the critical line. It is on by default in our
> RPMs
In which case, my position is that this should be part of the default
configuration (at least for Linux systems). But I've argued that
one before, too....
Dave
-------------------------------------------------------
SF.Net email is sponsored by:
Tame your development challenges with Apache's Geronimo App Server. Download
it for free - -and be entered to win a 42" plasma tv or your very own
Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php
_______________________________________________
Net-snmp-coders mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/net-snmp-coders
