[OpenAFS-devel] warnings fix

[Derrick Brashear]

On Wed, Jul 15, 2009 at 3:30 PM, Tom Keiser<tkeiser@sinenomine.net> wrote:
> On Wed, Jul 15, 2009 at 12:01 PM, Jeffrey
> Altman<jaltman@secure-endpoints.com> wrote:
>> On a separate not, I take great offense at your comment about "shaming"
>> the gatekeepers into doing something. =A0Take a look at the analysis of
>> who actually has done something to improve OpenAFS
>>
>> =A0https://www.ohloh.net/p/openafs/contributors
>>
>> What you will notice is that it is the gatekeepers that do the vast
>> majority of the work. =A0That list is measuring authorship not commits.
>
> I find your interpretation of those numbers dangerously distorting.
> Firstly, what is meant by "to improve OpenAFS"? =A0If you're referring
> strictly to contributions accepted by the gatekeepers, then fine.
> However, I consider that a rather narrow interpretation given the vast
> amount of development work that goes on in the hopes that it will
> someday be accepted upstream.
>
> Secondly, commit count is hardly a useful metric, as it has only a
> slight positive correlation to actual work. =A0In addition, things would
> look quite different if the stats took into account the efforts of
> various community members whose major contributions have been going
> through the review process for the last several years. =A0Especially
> with the major development efforts, the current fiat-grounded power
> structure is biased in favor of contributions by the gatekeepers.
> Until a few years after we move to a just and meritocratic decision
> model with elected leadership, I don't think such metrics are likely
> to yield definitively useful information.

Which is exactly the sort of response I expected when I made the
followup comment I did, which I notice you
ignored.

I could say nobody else is bending over to clean up cross platform
build issues, but that would be a grave disservice to the people that
are. But no one was bending over, for instance, to fix the build issue
stemming from DAFS which was segfaulting the HPUX compiler; I had to
do it, despite not even having hardware to test on. Gerrit has made
contributing easier; talk is cheaper.

The rest of your point, well, anyone who says they could do a better
job reviewing large patches provided as simply a blob is either a
savant, or lying to you, so I'm not sure what sort of miracle you
expect from your proposition; If you'd like to get that person or
persons to start reviewing, they'd prove themselves in short order.

--=20
Derrick