[AUUG-Talk]: What's wrong with the AOSA Awards.
db at dawnbreaks.net
Tue Dec 7 01:41:49 EST 2004
I'm putting this out in the open (where the relevant AUUG
committees also can read it) - I think the AOSA awards are
bit of a waste of time, as presently structured.
Yes, it's nice to acknowledge individual effort. A pat
on the back is always appreciated.
I believe there are better goals, but I don't believe
that AOSA even achieves this one very well. When asked to
vote recently, I discovered that my vote would essentially
have boiled down to familiarity with a given person or project
or technology ... or at least be directed approximately self-
interest in what I felt would be nicest to have more of. I
didn't really have the right information to judge the degree
of an individual's effort or ingenuity towards their alleged
goals. I couldn't determine which contributions to open-source
were really meritous. Yet awards are intended to acknowledge
merit. In any case, I didn't vote. I was sorry to see that
the clean-room BouncyCastle Java crypto implementation didn't
get a jersey, but since my grounds for preferring it do have a
lot to do with my enthusiasm for Java, and as this does not appear
to be a shared cultural value (which was underscored to me at
AUUG 2004) I was not too upset about it. Indeed - any complaint
would have been based entirely on values, and not on any comparison
of merit - I did not have the information for making an *argument*
that the BouncyCastle's contribution to open-source was more
Oh, I felt the categories were a bit artificial, too.
But these are minor quibbles - my real point is that I think the
idea of rewarding some vague culturally-based appreciation of merit
*post-hoc* is to miss a great opportunity - that of setting a direction and an
incentive structrure for doing the things that are *significant* for the
open-source catalogue: gaps in the lineup, unwritten but tricky libraries that
would enable new classes of applications, refactorings of code that would be
useful if it were better organised [decent dynamic word-wrapping for
editors, for example. Sheesh.] ... and so on.
Forget the gold tuxes and besties - let's put up some research
money to encourage projects which promise progress towards some
problem which AUUG thinks is significant. For example, a domain-
specific-language for writing video-drivers [I know nothing of device drivers,
but I know that C is a monster], some good-quality
tools, documentation, and a registry-style website that engaged
device manufacturers [in multiple languages] might make a difference
to availability of drivers for various open-source operating-systems.
Heck, if the DSL was good enough, hardware vendors might even stop
regarding their drivers as expensive secrets.
Sure, quibble with my example as being less important than something
else. Certainly, suggest improvements to the mechanism or deliverables
which would increase the chance of success. But please don't tell me
that the goal itself is not worth pursuing, or is not achievable for
some pessimistic reason. Such an (dare I say it) *innovative* endeavour
would be effort well spent. AUUG should not so much reward it as motivate
More information about the Talk