Lynn W. Taylor wrote:
> P.S. the short version of everything I just wrote: credit is not 
> science, it's economics (money).
> 
> Martin wrote:
> 
>> For myself, I much prefer the idea of scientific rigorous direct 
>> measurement certainty over that of a complex statistical guess...


I like the 'short' version! Much more readable even if not fully 
understandable...


An important point is that the current credits are many different things 
to different people, and they /appear/ to keep /changing/ ... And I have 
no respect for what the current credits might represent. I don't believe 
they truly represent Cobblestones.


Do we really want to keep them as a /variable/ that is some vague 
abstraction that gets twisted as projects and computer hardware and 
algorithm optimisations progress (and change)?


Hence, let's *make* the credits a _defined measurable_ that can then 
/also/ be used as a 'currency' for 'bragging rights'.

That means we can then also say meaningful things such as project X 
required Y amount of compute resources to get it's final results. This 
is more/less efficient by ZZZZZZ dollars over using a dedicated 
supercomputer.

Or we can say algorithm A is in whatever way 'better' wrt compute 
resources used as compared to algorithm B.


For the sake of taking a little care over a general purpose reference 
computer, that appears to give a very good scientific 'plus'.

We also hopefully get rid of a LOT of futile arguments and bad feelings.


Regards,
Martin

-- 
--------------------
Martin Lomas
m_boincdev ml1 co uk.ddSPAM.dd
--------------------
_______________________________________________
boinc_dev mailing list
[email protected]
http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev
To unsubscribe, visit the above URL and
(near bottom of page) enter your email address.

Reply via email to