[Shootout-list] demoting marginal languages

[Bengt Kleberg]

Brandon J. Van Every wrote:
> Isaac Gouy wrote:
> 
>>Yes, I suggest removing languages that have so little interest no one
>>has bothered to implement 15 of these trivial programs.
> 
> 
> There is a slippery slope here, however.  In open source land, labor for
> doing this kind of work is not plentiful.  I think we should be
> repositing people's incremental results, not requiring some bold
> volunteer to step forward and write all the tests for a given language
> in one fell swoop.

this is correct. if there is no result from writing 14 tests, just 
because number 15 remains, it will be difficult to motivate unpaid 
programmers to write any tests.


> I think it's reasonable to deny a LCD score to any language that can't
> complete all LCD tests.  i.e. effective LCD score of 0.  However, I
> don't believe the language should be removed from the Shootout, and I
> think the data on individual tests should be available on the language's
> webpage.  It's the composite LCD score I'm talking about suspending.
> The "front number."  Can't complete the tests, then you don't get a
> ranking.  That's the penalty for a language community being asleep at
> the wheel.
> 
...deleted

> Regarding the issue of too many languages to look at: if a LCD benchmark
> is fairly agreed upon, then it would be most fair to list languages in
> order of their performance.  A tab to re-list in alphabetical order
> would be desireable, but 'by performance' should be the default.  That

...deleted
> Thus the problem of information overload scales up and solves itself.
> We aren't really going to end up with issues of "which Scheme
> implementation."  The best implementations will rise up the ladder.

lots of good thinking here. i agree.

however, what is it that stops us from giving a language that does not 
have an implementation of a test a very bad score (maximum time 
exceeded) on that test in the basic benchmark. the problem of 
information overload scales up and solves itself. we aren't really going 
to end up with issues of ''which language should be in the basic 
benchmark score card''. the best implementations will rise up the 
ladder. :-)


bengt