[Shootout-list] process creation & message passing

Bengt Kleberg bengt.kleberg@ericsson.com
Tue, 19 Oct 2004 17:17:58 +0200


Isaac Gouy wrote:

 >> the time it takes to create, and maintain, a variable amount of 
processes, not something else.
 >
 >
 >
 > Why a "variable amount of processes"?


because i think that is valuable information.


 >> i disagree. even languages that use kernel threads can create
 >> threads. it is not impossible.
 >
 >
 >
 > Yes they can, and the number of kernel threads that can be created is
 > limited by the available memory on the test machine.


i agree. the number of threads (kernel or non-kernel) is limited by 
available memory. i am interested in the limit.

what is your point in this discussion?
the original quoute that started my argumentation was from you:

''The new process creation & message passing tests will not provide a
measurement of the time to create a thread, or the time to pass
messages between threads, for most of the languages in the Shootout.''

i say the new tests will provide measurement of the time to create a 
thread, and the time to pass messages between threads. you say?


...deleted

 >> otherwise we end up with test results like the current objinst. i 
think objinst is a failure.
 >
 >
 >
 > Why?


because it claims to test
''the speed of object instantiation in OO languages''
and it has several entries of non-oo languages.


...deleted

 > Can you read Java?


slowly, if i have to. it was 5 years ago i stopped doing java 
programming. even then i was not doing advanced things.


 >> how do we get a graph over n from this test?
 >
 >
 >
 > You deleted the series of measurements given in the original posting!?


i was trying to interpret the text. the series of measurements was the 
desired result. i wanted to know how to get to it. i did not then, nor 
do i now, doubt its existence.

i think you have answered the question in other email in this thread 
(process creation & message passing)


bengt