commons-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mark R. Diggory" <>
Subject Re: [math][functor] MathBeans was: (Re: [math] ... just one more reference...)
Date Thu, 23 Oct 2003 06:00:43 GMT
Ok, here's an implementation with an EvaluationContext, let me know what 
you think. Its almost java.lang.Number free, I cut corners and extended 
Number in the DefaultContext.DefaultValue InnerClass.

It doesn't depend on Functor because the Evaluation class doesn't really 
match up to that API now. :-( Oh, Well.


Paul Libbrecht wrote:
> Mark R. Diggory wrote:
>> I understand you logic behind maintaining a Context with info 
>> concerning precision. It is sensible.
> Note that this was not the context in the sense of a "configuration" or 
> such (e.g. that would say which algorithm should be used to do this or 
> that)... It was more the description of an evaluation-process.
> A more funky EvaluatingContext would be the computation of types or 
> return types....
>> precision is always going to be an important issue.
> That can all make it hard...
>> I've use commons pool before. The important thing to point out is that 
>> you need to always return your objects to pool after your done with 
>> them. It may be tricky to maintain if things like MathObjects can get 
>> handed outside of the EvaluatingContext in any way.
> At worst, that's done in the finalize method...
> Only routines that will return the object to pool will be efficient, 
> that's about enough for a rationale.
> Commons-pools is probably a good choice...
> Still not wanting to have this in commons-math ?
> Anything about this policy of having a single artifact per commons 
> sub-project ?
> Paul
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:

Mark Diggory
Software Developer
Harvard MIT Data Center

View raw message