Re: [Factor-talk] A performance challenge

2015-01-25 Thread John Benediktsson
If you want to avoid those dynamic type checks and not use TYPED, another 
alternative is declare:

{ object array fixnum } declare



 On Jan 25, 2015, at 9:12 AM, Björn Lindqvist bjou...@gmail.com wrote:
 
 2015-01-18 2:44 GMT+00:00 John Benediktsson mrj...@gmail.com:
 Also, minor comment, instead of:
 
length [ 0 ] { } replicate-as ;
 
 You can just do:
 
length 0 array
 
 Thanks!
 
 And instead of the array-nth stuff, you can just do some type declarations
 and the compiler should make it the same as your array words:
 
{ fixnum array } declare nth-unsafe ;
 
{ array } declare first2-unsafe ;
 
 Oh of course. When I wrote the code I was trying to stay stay close to
 the metal and tell Factor exactly what code it should generate. Even
 if it figures it out equally well on its own anyway.
 
 I recommend using TYPED: or TYPED:: declarations to specify neighbors is an
 array and normal sequence words?
 
 I was trying that, but Factor appears to add dynamic type checks to
 words which call typed words which increase the size of the generated
 code.
 
 
 -- 
 mvh/best regards Björn Lindqvist
 
 --
 New Year. New Location. New Benefits. New Data Center in Ashburn, VA.
 GigeNET is offering a free month of service with a new server in Ashburn.
 Choose from 2 high performing configs, both with 100TB of bandwidth.
 Higher redundancy.Lower latency.Increased capacity.Completely compliant.
 http://p.sf.net/sfu/gigenet
 ___
 Factor-talk mailing list
 Factor-talk@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/factor-talk

--
New Year. New Location. New Benefits. New Data Center in Ashburn, VA.
GigeNET is offering a free month of service with a new server in Ashburn.
Choose from 2 high performing configs, both with 100TB of bandwidth.
Higher redundancy.Lower latency.Increased capacity.Completely compliant.
http://p.sf.net/sfu/gigenet
___
Factor-talk mailing list
Factor-talk@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/factor-talk


Re: [Factor-talk] A performance challenge

2015-01-25 Thread Björn Lindqvist
2015-01-18 2:44 GMT+00:00 John Benediktsson mrj...@gmail.com:
 Also, minor comment, instead of:

 length [ 0 ] { } replicate-as ;

 You can just do:

 length 0 array

Thanks!

 And instead of the array-nth stuff, you can just do some type declarations
 and the compiler should make it the same as your array words:

 { fixnum array } declare nth-unsafe ;

 { array } declare first2-unsafe ;

Oh of course. When I wrote the code I was trying to stay stay close to
the metal and tell Factor exactly what code it should generate. Even
if it figures it out equally well on its own anyway.

 I recommend using TYPED: or TYPED:: declarations to specify neighbors is an
 array and normal sequence words?

I was trying that, but Factor appears to add dynamic type checks to
words which call typed words which increase the size of the generated
code.


-- 
mvh/best regards Björn Lindqvist

--
New Year. New Location. New Benefits. New Data Center in Ashburn, VA.
GigeNET is offering a free month of service with a new server in Ashburn.
Choose from 2 high performing configs, both with 100TB of bandwidth.
Higher redundancy.Lower latency.Increased capacity.Completely compliant.
http://p.sf.net/sfu/gigenet
___
Factor-talk mailing list
Factor-talk@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/factor-talk