beneroth changed the topic of #picolisp to: PicoLisp language | The scalpel of software development | Channel Log: https://libera.irclog.whitequark.org/picolisp | Check www.picolisp.com for more information
iacore has quit [Ping timeout: 244 seconds]
bjorkintosh has quit [Remote host closed the connection]
bjorkintosh has joined #picolisp
bjorkintosh has joined #picolisp
bjorkintosh has quit [Changing host]
ello has quit [Ping timeout: 252 seconds]
rob_w has joined #picolisp
chexum has quit [Remote host closed the connection]
chexum has joined #picolisp
iacore has joined #picolisp
<tankf33der> hi all
<tankf33der> libgmp is ok with mul and div on range 50-100M digits.
<tankf33der> next challenge is pil vs. libgmp
<tankf33der> on huge digits length
<abu[7]> Remember that picolisp is not optimal for big numbers
<abu[7]> ie stored in cells
<tankf33der> sure. i do optimizations.
<abu[7]> Yeah, but direct bignum buffers in C are easier to make fast
<tankf33der> is it easy to demonstrate ?
<abu[7]> The speed diff?
<tankf33der> how to use buffers
<abu[7]> malloc()
<abu[7]> PicoLisp does it with cell operations
<abu[7]> so needs double spake
<abu[7]> space
<abu[7]> and the cells are built dynamically for all calculations
<abu[7]> also, it allocates *new* cells instead of re-using large allocated buffers
<abu[7]> (8
<abu[7]> (+ Big1 Big2) makes a new number
<abu[7]> so a lot of gc'ing
<abu[7]> GMP needs no gc
<abu[7]> So I think it is rather useless to compare these two strategies
<tankf33der> ok
<abu[7]> Per it would be interesting to make an interface lib via native() calls to GMP? Tosupport features not supported by PicoLisp?
<abu[7]> ie. big floats
<tankf33der> i already did for big int
<tankf33der> should be easy
<abu[7]> ok
<tankf33der> in this link:
<tankf33der> did you mean chapter - "Floating-point Functions" ?
<abu[7]> Yes, something like that
<abu[7]> A general correct floating point library is not trivial
<tankf33der> and what next?
<tankf33der> do you have something specific in head ?
<abu[7]> Outside of bignums?
<tankf33der> no, inside
<tankf33der> next is test something ?
<abu[7]> Hmm, you already did almost all kinds oft tests :)
<tankf33der> ok
<abu[7]> GMP is really big
<tankf33der> bigints are 160 functions in total, i binded 60% of them
<tankf33der> rest are useless imho
<abu[7]> Wow
<tankf33der> all these ported
<abu[7]> Perfect!
<tankf33der> 60%
<tankf33der> away in 10mins
<abu[7]> 👍
rob_w has quit [Remote host closed the connection]
rob_w has joined #picolisp
<tankf33der> abu[7]: here ?
<tankf33der> afk
<abu[7]> Here
<tankf33der> back
<tankf33der> so
<tankf33der> scl function call should be in the beginning of file, right ?
<tankf33der> because this one already does not work
iacore has quit [Ping timeout: 245 seconds]
<tankf33der> afk
rob_w has quit [Quit: Leaving]
poosh has joined #picolisp
poosh has quit [Remote host closed the connection]
iacore has joined #picolisp
ello has joined #picolisp
ello has quit [Client Quit]
ello has joined #picolisp