1. Re: Euphoria bug (Robert)
- Posted by Vincent <darkvincentdude at yahoo.com>
Sep 15, 2005
-
Last edited Sep 16, 2005
Bob Thompson wrote:
>
> > >Matt Lewis wrote:
> > >What I find interesting is that you only have to assign something to a
> > >sub-sequence:
> > >}}}
<eucode>
> > > s[1][1] = 0
> <font color="#330033">> <font color=</font><font
> color="#00A033">"#330033"</font><font color="#330033">>></eucode>
{{{
</font></font>
> > >...before the loop, and it runs quickly.
> > That does not seem to make a difference here.
>
> Are you sure Pete? I thought a variation on Matt's code would be a neat
> temporary fix.
>
> sequence s
> atom t
> object data
> constant repeats = 100000, len = 1000
>
> s = repeat(repeat({}, len), len)
> s[1][1] = {}--doesn't change initial structure and makes it fast
>
> t = time()
> for n = 1 to repeats do
> s[999][999] = 1
> data = s[999][999]
> end for
> ?time()-t
>
> machine_proc(26,0)
>
It's pretty strange how adding "s[1][1] = {} or 0" makes the bug disappear.
Robert, have you tracked down the location of the bug in the backend?
Any chance of finding other bugs in the same league as this one?
Don't forget to add this to the bug list on the "next release" page, so you dont
forget to fix it. I find it very alarming that a bug like this can slow down
program execution 140-170 times {according to time(): 8.50/0.05}. It seems this
bug has been around for a long time.
I want it smashed! 
Regards,
Vincent
----------------------------------------------
___ __________ ___
/__/\ /__________\ |\ _\
\::\'\ //::::::::::\\ |'|::|
\::\'\ //:::_::::_:::\\ |'|::|
\::\'\ //::/ |::| \::\\ |'|::|
\::\'\ //::/ |::| \::\\|'|::|
\::\'\__//::/ |::| \::\|'|::|
\::\','/::/ |::| \::\\|::|
\::\_/::/ |::| \::\|::|
\::,::/ |::| \:::::|
\___/ |__| \____|
.``.
',,'
