Subj : Re: Possible Bug in js_NumberToString?
To   : netscape.public.mozilla.jseng
From : "Stephen Jones" <Stephen@Jones-net.com>
Date : Sat May 10 2003 12:51 pm

I'm pretty sure it's 1.5-rc4 . . . could be rc3a, because I have both
archived on my machine, but I'm pretty sure it's rc4 . . . any place I can
look in the code to be sure?

BTW: This only seemed to be happening on Windows, the MacX platform seemed
fine.

Side Note: We're extremely pleased with SpiderMonkey. We were using the
Microsoft Active Scripting engine in our product and when we ported it to
the Mac, we needed another solution. The speed increases we're seeing over
the Microsoft engine are incredible. The static methods and garbage
collection make SpiderMonkey a little difficult to deal with, but once
integrated, the engine seems very solid.

I heard rumor at some point of a 2.0 version. Is this in the works? We would
love to see a C++ solution. Accepting any design input?

Thanks for all the hard work,
- Stephen


"Brendan Eich" <brendan@meer.net> wrote in message
news:3EBC2BBB.9050401@meer.net...
> Stephen Jones wrote:
>
> >In js_NumberToString, for doubles with large mantissas, JS_dtostr
ASSERTS. I
> >increased the size of the buffer passed into JS_dtostr and it works
> >perfectly.
> >Comments?
> >Thanks,
> >- S
> >
> >
>
> What version of the code are you using?
>
> /be
>
> >
> >
> >
>

.