Re: [Geiser-users] possibly geiser bug

[Jose A. Ortega Ruiz]

On Thu, Jan 31 2013, Michael Filonenko wrote:

> Hi.
>
> First of all thanks for geiser, it is very usefull tool for "racketing".
>
> I use unsafe float point operations (unsafe-fl+, etc) for perfomance
> reasons; also I use racket library `disassemble' to check "jitted"
> function code.
>
> So, the problem is that when I work and decompile jitted function in
> geiser repl session the resulting assembler code is not equal when I
> do the same things in built-in repl session (just racket with xterm).
>
> Is this issue relative to geiser?

Could be.  Geiser creates its own REPL and one of the things it does is
requiring errortrace/errortrace-lib (in eval.rkt) at startup to provide
better diagnosis: that's possibly having an impact.  A second thing it
does is setting the parameter enforce-module-constants to #f, to allow
redefinitions inside modules without restarting the REPL, and i'm pretty
sure this is also going to affect the generated bytecode.

HTH,
jao
-- 
Humans think they are smarter than dolphins because we build cars and
buildings and start wars etc., and all that dolphins do is swim in the
water, eat fish and play around. Dolphins believe that they are
smarter for exactly the same reasons. -Douglas Adams