[Techtoolslist] OT: FIDE x64 and the future

Tim Matthews trmatthe at gmail.com
Mon Feb 23 13:38:02 EST 2009


Hi all,

I just wanted to chime in here. First, thanks to James for FIDE and the many
hours of easy 9010A programming it has provided me.

Secondly, I recently upgraded to Vista x64 and found that FIDE no longer
works. I was planning to mail James separately but this seems an appropriate
forum. The issue as Martin has alluded is 9lc.exe and its status as a 16bit
exe. As there is no 16bit run time or thunking on x64, it fails badly. I was
happy to accept this as recently I've just been using FIDE to suck scripts
from the Fluke for storage. Unfortunately, it seems FIDE tries to run 9lc
when most FIDE - Fluke IO operations occur. These fail and FIDE refuses to
go further. Is it possible to provide a short-term fix build which allows
the FIDE - Fluke IO without calling 9lc? A sort of compiled object only
mode?

The second part, and that which I think is much more important for the
future when x64 and/or non-intel CPUs become ubiquitous is to define a
replacement for 9lc. I've not performed any analysis yet, but I would
anticipate it's a simple, non-relocatable compiler. Are there any details
'out there'? Does anybody have any Fluke documentation or tech notes on the
format? I've written compilers before and hopefully in the reaches of my
brain, enough information remains to get a basic replacement coded. I don't
think I have enough time to reduce 9lc back to source to determine the inner
workings, but I am confident I can find enough time to code a replacement
from a partial spec.

To make a production grade replacement, I'd need plenty of varied source
scripts too; if this takes off I'll be back for those later. It would of
course be released to the public domain.

thanks,
tim



More information about the Techtoolslist mailing list