Large processing tables

Kenneth Brody kenbrody at bestweb.net
Fri Feb 23 07:36:55 PST 2007


Quoting George Simon (Fri, 23 Feb 2007 10:01:44 -0500):

> I've received a couple of private messages from other that this has
> happened to, so at least I know I'm not alone.
> Short of upgrading to 5.6 and the obvious solutions like moving some
> code to called tables, is there any solution to this problem?

What's wrong with upgrading to 5.6?  :-)

> Here is what I have done so far:
>
> I've eliminated some long-winded remarks and moved all declarations to
> the bottom of the table, but it is still not enough.

Comments and declarations do not generate any code.  (Aside from a
blank source line generating one byte of compiled code for the line
itself.)

Eliminating comments has zero effect on the compiled code, with the
exception of eliminating a comment-only line entirely will "save"
one byte of code.

Moving declarations to the bottom will only "save" one byte for each
line that was eliminated from the top of the source.

Unless the code was just a few bytes over 128K, the above are not
likely to have any noticable effect.

> Any other suggestions?

You could make the code smaller by splitting it up and using CALL.

How large is the compiled code?  If you look at the tok file, the
32-bit little-endian value at offset 28 (1C hex) is the size of
the compiled code.

> -----Original Message-----
> From: Kenneth Brody [mailto:kenbrody at bestweb.net]
[...]
> Pre-5.6, filePro processing was limited to 128K of compiled code.
> This was increased to 2MB in the 5.6 release.  It's possible that
> you have run into some problem related to this.
>
> (Note that the tok file includes more than just compiled code -- it
> also includes all variable definitions, lookup descriptions, string
> literals, and so on.  There is no direct correlation between the
> size of the tok file and the size of the compiled code.)


--
KenBrody at BestWeb dot net        spamtrap: <g8ymh8uf001 at sneakemail.com>
http://www.hvcomputer.com
http://www.fileProPlus.com


More information about the Filepro-list mailing list