Hi Gabriele
At 04:22 PM 7/5/00 +0200, you wrote:
>[EMAIL PROTECTED] wrote:
>
>> My first step is to step through the "C" code in the DB Kernel,
>> "clone" the code and the data structures in rebol. Make it work,
>> and THEN
>> optimize with rebol's own features.
>
>This way you'll get very ugly REBOL code, and you may have trouble
>converting most data structures. Also, you will not be able to
>optimize later, because the problem is with data structures
>design...
>
>You'll end up having to rewrite the code from scratch anyway. So,
>I suggest you to get confident with the database code (perhaps
>doing some tests in C) and then, after you know WHAT you should do
>(and not only how the C code does it), write the REBOL code
>directly.
I will wite one function call create, which creates a database.
It will be difficult, and it will take a long time.
Then it will be rewritten.
Everything else will flow from that. Easily.
I am already somewhat familar with the kernel and of course,
am intimate :) with the wrapper.
-T
>Regards,
> Gabriele.
>--
>Gabriele Santilli <[EMAIL PROTECTED]> - Amigan - REBOL programmer
>Amiga Group Italia sez. L'Aquila -- http://www.amyresource.it/AGI/
>
>
