Atleast I learned something today, thanx aurora.

Well u don't HAVE to use dos2 (is was an example)...
ofcourse all this is while not using BASIC (as I stated.. oh, a few times) (atleast for beginners?!)
why not put an init routine somewhere in hi-mem which sets pages and then jump to your code. well these are examples to learn from. they're absolutely not mandetory ofcourse..

2 cents I guess.

Quote:

ofcourse all this is while not using BASIC (as I stated.. oh, a few times)

Ah, but you also said:

Quote:

Wbass2 is a great development environment, especially for all the BASIC peeps (wb2 + basic runs smooth)

...which, to me, sounded like WBAss2 (and hence the program under development) runs in a BASIC environment. If that's not the case, I stand corrected. As said, I've never seen or used WBAss2.

2 cts, 2.

Assembling to another page will allow your source in WBASS to grow some more, but there is a point at which it is simply too large for WBASS to handle. When that happens, you either have to divide your code into smaller modules that can be assembled one at a time, or switch to another assembler.

If you want to have your assembler running on MSX, Compass is probably the best assembler for large projects. Otherwise, you can choose one of many cross assemblers (for C-BIOS we support Pasmo, SjASM and tniASM). I really prefer a cross assembler because it's much faster and has fewer limitations. Also you don't have to spend time setting up your development environment again after your program crashes the MSX.

I still use WBASS sometimes: it's very handy for small experiments, but not for large projects.

it's ALWAYS good to split sources in modules... why?
- faster assembling time
- better debug posib.
- better overview
- less mem use in development mode
- keeps you focusted (!)

good luck.

hmmm, I see,

So maybe I should consider a cross assembler.
I made a routine which loads several files into memory (including an .sc5 file)
All files have to be loaded from page 2 on....... (which sucks !!)

So this means when I start Wbass, I type
Page 2,3,0,3 and then I can use memory locations

$8000-$C800 (more or less)

Because I load the files in page 2, I cannot use page 1 any more (I used to code from $4000 -

So in total I still have only 16 kb of memory to use.....
Is there a better way in Wbass (Im not giving Wbass up right away)
to clear more memory.........

Quote:

Otherwise, you can choose one of many cross assemblers (for C-BIOS we support Pasmo, SjASM and tniASM). I really prefer a cross assembler because it's much faster and has fewer limitations. Also you don't have to spend time setting up your development environment again after your program crashes the MSX

that sounds better !!!
How does it work, under windows? and.... you dont need an msx emulator anymore???

Quote:

hmmm, I see, So maybe I should consider a cross assembler. I made a routine which loads several files into memory (including an .sc5 file) All files have to be loaded from page 2 on....... (which sucks !!) So this means when I start Wbass, I type Page 2,3,0,3 and then I can use memory locations $8000-$C800 (more or less) Because I load the files in page 2, I cannot use page 1 any more (I used to code from $4000 - So in total I still have only 16 kb of memory to use..... Is there a better way in Wbass (Im not giving Wbass up right away) to clear more memory.........

You clearly didn't grasp the whole concept yet.
A memorymapperblock (16k) can be set on ANY page like:
- page 1,3,0,3
This will set primairy slot 3 and secondairy slot 0 (for turbo R) to RAM mode and maps memorymapperblock 3 to it.
- page 2,3,0,3
does the same except this time on page 2
So now you'll have both page 1 and page 2 mapped with memorymapperblock 3 which basically means you'll end up having the SAME DATA on both pages.

So you're NOT limited to 16k. no mam! When using a memorymapper of, say, 128k you'll have the option to place memmap 0-7 on each page.

just remember that page 3 has page 0 mapped (standard variables etc) I should not ever change that if I were you. for the rest .. go ahead.

typically mem layout for me is:
page 0,3,0,3 ;F-kernel bios
page 1,3,0,2 ;F-kernel system (which will be called in BIOS)
page 1,3,0,4 ;CODE page (user)
page 2,3,0,1 ;Temp page for data transfer
page 2,3,0,5 ;DATA page (user)
page 3,3,0,0 ;Standard MSX vars etc

so the Kernel needs atleast:
page 0,3,0,3
page 1,3,0,2
page 2,3,0,1
page 3,3,0,0

whil MY program will use:
page 0,3,0,3 (partial left over space)
page 1,3,0,4
page 2,3,0,5
page 3,3,0,0

There's a lot of page switching involved during F-kernel calls. My code and data page are always in front. When using ANY f-kernel routines these pages MIGHT be changed temp. (but always reset while returning)

so you see; while using 128k I would STILL have 2 memmapblocks (32k) left for my RAMdisk! (block 6 and 7)