exxos's Atari, Amiga & retro forum

Another odd one.. Pressing any key or clicking the window causes the AES to send multiple WM_REDRAW events to the GB7 main window.. WHY ?!

Always seems to be

X,Y,W,H. Main window pretty much.

exxos wrote: 02 Dec 2022 22:08 So it may not like it but the menu was drawn over the top and then turned into a white square and then restored correctly :D

Huzzah!

Well done. :)

BW

Four hours later I managed to get my config set up again to compile TOS. my setup had completely disappeared and I don't understand why. I half wonder if it was when I was installing MiNT the other day it might have deleted stuff or something. Anyway.. But its own drive and config setup so I should not have to go through this faffing about again next time around.

So will see if I can speed up the TTram test next and compile different languages for ST536 TOS :)

On back to fiddling with GB7 code. It is pretty bonkers because TOS and MINT behave differently. It is like MINT ignores wind_update(1).

The original GB4 actually suffered with this problem and I think it's original creator was struggling with this as well looking at the code..

Is not really a problem but as the window is redrawing and moving to the right it leaves the black lines on the left..


In TOS I can use FMD_FINISH in the loop and it does not happen. HOWEVER, in MINT , the black lines return and it seems to play catch up and redraws the right-hand side of the background line by line which can take a few seconds. Ordinarily I just refresh one large block. So it is a bit confusing what is actually going on there.

Technically the AES should be able to redraw the background in real-time as the window moves to the right. But the problem is, if I process WM_REDAW messages in the loop, the dialogue box appears once then vanishes and the system totally locks up :roll:

I guess when you mention MINT technically you should be saying XaAES since there are multiple different AES's you could be using which handle this issue independently. Perhaps you could try MyAES since the creator OL is very active and could give you good insight, you might also find a bug in XaAES and @mikro might be able to help with that.

Steve wrote: 06 Dec 2022 21:15 I guess when you mention MINT technically you should be saying XaAES since there are multiple different AES's you could be using which handle this issue independently. Perhaps you could try MyAES since the creator OL is very active and could give you good insight, you might also find a bug in XaAES and @mikro might be able to help with that.

Yeah I am saying MINT and shouldn't be. @Badwolf sent me this all setup so no idea what it is really.

V4 video DAC board finally sorted :)

So as some people know I have been busy with assembly code to patch ST536 TOS.

viewtopic.php?f=39&t=6011
viewtopic.php?f=39&t=6019

I am really building this for my personal preference more than anything but well release is updated version when it has been tested a little bit better.

Mostly it will now "politely" inform you you are trying to run it on a 68000 CPU. Somebody probably will, and Hatari is hellbent on setting this as the CPU all the time which causes confusion as to why it has crashed :roll: There is a not very well hidden Easter egg in that routine as well :lol:

I have added in some text after the logo because sometimes when things go wrong, it totally locks up and there is nothing on the screen at all. So you are left waiting normally to see if it springs to life or not.. So anyway, he now displays ROM CRC checking message and flashes the cursor so you know it has not locked up until the main memory test starts.Tests should go faster but are currently on untested.

I have also done away with the hard drive delay wait. Most of us are using flash media these days anyway. But of course, all you have to do is switch on your old spinning media and wait a few seconds before turning on the ST anyway.

Given how long TOS takes to get to the memory test, even the slowest spinning rust drives would be fully spun up by the time the memory test finishes anyway.

(OK, maybe not some of the old Fujitsu Eagle 500MB IPI drives, they took 2 minutes and dimmed the lights whilst spinning up, but you couldn't connect those to an ST anyway.)

stephen_usher wrote: 08 Dec 2022 22:53 Given how long TOS takes to get to the memory test, even the slowest spinning rust drives would be fully spun up by the time the memory test finishes anyway.

Exactly.

stephen_usher wrote: 08 Dec 2022 22:53 (OK, maybe not some of the old Fujitsu Eagle 500MB IPI drives, they took 2 minutes and dimmed the lights whilst spinning up, but you couldn't connect those to an ST anyway.)

:lolbig: