Atari-Forum

I was going to do this for Starball but it was too much faff to be worthwhile, so all it does is scroll the table top to bottom. Everything in the game is running as normal - just there's no ball...

C was extraordinarily unreadable before ANSI C came along. I had three or four goes at getting it in the 80s and just couldn't make it fit in my head until I started working on projects in ANSI format.

ext.l takes a word and sign-extends it to a long.

Another advantage of nyh's routine is that it doesn't corrupt a0 or d0.w, so this may allow the calling code to be more efficient.

512 clock cycles per 50Hz line x 313 lines == 160256.
508 clock cycles per 60Hz line x 265 lines == 134620.

FedePede04 wrote:it was something like this i was looking for. just to be on the safe side, for me it look like that popping / pushing a long word to the stack takes 12 clk cycles, is that right

Yes.

I think the problem is add.w d0,d3 640*(something) approaches 16 bits after just over 100 lines, and if the 16-bit add fails to carry you'll get a bad address. I think the problem shows at any pixels on line 102 with an X greater than 127. Change it to adda.w d0,a0, or add.l d0,d3 and I think it wil...

ljbk wrote:The information provided is just to find out which registers are used/trashed for sure.
I already told Peter that i do not advice such type of coding.

Apologies. I should read previous posts more carefully .

For case 1, you just need to fill your registers with a specific value like $12345678, make the trap calls and check which registers have been modified. There's no 'just' about it. You must ensure that you test every codepath through every TOS version to be certain, which is an immense amount of ef...

Here's the tables I generated from timing a real ST. This is for all instructions and operands in RAM - registers and the like take different hits depending on the location. The first table is the base cycle count, the second is the pairing table, whether it accepts or leaves an odd phase (if you fo...

Because it seems to me, that whatever criteria you apply, the joystick ports seem highly unlikely to meet them since they were universally unused. They were universally unused because next model up didn't have them. Writing for dropped features would be a mistake. If TT030 had those ports, you woul...

Dio two joystick ports = 700DM????? And I said that where? It was an extremely rough estimate as to how much cost would be required to bring the TT to 100% STE compatibility. If you're not going to insist on every feature, what's your criteria for insisting on any feature? Because it seems to me, t...

Tell me again how TT030 missing features lower end machines have is a good thing. When adding them adds 10% to the cost of the machine and delays the release schedule by six months when it's not going to make you dramatically more money. I mean, reductio ad absurdum, it was a bad thing that the ST ...

12 bits of software of which only two I've heard of, and one of those (Llamazap) is Falcon-only? Yeah, that must certainly have not-sold a couple of million TTs... I mean, clearly, their absence was a far bigger problem for software compatibility than the change of CPU stack frame format or the priv...

Is there a single piece of software that uses the STE joystick ports? Is there even a joystick to connect to them? Sorry chap, but you're focusing on irrelevant details. Tiny (and I do mean tiny) technical changes are completely unimportant. Maybe five people on the planet chose a Mac over a TT beca...

Better debugged (more efficient?) OS OS upgrades No games to distort brand Dozens of halo effects: far stronger business brand, excellent US distribution network, fanatical userbase. Extraordinarily hard to credit how mental people went over Hypercard To the Atari pluses you can add 'Disks you can a...

Really, 30 pieces of software, and thats 'all major packages'...? :) I'm not a wide user of software - it's too much investment to learn a complex piece of software, and most software is very expensive if you're not going to extensively use its facilities, so I generally avoid using non-free (as in ...

I can't name one piece of software on the PC I am using that was not originally developed on PC (arguably a few are Unix workstation originally rather than Windows, but not many of those - plus, workstations are if anything further from the command-line-free ST heritage than the PC is). Indeed, even...

I see your point there, they should not have even tried to go to the commecial market and don't need to spend the money to advertise then. Some sort of USP and differentiation was needed to target the business market. I actually think the ST came fairly close with the SM124, the Megas and then the ...

But that's because the PC was a tool and the Spectrum was a toy. Nobody remembers fondly the spade they dig the garden with, but they have quite an attachment to those Star Wars figures.

1.5ms is 23 video lines, and sync-scrolling consumes a lot less than that. So you can happily disable interrupts for 20 lines plus without any risk of missing an interrupt.

And he was absolutely right.

Dark Willow wrote:Atari never really understood the professional market, and Tramiel's "box shifting" approach pretty much precluded a strategy of getting out of the low-end home gaming market. The mess of the AppleTalk ports really says it all.

Yeah, excellent points.

On both Atari and Amiga, coders were playing with graphics and music ... while the PC platform was vastly investing in it's programming environment. True Basic, Turbo Pascal 3, and later Turbo Pascal 5.5 with it's object oriented base. Made reusing code easy and it was more important on how code wa...

Yes - the Timer B is higher priority than the ACIAs so as long as the MFP is in SEI mode the ACIA can't interrupt a Timer B routine.