@Hammer

So what? What does any of this have to do with whether or not parallel bitplane writes may have been good for the Amiga?

Every single thread you enter, you pollute with endless pages of frankly irrelevant intel/AMD guff. Literally nobody cares. You have a PC. Amazing. Almost everyone else does too including almost everyone here (at a guess). There are lots of other forums / social media groups that specialise in that.

Look, there are times when architectural comparisons between AMD64 / PPC / 68K are relevant, but it's a small minority of cases.

You have a number of Amiga systems yet you barely talk about them in comparison. Maybe you post endlessly about those on a PC master race forum we don't know about. Try swapping the two.

_________________
Doing stupid things for fun...

@Hammer

Quote:

Hammer wrote: @cdimauro Quote: But I can tell you that I consider complete idiots the engineers that have invented the above Slow-Mem: it's NOT usable as Chip-Mem NEITHER as Fast-Mem. It has ALL DISAVANTAGES and NO usage (despite as being a storage). That was the most idiotic thing on the Amiga land, since it severely crippled the games. 1989-era A500 Rev6A's ECS Agnus can address Slow RAM as Chip RAM at different address ranges e.g. copper pointer set to 0x090000. You can have some kind of 1MB Chip RAM with 512KB+512KB configuration with an ECS Agnus chip. For Amiga 500/2000 models from 1987 to 1988, Slow RAM design wasn't a good move.

This doesn't work, really! You've to understand that software houses and their developers could only produce software for the most common platforms. So, Amiga 500/2000 for OCS (512kB of Chip + 512kB Slow) and Amiga 1200 for AGA (2MB Chip).

If a machine was more powerful then the extra resources could be used (more RAM -> disk caching), but it was NOT guaranteed, since games were developed for the base machines.

For example, I've developed the parallax of the floor (bottom 64 lines of the screen) for Fightin' Spirit, but only for machines having at least 1MB of Chip Mem (that wasn't implemented on the final version because backgrounds were already almost done and the graphic artist didn't want to change them again).
That was an EXTRA thing, but the core of the game was exactly the same.

So, having SOME machines with 1MB of Chip even with those tricks was good, but definitely NOT the target for the developers.

@Gunnar

Quote:

Gunnar wrote: Hammer you posts imply that SNES would be better than AMIGA ... Is this true? Or is this just misunderstanding?

I would say that it's mostly true. Well, SNES came on 1990, so it makes also sense: Amiga (1000) arrived 5 years before...
Quote:

If you understand the SNES than you know that its a pretty wicked system

Depends on what you want to do with it. For its scope it's an awesome system.
Quote:

which very little memory,

There's plenty of memory for what he needs to do.

The processor can address the same amount of memory of a 68000 or A1200's 68EC020: 16MB.
Quote:

with good features and with funky features but at the same time also many limitations, and with a CPU coming from the C64.

Coming... not exactly, but it's a BIG evolution from it, anyway.
Quote:

SNES and Amiga 1200 have different features and different attributes. There are very many aspects in which the A1200 beat the SNES hands down.

Which ones? Remember that the topic is games. 2D games, specifically.

Which advantages has the A1200 over the SNES?
Quote:

And there are also some aspect where the SNES has advantages.

I would say... A LOT of advantages.
Quote:

You can make nice 2D games for both. Of course the Amiga has the much better CPU.

Which is mostly irrelevant for the scope. On both systems the CPU is mostly the slave of the hardware: it's there to load the custom hardware's registers.

The business logic part takes very little time and the SNES's CPU can easily do it, even on a very efficient way, since it can access memory in a single cycle (OK, the CPU clock is lower than a 68000's one, but the latter needs 4 cycles for it. The A1200/68EC020 is better, but it's crippled by the chipset for the memory access).
Quote:

And the Amiga has a lot more options in software and offers also real computer usage with Workbench, Dpaint, Music making, Surfing the Internet. The Amiga has a lot less limitation and has excellent upgrade options also upgrades with good price like 68030 CPU card with plenty memory.

Correct, but irrelevant when sticking to the specific topic (2D games).

Now... work time.

@cdimauro

GUNNAR:
Quote:

which very little memory,

Cesare:
Quote:

There's plenty of memory for what he needs to do. The processor can address the same amount of memory of a 68000 or A1200's 68EC020: 16MB.

The SNES has 128 KB RAM.

The Amiga is a real computer. While you can play very nice games on Amiga.
You can also use the Amiga for "normal" Computer uses:
writing letters, programming, painting images, composing music, surfing the internet, coding demos, do 3D rendering ... whatever you want.


The SNES is not a computer. Its a ROM based game console with only 128K 8bit RAM
The SNES GFX system is not possible to show something like Amiga workbench.
You can also not code demos on it like on Amiga.
Nevertheless the SNES can do pretty decent 2D games.

For playing 2D games both systems are nice.

The AMIGA 1200 has of course a lot more options what you can do with it.

--
SNES GFX is tile based (like C64)
SNES supports several playfields = nice for scrolling
SNES supports 16 colors sprites.
SNES DMA can show 256 sprite pixel per row
SNES screen mode is 256 pixel wide (lowres)
SNES has a 3.5 Mhz 8/16 bit CPU

AMIGA AGA GFX is plane based.
AMIGA supports 2 playfields
AGA sprites can be 4 color or 16 color.
AGA Sprite DMA show 512 pixel with 4 color per row or 256 pixel with 16 color per row
AGA supports lowres and also hires screenmodes
AMIGA 1200 has a very nice to code 32bit 68020 CPU @14Mhz


I think SNES has some advantages, sprites are more flexible, playfield have more options, the tiled GFX needs less composition copies. The Amiga 1200 has a lot more CPU power, offers more screenmodes, has 24bit color palette.


I think the SNES is a little stronger for making lowres games.
The Amiga is more flexible.
And both can do very nice games.

Last edited by Gunnar on 14-Mar-2024 at 09:45 AM.

@cdimauro

GUNNAR
Quote:

with good features and with funky features but at the same time also many limitations, and with a CPU coming from the C64.

Quote:

Coming... not exactly, but it's a BIG evolution from it, anyway.

The CPU is a 6502 family CPU, its 8bit with 16bit register.

- It has only 8bit Databus
- It has only 3 register: A,X,Y
- The CPU has only the 6502 instructions this means the CPU has neither MUL and nor DIVIDE.

The Motorola 68020 of the Amiga 1200 plays in a complete different league
The 68020 CPU has 8 Data register each 32bit
The 68020 CPU has 8 Address register each 32bit
The 68020 can do 8bit/16bit/32bit and even some 64bit calculations. (MUL/DIV in 64bit!)
The 68020 CPU has a full 32bit bus, and even a small instruction cache.


Quote:

SNES's CPU can easily do it, even on a very efficient way,

The SNES CPU has an 8 bit DATA bus only
The Amiga 1200 Motorola 68020 CPU has a 32bit Bus and even a small instruction cache.
The 68020 CPU is performance and feature wise a complete different league.


The Amiga 1200 CPU can be much nicer programmed
and with its much higher performance can do a lot more complex game logic.

@Gunnar

Quote:

Gunnar wrote: @cdimauro GUNNAR: Quote: which very little memory, Cesare: Quote: There's plenty of memory for what he needs to do. The processor can address the same amount of memory of a 68000 or A1200's 68EC020: 16MB. The SNES has 128 KB RAM.

No, it has 256kB of RAM.

Anyway, you talked about "memory". And I've correctly replied about memory.

Talking only about RAM on a system brings nowhere: it's the total amount of usable (not only addressable) memory which is relevant, because memory is used depending on the context.

In fact, the SNES does NOT require a lot of RAM, like Amigas, because it has A LOT of space used as ROM for storing code and static (read-only) data, and game assets (graphics and sound).

That's without taking into account the possibilities offered by its features. For example, X & Y flipping of tiles and sprites allows to double up to quadruple the memory space, compared to the Amiga which lacks this possibility. Same thing for the audio samples with the ADCPM.

At the end the concept is always the same: a system should be evaluated depending on its scope. And for its scope, SNES has WAY MORE memory compared to any Amiga doing the same things (games).
Quote:

Gunnar wrote: The Amiga is a real computer. While you can play very nice games on Amiga. You can also use the Amiga for "normal" Computer uses: writing letters, programming, painting images, composing music, surfing the internet, coding demos, do 3D rendering ... whatever you want.

Sure, and I've nothing to say about that.

However, this specific part of the discussion was about 2D games.
Quote:

The SNES is not a computer. Its a ROM based game console with only 128K 8bit RAM The SNES GFX system is not possible to show something like Amiga workbench.

Same as above: it was made for games. Where is simply shines.
Quote:

You can also not code demos on it like on Amiga.

You can do them with the SNES as well. There are plenty of demos.
Quote:

Nevertheless the SNES can do pretty decent 2D games.

Well, SNES did AWESOME games. Not only "decent". It simply shined at making 2D games.
Quote:

For playing 2D games both systems are nice.

I beg to differ: you're trying to put a Ferrari and a Volkswagen on the same level. You can guess who is the Ferrari and who the VW here.
Quote:

The AMIGA 1200 has of course a lot more options what you can do with it.

For doing 2D games? I don't think so. Rather, it's the opposite: SNES has way more options.
Quote:

-- SNES GFX is tile based (like C64)

Very good for 2D games.
Quote:

SNES supports several playfields = nice for scrolling

Up to 4 playfields, with playfields up to 256 colours and with rotating & scaling graphics with the famous MODE 7.
Quote:

SNES supports 16 colors sprites.

With sprites selecting independently the palette from any of 8 (and the same for the playfields: the 16 colours fields have 8 different palette that can be selected, independently).
Quote:

SNES DMA can show 256 sprite pixel per row

Correct, but you missed that there are 128 sprites and that 32 of them can be displayed per each scanline. That's a very important technical detail.
Quote:

SNES screen mode is 256 pixel wide (lowres)

That's not accurate. SNES can also do hires graphic. But for 2D games, the normale = lores resolution is the best. As it is for an Amiga 1200 for the same scope.
Quote:

SNES has a 3.5 Mhz 8/16 bit CPU

Which is more than enough for what's needed.

Anyway, you're reduced a lot the things that a SNES could do. What the about the audio section, for example? Isn't it important? And about the 8 independent DMA channel that the CPU can use to quickly transfer memory?
Quote:

AMIGA AGA GFX is plane based.

Not good for 2D games.
Quote:

AMIGA supports 2 playfields

With max 16 colours each.
Quote:

AGA sprites can be 4 color or 16 color.

Only 8 x 4 colours or 4 x 16 colours.

But if you enable the horizontal scroll they are reduced. Especially with AGA and 4 x fetch mode.
Quote:

AGA Sprite DMA show 512 pixel with 4 color per row or 256 pixel with 16 color per row

That's theory. The practice is very different: see above.

And 64 pixels wide sprites aren't so flexible as 8 or 16 pixels wide sprites of the SNES. And the SNES has MANY sprites which can be used even with any scrolling.
Quote:

AGA supports lowres and also hires screenmodes

Irrelevant for games: lores is almost always the only option. Plus, SNES can do hires as well (and with many sprites usable).
Quote:

AMIGA 1200 has a very nice to code 32bit 68020 CPU @14Mhz

Sure. Nothing to say here, besides that it's crippled by sharing the memory bandwidth with the chipset AND crippled by the 68k-only chipset design.
Quote:

I think SNES has some advantages, sprites are more flexible, playfield have more options, the tiled GFX needs less composition copies.

And even more: look at the missing features.
Quote:

The Amiga 1200 has a lot more CPU power,

Absolutely.
Quote:

offers more screenmodes,

It offers only two screenmodes, effectively: a single and a dual playfield.

HAM for games? Leave it alone. EHB? Not required with a 8 bit palette.

Take a look at SNES modes and you can see what does it mean having different screen modes. And, especially, special effects, like transparency windows which are simply awesome.
Quote:

has 24bit color palette.

Good if you load it once and forget it. But if you've to change it... it's a pain, because the Copper is crippled both by its speed and the absurd mechanism for loading a 24 bits colours.
Quote:

I think the SNES is a little stronger for making lowres games. The Amiga is more flexible. And both can do very nice games.

Let's put it this way: the SNES simply OBLITERATES an Amiga 1200 at making 2D games.

The A1200 can do decent games, but its hardware is NOT tailored for making games.
Quote:

Gunnar wrote: @cdimauro Quote: Coming... not exactly, but it's a BIG evolution from it, anyway. The CPU is a 6502 family CPU, its 8bit with 16bit register. - It has only 8bit Databus

Enough for mostly loading 8 bit data.
Quote:

- It has only 3 register: A,X,Y

Enough for a slave of the chipset.
Quote:

- The CPU has only the 6502 instructions this means the CPU has neither MUL and nor DIVIDE.

The SNES has ad-hoc registers for this.
Quote:

The Motorola 68020 of the Amiga 1200 plays in a complete different league The 68020 CPU has 8 Data register each 32bit The 68020 CPU has 8 Address register each 32bit The 68020 can do 8bit/16bit/32bit and even some 64bit calculations. (MUL/DIV in 64bit!) The 68020 CPU has a full 32bit bus, and even a small instruction cache.

Absolutely nothing to say here. The 68EC020 is a wonderful CPU and very very nice to code.
Quote:

Quote: SNES's CPU can easily do it, even on a very efficient way, The SNES CPU has an 8 bit DATA bus only The Amiga 1200 Motorola 68020 CPU has a 32bit Bus and even a small instruction cache. The 68020 CPU is performance and feature wise a complete different league.

You've to consider the context: for making 2D games the SNES CPU is very very good.

Yes, it has an 8 bit bus. But most of its instructions are only 8 bit. And loading or storing 8 bit is done in ONE clock cycle.
Quote:

The Amiga 1200 CPU can be much nicer programmed

Nothing to say here.
Quote:

and with its much higher performance can do a lot more complex game logic.

Which is mostly not needed.

However and per what I've said before, the SNES can also do complex logic. In fact, it has the complete frame time available for making even complex computations while the chipset is doing almost all work (there's a programmable horizontal DMA for making effects like the Copper. And a separate CPU only for the audio).
And with its very quick memory access it can really do a lot of stuff.

Whereas the on the A1200 the CPU should do the computation before that the new frame starts, and then it has to program the Blitter to squeeze the most of the system. While the memory accesses are blocked by the system, the Blitter and the Copper, which have precedence over it.

P.S. As usual in mornings, I've no time to read again. I've to go work...

@OlafS25

Quote:

if a company makes so many wrong decisions as Commodore it deserves to disappear. As I wrote no idea what to do, no idea of the future

Well no one really has an idea of the future.
But still you need to make decisions.

Its very easy to condemn a decision 30 years later with 20/20 hindsight.


But were Commodores decisions really all terrible bad?


Lets look at some "arguments" here.

a) Some would say that Commodore working on the C64+ upgrade was a terrible idea of Commodore.

But lets look at the SNES.
The C64 came out in 1983 - the SNES came out in Europe in 1992
The SNES was very successful.

The SNES is technically build a lot like an upgraded C64.
SNES has tiled GFX - like the C64.
SNES uses Hardware sprites - like the C64
SNES has an own Sound processor - like the C64
SNES uses an upgraded version of the C64 CPU (6502).

Of course the Sprites and GFX and sounds features of the SNES are stronger compared to C64.

The SNES looks to me a lot like an upgraded/improved C64 design.
The SNES was very successful.

I think the SNES shows us several things:
- millions of people happily bought an design similar to an upgraded C64 in the mid 90th

Does the whole discussion you could not sell a system without Chunky support hold water?
- SNES sold very nice even without Chunky Pixel support




Regarding the Amiga was "DOOMed" without Chunky Support argument.

Does this argument hold water?

- While AGA did no had chunky build in: you could run DOOM with an 68030 CPU on A1200.
Software C2P was possible and worked fast enough to play games.

- AKIKO made C2P a little faster on the CD32. This was better than nothing.

- If Commodore would not get bankrupt, could they not easily have sold
an external adapter like GRAFITI to people - to give all AGA machines Chunky as upgrade?

- Last but not least the Amiga dev-team was working on AAA chipset - which included Chunky support

@Gunnar

Quote:

Gunnar wrote: Its very easy to condemn a decision 30 years later with 20/20 hindsight. But were Commodores decisions really all terrible bad?

No. Like all companies they made some good decisions and some bad ones, and some that would have been good but for circumstances. Buying and developing the Amiga was a good decision IMO, but some of the things they did with it (A3000, CDTV, A600) were not. They couldn't afford to make unprofitable models. Commodore's biggest problem was not appreciating the precarious situation they were in financially and in the marketplace. By the time they did it was too late.

Quote:

Some would say that Commodore working on the C64+ upgrade was a terrible idea of Commodore. But lets look at the SNES. The C64 came out in 1983 - the SNES came out in Europe in 1992 The SNES was very successful. The SNES is technically build a lot like an upgraded C64.

The C65 was specifically designed to be a 'NES killer'. The plan was to release it in 1990. If this had happened it could have been very successful.

Quote:

Does the whole discussion you could not sell a system without Chunky support hold water? - SNES sold very nice even without Chunky Pixel support

No, chunky pixels are not needed for 2D games, and only make sense for 256 colors. But VGA had that, and since the PC was the Amiga's major competition there was a perception that the Amiga was inferior for lacking them.

Quote:

Regarding the Amiga was "DOOMed" without Chunky Support argument. Does this argument hold water?

No. Doom was released on December 10, 1993. In the most optimistic time frame it wouldn't reach the Amiga until early 1994, by which time Commodore was already 'doomed' for unrelated reasons. Doom was intended to run on a fast 486 so it wouldn't have mattered if the 'the Amiga' had chunky pixels, the A500 was the only model with a sufficiently large user base and it was nowhere near powerful enough to run Doom anyway (as John Carmak pointed out).

PC owners who had less than a 486 in 1994 were willing to plonk down the cash for one in order to play Doom. Amiga owners were the same, ie. they too would buy a 486 rather than a new Amiga. Why? Because with a 486 you got all the other stuff like Microsoft Windows and Word and Excel and the latest PC games etc. PCs were made by everyone and sold everywhere and the industry standard and obviously the future.

The Amiga was made by one company with a shaky future which for some strange reason was still trying to go it alone with a non IBM-compatible computer - the kiss of death for selling to the broader market. Unlike Apple it hadn't managed to cultivate a following among the elites, so how would it fit into the picture? Answer, it wouldn't, and chunky pixels were only a tiny part of the problem.

Also, with 486 users had many operating systems to chose from. I had Amiga friends who moved to OS/2, as well as BSD and Linux during that period.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

https://youtu.be/YdeUebxTt5M?feature=shared

https://youtu.be/UtWZEqmO-aA?feature=shared

https://youtu.be/1pAKQGP0cKA?feature=shared

Excellent channel by the way.

Last edited by kamelito on 15-Mar-2024 at 08:42 PM.
Last edited by kamelito on 15-Mar-2024 at 08:39 PM.

@Gunnar

Quote:

Gunnar wrote: @OlafS25 Quote: if a company makes so many wrong decisions as Commodore it deserves to disappear. As I wrote no idea what to do, no idea of the future Well no one really has an idea of the future. But still you need to make decisions. Its very easy to condemn a decision 30 years later with 20/20 hindsight. But were Commodores decisions really all terrible bad?

Yes, most of the times.
Quote:

Lets look at some "arguments" here. a) Some would say that Commodore working on the C64+ upgrade was a terrible idea of Commodore. But lets look at the SNES. The C64 came out in 1983 - the SNES came out in Europe in 1992 The SNES was very successful. The SNES is technically build a lot like an upgraded C64. SNES has tiled GFX - like the C64. SNES uses Hardware sprites - like the C64 SNES has an own Sound processor - like the C64 SNES uses an upgraded version of the C64 CPU (6502). Of course the Sprites and GFX and sounds features of the SNES are stronger compared to C64. The SNES looks to me a lot like an upgraded/improved C64 design. The SNES was very successful. I think the SNES shows us several things: - millions of people happily bought an design similar to an upgraded C64 in the mid 90th Does the whole discussion you could not sell a system without Chunky support hold water? - SNES sold very nice even without Chunky Pixel support

Yes, but it's a technology from 1990 and purely oriented to 2D games.

In fact, some special chips were introduced for 3D games. But it was some years late.

BTW, 3D games existed since very long time, but they were wire-frame or just using polygon filling. Even the C64 had 3D games.

However, on early 90s technologies changed according to the new needs, but the SNES project was already started on the late 80's and went on production on 1990: the hardware was done.
Quote:

Regarding the Amiga was "DOOMed" without Chunky Support argument. Does this argument hold water? - While AGA did no had chunky build in: you could run DOOM with an 68030 CPU on A1200. Software C2P was possible and worked fast enough to play games.

What you're missing here is the context. Specifically, sticking to the context, so in the PRECISE historical period.

So, the question was and should be: was a good C2P routine available for the A1200 when Doom was released? I mean: good, so that a possible Doom conversion could have decently run using a 68030 accelerator board.
Quote:

- AKIKO made C2P a little faster on the CD32. This was better than nothing.

Lazy engineers...
Quote:

- If Commodore would not get bankrupt, could they not easily have sold an external adapter like GRAFITI to people - to give all AGA machines Chunky as upgrade? - Last but not least the Amiga dev-team was working on AAA chipset - which included Chunky support

Both were too late. And not competitive.

@bhabbott

Quote:

bhabbott wrote: @Gunnar Quote: Gunnar wrote: Some would say that Commodore working on the C64+ upgrade was a terrible idea of Commodore. But lets look at the SNES. The C64 came out in 1983 - the SNES came out in Europe in 1992 The SNES was very successful. The SNES is technically build a lot like an upgraded C64. The C65 was specifically designed to be a 'NES killer'. The plan was to release it in 1990. If this had happened it could have been very successful.

ROTFL Have you ever compared the C65 and SNES technical specs? C65 is not even a toy for newborns compared to the SNES.
Quote:

Quote: Does the whole discussion you could not sell a system without Chunky support hold water? - SNES sold very nice even without Chunky Pixel support No, chunky pixels are not needed for 2D games, and only make sense for 256 colors.

You don't know of what you're talking about. Packed/chunky graphics is almost always better than planar graphics even purely sticking on 2D games.
Quote:

But VGA had that, and since the PC was the Amiga's major competition there was a perception that the Amiga was inferior for lacking them.

https://www.mobygames.com/game/1660/star-trek-v-the-final-frontier/screenshots/dos/
Quote:

PCs were made by everyone and sold everywhere and the industry standard and obviously the future. The Amiga was made by one company with a shaky future which for some strange reason was still trying to go it alone with a non IBM-compatible computer - the kiss of death for selling to the broader market. Unlike Apple it hadn't managed to cultivate a following among the elites, so how would it fit into the picture? Answer, it wouldn't, and chunky pixels were only a tiny part of the problem.

https://en.wikipedia.org/wiki/Macintosh_II

Release date March 2, 1987

Graphics: The Macintosh II includes a graphics card that supports a true-color 16.7 million color palette[14] and was available in two configurations: 4-bit and 8-bit. The 4-bit model supports 16 colors on a 640×480 display and 256 colors (8-bit video) on a 512×384 display, which means that VRAM was 256 KB. The 8-bit model supports 256-color video on a 640×480 display, which means that VRAM was 512 KB in size. With an optional RAM upgrade (requiring 120 ns DIP chips), the 4-bit version supports 640×480 in 8-bit color.[15] The video card does not include hardware acceleration of drawing operations.

@kamelito

Quote:

kamelito wrote: https://youtu.be/YdeUebxTt5M?feature=shared https://youtu.be/UtWZEqmO-aA?feature=shared https://youtu.be/1pAKQGP0cKA?feature=shared Excellent channel by the way.

Well, the first video is a mixture or ridiculous and hilarious: it's clearly evident that the guy doesn't know the SNES specs and never seen games for this console.

Here's one that shuts up his very large mouth: https://www.youtube.com/watch?v=PfkJctxcMBs

But, hey: he can try to do something similar, even for an Amiga 1200 which is more powerful compared to a classic Amiga where Leander was running.
I want to see how multiplexing the sprites could help him.

@cdimauro

This man is a successful game programmer, he knew the Amiga and Megadrive perfectly, he founded Tt games (tt for Traveler Tales) he’s probably now a multi millionaire by selling TT Games to Warner. If you had a look at his videos you’ll know that he is very clever and talented and he delivered many games not just talks.

Now a reply to some old comment.

@Gunnar

Quote:

Gunnar wrote: @Hammer Quote: My last post's point is about Commodore's management issues that affected the engineering results. I provided other management examples with fewer fckups. Let to sum up the positions: Cesare: "Amiga not good because = Amiga engineers were un-creative and did a bad job"

Wrong: I NEVER stated it.

Amiga (1000) was a good machine and engineers did a great job with it, which we enjoyed (in its variants: 500 and 2000). Which doesn't mean that everything was good: they could have improved in some points.

Unfortunately the introduction of Slow-Mem was a totally idiotic thing.
I don't know how did (I'm guessing not the original creators), but it SEVERELY CRIPPLED the platform for game development!!!
512kB + 512kB of ALL Chip Mem would have made BIG, BIG CHANGES!

ECS + AGA + Akiko chips developers -> lack of creativity.
Quote:

Hammer: "Amiga not good because = Commodore Management was bad" Gunnar: "Nothing is ever perfect. Hardware engineering is hard = Amiga was OK." Are these our views?

No. See above.
Quote:

Did I understand the argumentation right, that one core point in the AGA discussion is : missing chunky support. This is correct, AGA lacks chunky support.

Not only AGA: OCS and ECS as well.
Quote:

For some games Chunky format is pretty beneficial.

For almost all games: chunky was/is alwast always better even for 2D games.

For more detailed (and technical) information you can read the 17 articles that I've written on the topic.
Quote:

I think one of the most succesfull games using Chunky was maybe DOOM. DOOM came out late December 1993. In 1994 DOOM was a big success.

Even Elite, Frontier, Stunt Car Racing, F1GP, F16-Falcon, etc. were big successes. When they were released?
Quote:

The Amiga original chipset design work began in 1983. In 1983 popular home-Computers had 16 KB Memory. When Amiga development began the firsts Commodore C64 came out.

Which had... packed/chunky graphics. As many other systems.
Quote:

A like DOOM which needs several Mega-Byte Memory was for sure difficult to foresee 11 years earlier in 1983.

If a flight simulator was the inspiration for the Amiga chipset / technology, then planar graphics wasn't a good design decision from the beginning...
Quote:

Commodore made 2 iterations, improvements on the OCS chipset. ECS = the A3000/600 chipset. ECS added programming of the screen modes. This means the chipset can switch between NTSC/PAL mode.

The ONLY feature that SOME game offered (demos / cracks used it much more).
Quote:

AGA = The A4000/1200 chipset AGA added more colors (64 out of 4096) => (256 out of 16 Million) AGA added wider sprites 16 = 32/64 pixel

Already replied several times: just listing the specs gives nothing! It's about how the AGA performed with REAL CODE, with REAL GAMES which is important.
Quote:

I'm not sure exactly but I think the AGA development started in year 1989. AGA came out in 1992.

Wrong. It was very well known that AGA was a quick (and dirty) port of some AAA features to ECS.

Anyway and even if it would have been true... THREE years to deliver this crap?!? Seriously?
Quote:

My understanding is that the design goal of AGA was to * play all existing Amiga games * give features to make more nicer looking Amiga games

No, see above: the reason is because Commodore was shitting itself from fear due to the competition which was eroding its market. Then they decided to give some small content to the its customer base, waiting for the next big thing (which would have been still too late).
Quote:

How did very succesfull Amiga games look like when work on AGA chipset started ? * Turrican * Shadow of the Beast Is it not imaginable that the "goal" of AGA was to run games like Turrican and to add features that help to make more such "successful" games?

No. It was too little, too late, with only a bunch of features added on top of the ECS.
Quote:

In my opinion AGA reaches this goal. AGA can run pretty good such 2D games.

Good or decent? I would say decent, but good... definitely not. As per other comments that I've written: the competition did WAY BETTER and even some years before that the AGA was introduced.
Quote:

Did they forsee in 1989 when AGA design work started the immense success of DOOM in 1994 ? Probably not.

See above: they have NOT started on 1989. If true then it would have been even worse: three years for a crappy chipset, with the most important things unchanged.

@kamelito

Quote:

kamelito wrote: @cdimauro This man is a successful game programmer, he knew the Amiga and Megadrive perfectly, he founded Tt games (tt for Traveler Tales) he’s probably now a multi millionaire by selling TT Games to Warner. If you had a look at his videos you’ll know that he is very clever and talented and he delivered many games not just talks.

He should take a look at the video that I've posted: it's a best selling SNES game.

I want to see how he could do something even barely similar with the sprite multiplexing that he trumpeted about.

He could have been very talented but the comparison with the SNES is simply ridiculous.

BTW, the SNES was running the game at 60Hz/FPS...

And now the last reply to an old message.

@Gunnar

Quote:

Gunnar wrote: @cdimauro Quote: What's not clear to you that the project wasn't mine? That was from a company? That the company's CEO was leading it? That ALL relevant things about the hardware were developed by the company's employees? That all declarations regarding this was coming from them? No this is not clear, as YOU Cesare did post a lot online about TINA.

Wrong: it wasn't clear ONLY TO YOU!

And that despite I've ALREADY reported SEVERAL TIMES what was my role, since the beginning, and especially TO YOU when you started the defamation campaign against me (that you restarted again now).

So, it's YOUR problem!
Quote:

You wrote in several Amiga forums

"Several"? Which ones? List them, so that we can count if "several" applies.
Quote:

and you posted also in the TINA Forum as admin yourself.

Wrong again: I was NOT the admin of the forum!

In fact, I had admin privileges ONLY TO MANAGE USERS. As you also reported below.

The forum's admin was the company's CEO. As you also reported below as well.
Quote:

In your posts, _you_ "explained"

Wrong again: I've explained nothing. I've just reported the features list, as it was provided me by the CEO.
Quote:

the fake features of the TINA system to people. Quote: CPU dedicate: 3 x FPGA one per core, 400MHz Memoria RAM: 256MB-1GB Ram Hard Disk: HDD IDE Floppy Disk: FDD controller Ingressi: PS/2 mouse and keyboard Ingressi: SDCard slot Collegamenti esterni: Ethernet Uscita video: HDMI out (or VGA) Uscita audio: AUDIO out I/O: 1 x Serial Port PRINCIPALI CARATTERISTICHE DI TiNA: - Extended AGA chipset with chunky pixel up to 32 bit and support for a FullHD (1920x1080) at 60Hz/FPS - Improved Copper and Blitter, especially the Blitter in order to manipulate up to 32bit colour - Improved Sprites (still 8) to support chunky-pixel mode up to 32 bit colour, with the possibility of horizontal and vertical flip - from 256MB to 1GB of memory (all chip-ram) with a theoretic bandwith of 3.2GB/s (roughly 450 times what OCS/ECS machines offered, and 112 times what AGA had available just for the display and sprites), later on there might be the possibility to quadruple it using DDR2 memory (reaching similar Nintendo Wii U values). - Minimum of 8 audio channels, 16 bit stereo at 48Khz (minimum). - IDE/ATA controller for legacy hard drives - SD card slot for firmware, ADF images and HDF to emulate up to 2 hard disks. - Compatible with AmigaOS3.x All these advertised feature look for very impressive -

As you also stated, this is advertisement: just listing features.
Quote:

BUT NOTHING OF THEM IS TRUE. NONE OF THEM IS CORRECT!

Then start EXPLAINING (AND PROVE!) why NOTHING/NONE (your words!) was true/correct.

I repeat again for YOUR convenience: YOU've to PROVE that NOT A SINGLE THING was true/correct.
Quote:

Fact: There NEVER was a TINA.

Wrong again: the project was ongoing.
Quote:

Fact: Tthe website was a hoax.

Wrong again: the website was about a project which was being developed on.
Quote:

Fact: The photos of PCBs, that were online on the TINA website - "showing TINA development" were taken from other websites, they were NOT from TINA project, they were pictures of other PCB, of other projects, of other people. (simply stolen from the internet)

I don't know this and I wasn't the website admin / content creator. Other people were doing it, as you also reported below.
Quote:

Cesare Di Mauro I understand that you say, that you were only part of the team.

Wrong again: I was part of team. But NOT of the company.
Quote:

And you did not design the specs.

Correct. Not the hardware specs (frequencies, bus width, etc.), to be fully precise.
Quote:

But the fact remains, that the TINA project was a 100% hoax. Nothing of it was real. Nothing of it was ever done.

Wrong again: the project was ongoing, as I've already stated.

You can tell that it never went to the market, like many other projects. Much more below on this.
Quote:

The "promised" features like 400Mhz and 128bit Bus - are technical totally impossible with the claimed to be used hardware.

See above: not my problem / responsibility. I repeat since several years that I'm NOT a hardware engineer.

Anyway, the issues that they faced with those goals were reported on the forum, during the last period (and I was still active / part of the project). Have you "forgot" it, or do you selectively recall only what's convenient for your propaganda?
Quote:

If you Cesare Di Mauro post about these "false" facts in forums and you advertise impossible features of a non existing hoax system ... Then what does this make you?

An ambassador? Have you understood my role or not yet?

Anyway, let me quote some "expert": https://amigaworld.net/modules/newbb/viewtopic.php?topic_id=43687&forum=2&start=60&viewmode=flat&order=0#868439

Quote:

That's called life. If you look around then you see that every company makes plans ... And _very_ often they fail to reach them. I myself worked at IBM on developing new CPUs that never came to market. Yes this is frustrating but this is totally normal. INTEL, IBM, SONY, AMD, ARM, MOTOROLA, SUN, ATARI, COMMODORE = every company on this planet made plans which they fail. As more ambitious your plans are - as higher is the risk they fail.

Do you agree with this? I think that it's true and it applies to TINA as well: too high expectations that were disappointed when the project was implemented, which led to its closure.

It's life. It happens.

You know that I worked at Intel. I was part of the Xeon Phi team (for the debuggers extensions). Xeon Phi was born from Larrabee's ashes.
With Larrabee Intel tried to enter the GPU market, but it failed miserably because it was too much optimistic about what possible to achieve using CPU cores instead of the specialized cores of a regular GPU, and as well as the fixed-functions which they implemented on specific parts.

Same thing happened some years before with the Pentium 4 project: the plans were to reach 10Ghz on 2010. Again, they were too much optimistic.

Now, can YOU state that Larrabee and (subsequent) P4 projects has to be considered hoaxes? That Intel's engineers were ill-equipped?
I don't think so. Their expectations were too high and the reality brought them back.
It happens. It's life, as the above "expert" reported.

Same thing happened with USA Racing: the project was so much big that it would have been impossible to complete it in reasonable (usual) time AND make profit. Even after that the "monster" virtual map (8192 x 65536 pixels) was halved (4096 x 65536 pixels). Was it an hoax? Were the people working on it ill-equipped? No. Same thing: too high expectations which lead the project to be abandoned.
It happens. It's life, as the above "expert" reported.

And I think that he's right (at least on that). What do you think about it? Do you agree with the above "expert"?
Quote:

@Gunnar Quote: Gunnar wrote: @Cesare Di Mauro [quote]What's not clear to you that the project wasn't mine? That was from a company? That the company's CEO was leading it? That ALL relevant things about the hardware were developed by the company's employees? That all declarations regarding this was coming from them? Cesare, is this true what you say?

Yes. Of course.
Quote:

Lets look at this Company: Quote: BERTOCAR (Italy) - System admin TiNA project. CDMAURO (Italy) - Amiga hardware expert. User admin forum TiNA project. SCHIUMACAL (Italy) - User admin forum TiNA project and maintainer TiNA project website. THEDADDY (U.K./Italy) - Official translator Team. DANIELE (Italy) - Morally supports the Team in time of difficulty. Who of them is the CEO?

BERTOCAR.
Quote:

Who of them are employees?

None: those were only volunteers.
Quote:

Who got paid?

None: those were only volunteers.

Of course, the company's employees working on the project had their salaries.
Quote:

And what was the TINA companys office?

See above: it doesn't apply to me and the volunteers. We had no desk / office: we were' working from our homes.
Quote:

CESARE: That was from a company?

I don't get it: what do you mean with that?
Quote:

Lets look at your TINA team own post: [quote]TiNA is a project to which we dedicate ourselves with seriousness and strong passion, but always in our free hours, precisely because at the moment we don't think about anything other than passion and pure fun in creating.

As above: we did for passion. Nobody of us was paid.
Quote:

Cesare the specs of the TINA project were from day 1 absolutely impossible FPGA which was claimed to run 400MHz CPU - is technically absolutely impossible to reach this clock. The 128 system bus between the 3 FPGA and the 128 memory bus is technically absolutely impossible to reach - as the used FGA not even have the PINS for this!

See above: not my domain. I don't know what's possible or not.
Quote:

Cesare you knew these project features and you wrote online about them.

Yes, as all of us: that's what was communicated to us from the company.
Quote:

Cesare and you advertised these hoax features for long time.

See above: it wasn't a hoax. It was a project which wasn't materialized despite the efforts. As many others. As it happens on the industry. It's... life!

You can speak with the above "expert" about this. If you don't agree, you can always argue with him and try to convince him that he was wrong.

And, as I've already stated, I've just reported what the company was expecting to achieve. Me, like all others. Just ambassadors.
Quote:

And which point did you realize that these features are technically totally impossible?

Never. What's not clear to you that I'm NOT a hardware engineer? Could you please tell me how many times I've to repeat it? How should I've known it, using a crystal ball?!?
Quote:

@Gunnar [quote] Gunnar wrote: @Gunnar [quote] Cesare you advertised these impossible hoax features for the TINA for long time. You posted in many forum, false numbers. At which point did you realize that these features are technically totally impossible? [/quole] At which point did you correct this? At which point did you tell people that the photos on the website were fake. At which point did you tell people that you NOT have a soft CPU with 800Mips as promised. At which point did you tell people that the advertise 400MHz were not true and are totally impossible. At which point did you tell people that the advertised 128Bit bus was fake and is in fact technically totally impossible. At which point did you tell people that nothing that was promised was real - was that all was a joke? At which point did you tell people that the people involved have no experience with FPGAs or PCB layout - and that the whole website was just a hoax? At which point did you tell people that the PCB schematics shown where all FAKE?

Same as above: NEVER.

I can only add that when I've left the team, then I've reported it on TINA's forum and also here. You also were well aware of it (albeit you pretend to do not understand).

After that, I wasn't in touch anymore with the project and contributed in no way (for a simple reason: I was an Intel employee and I had a contract to fulfill. Basically I was "Intel's property". Who work/ed with companies like that knows exactly what I mean).

So, I don't know what happened. A certain point in time one of the volunteers (the web master) approached me and reported that the site was gone. Which means, that I don't even know when the CEO decided to close the project.

That's all.

But while the project goals were too high and could not be satisfied, much probably because the hardware engineers haven't had enough experience for such high-performance designs, at least there was good faith about it. In fact, not a single cent was gained by anybody. And the company wasted time / resources on that without a ROI.


Now let's talk about another project which went to the market, but it was advertised and it's still advertised with wrong information with the precise purpose to capture the attention of passionate people and let them buy the product: http://apollo-core.com/index.htm

It was already reported several times by different people that the information shown there are NOT correct and misleading, since the product is missing features whilst claiming full compatibility with the 68k family.

Specifically (and to be short, only listing the most important points):
- instructions are missing;
- PMMU (usable from the existing software) is missing;
- some instructions aren't implemented in hardware but only in software (hence: they are slow);
- FPU is only supporting up to double-precision and is lacking extended precision.

Nevertheless, nothing was changed even after that this was highlighted so MANY TIME.

So, it's crystal clear that this information is left wrong/misleading ON PURPOSE, and NOT by mistake! So, there's NO good faith: at all!

It means that this is a CRIMINAL plan -> FRAUD.

It means that, BY DEFINITION, the product is a SCAM.

It means that, BY DEFINITION, who's selling this product is a CROOK.

What do you think about that?


Now that I've fully clarified my (and of other volunteers) role in the TiNA project and how the things went (as I've said: in GOOD FAITH! No customer lost money about that) you should be satisfied.

And I don't want to waste my time anymore on that.

It means that if you'll continue again with your defamation campaign then I'm make public and I will spread around what your are doing with your project, so that a large audience will get the right knowledge about it and if people feels swindled then they can sue you and bring you to the court.

Understood? It should be crystal clear, right?

@cdimauro

Quote:

I think that it's true and it applies to TINA as well: too high expectations that were disappointed when the project was implemented, which led to its closure.

TINA was hoax

Did you ever do a PCB of it?
No you did not

But you showed PCB pictures on the TINA website.
And these pictures where stolen from other project in the internet.
Why did you guys steal pictures from other websites?

The advertised numbers of the project were completely technically impossible.
This has nothing to do with to high expectations.

This means you have no clue about hardware and you not cared to look in the FPGA manual.

The TINA was advertised with 128bit memory and with 400MHz clock.
Both values are impossible.

The FPGA does not have enough IO pins to connect 128 Bit of memory.
And the FPGA can not reach 400MHz.

What does this show us?

You guys made a nice looking website, with stolen pictures from the internet.
With schematics with 128bit bus you painted and with 400MHz.

If you read the FPGA manual than you see immediately that both values ARE IMPOSSIBLE.

Do you say that you guys did all the website work and painted the schematics
without anyone of you reading the FPGA handbook?

@cdimauro

Quote:

- FPU is only supporting up to double-precision and is lacking extended precision. Nevertheless, nothing was changed even after that this was highlighted so MANY TIME.

You are simply not able to read.
The website clearly states that the FPU is 64bit.

Quote:

Apollo Core 68080 advantages: Fully Pipelined Superscalar Executes up to 4 instructions per clock cycle Two address calculation engines Two integer execution engines Market leading code density Optimal cache utilization Separate data and instruction caches, supporting concurrent fetch/read/write per clock cycle Automatic memory prefetching Memory stream detection Store buffer Branch prediction Fully pipelined, 64bit SIMD AMMX Vector unit Fully pipelined, double precision FPU