AmigaOS 3.2 and beyond

[Thomas Richter]

Originally Posted by Rotzloeffel

Screen dragging will be available in p96 2.5

Yes, but... There are a couple of restrictions due to the VGA hardware RTG cards are based on. First, there needs to be enough video memory available to keep both front and back screen on the board. For example, with 2MB graphics memory available, two 1280x1024 8 bit screens will *not* fit on the board (1280x1024 > 1024x1024 = 1MB, thus more than 2MB for the two screens needed).

Furthermore, VGA chips can only adjust the frame pointer mid-point, thus they cannot switch the palette, and they cannot switch the graphics mode. Thus, a chunky 8-bit screen cannot be dragged over a true-color screen, and if two screens with with a lookup table share the monitor, only the lower screen will get the correct colors. Also, native screens cannot be mixed with RTG screens (for obvious reasons).

What will work, however, is mode-coercion. That is, if the screen sizes are different, but the modes are identical (i.e. twice hi-color), then the upper screen will be "coerced" to the mode of the lower screen, and will become an auto-scroll screen (as for native screens).

Finally, there can be only two screens at once on the display, and there are some chips that do not support screen dragging at all since they lack the necessary feature, namely a line compare register.

One final word: P96 2.5.0 works different than the Os 4 implementation. The latter "moves bytes through the video RAM", but then requires more disciplined programs that do not write into video ram themselves, and it requires somewhat more CPU power. Neither of that is the case for the 68K, hence we have a different technology here.

[hUMUNGUs]

Originally Posted by Rotzloeffel

Screen dragging will be available in p96 2.5

Really ? Awesome :-)

[Rotzloeffel]

Originally Posted by hUMUNGUs

Really ? Awesome :-)

Yes! As Thomas wrote above, with some limitations. But these "limits" are technical nature and can not be fixed.


As in CybergraphX the Cybervision64 was the only card which supports screen-dragging by setting the responsible Tooltype, I actually tested it on CV64, CV64/3D, GVP-Spectrum and it works!

[Tigerskunk]

Originally Posted by Rotzloeffel

Yes! As Thomas wrote above, with some limitations. But these "limits" are technical nature and can not be fixed.


As in CybergraphX the Cybervision64 was the only card which supports screen-dragging by setting the responsible Tooltype, I actually tested it on CV64, CV64/3D, GVP-Spectrum and it works!

I hope I am still alive when this gets released...

[Storm]

Lol, I’m sure you will be

[Thomas Richter]

Originally Posted by Steril707

I hope I am still alive when this gets released...

I'm not so foolish to give a release date at this time, and to make announcements of things that are not yet working to my full satisfaction. Things will be ready when they are ready. However, to give you a tentative estimate, a release this year is highly likely.

To make another announcement of things that are in a state "ready enough to make an announcement", we will have a CVisionPPC driver as well. However, the chip the cvision is based on lacks a necessary hardware feature to support screen dragging. Thus, no screen dragging support for this card, sorry.

[IanS]

Originally Posted by Steril707

Your ideas are not in any way new, and the debates about this have raged all over Amiga forums since the internet was in its infancy.

Basically it comes down to:
* AOS is a hobby OS for the remaining few people using Amigas in 2020
* there is no interest for an OS with these modern features. MacOS covers the arty and trendy people faction, Windows the majority of business users, and Linux the nerds and server shit. There is simply no need for another OS in a world where classic tabletop and laptop OSes are on their way out anyway.
* there have been many tries to modernize AOS. Like Amithlon, or even within sub branches like AROS there once was "Anubis", which was supposed to be a Linux with an AOS UI. Guess what? No one gave a fuck, so it died a year later.

What you are doing here is proposing modern people should use flying cars, and thinking you are a genius for doing that. When the reality is a lot more difficult, and there are reasons why this is not happening.

Plus, talk is cheap.
If you think it's a great idea what you are proposing, put in the man hours, or shut up.

Before telling me to shut up, you might want to actually read my post.

I CLEARLY stated that these ideas aren't new, they are not "my" ideas and that nobody is likely to do any of them. I was around when all these ideas first appeared in the Amiga world, and they all had their proponents and detractors. Guess what? They are all still being worked on to some degree. I didn't suggest they were great ideas, just that those that want such modern features will be better off adapting something like Linux, rather than starting a whole new OS from scratch.

I am not suggesting we all use flying cars, although in fact a great many people have made individual flying vehicles, including multi rotor vehicles akin to flying motorcycles. Essentially full size versions of the remote control drones in common use. I think they are interesting. Certainly cheaper and easier to fly than a typical helicopter.

I already mentioned Aros, Morphos and Amithlon, and to say "nobody gave a fuck" about Amithlon is about as wrong as you can get. Lots of people were upset, and begged for it to continue development. Even today, people are trying to modernise it. The reason it died? Typical legal issues, causing the author to essentially leave the Amiga scene. You might want to do some more research on that subject.

There's no interest in an OS with these 'new' features? Really? Then why is it such a common topic in Amiga circles? Or have you missed that? Tabletop OSs are dying? No, the market has changed, but they're not dying. None of the mobile OSs is as capable as ones running on laptop and desktop systems, mainly because of heat and power demands. Besides, most mobile OSs are derived from bigger, more mature bases. Apple use their Mac OS as the base for IOS, Android is based on the Linux kernel, Blackberry 10 was based on QNX and Windows phone 10 was based on Windows 10. Those core OSs are still being developed. Apple are even moving their desktops to custom ARM chips... so not quite as dead as you propose.

So instead of telling me to do it myself or to shut up, you perhaps ought to learn basic comprehension skills, do some of your own research and stop being an idiot. Don't like being called an idiot? Well when you're rude to others, expect them to return the compliment.

[bubbob42]

Well, we were never short of people having ideas, weren’t we?

[Rotzloeffel]

Originally Posted by Thomas Richter

To make another announcement of things that are in a state "ready enough to make an announcement", we will have a CVisionPPC driver as well. However, the chip the cvision is based on lacks a necessary hardware feature to support screen dragging. Thus, no screen dragging support for this card, sorry.

[Tigerskunk]

Originally Posted by IanS

just that those that want such modern features will be better off adapting something like Linux, rather than starting a whole new OS from scratch.

It has been tried. No one gave a shit.

Originally Posted by IanS

I already mentioned Aros, Morphos and Amithlon, and to say "nobody gave a fuck" about Amithlon is about as wrong as you can get. Lots of people were upset, and begged for it to continue development. Even today, people are trying to modernise it.

If there would have been money to be made, this thing would still exist.

Originally Posted by IanS

There's no interest in an OS with these 'new' features? Really? Then why is it such a common topic in Amiga circles?

MorphOS and AOS 4.1 are barely alive.

Originally Posted by IanS

Or have you missed that? Tabletop OSs are dying? No, the market has changed, but they're not dying.

The market has changed to a large percentage of the use of mobile OSes over the last ten years.

Desktop OS is kept around for coders and professionals. No casual (probably at least 95% of the market) will build a pc and put Windows, MacOS or god forbid, Linux on it anymore.

Amiga is barely a market for a few thousand people left.
That's niche hobbyist, but nothing more.

Originally Posted by IanS

None of the mobile OSs is as capable as ones running on laptop and desktop systems, mainly because of heat and power demands. Besides, most mobile OSs are derived from bigger, more mature bases. Apple use their Mac OS as the base for IOS, Android is based on the Linux kernel, Blackberry 10 was based on QNX and Windows phone 10 was based on Windows 10. Those core OSs are still being developed. Apple are even moving their desktops to custom ARM chips... so not quite as dead as you propose.

See above.

Originally Posted by IanS

So instead of telling me to do it myself or to shut up, you perhaps ought to learn basic comprehension skills, do some of your own research and stop being an idiot. Don't like being called an idiot? Well when you're rude to others, expect them to return the compliment.

Call me whatever you want, you are still wrong.

[Thomas Richter]

Originally Posted by IanS

There's no interest in an OS with these 'new' features? Really?

Look, the usefulness of a piece of software depends on its "use cases". So tell me, who wants to use such an operating system, and what is the target market? Which competitors operate in the same market, and how is this idea ahead of the competitors?


Frankly, I see nothing. There is the "niche retro market", which is defined by old legacy Amiga applications. This is the market 3.2 and following address. There is some interest, and none of the competitors address this. As soon as you break any sort of compatibility, you loose the market.



So why do you need new features? We are not talking about a "productive system" that would require such features. Actually, if you really need that today, your best bet is to run a variant of UAE on top of a stable system, yet loose an important aspect of the whole story, namely the original feeling of the machine.


So, try to answer this important question: Who would use it, and why would someone use it?

Originally Posted by IanS

Tabletop OSs are dying?

No, but that market is saturated. There is Win, there is MacOs, there is Linux and its variants, Android for some particular markets.. Several attempts to establish more alternatives have failed (FirefoxOs - anyone?).


So, why another attempt?


Do not run behind your competitors, and do not attempt to play by somebody else's rules. The best way how to win a game is to define its rules yourself.

[Warty]

Originally Posted by Thomas Richter

There is the "niche retro market", which is defined by old legacy Amiga applications.

I absolutely agree with the sentiment that the only real market here is the niche retro market [for legacy Amiga users and a handful of younger folks who are interested in old machines, and another handful of old Atari/etc users who are interested in other machines of the era, not just the ones they used].

but I wonder if there are perhaps 2 sub-markets here (with lots of cross-over):
1. People who want to use their old Amiga software.
2. People who want to use their old Amiga hardware.

Originally Posted by Thomas Richter

This is the market 3.2 and following address. There is some interest, and none of the competitors address this. As soon as you break any sort of compatibility, you loose the market.

I'm thinking that just about everyone in the retro market has already figured out how to multi-boot their system to different kickstarts. And definitely, 100% of the time, also have multiple UAE configs for various KS versions.

So I'm wondering if it's really a hard rule that you can't break compatibility with ANY legacy software, in the name of making a better OS for retro Amiga hardware. It's not like all games ever ran on every KS/WB combination. As long as you didn't break WHDLoad, I think you might have a little more room to make a change or two that breaks some legacy app. If the user really needs that legacy app, they already have 5 ways to run it. They don't need 3.2 or 3.3 to be the only way to run it.

I don't know that for sure of course. This effort is definitely way off the beaten path, there is no precedent as far as I'm aware of. Apple, Microsoft regularly broke old apps moving forward (Apple lots more than MS of course, until Win7/win10 era). But that's a different paradigm altogether. As a purely retro platform, you definitely have more restrictions about what will be acceptable.

But maybe it would be worth looking at, for 3.3 or something, whether some old KS code could be dumped in favor of squeezing in more of your new stuff. The more stuff that gets pushed out of ROM and into regular RAM, the tougher it gets to run on some older systems. But that code has to go somewhere...

Anyway, thanks for keeping us up to date on the progress!

[Thomas Richter]

Originally Posted by Warty

But maybe it would be worth looking at, for 3.3 or something, whether some old KS code could be dumped in favor of squeezing in more of your new stuff. The more stuff that gets pushed out of ROM and into regular RAM, the tougher it gets to run on some older systems. But that code has to go somewhere...

Well, this sounds more like a market for a piece of hardware, consisting of a flash-ROM with an additional "switcher" mechanism on top that allows you to load several kickstarts into the same system, then switch at boot time between them. Thus, some sort of "intelligent kickstart switcher", which can be udpated by software, so you don't have to exchange physical ROM chips on your machine.


If you ask me: Is there a market for such a thing? Yes. Would I buy one? Yes.

[steenl]

Originally Posted by Thomas Richter

Well, this sounds more like a market for a piece of hardware, consisting of a flash-ROM with an additional "switcher" mechanism on top that allows you to load several kickstarts into the same system, then switch at boot time between them. Thus, some sort of "intelligent kickstart switcher", which can be udpated by software, so you don't have to exchange physical ROM chips on your machine.


If you ask me: Is there a market for such a thing? Yes. Would I buy one? Yes.

Like the Kipper2K Romulator, unfortunately only for a500/600/2000

[lmimmfn]

Originally Posted by Thomas Richter

Well, this sounds more like a market for a piece of hardware, consisting of a flash-ROM with an additional "switcher" mechanism on top that allows you to load several kickstarts into the same system, then switch at boot time between them. Thus, some sort of "intelligent kickstart switcher", which can be udpated by software, so you don't have to exchange physical ROM chips on your machine.

If you ask me: Is there a market for such a thing? Yes. Would I buy one? Yes.

Lets not forget within reason the OS can drive the market when it comes to convenience. I would be ok with this if theres an advantage. We're all using multiple rom versions for whdload, slightly different i know.

[bubbob42]

Originally Posted by lmimmfn

Lets not forget within reason the OS can drive the market when it comes to convenience. I would be ok with this if theres an advantage. We're all using multiple rom versions for whdload, slightly different i know.

I think this has happened or is happening already, even before we brought 3.1.4.

Almost all recent accelerators support MapRom and many (HC-series, ACA500+, Vampire...) offer at least one slot to be flashed with a Kickstart using command line tools on the Amiga directly.

True: The AGA machines have been left behind concerning the introduction of the latter, but only until recently - both Vampire and Warp accelerators have flash rom and the upcoming IComp card should have it as well.

[Storm]

I have done some more testing and I just can't get SuperFloppy to work on my A2000 with a GVP 030 Combo Card (latest firmware). I do have it working on my A1200 with a Blizzard SCSI addon to my 1260 (using a mount file). I followed the FAQ and tried to set up a mountlist and also tried a mount file. I feel it has something to do with the SCSi device, but I can't be sure. Tested with SyQuest drive with 88MB and 200MB media and a SyJet. Will test with Iomega equipment later on.

[Tomislav]

Originally Posted by steenl

Like the Kipper2K Romulator, unfortunately only for a500/600/2000

There is also Romy for A3000/A4000.

[lmimmfn]

Would booting from pcmcia be a possibility?

[Rotzloeffel]

No,.. Therefore the driver must be in ROM, which is not