If you had the chance to build a new Amiga

[Samurai_Crow]

What got me started on Amigas was that it was a computer with a console-class chipset. It had all the tricks of both! It also didn't hurt that I had used C64s and C128s for years either.

If Sony would quit hobbling the homebrew scene with rediculous DRM on the Playstations and come out with a desktop class OS for them they'd put Microsoft in a position of trying to merge the XBox consoles with the PC. Of course the PC manufacturers wouldn't like that, even though it's a short jump for Microsoft. Why Nintendo doesn't continue the tradition of Mario Paint on their current platforms astounds me.

Any one of those scenarios would satisfy me.

@BippyM
Have you looked into the RockPro64? It has a hex core ARM64. The 2 bigger A72 cores have good single and dual-threaded performance while the other 4 threads are lower power consumption for background processes. The ODroid N2 has the same SoC also (the RockChip 3399) for a comparable price. The only things weighing those down are the Linux OS being heavy.

[BippyM]

I think that is what cannot be recreated though.. The magical mix of hardware and software and os. Get that right

[ReaperX]

I think that the only aspect of the Amiga that remains relevant to the modern computing world, outside of retrocomputing and gaming as a hobby, is the philosophy: an easy to use, GUI-based system that empowers the user and encourages artistic and programming creativity, including direct hardware programming.

Modern hardware and OSs are way more powerful and capable than the old home computers of the 80s, but we have been paying a price for this progress. There are now greater access barriers to programming than there used to be, there is less software efficiency, more bugs and more mystery behavior that no one can fix or even understand due to excessive complexity and stacking of abstraction layers, and less "ownership" in the true sense.

For the beginner, there is no experimenting with simple graphics on a modern PC. If you want to program graphics, you need to choose a programming language first, but then you'll be disappointed to find that there is only text-based output (think Python). Just being able to SetPixel(x,y,color) or something like that requires a library. There will be many choices for that, just like there were many choices for the language. Here, choice = barrier. Once you got the library, you may need to navigate an obstacle course of software abstractions: create a window object, drawing context, color object, point object, and similar. Many people will have given up before they get to that point.

When you've mastered the required prerequisites, you may find that the performance of your program is disappointing, and there are some things you just can't do. For example, you may be able to create graphics, but you can't animate it with sufficient fps. You're working with abstractions that call other abstractions, which call yet other abstractions, and none of them are necessarily meant or optimized for what you want to do. These structural problems affect the beginner as well as the professional.

On the Amiga, you had AmigaBASIC, and all it took was CLS and PSET, and a little math, to explore and experiment with graphics. It was a huge success experience to make the Mandelbrot set appear on your screen, with your own program. Those who had progressed beyond BASIC and learned C had direct access to OS and hardware. There was easy access for programming beginners, and unlimited agency and greatest possible performance for advanced coders. This encouraged a culture of excellence and innovation, of pushing the limits of the hardware, of doing more with the same.

This is what we've lost, and it's been to the detriment of software engineering culture and computer technology in general. Young people know less about how computers work than the 1980s/1990s home computer generation, and that includes programmers. There is ever less efficiency in programs and too much Connect Part A to Part B and Hope It Works software design. Moore's law enabled all this, but it's ending.

On the flip side, and this portents well for the idea of a New Amiga, there is huge optimization potential now in better coding, cutting out the middle men of resource-eating frameworks and abstraction layers, and even programming in assembly. Custom hardware could make a comeback, based on FPGAs. Read "Software disenchantment" by Nikita and "The Death of Moore’s Law Will Spur Innovation" by Andrew Huang.

If I was this multi-billionaire that some posters have imagined, I'd buy up the Amiga IP and release it as open source, just for good measure. Then I'd start a non-profit similar to the Raspberry Pi foundation and give it a few billion dollars, the Amiga naming rights, and the mission of building and continuously developing a new (open source) GUI OS from scratch, in the spirit of the Amiga. Let's call it AmigaOS 5.

AmigaOS 5 would run on a standardized hardware platform, so it and its applications could be more specifically optimized for that hardware, which reduces complexity and abstraction. With the near end of the exponential performance growth of hardware, this approach makes sense again. Ideally, the hardware would be open as well, without mystery BLOBs, "management engines", "platform security processors" and similar, but that wouldn't be a realistic option right now, even with a billion dollar budget, so today I'd use standard PC components.

AmigaOS 5 would be based on an Exo Kernel, so that applications could have as much direct hardware access as security considerations permit. Unlike Linux, it would be a complete desktop OS, with the GUI an integrated part. There would be no fragmentation of the ecosystem and developer community along any number of distributions, desktop environments, toolkits and display servers, nor any of the instability and inefficiency you get with having an OS that is pieced together from independently developed components.

Some of my design goals that would set this system apart from existing ones would be:

- GUI and standard applications are extremely responsive, with low latency and completely smooth, delay-free scrolling. This would not just be for gaming. High visual latency makes typing slower, more prone to errors and causes more eye and muscle strain. Similarly, choppy scrolling in productivity applications causes more fatigue and reduces performance, because it's disorienting to human perception.
- Fast startup: after power-on, you're on the desktop in 2 seconds.
- Security, privacy, reliability and performance as development priorities, enabled by less OS complexity and fewer abstraction layers.

A major distinguishing feature would be usable security and privacy. For example, there would be a button on each application window to persistently enable/disable WAN access, at the OS level. There would be a per-application button where you can make its internet traffic go through a VPN connection. There could be a set of buttons to restrict an application in other ways, basically offering common sandbox functionality, but without the hassle of having to edit configuration files, and with permissions changeable while applications are running. Basic, but secure, password-protected and GUI configurable LAN filesharing would be another feature. (Linux doesn't have this, Samba is rather a mess.)

I'd re-introduce some of the nice lost features of the Amiga to modern desktop computing: scrollable, independent screens that can have different color depths and resolutions, .info files, arbitrary icon sizes, datatypes.

There would be a transparent emulation layer so classic Amiga software can be run seemingly natively.

I'd like the desktop to improve on the core functionality, which neither the commercial OSes nor the Linux desktops can't ever seem to manage.. for example, ability to rename/move a file or folder while they are being copied/written to; automatic queuing and starting of multiple copy operations; a copy process of many files that's going to take many hours doesn't stop with a popup as soon as you step away because it encountered one file that already exists, but continues in the background; copy isn't denied because there is not enough file space (even though most files already exist and you intend to overwrite them..). There are many small nuisances like that in existing desktops systems.

The OS would ship with an IDE, support for just a few "official" programming languages (say, Python, C or C++, assembly) and standard libraries for programming GUIs, multimedia, networking, encryption, 2D and 3D graphics, powerful enough to create games and professional GUI applications. Most importantly, it would be easy and fun to work with. A few lines of code should enable you to create and display simple 2D and 3D graphics and animation.

Since we have a billion dollar budget, there should be enough to write or port a full set of open source applications to AmigaOS 5: web browser, end-to-end encrypted chat and videoconferencing, media player, an office suite, drawing/painting/sound/music/video editing, 3D modeling/rendering.

For those who don't want to program, the New Amiga would just be an amazingly fast, responsive desktop computer with a complete suite of communications, productivity and creative applications, all with consistent look and feel, and superior usability, privacy, security and reliability. People would be amazed how much better they feel working with a computer that doesn't constantly stress them with high latencies, unresponsive scrolling, random freezes, etc.

[Hedeon]

Seeing you registered just to make this posting....are you a billionaire by any chance? :-D

[ReaperX]

Originally Posted by Hedeon

Seeing you registered just to make this posting....are you a billionaire by any chance? :-D

Sadly no

As unlikely as a return of the Amiga may seem, I think that some sort of revolution in personal computing along the ideas that inspired the original Amigas is inevitable. We've gone down the wrong path in the last quarter century.

The Jay Miner society wrote in 1998,

"Like all Amigans (and not a few users of other platforms too) we had real concerns about the direction personal computing was moving - more complexity, extravagant waste of resources (with consequently less stability and functionality), fewer alternatives, 'dumbed-down' software with little flexibility. This is the exact opposite of the elegant promise of the Amiga envisioned by Jay Miner."

Much of this remains true today. Complexity and computational resource waste have tremendously increased since these words were written. The other criticisms - lack of stability and functionality, lack of alternatives, dumbed-down and inflexibe OS and software - apply at least to the computers that the majority of humanity now uses, i.e. Android-based phones. Millions use Windows on their PCs somewhat unhappily because they have no choice or feel that they have no choice.

To all of that, we have now added a problem that was barely imagined in the 90s - lack of privacy due to pervasive data transmission and collection.

Yes, there is Linux, and many Linux OSes are great. I've been mostly happy with Manjaro for several years. Still, Linux OSes come with their own set of issues. It's impossible to do justice to that topic in a single post, so I won't try.

Humanity could use a new Amiga, in the spiritual sense, because nothing we have right now is as good as the Amiga was, relatively speaking, back then.

It may not be by that name, and it may not happen due to the charity of a single individual. Perhaps a bunch of developers will one day get together and build a modern, Amiga-inspired OS as a grassroots project.