AmigaOne X5000 – first impression | Hacker News Reader

AmigaOne X5000 – first impression

amigapodcast.com

78 pointsbane58 comments

I don't know about the Amiga stuff, but as PowerPC board for running Linux looks great:

"The e5500 is a superscalar, dual-issue core with out-of-order execution and in-order completion. It maintains the seven-stage, four-issue pipeline that is a trademark of the e500 family, but is able to provide additional performance, by scaling up to 2.5 GHz." [1]

[1] http://www.nxp.com/assets/documents/data/en/white-papers/64B...

Yeah, my first thought was "this board looks great for testing big-endian". Would be nice to see comparisons to e.g. the G5.

Sure. I bought back in the day a G4 Mac Mini just for that purpose (running both Linux and OSX) :-)

In which case do you need to test big endian? Compiler or low-level library development?

TD-linux develops codec libraries.

I develop crypto code. Testing on BE is nice, not just because of endian neutral code but testing on something 'weird' can turn rare faults into common ones.

Beyond endianness I've found interesting bugs by running code on HP-PA (stack grows in the reverse direction) and on Itanium (misalignment causes a trap to the kernel)...

I wish there were a qemu supported fake architecture that made every optional decision differently (or at least every one that works with most code...).

Mainly for networking or disk storage, and in general for anything involving storing integer data into raw packets that will reach different CPU architectures (e.g. load or store an integer from/to a memory buffer into/from a CPU register).

At what point will all big-endian usage be just developers testing big-endian with no users to use the results, and at what point shall we stop even testing?

(Yesterday, I finally got around to making encoding_rs work on 64-bit big-endian systems by testing PPC64 in qemu. Then I timed out trying to get a 32-bit big-endian system with recent enough glibc to run Rust up and running in qemu, so I still haven't tested the 32-bit case... Unclear if anyone even cares.)

With IBM's huge investment in better support for little-endian PPC (in hardware, in software development toolchains, and in software itself) it seems likely that many PPC users are going to slowly move away from little-endian. I can imagine some embedded uses, e.g. in network hardware, may well stick with big-endian because it fundamentally makes sense there, but I imagine it's going to become increasingly obscure.

My above comment is of course wrong, they're going to move away from big-endian to little-endian. Pretty sure I changed what I was saying half-way through. :)

MIPS is the other large big-endian case that's still in wide use (especially in home routers), however the machines are even more painfully slow to test on.

Most MIPS CPUs can run both BE and LE (MIPS-LE), although is usually tied to the kind of applications (e.g. most MIPS routers run in big endian mode)

It's a pretty old chip, and there's a newer generation of Freescale stuff -- the e6500 -- with more cores, smt, pcie 3, etc. Plus, it's kinda expensive compared to other PPC boards. Even the Qoriq reference board is cheaper [1], and reference boards tend to be rather high priced themselves. If you're gonna run Linux, your best bet is elsewhere.

[1] http://www.nxp.com/products/microcontrollers-and-processors/...

Wow, thank you very much! :-)

Can someone tell me what makes "real" Amiga so wonderful that it's worth paying $1500 for an objectively outdated computer? What does legit Amiga do that something like, for example, AROS doesn't do?

I also do not quite understand. As someone who loves 68000-based powerhouse art computers from the '80s, I can understand the appeal of the original Amiga line, but the PowerPC ones confuse me a bit because they no longer bring a unique graphics chipset to the table. Without the (potentially) interesting graphics chipset, does the appeal really extend beyond the OS interface design and the brand itself?

There where some pretty efficient workflows in many of the tools, it was a really logical layout to an OS, and it seems like much of that thinking infected many of the developers that built apps for it. The video toaster comes to mind, while the Amiga hardware was obsolete by the time NT and SGI where in full swing for graphics, you still saw the real time stuff being done on toasters. It took advances in software to finally put the final nail in it's coffin. A lot of people that did not use Amiga's don't really realize how far ahead of their time they where. Through a lot of ingenious software and hardware designs and hacks they really built something that was from the future. That being said, after the toaster was surpassed, the present passed the Amiga up and now it's just nostalgia.

I spent more money than that on a bicycle just to learn to build bicycles. As hobbies go, $1500 isn't crazy.

Yep, it's just a hobby and not a terribly expensive one as you say.

I loved my Amiga 500 & 2000 back in the day, and spent more on them than this costs although it was obvious they were never going to be main stream.

I wouldn't do it now, but good luck to those who do.

The Amiga has a small but really enthusiastic fanbase that doesn't want to admit that it's dead. That's it, as far as I can tell. All of the modern Amiga hardware seems to suffer from similar issues to the ones mentioned here - it's expensive, slow by modern standards, has strange and annoying hardware and driver issues, and there's just not much software for PPC Amiga. It can't even run modern web browsers. (Unless I'm mistaken, the second core and a lot of the motherboard peripherals are entirely unused because AmigaOS doesn't support them.)

I definitely think that, when it was released, Amiga was ahead of its time, at least compared to the equivalently priced Mac and DOS/early-windows systems. I just feel that the system hasn't really kept up since commodore left the picture.

As a child, I remember walking into a "World Computers" retail store and seeing a Commodore Amiga running some graphics demo. I was utterly blown away. It was far beyond what I had ever seen with a home computer, which at that point, was limited to a Macintosh SE and an 80286 PC running DOS 6.

Bouncing Ball?

https://www.youtube.com/watch?v=-ga41edXw3A

My first store demo of an A500 was an educational title that taught German with full audio samples using a point and click interface. My dad bought the thing with practically zero convincing from his excited kids.

You've hit the nail, hard. I grew up on Amigas and SGIs and owe those platforms everything I am today professionally, but what is going on with Amiga post 90's is catering that nostalgia feel and that's about it. It's dead, Jim - but let's sell flowers to people worshiping the grave.

Rationally you are 100% on the money, but people are not always rational and I actually love it that they did this simply because they could. Bringing a machine to life at this level is an adventure all by itself and given the rather boring nature of the compute landscape I'm all for people taking the long way home and enjoying the scenery rather than just being driven by price and practicalities.

As a nostalgic Amiga user I agree on principle, sadly this platform has little to do with original Amigas; I applaude their effort but wouldn't spend even 100 euros to buy one. The Amiga success was a result of genius, money and time constraints, being in the right place at the right moment and a huge users niche still untouched by IBM/Apple/Atari; it's impossible to repeat, period. However I we wondered what if the surviving original developers banded together again and designed something new to be produced by some Chinese fab in Shenzhen, like small companies like Friendly Elec or the guys behind the Orange Pi can do today like twice a month. That could be an interesting move to give new life to a concept.

To the point!

What I don't get: Why don't they use at least a customized case which doesn't look like a typical PC case (such as Apple does with the Mac Pro)?

I had a maxed out Amiga 500 (with sidecar expansion) back in the day and it was awesome (for games and 2D graphics) in 1988, amazing in 1990, and then Commodore ate it owing to monumental incompetence.

Even at the height of its technical superiority, it was crippled for serious work. Its pixels weren't square in any graphics mode, and in the highest resolution mode you needed to endure flicker unless you had a special monitor.

I remember when our 512k mac's power supply died and we tried to use our (maxed out) Amiga for word processing we discovered the top-rated word processors became unresponsive with documents over 4-5 pages (WriteNow on the Mac was cheerfully handling hundred page documents with graphics). And its print quality was atrocious even setting aside the whole non square pixel problem.

I don't know why you were downvoted because you're describing a major problem the Amiga had in the creative industries of the '80s and '90s.

Back then almost everyone was still designing for print, and the Amiga just wasn't cut out for that. It was a multimedia machine before there was any way to distribute multimedia to anyone else than other Amiga users.

Video production was a niche that was compatible with the Amiga's design, but working with analog video wasn't going to be mainstream ever... and for digital video, the Amiga was fundamentally underpowered.

Designers were a tiny minority of computer users back in the heyday of Amigas, and they're still a tiny minority of computer users now. Designing for print on computers back then was as much a nice activity as using computers for video post production. So if the Amiga sucked at printing, that was not what held it back from succeeding, as desktop publishing was just in its cradle back then. In fact, as an avid Amiga user back in the day, I never ever heard anyone complaining about the Amiga's printing chops.

These are the kinds of arguments against the Amiga that I actually heard: "Who would ever need 4096 colors?" "I don't need to use all those sounds" (coming from IBM PC owners whose computers could only display 16 colors and only produce bleeps and blops for sound).

The Amiga's real forte was games, pure and simple. It was a monster of a gaming rig, for the day, and a ton of gaming innovation happened on the platform. Gaming was a pretty mainstream activity for computer users even then, and Commodore completely flopped by not marketing the Amiga as primarily a gaming machine.

Word processing was a huge market even if design was not.

I totally understanding the beauty of the system; what I don't understand why they don't move to something like AROS, which is open source and runs on (substantially) cheaper hardware.

What makes the Amiga so wonderful is a great question, but what makes it worth $1500 depends very much on the buyer. Some people will pay far, far more than that for a stamp or a coin or a rock, which do absolutely nothing. Beauty and value are in the eye of the beholder.

There's a large number of electronic music producers that still use these things.

I'm absolutely convinced they can't produce anything on their Amiga 500 that can't be made on a $50 sbc just as much as the people I've spoken to are absolutely convinced that I'm wrong and just don't get it.

I think it's more a sign of prestige for them; like trying to argue that the durability and workmanship on the $1,000 designer t-shirt is indistinguishable in quality from the generic $10 one - this is also a losing battle.

Does objective criteria matter if it is a machine that does what you need of it and you find pleasure in using it?

I write this while sitting in front of a modern machine that has as much appeal as a kitchen appliance, yet I regularly dig out an old computer to revive old interests. It is irrational, since that modern machine can easily be used to explore those old interests. Yet the old computer gives me pleasure while doing so, so that is the route that I have choosen.

Personally, I like the diversity of ideas and implementations. It can lead to new explorations by researchers/developers, compilers detecting unforseen errors among other things, and immunity for a subset to untargeted malware.

Me, I'd rather emulate a few classic games. But it had me wondering for a while if I'd get any value from this, before deciding "no".

I find it interesting that so many of the people making YT videos and in the current Amiga community don't seem old enough to have been there first time around! So it can't be just nostalgia.

What I really want is someone to reimagine a 3000+ prototype[0] for this decade and a modern Intuition. Then I'd probably impulse spend a lot of $. Of course that would cost someone a lot of R&D spend! It's almost inconceivable anything could recapture the magic though. :)

There's an interview from a couple of years ago with the guy who made the X5000 https://www.youtube.com/watch?v=Mb6nbQSOAA0 and his Amiga addiction.

I started out doing real programming on an Amiga and I took a LONG time to retire it. Wasn't for lack of trying...

For years after if I needed to do something useful I used the Amiga over Win, on my first couple of very expensive PCs, because it was an order of magnitude more productive. It felt faster, with endless virtual screens, despite the ludicrous MHz/graphics gap. It was actually faster at most real world tasks except those needing raw CPU loads. It was far more reliable ripping or writing CDs etc. You generally plugged any Zorro card in and it just worked - none of the faffing with interrupt jumpers and Plug n Pray/BSOD. I used it over Linux/BSD as the sw wasn't there.

It took until the XP era, and much expense, before the Amiga became pointless. I was delighted when the Macbook Pro arrived... I'd sold the 4000 well before then.

I still miss many aspects of it. Not because I am somehow stuck in nostalgia for the 90s, MHz over GHz or scanlines and flicker fixers, but because I feel it was, in many ways a far better starting point for modern computing - mostly Intuition, but also the approach to everything in the hardware design (pre 4000 anyway). If only they had developed over the years since...

I suspect, had Commodore lasted, that they'd have gone an OSX-like route - new processor (this was already planned as there was no future in 68k series) and reimagining of Intuition on top of nix, maybe BSD. It already had a rather nix flavour, ARexx and so on. Hopefully on something more interesting than just PC clone hardware, though I doubt it these days.

[0] Dave Haynie designed a machine that was another leap forward in 90/91 whilst AA was still being worked on (I think): early spec PDF: http://www.thule.no/haynie/research/a3000p/docs/a3000p.pdf

Medhi Ali wanted to cut cost above all, and make some crap PCs, so the world got the piece of junk A4000 and AGA instead. Oh, and a cost-cutting A600 that cost more to produce than the A500.

I owned an original Amiga way back when. In fact, my was from the batch that had the design team signatures injection molded into the inside cover.

It was a fantastic machine at the time. I enjoyed using it. It was a refreshing new take.

Today? Nah. There is no reason to revive almost anything from that era. The only things I have that survive the test of time are my HP-41 calculators and a rugged-as-f--k all metal shop vacuum cleaner that still works as new thirty years after I bought it.

When we were kids the Amiga was for quite some time the center of our life.

When it was launched there were PC clones with EGA graphics (16 colours), the Atari ST just launched which was a good computer and we were also outgrowing our 8-bit computers (C64, Amstrad CPC, ZX Spectrum). Then, the Amiga was gem-like compared to the alternatives and I remember I was so in love reading colorful Amiga ads over and over: https://s-media-cache-ak0.pinimg.com/originals/ce/8a/49/ce8a...

But to be honest, I don't get why people still stick to it nowadays.

I was going to be short. Still, even if large, I won't price this above two cents.

I am saddened to say, but everything that made the Amiga great when it was launched seems meh by today's standards.

Amigas had preemptive multitasking before any other home computer. Today it's the norm. Amigas could switch video modes every scanline. Today, even the concept of a scanline takes a while to explain to someone born in the 90's. The same happens with the careful NTSC timing ("what's NTSC, grandpa?") that, ultimately, doomed the machine to low-resolution graphics until it was too late. Hardware accelerated graphics (their blitter was cool) is also the norm (and many orders of magnitude faster now). We don't even talk about audio quality anymore.

Everything that defined an Amiga is commonplace. To experience a UI that looks like a cross between Gnome 2 and BeOS running on a processor that once replaced 68Ks is not what would define an Amiga today.

So, what I would expect an Amiga for the 2020's to be?

I'd imagine, for starters, a uniform CPU/GPU architecture. Lots of cores accessing a single memory pool (let them have some scratch memory for themselves to save bandwidth) running the same software doing the job of both CPU and GPU. Make this CPU/GPU hybrid monster capable of combining multiple 4K inputs piped into textures and composite them in real time on any display surface available. A radical idea should set the tone.

Maybe, for the GUI, explore the idea of depth to convey information about windows. A simple sensor could locate your head and adjust the screen perspective accordingly. To the programmer, the GUI should be simple. Reminisce about Plan 9 a bit. If the head detection thing is capable of dealing with hand gestures, have a couple gestures to zoom in/out, switch task/workspace and so on. Make the visual language elegant for the 2020's. Let go of the 80's.

Embrace the idea of "the network is the computer". Again, think Plan 9.

Obviously, make the OS developer friendly, not only to develop for it, but for other platforms as well (whoever uses this computer needs to be able to work). Don't bother with the Amiga CLI - we have better stuff now. Ports of Clang/LLVM, OpenJDK, the usual browsers and, of course, Emacs, are the bare essentials. Make it POSIX so it's easy, but favor languages that express concurrency and multiprocessing easily because you'll have a whole lot of cores.

Obviously it should be able to read most common filesystems, even if its native filesystem is something alien. Snapshots, implicit RAID, part of the computer address space... Something new, please. Maybe don't have a disk and treat RAM as cache for a persistent store. It'd need to be radical or it'd be just another cute computer.

Ditch the mouse. Use a trackpad. Just make it right (as good as Apple's).

The system enclosure can't look like a PC. If at all possible, build it with no cables - edge connectors to drives, power, external ports. If it's pretty enough, you get the chance of using transparency to show it off. Again, 2020's, not 80's.

Use the ball as soon as possible at boot, at the highest resolution the display allows. Use the bounce sound to signal successful POST. 2020's, but still Amiga.

Finally, the keyboard. Whatever you do, don't make it a generic (or worse, cheap) PC keyboard. An Amiga should feel different (even if the original was so-so). From layout to switches to keycaps, it'll have to convey uniqueness.

That ended up longer than expected...

I thought I read the Amiga's also had dedicated hardware for specific functions. That happens on desktops with graphics, sound, and networking cards. However, a modern Amiga might integrate a FPGA (or several) and standard API somewhat like HPC servers are doing. Then, apps can take advantage of it. Might also justify its cost.

Pico Computing already sells desktops that are similar to what I'm thinking about. They're just non-standard & you customize stuff to them where this would be standard with flagship apps (even FOSS) using accelerators by default.

EDIT: Or even heterogenous processors with varying strengths for great power/performance ratio. Similar to what phones are doing.

I'd buy one, just to motivate the producers. Since SGI died the computers are a bit boring. (And SGI was way too expensive but that was some pretty amazing hardware for the time.)

They seemed to stay technologically ahead of their time in various ways. They did NUMA systems with great performance. Did them in desktops, too. Tried stripped OS's for the compute nodes that redirected things like I/O to others. Eliminated most bottlenecks that were in graphics buses. Supported open standards for graphics. The Bill Gates and SGI demo of modifying live video. Dual mobos in the desktops. Threw in RAS, Trusted extensions to Irix, and other things competitors were doing. Let you run Quake, though. FPGA's as a compute node later on with NUMA latency and bandwith. Took forever for competition to do that.

All that said, the cases for the desktops and Origin especially looked cool as hell. There was definitely a wow factor showing off SGI hardware even to laypeople. Then there was one seeing what it could do.

Where's a VC when you need one?! Someone get this plan in motion.

I don't think there is a single VC able to invest enough to see this idea trough. The CPU alone is something Intel attempted and couldn't make perform well as a GPU. The Xeon Phi is (like the SPARC Niagara was before it) very interesting (and I'd love to see a similar part meant for desktops) and a model of what I think future CPUs will be.

That's more or less what I wanted to say (as a former Amiga owner and enthusiast, back in the '90s). These nostalgic operations that keep popping up every few months or years are basically denying the spirit of the original Amiga, which was that of a machine ahead of its time. A completely new machine instead, entirely different and totally incompatible with the old Amiga, based on some radically different architecture from the current desktop computers, would be truer to its original spirit.

I thought Be would be the spiritual successor to the Amiga and attractive to the original Amiga user base. Having said that I didn't buy one because I didn't want to spend time and money on a platform that was obviously not going anywhere. Even a Mac would be a reasonable place for an ex-Amiganaut to end up. I didn't go there either because while I thought the original OS-X was a great idea I think it got more fragmented and further from the lauded Apple design brilliance with every release. Plus Apple stuff is stupid expensive for what you get.

> Hardware accelerated graphics (their blitter was cool) is also the norm (and many orders of magnitude faster now).

Not really. Current GPUs don't have 2D hardware, everything is just done in "software".

> I'd imagine, for starters, a uniform CPU/GPU architecture. Lots of cores accessing a single memory pool (let them have some scratch memory for themselves to save bandwidth) running the same software doing the job of both CPU and GPU. Make this CPU/GPU hybrid monster capable of combining multiple 4K inputs piped into textures and composite them in real time on any display surface available.

Sounds rather like Larrabee [1].

[1] https://en.wikipedia.org/wiki/Larrabee_(microarchitecture)

For the blitter, this re-imagining is doing everything in software anyway. We can add a couple instructions for that.

I was thinking about Xeon Phi (which is the child of Larrabee), but with cores able to perform better than x86's as GPU cores. There seems to be a great deal of ground between CPUs and GPUs and a lot of spots one can be to function reasonably as both.

I just can't swallow that it looks like a generic tower PC. What if they actually went with the A1200/A500 form factor?

Does anyone remember CygnusEd? To this day nothing I use scrolled as nice as that text editor did. Such fluid scrolling. Sigh.

Yeah, it was a very nice experience.

Also, a mouse pointer that was synced to vertical blank together with a mouse that seemed to feed the computer with events fast enough to make it non-laggy and fluid.

LOL. Yes. That's what inspired me to write my own Windows text editor with smooth scrolling: http://deadfrog.co.uk

It's a bit rubbish.

It is written in the Amiga style of every-byte-matters though. If it wasn't for the embedded Python interpreter, the entire install would be about 300k.

> It is written in the Amiga style of every-byte-matters though.

This needs to be in the marketing of your website!

It sounds like you're focusing on programmers, and I think they'll get the nostalgia of "remember the days when every byte mattered?" With the 1TB drive in my laptop I don't actually care about app size, but I love that you have a 1MB text editor in the days of Electron-based hipster minimalist text editors that are over 400MB.

The story in your release notes about resisting adding a hi-res Vista icon because you thought 260KB for an icon was obscene, that story is both hilarious and awesome :)

It looks like this is mostly just a fun project for you, but as a potential user I'd love to see more screenshots, an EXE installer option instead of .zip, and a big download button on the front page so I don't have to dig through the site to find how to download.

But most of all, I think it's hilarious/awesome to see a text editor with a "Frames Per Second" meter! :)

Thanks. Hilarious/awesome is about what I was aiming for. Your comments are very sensible. I'll see what I can do.

I do. CED was simply amazing both feature wise and for its speed and smoothness. Text searching and replacing was really fast too. Later I became a registered user of GoldEd which was more programmer oriented and incredibly rich in features, scriptable etc, still it could not reach CED speed.

>2 GB of RAM for the system (and separate 2 GB for Ram Disk)

Pathetic operating system without dynamic filesystem page caching or SMP. I honestly don't understand why people obsess over Amiga.

Interesting that the economics are there for this computer to get made when the TALOS POWER8 desktop failed on the level of economics.

Reply on news.ycombinator.com