@sll

While you are looking for crappy substitutes for the Amiga, Hyperion back to sells 3.1.4.

Virginio

@matthey

I agree with you that SysInfo is flawed and that better benchmarks should be done. However, in the meantime, these are concrete numbers, not hypothetical values based on "ifs". (Besides, using SysInfo for both benchmarks should make them comparable no matter the limitations of SysInfo!)

Last edited by tygre on 21-Nov-2018 at 02:14 PM.

_________________
Tygre
Scientific Progress Goes Boing!

@kolla

Quote:

kolla wrote: @ne_one Quote: If the objective is to offer an inexpensive Amiga retro experience, we're getting within firing range. With a Vampire? Not really no, there is no cheap Vampire standalone in sight. They could of course try to make some cheap alternative, with a cheaper FPGA, but then it is quite likely that the current Apollo Core and SAGA will not fit, or would require serious "stripping" of the core. But then one might just as well do a MiST clone and use TG68020+MinimigAGA.

How can SAGA not fit? Isn't is just a bunch of software routines running on the other cpu thread since the '080 is dual-threaded and the OS can't normally do anything with the other thread?
I tried a discussion about implementing actual 3d functions into the chipset hardware then it came out that the blitter was being emulated on the other cpu thread then things started to make sense on what was taking so long to release a standalone unit: they haven't truly implemented the chipset. It's basically a DRACO...

...as time goes on, I find myself agreeing with you more and more on things...

Last edited by Lou on 21-Nov-2018 at 02:34 PM.

@tygre

Sysinfo does not show the gain with the Vampire MMX-instructions used with the Riva videoplayer e.g.

@bennymee

Can the Riva player work on a "stock" Blizzard 68060?

Edit 1: Ah, I should have RTFM From Aminet: " Two binaries are provided in this archive, a classic m68k compatible build (68040+) and an AMMX build, exclusive to Apollo Core Gold 2.7 and newer."

Edit 2: Will try later!

Last edited by tygre on 21-Nov-2018 at 02:38 PM.
Last edited by tygre on 21-Nov-2018 at 02:37 PM.

_________________
Tygre
Scientific Progress Goes Boing!

@BigD

Quote:

macOS is getting more like iOS by the year and ALL desktop OSes seem to want to be more like phone OSes!

I haven't used macOS for about three years now, so I didn't know that was happening. Gnome 3 on Linux tries to turn a 24-inch monitor into a phone, so I don't use that either. I really don't think UI design has progressed much since Windows 95.

_________________
"Unix is supposed to fix that." -- Jay Miner

@simplex

Quote:

AmigaOS was already modern in a lot of ways back in the 80s. A lot of the ways it's not modern now can probably be done without making it feel like macOS or Linux. That would also break a lot of compatibility.

It depends on what is being updated. Adding SMP would be a win; I don't think anyone would complain about that.

Adding memory protection, which it desperately needs, would break compatibility, so that's a problem either way.

And there are some places it just shouldn't go: I think systemd on Linux is a good example of something that Amiga is better off without.

_________________
"Unix is supposed to fix that." -- Jay Miner

Quote:

kolla wrote: The biggest obstacle is lack of modern development tools for Amiga, and the fact that the OS is seriously lacking in just about everything that developers expect from an operating system today. It does not at all help that all the "important" codebase also is closed behind people who are shit scared of losing any kind of perceived ownership and control over code. You cannot really expect to recruit developers in such an environment. The best you can hope for, are embedded people and cross-development tools. But they are rarely interested the type of programs that Amiga users crave for.

The Amiga could have all the modern development tools but a user base of only a few thousand isn't going to attract major developers. It's all about installed user base size and the way to achieve this is with hardware. An OS with modern features and source code available can be given away but is highly unlikely to significantly expand the user base. Most users buy hardware because it is cool but won't replace their current OS. A-Eon realized hardware is necessary for AmigaOS 4 survival but it has targeted an expensive specialized niche market (instead of a mass produced market) so far which has failed to significantly expand the user base. Of course part of the problem has been that AmigaOS 4 PPC aims for a desktop market where the AmigaOS is not competitive instead of 68k retro, hobbyist and embedded markets where the AmigaOS 68k is more competitive. The Raspberry Pi is an example of hardware like what the Amiga should be creating but with compatibility to run existing 68k software (especially games). The Pi has some boring embedded software but it has healthy software development unlike the Amiga.

Quote:

Porting to Amiga is more and more becoming a nightmare, as the world is moving on, while Amiga "ecosystem" is stuck in time. If you want to port some "modern" (less than 15-20 year old) networking code, you also very often end up having to "back-port" large chunks of a TCP stack, or "re-implement like it was 1995". Just think about it - the situation is that _ALL_ IP-stack development for AmigaOS depends entirely on ONE SINGLE person! And the situation is similar for other parts of the OS.

The Amiga is different (not Linux/BSD) which makes porting more difficult. The Amiga way has advantages but the user base has to be large enough to support it. From a marketing standpoint, a unique product is better able to differentiate itself than similar products in a saturated market and can support higher profit margins.

Quote:

tygre wrote: I agree with you that SysInfo is flawed and that better benchmarks should be done. However, in the meantime, these are concrete numbers, not hypothetical values based on "ifs". (Besides, using SysInfo for both benchmarks should make them comparable no matter the limitations of SysInfo!)

The results of SysInfo are *not* comparable if done on different CPUs. It is better to look at clock frequency differences if using the same CPU. The SysInfo benchmark is that bad!

Better Amiga benchmarks are AIBB (may need a patch or specific tooltypes to keep from crashing on startup with 68060 and Apollo Core), SysSpeed (flawed tests also but much better than SysInfo) and Gunnar's minibench (CLI only and measures specific CPU results).

@bison

Quote:

It depends on what is being updated. Adding SMP would be a win; I don't think anyone would complain about that. Adding memory protection, which it desperately needs, would break compatibility, so that's a problem either way.

SMP and memory protection are precisely two things I had in mind. Routine trashing of the entire system during development is simply not acceptable on a desktop system for the vast majority of developers, especially when C essentially requires you to use pointers.

I'm not actually familiar with systemd, though I know it's something on my Linux systems. I'm wondering if that's one of those things that Linux users also hate; I remember reading about something similar.

_________________
I've decided to follow an awful lot of people I respect and leave AmigaWorld. If for some reason you want to talk to me, it shouldn't take much effort to find me.

@bison

Quote:

Gnome 3 on Linux tries to turn a 24-inch monitor into a phone, so I don't use that either.

KDE FTW!

Quote:

I really don't think UI design has progressed much since Windows 95.

macOS about 10 years ago was actually quite good, far better than W95. Then again, it was basically a prettified NeXT.

_________________
I've decided to follow an awful lot of people I respect and leave AmigaWorld. If for some reason you want to talk to me, it shouldn't take much effort to find me.

@matthey

Thanks for the pointers! I will try these benchmarking programs! I could not find minibench, though, because the link on on this page is broken. Would you know where I could find it?

Edit 1: minibench is available only for Linux X86 and PPC here...

For my own education , could you explain me (maybe in another thread/by PM) why is SysInfo (and SysSpeed) so broken?

Cheers!

Last edited by tygre on 21-Nov-2018 at 10:40 PM.

_________________
Tygre
Scientific Progress Goes Boing!

@matthey

Quote:

The Amiga could have all the modern development tools but a user base of only a few thousand isn't going to attract major developers.

Amiga could have millions of users, and developers would still not be bothered because the OS simply suck for any kind of modern use. Number of owners do not attract developers, number of *paying* users attract developers, and developer tools attract developers. Amiga is running short on both. Ironically, it seems that most Amiga hardware is in the hands of horders, like Trevor.

Quote:

It's all about installed user base size and the way to achieve this is with hardware.

And exactly what can one achieve with Amiga?? Not so much, if anything at all. Unless you are willing to spend considerable amount of time hacking on your own, something that apparently not so many users are willing to do. They want finished solutions, plug and play. With any hardware locked to AmigaOS, it is very limited what users can achieve, and one can achieve much more and cheaper using any sort of hardware that runs for example Linux. Isn't it ironic that most of the development and progress for AEON hardware happens not on AmigaOS, but on Linux? Why is that, do you think?

Last edited by kolla on 22-Nov-2018 at 08:46 AM.
Last edited by kolla on 22-Nov-2018 at 08:43 AM.
Last edited by kolla on 22-Nov-2018 at 08:43 AM.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

Quote:

And exactly what can one achieve with Amiga?? Not so much, if anything at all.

Did you not read this thread? The Amiga still has its uses.

https://amigaworld.net/modules/newbb/viewtopic.php?topic_id=42975&forum=27

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@BigD

Yes, I did... wasn't it about MorphOS?

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

It was about ImageFX and PageStream which are available for use on multiple Amiga-like OSes.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@kolla

Quote:

there is no cheap Vampire standalone in sight

We're discussing this in hypothetical terms - it's eminently doable.

No one is suggesting that you could take the existing Vampire and package it up as a marketable solution.

@bison

Quote:

A thoroughly modern update of AmigaOS would be a lot more like macOS or Linux than it is now, and a lot of people wouldn't like that at all.

That's silly.

No one ever wanted the Amiga to be stuck in 1993 forever and anyone fixated on that experience already has options.

We've hit the point where preserving the legacy is a given. It's time to start forward moving forward.

@ne_one

Quote:

ne_one wrote: @kolla Quote: there is no cheap Vampire standalone in sight We're discussing this in hypothetical terms - it's eminently doable.

I doupt very much that it is doable to make a cheap stand-alone with Apollo Core, unless you move out of FPGA and onto ASIC, but that requires quite an economic effort and guaranteed sales. I seriously doubt that there is a large enough market for such a product.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

Looks like it could be back again

https://www.amikit.amiga.sk/os31update?fbclid=IwAR3_zVEUeUPcp7P0JypFE8Qfb0a4b4r-7HGnik6WIqn-fy8wLg4oGyjU5vo

_________________
A1200 ACA1232 128MB Indivison MkIICr
Elbox empty Power Tower
RPi3 Emulating C64 ZX Atari PS BBC
Wii with Amiga emulation
Vampire v4 SA