OK...this is absolutely going to get so negative feedback, particularly considering my oft stated opinion of Tabor (which is a bit of a 'turd sandwich', and not apologizing for the 'T' word).

The X1000 was introduced when Ben Herman's was part of Aeon, but he isn't now. Nor would he stand to gain anything if our team was to extend support the the X1000.

AND, its still a damned nice system, with better PCI-E expansion capability than either the X5000 OR the PCI-E G5.

Then there's Tabor...yeah...its definitely got a poorly chosen cpu with e500v2 core and its non-standard FPU, but Aeon has 1000 of these already made and ready to sell (and plans to build thousands more), with a selling price of about $500.

Between to two of these, that's almost two thousand Amiga platforms that could run MorphOS. With more A1222 depending on the success of that system.

Look...I've already told Mark and Frank that I don't like Tabor, and that we probably shouldn't support it, but you know...I could be wrong (hey, it happens ).
Its much faster than the Efika, and even if it had NO FPU, those instructions could be trapped and emulated.
As it is, it has a crappy FPU, but with trapping and FPU emulation, the A1222 would still be light years faster than the 400MHz cpu in the Efika.

Again, we'll have the argument about the number of systems being supported.
We COULD just stop supporting the oldest, slowest hardware.
After all, legacy systems with PPC accelerators are no longer supported, and the fastest of those was probably on par with the Efika (366MHz 604e vs 400 MHz Freescale MPC5200B?).

Rethinking all my prior negative posts, supporting ALL Aeon platform makes sense to me.
People like AmigaDave would get to use a better OS (one that I know at least David would prefer).
And Tabor would offer a low cost NEW entry platform. Yeah, not ideal, but serviceable.
Sure, a G4 PowerMac would make more sense, except it won't take PCI-E video cards, it isn't available new, and its an Apple (I for one, am not very fond of Apple, and I'm sure many other aren't as well).

So, while I await X5000 support for 3.10, I thought I'd just throw this idea out.
Some of you may think of it as just another brain fart on my part, but I think its worth consideration.

My opinion is to support 1222 just because of it the lower cost entry point into NG platforms and they plan to make a lot of them once 4.1FE is available publicly for it. I dont like the biard much, but I do see a place for it. It would be silly not to support it just because of spite.

Tabor is still way more expensive than buying a few ppc Apple computers off ebay.
If support was costing close to no development resources - yeah , bring it on then.
But actually I don't see much of a point in supporting more and more hardware instead of expanding he OS feature wiseor improving applications. Hardware is not our major limitation these days, but useful features and applications. And I actually doubt that many OS4 nuts will start using MorphOS on a regular basis (i.e. bcome paying users or new application developers or so).


IMHO no priority for support of new ppc hardware, but roll out 3.10 ASAP and then focus on MorphOS NG on x64

Quote:

I understand the drive toward X64, believe me. I have an i7 laptop, an AMD AM3+ motherboard, and I'm considering an upgrade to an AM4 Ryzen system. X64 is powerful, available, and cheap.

But we are getting an X5000 port, and I could see X1000 and A1222 users adopting MorphOS PPC.
Its a better operating system than OS4.

And many of us want the PCI-E G5 port released for general use (I'll buy a damned network card, honestly, that's not a problem).

Further, I anticipate continuing to use MorphOS PPC even after an X64 shift, if only for nostalgia value (after all, I still use OS3.1 on an A2000, and that is WAY too uncompetitive).

As to OS4 'nuts' that might like to use MorphOS, ask AmigaDave if he'd like the option on his X1000.
And do you think all those inquiries about MorphOS on the X5000 are just from our community, because I know they're not.
Finally, would the availability of this support lessen our user base? Of course not. We will have an option for both X5000 models, so why not Tabor as well?
We could offer the AmigaOne community an OS worth using.
So, again, why not cover EVERYTHING from X1000 to Tabor?

[ Edited by Jim 23.09.2017 - 16:47 ]

Quote:
He has a MorphOS system already. Of course less boxes are a bonus (IMHO at least), but well, he knew that MorphOS will not support X1000 before purchase. And he wouldn't be an additional user, but only an additional registration.

Quote:
Very simple anser: Because development resources are very limited and I _very much_ doubt spending these limited resources on supporting some obscure expensive (Tabor) or überexpensive EOL (X1000) hardware is actually somewhee in the reagion of the top 10 of all sane choices.
If adding support costs one weekend, e few pizzas and a beer (or something the like), go on. But I very much and seriously doubt. Hence, I don't support this idea.

Well, I'm not sure the PCI-E G5 would be more than a few pizzas, little beer, and a couple of weekends work (I've exchanged e-mail with Mark when he's working on something...and frankly, he's freakin amazing).

The X1000 is much more problematic as its based on old PA Semi stuff, so it might not be practical.

Tabor, our developers either already have these or they could obtain them (probably without paying for them).
Trevor's been pretty generous about that, and if not Aeon, I know some vendors that would consider it.

Think about this objectively, in October 1000 new AmigaOne A1222 boards will be released to the community. And when those are sold, more will be produced. We already know there are X1000 owners that would have liked to have MorphOS as an option.
So, if just 10% of Tabor buyers adopt MorphOS, that's 100 new users. Charge those users a $100 license fee, that $10,000 or more than twice what Bigfoot was paid for the G5 port.
And that's just the customers for the first batch of Tabor boards.

I think they'll sell more at a cost that's less than a third of what the X5000 costs.

Say they sell 5000, that's not unreasonable considering how many Vampire boards have been sold (or considering how many more will be sold once the Vampire4 is released).
At again a 10% MorphOS adoption rate and a $100 license fee, that is $50,000.

Hey, I have about 200 college credits, 15 of those are in Economics alone. I have a BS in Business Management and I'm working on an MBA. I've been in retail management, and when I consider business opportunities, I always think about economic considerations, not what I want to do.

And this looks like an opportunity.

The X5000 will only bring us a few licensees.
The PCI-E G5 might bring us a few more.
But there's going to be a shit load of Tabor owners.

You don't need five semesters of Economics to work this one out, its an real chance to sell some licenses.

[ Edited by Jim 23.09.2017 - 18:12 ]

i'd love to buy some newly produced fresh hardware with warranty. There's only one argument speaking for the old apple products: mobility.
Theres no other way to get Amiga NG on a mobile device like an ibook or powerbook, except for AROS probably, but im not interested in that and most of the users neither.
I dont know what the price will be of a complete Tabor system, but the specs are nice and with a dual cpu (in the sense of a co processor) support this little thing should be fast enough to handle every day stuff like browsing, email, video, etc.

> the X1000 [has] [...] better PCI-E expansion capability than [...] the PCI-E G5.

Really? Powermac11,2 has one x16, one x8 and two x4 while the X1000 has one x16 (or two x8) and two x1.

> with trapping and FPU emulation, the A1222 would still be light years faster
> than the 400MHz cpu in the Efika.

...except for standard FPU code ;-)

> the fastest of those was probably on par with the Efika
> (366MHz 604e vs 400 MHz Freescale MPC5200B?).

...whereas the fastest stock PowerUP card as delivered by Phase 5 had only a 233 MHz 604e. Faster chips were put in place by 3rd parties on individual basis only years later.

>Really? Powermac11,2 has one x16, one x8 and two x4 while the X1000 has one x16 (or two x8) and two x1.

Good point, I'd forgotten about the X8 slot in the G5, God only know why, since its where I have my secondary video card (its useful to keep an Nvidia 6600 in that slot while running a Radeon HD in the primary slot so that you still have access to firmware).

Point one to the G5, a few more PCI-E lanes than the X1000 (and a shitload more than the X5000).

>> with trapping and FPU emulation, the A1222 would still be light years faster
>> than the 400MHz cpu in the Efika.

>...except for standard FPU code

Do you really think emulated floating point code is going to slow a 1200MHz computer down below the performance of a 400MHz computer? I suppose it would depend on the balance of the instructions in the programs they were running, but in most cases I doubt the Efika would be anywhere near the A1222.

>> the fastest of those [603/4e] was probably on par with the Efika
>> (366MHz 604e vs 400 MHz Freescale MPC5200B).

>...whereas the fastest stock PowerUP card as delivered by Phase 5 had only a 233 MHz 604e. Faster chips were put in place by 3rd parties on individual basis only years later.

Yes, and we both know the curmudgeon that was foolish enough to sell his upgraded 366MHz 604e powered Amiga4000 for the funds to buy an X5000.
Which do you think would be faster, an Efika or that 366MHz 604e power A4000? My guess is that with a PCI RTG card it would be the A4000, since a 604e is going to be faster than the 603e related processor in the Efika.

Anyway, sure, the X5000 would perform better than the A4000, but a 366MHz A4000? Why would you ever sell that? Running OS3.1-3.9 it would actually be fast enough to compete with our slower offerings.

Finally, can you offer anything to counter my economic argument?
I think its the soundest part of what I've presented here.
A thousand initial new potential users, with possibly thousands more in the years to come.



[ Edited by Jim 23.09.2017 - 18:53 ]

> many of us want the PCI-E G5 port released for general use (I'll buy a damned
> network card, honestly, that's not a problem).

...and you'll "alter the water cooling system of a G5 with a hardware monitor to control fan speed" :-)

http://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=12150&forum=18&start=4

> all those inquiries about MorphOS on the X5000

How many of the 500 potential X5000/20 owners are these? :-)

Quote:

Are you lumping me into the group of "OS4 'nuts' that might like to use MorphOS"???

I can and do use MorphOS, and on a system much faster than my X1000, as I own a dual 2.7GHz G5 PowerMac (which you know about).

As much as I appreciate your passion for MorphOS, and appreciate a bit less your passion for PPC, I think your logic on this argument is all fouled up, and I have to disagree. I would love to be able to run MorphOS on my X1000, but not if it took a single hour away from our transition to x64 hardware and an improved MorphOS with memory protection and true SMP.

As for the Tabor, aka A1222, and the 1000 motherboards that A-Eon has sitting in storage (except for the 20+/- beta test boards currently in use), while I like to support Trevor and Matthew in their efforts to improve our community, I also have to disagree with your assessment of how well the A1222 will sell, and specially with your comparison of the Tabor to the Vampire products. They are apples and oranges and can't be compared, plus, most of the remaining "Classic" Amiga users are NOT interested in any NG options, and won't be buying an A1222, no matter what it's final selling price ends up being. The size of the NG community is tiny compared to the 68k Amiga community. This is something I did not realize until about a couple of years after becoming active again, and buying my first MorphOS system. I think the only chance A-Eon has of converting more of those 68k Amiga users into OS4 users, is to continue to refine and improve the "Run-In-UAE" functionality, so that running 68k Amiga software becomes fool proof and transparent to the user, and they actually get a more compatible system than any of the real original Amiga computers, or any of the FPGA alternatives. Even then, there will be a large part of the existing 68k Amiga community, who will never convert to any of the NG options.

So, let's see how many of those 1,000 Tabor motherboards sell, and how quickly. Do you think that when they are finally released we will see sales numbers even 1/10th as fast as the sales of the Vampire v2? Personally, I don't think so, but for Trevor's and Matthew's sake, I hope enough of them sell to recoup most of their investment.

> in October 1000 new AmigaOne A1222 boards will be released to the community.

Hasn't this already been postponed to early 2018? They must be 3 years old by then.

> Charge those users a $100 license fee

I wouldn't find it fair to charge them more than Sam460 user (49 EUR).

> with a dual cpu (in the sense of a co processor) support this little thing should be
> fast enough to handle every day stuff like browsing, [...] video

AFAIK, using a second core or CPU as co-processor has not been announced for OS4 or MorphOS. And don't underestimate the workloads of modern browsing without JavaScript JIT and modern video without GPU-assisted decoding.

No David, you know I don't lump you in with the 'nuts'.
You're here aren't you?

So tell me, wouldn't you love to have MorphOS for your X1000?
When you compare the X1000 to the X5000, its actually got some advantages (like a lot more PCI-E lanes and much higher video card bandwidth).

Then there's the G5, which we both adopted pretty early, and I KNOW you supported (because shipping one all the way to Mark IS a little 'nutty', but I LOVE you for doing that, you're sentiments are right in line with mine - I have a limited income, but I'm going to do what I can to help this succeed).

So tell me, if the PCI-E G5 was supported, wouldn't you go out and find one? I already have, the AGP G5 sold me on that idea, and the 11,2 is all the earlier G5s were, but BETTER.
Imagine if you could install a Radeon HD 6000 series card in your current G5, because with an 11,2 you CAN.
You can also install SATA3 controllers because the PCI-E slots have enough bandwidth to support this (PCI is too slow).

And, um, ugh, Tabor...yeah, I find it a hard sell too, but I think the initial 1000 should be pretty easy to sell at $500. Amigans can be weird about pricing. After all, how many A500 and A600s were sold after the A1200 was introduced?

Personally, if I'm going cheap, I still look for a good mid-point between price and performance.
As an example, I kind of like AMD's RX560 video card, cheap but powerful enough.
Same thing for most of the Ryzen cpu line, cheap but kickass.

So, why not support the 11,2 G5? Really cheap. Faster than any X5000.

And I'll probably buy an X5000/40 anyway, but I'm not really interested in Tabor.
And I don't think most of our existing user base is interested in Tabor.

I just know that they've got 1000 of these ready to sell, and they're not going to hold on to them.
If the don't sell at $500, count on discounts.
After all, its based on a processor that cost about $35.

An the Amiga fanatic ARE going to buy it, because they don't understand why people like us would prefer to spend a extra $1200 and get a system that performs adequately.

Anyway, I don't want to offend any of you, I'm just going on gut instinct here. And the salesman in me is telling me we are missing an opportunity.

> Point one to the G5, a few more PCI-E lanes than the X1000

Yes, 32 vs. 18.

> Do you really think emulated floating point code is going to slow a 1200MHz computer
> down below the performance of a 400MHz computer?

We know that floating point performance of code compiled natively for the SPE of the 1.2 GHz e500v2 is about that of a 600 MHz G3. Depending on the efficiency of the FPU emulation, I think it could end up undercutting a 400 MHz G2. But we'll see, I guess.

> I suppose it would depend on the balance of the instructions in the programs they
> were running

Yes, absolutely. My rather pessimistic estimation is about pure floating point code, of course.

> Which do you think would be faster, an Efika or that 366MHz 604e power A4000?

http://morph.zone/modules/newbb_plus/viewtopic.php?forum=11&topic_id=11137&start=744

> a 604e is going to be faster than the 603e related processor in the Efika.

Indeed, but the Efika is running at 9% higher clock and isn't crippled by a slow memory interface (which the PPC G2 CPUs are on the PowerUP boards in order to accommodate to the m68k CPU's needs).

> but a 366MHz A4000? Why would you ever sell that?

Well, I sold my 200 MHz A4000 when I found myself not using it anymore after having bought a 600 MHz Pegasos I. I never regretted it :-)

> Running OS3.1-3.9 it would actually be fast enough to compete with our slower offerings.

I doubt it. Being an m68k-based OS, OS3.x uses the PPC CPU only as kind of co-processor through specially written software. Beside the crippled memory interface, there's a rather big speed penalty caused by context switches (cache flushes) whenever a PPC program goes through the m68k OS API.

>I doubt it. Being an m68k-based OS, OS3.x uses the PPC CPU only as kind of co-processor through specially written software. Beside the crippled memory interface, there's a rather big speed penalty caused by context switches (cache flushes) whenever a PPC program goes through the m68k OS API.

Good point, you'd need a native PPC OS to get any performance out of that.
Under something like OS4 that A4000 might be interesting, especially with legacy code that hits real hardware locations instead of going through drivers.
But that would make comparison with the Efika difficult as that would be running MorphOS, which doesn't support Amiga native graphics or sound.

In either case, I'd miss my PowerMacs.

And that would be an issue with Tabor as well.
Outside of PCI-E video card support, it hasn't got much to offer.

At 1200MHz, a top end G4 PowerMac will dust it.
If Mark chooses to add AGP support to the R600 and R700 driver, even our older hardware will compare quite nicely to Aeon's lineup.

This dissent about the 11,2 baffles me. 32 PCI-E lanes?
Why aren't we on that like white on rice?

> When you compare the X1000 to the X5000, its actually got some advantages
> (like a lot more PCI-E lanes and much higher video card bandwidth).

...and much better memory bandwidth, and AltiVec SIMD :-) And the PA6T's FPU might be stronger than the e5500's, at least per clock. The X5000 excels in integer performance. What else?

> how many A500 and A600s were sold after the A1200 was introduced?

At least the A500 (including ECS-based "Plus" model) was discontinued half a year before A1200 introduction.

> people like us would prefer to spend a extra $1200 and get a system that
> performs adequately.

...or spend next to nothing for a used PPC Mac ;-)

>...or spend next to nothing for a used PPC Mac

Well...my G5s have both cost me $30 and $50 respectively, and I did have to rebuild the cooling system ($5 for four o-rings and you only need two)...so its not nothing , but it IS the best value I've ever had in computer hardware.

And I've got this 11,2 sitting here humming away running Linux, which it does almost as well as my X64 hardware, which makes the arguments against porting MorphOS to these systems doubly frustrating because its 'low hanging fruit' and it would leap frog us over A-eon's offerings in performance and price (the highest price I've seen for an 11,2, which was completely refurbished, was $300, so we are talking about half the price of a Tabor board alone).



[ Edited by Jim 23.09.2017 - 22:18 ]

I support Jim's views about the prospects of A1222/Tabor. Especially after I have seen an EFIKA in action (at GamesCon 2017). It's just so *sloooowww*... I mean not MorphOS itself, it runs pretty well and seems well optimized! But as soon you start any kind of application or game which is at least a little demanding in terms of CPU or RAM it crawls. No point in running SDL games, no point in running Battle Squadron, no point in running OWB... IMHO the EFIKA support could be dropped (if this makes sense for the greater good).

On the other side the SAM460 did well at running games & applications with ease. The only downside is, that it's not for sale any longer... IMHO the Tabor will do very well as SAM/EFIKA replacement, at least that's what people report using it for OS4/Linux. And it's cheaper too.

The PPC-Apple systems are best at mobility as Cego said. I use my PowerBook G4 much more often than my G5. What I dislike about the PPC-Apples is the "OpenFirmware graphic cards mess". It REALLY is annoying to have to rely on an old OF-capable card if you want to see any boot messages! Maybe this is less of a hassle when you don't use your machine in a Linux dual-boot environment, but if you do it certainly is. Even if you got through the yaboot-bootloader mess it is. I can hardly imagine that getting a proper GPU and using a dual or triple boot environment is that laborious on a SAM/X1000/x5000.

> as soon you start any kind of application or game which is at least a little
> demanding in terms of CPU or RAM it crawls.

Crawling because of RAM sounds weird. If the amount of RAM is not enough the program shouldn't start at all or just crash, but not crawl. And RAM speed of Efika is not that bad.

> Tabor will do very well as SAM/EFIKA replacement [...]. And it's cheaper too.

I'm sure it won't beat the Efika 5200B in price. And we will see if it's cheaper than the Sam440.

Quote:

In July, 2017 interview Trevor said that he was "hopeful" that there would be an announcement
by the end of 2017. If I can find anything more recent from the source, I'll post it.

Source

#6

Quote:

Who cares about efika? It's eol and not wide spread. It has been a fun Board in its time with a pretty good price, but today it's relevance is close to zero.

Quote:
Which was my point too.

[ Editiert durch ernsteiswuerfel 24.09.2017 - 20:19 ]

Quote:

In my entire time here, I've only been in contact with one Efika user, who has since moved on the a better system.
At $100, the Efika was a pretty good bargain, but it didn't offer enough memory.
These day its possible to find G5s for $100 (even the occasional 11,2 mine was CHEAPER than that).

Btw - With the 11,2, you install an Apple video card that is compatible with open firmware in the X8 slot, and your better spec PC card in the X16 slot.
If you really want to get tricky, you set up Linux to use only the higher end card, which you then connect to the DVI or HDMI connector of your monitor, then you connect the Apple card to your VGA (or other available connector).

Viola, open firmware prompts until Linux initializes the PC card. Its a best of both worlds option.
So, don't like not having access to open firmware if your using a PC video card?
There are work-arounds.

Under MorphOS (were it to be ported to the 11,2), the best Apple card would be something like the Nvidia 6600 (its cheap, and MorphOS won't touch it), then the primary card could be something like the Radeon HD 6750 I've previously mentioned. Again, firmware prompts and a Radeon HD display.

As to Tabor, its NOT my 'cup of tea', but there's going to be a lot of them floating around, maybe we can lure a few users away from 'the dark side'.

[ Edited by Jim 24.09.2017 - 14:54 ]

Quote:

i actually do understand your reasoning, but I guess it's easier to get more users by progressing the OS and its applications. MorphOS on ppc and 32 bit and so on is a dead end. I think we only have the option for hop or top. Let's do the inevitable step ASAP and offer a product that has significant added avlue compared to the competion (primarily, but not only, the other Amigaish systems).
A MorphOS with SMP, 64 bit, 100% resource tracking, maybe MP and a few old things (AHI or so) updated, but still a kept MorphOS/Amiga identity. I really would love if this would go seamless, but it probably will not work. Hence, do the cut now. Yes, it's hurting, it's a sh*tload of work, but postponing the cut will not help - quite the contrary. And hence, don't waste develoipment time on supportimg more ppc hardware (G5 PCIE may be an exception since the port is progressed already and it's the final ppc deskop computer), but go for real progress.