As we know, about 7 millions of amiga where sold in the good old days.
Amiga is now slowly progressing.

What your proposal or idea to increase the current userbase?

I just saw that a software on Aminet (http://aminet.net/mus/play/playOGG.lha) has been downloaded more than 20.000 time. This mean that there are a lot of classic Amiga users around. Maybe something could be done to bring those to next gen hardware. What do you think about?

Have you other ideas to share?

Last edited by AmigaBlitter on 27-Feb-2013 at 11:06 AM.

_________________
retired

@AmigaBlitter

Best way would IF Jens can make ACA500 affordable, beautifull enough to call back old amiga users. A500 is best selling amiga, 95% of those 7 000 000 has good memories of it.

Second best would be low spec/affordable PPC amiga, wich need to have at least 800mhz CPU and price should be under 300€ and it would need to be whole computer, not a do it your self system.

It should be doable, it is proven to be possible to make mobo with those specs less than 50€

800mhz is somekind of limit to fluent net surfing etc.

And it really must be Amiga, there is no way that MorphOS or Aros could gain enough publicity to make more users, exect from those who already are Amiga hobbyists.

Last edited by utri007 on 26-Feb-2013 at 01:25 PM.

@AmigaBlitter

good question

@AmigaBlitter

This has been discussed many times before...

There are some fundamental issues that prevent NG Amigaoids from having a larger user base. Like insufficient software support. Currently, I do not see a way around them.

_________________

SW support is improving all the time.

Affordable new low end HW is what we need most now. IMHO.

IMO, most needed in priority order:
-200eur PPC HW (1ghz, L2, 3D GPU, etc.)
-multicore support and other key OS refinements
-bringing ease of use back to AOS
-no silly shouting etc on forums
-openness
-mid range PPC HW for about 500eur
-even more SW applications
-keeping AOS small+efficient+simple+fast+fun
-innovating new features or at least improving over others
-another super high end 1000+eur HW for performance hungry & hevy duty needs
...

Last edited by KimmoK on 26-Feb-2013 at 01:29 PM.
Last edited by KimmoK on 26-Feb-2013 at 01:28 PM.

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@KimmoK

Quote:

SW support is improving all the time.

Not enough.

Also, the current focus for devs and users is mobile.

_________________

@AmigaBlitter

Quote:

What your proposal or idea to increase the current userbase?

Essentially, first step is funding.

The balance is "chicken and egg", which is easily explained.

The s/w end is not mature enough to attract more people on its own. Ergo, it can not be a means to raising money by selling it. This is one reason why the focus shifted somewhat to h/w, despite the obvious fact that the h/w was also needed to run the operating system.

Therefore, under the current setup, h/w is the only means to build a small profit to use for funding. But...to make that profit, it requires that h/w must be priced over what most would consider fair market prices. That, in turn, reduces sales and ergo profit from both the h/w (our h/w developers and sellers) and the bundled OS sales end (Hyperion).

I realize many people understand the above, but I believe that to come up with a plan, one must first accept the fact that the project can not be funded internally.

So, there's your starting point for a plan. Try not to shout "angel investor" for the answer. Heh.

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@number6

At least 3 good opinion till now:

* Low cost Hardware (low spec)
* Software
* Fundings

For the low cost hardware, maybe a partner could be found. I had a look at servergy low cost developer platform, but reported to have an incompatible FPU with AmigaOS. Maybe the incompatible chip could be swapped with another low cost chip.
The software should coming soon after. There are a lot of software that can be ported on Amiga NextGen.

_________________
retired

release an netbook and a more high end laptop too maby a middle range for a reasonble price too....

make an softwere shop like metrpolis, noo in two days hopefully metropolis will be there. and i just hope the commnunaty will suport it.


some amazing videos of an furturristic os on youtube.

@utri007

Quote:

utri007 wrote: Best way would IF Jens can make ACA500 affordable, beautifull enough to call back old amiga users. A500 is best selling amiga, 95% of those 7 000 000 has good memories of it. Second best would be low spec/affordable PPC amiga, wich need to have at least 800mhz CPU and price should be under 300€ and it would need to be whole computer, not a do it your self system. And it really must be Amiga, there is no way that MorphOS or Aros could gain enough publicity to make more users, exect from those who already are Amiga hobbyists.

I don't believe it helps that much if it's called Amiga, if it doesn't run those games the 7M people remember and do it without UAE.

Next gen systems mainly attract people who did use their Amigas for productivity and especially for internet late 90s. Amount of those people are just a fraction and those know it that well that it doesn't matter if it's any of the three main choises nowadays.

Others just don't have any clue of application usage on Amiga and it's difficult to tell them what makes it good in that when they have certain expectations with current systems. If they buy ACA500, they buy it for gaming. It won't get them interested about next gen systems I'm afraid.

_________________
- The wiki based MorphOS Library - Your starting point for MorphOS
- Software made by jPV^RNO

@number6

I think you're onto something here. If the X1000 had been in a position where cost would have been going DOWN as volume goes up rather than vice versa, it would probably have worked much better for us already.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@retro

Quote:

release an netbook

This would speed up software development. I know certain dude whose job it is doing backup and other similar things, and in the spare time they can code...

_________________
retired

@olegil

I'd like your opinion.

Do you believe s/w (ideas/uses) drives h/w development in general?

If so, and the goal is to provide a more powerful h/w base to execute those plans, then how would a lower power entry point answer to those higher requirements?

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@AmigaBlitter

From Freescale, these are the chips we can use (now):
400MHz e300: MPC5121e/23/25, 35USD. Would make for a working iPad-lookalike, nothing more. Not very fast, even though it does have very fast memory.
667MHz e300: MPC8349 et al, 55USD. Single PCI bus, no onchip graphics. A PCI bridge would be needed to run graphics at 66MHz while things like sound and USB runs at 33. Could be done very cheaply if community would accept not having stuff onboard. A PCI southbridge sounds ideal but it's getting hard to find them.
667-1500MHz e500mc: P2040/41, 100-230USD. Quad core without Altivec, 3xPCIe 2.0,2xSATA 2.0 and 3xGbE should be possible, single layout can cater to wildly different performance classes over large range of prices, maybe three versions from 250 to 350 EUR could be within reach including a LITTLE bit of profit (would need to get volume up for this price, but there are features which could make it easy to manufacture and test).
1200-1500MHz e500mc: P3041, 163-215USD, same as above, just wider PCIe ports (x4,x4,x1 instead of x2,x1,x1 if I remember correctly). Might be possible to squeeze a board in under 400EUR if one was dedicated to cost savings (nothing onboard that doesn't direcly pop out of the chip ).
1600-2000MHz e5500: P5010/20, 250-350USD, same pinout and feature set as P3041 but one or two higher clocked 64bit cores with Altivec.

Anything else is either too far off in the future or wrong feature set for AOS4.

The only one in the list that lacks USB is the 8349, btw. So even without sound/etc it's not a TOTAL disaster to make stripped motherboards.

To start off, I would ignore the e300 parts, and focus on either a P2040 or a P3041 as an entry model. Design with minimum features, to avoid people paying for stuff that don't work (a common enough gripe these days).

The problem is getting people to accept that premise, the less stuff onboard you paid for that doesn't immediately have 100% working drivers, the better. Then the motherboard basically just becomes the CPU board, and the plug-in cards become the system.

Otherwise, one must make a common carrier board with plugin CPU boards, which could also work but would mean a LARGE amount of extra work and cost. So for entry level, no.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@number6

No, in general I don't think that is true. Rather opposite, as having a limited amount of some resource (memory, CPU, GPU etc) has traditionally been a driving force in the Amiga community. It's the brag factor of the demo scene. Having a lot of resources stimulates the masses into sloppy programming (basic etc), which can be ok but doesn't quite count as a driving force.

However, with limited amount of resources for general computing, it's important to push those computing fields that require more resources into specialized domains like hardware assist (again, this was the fun of the original amigas, making the system work well as a whole. we had a h.264 hardware discussion last year, about some dongles). So we need to allow people to scale their GPU to whatever they need, and have support for using the GPU to off-load tasks from the CPU. GPUs are, FLOPS for FLOPS, MUCH cheaper than CPUs. But this is something I think would solve itself, as long as Hans can make the interface for it so people can hack in whatever they wish.

So in essence, we need to attract ONE (1, uno, ichi etc) developer who is interested in GPGPU, and provide this person with a board with limited amount of CPU

The algorithms are out there already, we just need somewhere to execute them

Also, an FPGA eval board with software would be an awesome thing to be able to plug in to an Amiga, a large-ish Artix7 or Spartan6 doesn't need to cost a lot but can give a LOT of cool projects. Unfortunately we can't run the Xilinx ISE (x86 windows/linux only).

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@olegil

Yes, I'm sure we both remember the threads on A/V, h264 in particular, and the GPU being key.

But let's say we were talking 3d, a big part of the Amiga past. Is there anything other than good programming and pure processor power that would give that category a boost? Are there clear areas of applications you can outline where GPU alone would not be enough to bring us up to accepted standards?

I'm asking this because of the following quandry:

One of the questions I still have is whether we are defining what we can be (and working towards becoming that) vs realizing our limitations and tailoring the definition of what we can be TO those limitations.

#6

Last edited by number6 on 26-Feb-2013 at 03:17 PM.

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@AmigaBlitter

What about netpad instead of netbook (I would put USB on it, to use normal keyboard/mice instead of weird docking solutions). Basically a normal board but mounted in the back of an LCD screen with touch. Limited to chips with LCD controller (or a simple FPGA LCD controller can be made on a localbus, it's not exactly rocket science).

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

I think you need to bear something in mind when quoting all these low low prices of possibilities.

When Acube made the Sam440 they were aiming for the best low-end, model they could. Costs were to be kept low so as to make a "budget" OS4 machine.

Acube know what they're doing. If the cheapest they could do an Amiga-compatible motherboard worked out at $600 or whatever, then there's a reason for it. If they could have done one for a hundred dollars, I think it's pretty obvious they would have done.

They're not going to make a $600 board and sell 1,000 with a $200 profit on each, if they could make a $50 board and sell 10,000 with $100 profit on each. They're not stupid.

Out of all of us, Acube are the only ones who actually have experience of what's involved in making a worthwhile OS4-capable platform.

@olegil

http://hdrlab.org.nz/projects/amiga-os-4-projects/radeonhd-driver/radeonhd-development-log/the-radeonhd-rm-resource/

_________________
Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card

@number6

Well, if you DO look at my list of plausible targets you WILL see both low end and high end chips there. So I'm not entirely certain I see the conondrum at all

Basically (afaik), CPU processing power is used for these areas:
-game dynamics, like physics and enemy AI (3D is handled by GPU, sometimes also physics but that's pushing our luck a bit)
-video decoding (can be offloaded to GPU)
-video editing (mixing, subtitling etc, more a question of data throughput than actual processing power as the maths are simple for each pixel, besides, most of this can probably be offloaded to GPU in any case)
-video encoding (can also be offloaded)
-3D rendering (no need for realtime, but extra power helps. Possible to offload)
-foreign CPU emulation, javascript, flash. Problematic, but a GPGPU fix for this is also plausible some time in the future.

So we're stuck with a few areas (enemy AI, physics simulation, emulation, javascript, java and flash) which would be difficult to off-load, the rest are already being done on other platforms.

The point I was answering was what can be done in the low-end segment, and if you go back and study my answer you should see that mostly I did the cost reduction not by taking away CPU but by taking away things like onboard sound/video/whatnots, then letting the user decide which CPU family to spend money on.

No more complaints that people cannot use the onboard this or that, let's just accept that flooding the devs with new chips is not the best way to move forward.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.