@Boot_WB

So... for example with T1022 one could use one SATA2 port and the second SATA2/PCIe#4 port perhaps could be used as miniPCIe ... especially if there exist miniPCIe SATA card with a chip that we already have drivers for. Also.... one could bridge 3xPCI out of one of the PCIe, then put same PCI audio, sata etc stuff that exist on SAM440 and is well supported.

And if the DIU of T1022 is proven useless somehow, see if a chip without DIU is cheaper.

But adding normal PCIe and PCI slots the board grows large and the price might start to rise up towards SAM460 prices and above... And even if the price would not rise, larger boards are more dificult to use in some purposes.

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

We need a more standard API between Amiga flavors with PowerPC and 68k. Hyperion messed up by dropping 68k support instead of creating a unifying API between PPC and 68k. Even a cut down AmigaOS 4.x for 68k would have made their APIs more common and development easier. It's looking more and more like PowerPC will suffer the same fate as 68k. Will PowerPC then become not interesting to the Hyperion brothers anymore also? AROS could unify PowerPC and 68k but there is a lot of work needed. It will be interesting to see what happens with AmigaOS 4 and MOS. SMT would probably make users happy for a few years.

AGA has kludges but it could be fixed up and used for the top layer of 2D gfx. A chunky 2D lower layer and 3D could be added. The slowness of AGA was due to C= not upgrading the speed. Boxer did finally make some improvements after it was irrelevant but it was never released. AGA is not slow anymore in more modern logic chips and memory. We can see that already with the fpgaArcade. Several of the early 3D gfx chips were added to existing 2D chip sets so that probably is possible. PowerVR would probably be more realistic for 3D than anything ATI. It would be faster in some ways on a SoC than through PCIe. Users wanting more performance may still want a gfx card in a PCIe slot.

Quote:

billt wrote: In what way is e6500 unsuitable for desktop type uses?

There are 2 problems with current embedded PPC processors for desktop use.

1) Many embedded designs are made to execute particular software from caches at high speed. Instructions and units not needed for this purpose may be removed to save cost. These processors are more powerful variants of the old DSP. They are more flexible than the old DSP and can execute general purpose code better but bottlenecks need to be avoided. Different embedded processors will likely have different weaknesses making it a pain to create one standard OS code for all of them. It wouldn't be so bad if one line of embedded Processors was chosen and didn't disappear but that leads us to problem 2.

2) PPC is under attack and it's losing. General purpose x86_64 processors are more powerful and are using less electricity as the die sizes shrink. ARM Holdings has a lot of influence in the embedded markets and ARM64 looks poised to gain major market share. PowerPC never lived up to it's promises. There isn't that much difference between ARM64 and PowerPC 64 bit but ARM has been more successful in the market. ARM is automatically associated with low power consumption even though the new ARM64 processors may not be but if ARM64 has problems then they will come out with Thumb64. PowerPC will be around in the embedded space for a few more years but if Freescale starts manufacturing ARM64 then PowerPC is toast.

@matthey

Quote:

matthey wrote: Hyperion messed up by dropping 68k support instead of creating a unifying API between PPC and 68k.

What 68k support?
Quote:

The slowness of AGA was due to C= not upgrading the speed.

Quote:

It would be faster in some ways on a SoC than through PCIe. Users wanting more performance may still want a gfx card in a PCIe slot.

Quote:

2) PPC is under attack and it's losing. General purpose x86_64 processors are more powerful...

Is that your opinion? Or just your rhetoric? More powerful? What ?

_________________
Tinkering with computers.

@Boot_WB

(billt struggled with several levels of nested quote tags before settlign on this)

billt wrote:
Quote:

The biggest missing bit is audio, so let's do a Nemo and stick a southbridge on one of them PCIe ports.

Quote:

Quote: olegil wrote: @KimmoK [...] Seriously, you have an amazing ability to drag in stuff to overengineer ANY idea. billt wrote: KimmoK, he tells me the same thing all the time. :)

Then Boot_WB wrote in reply to above:
Quote:

How could that be? To keep the cost down, why not just keep the board as basic as possible - maybe taking advantage of on-SoC peripherals where that makes sense (see revisions to Nemo ethernet for how such simple things can add cost and time...), but adding as much as possible via PCI/e expansions. When you're the size of Sony or Apple you can start playing around with the more complex on-chip peripherals. but when you're NRE costs are passed on to each customer in >0.1% chunks, and board revisions cost >$50,000 a pop, it just doesn't make sense for a few watts of power saving, or a few MB/s saved on the PCI bus (imho)*. * Especially when you consider that drivers for such expansion cards already exist, whereas drivers for on-chip peripherals would have to be written from scratch.

Understand that what I'm interested in designing, I do not expect to happen several times. Nor do I see later expandability as a reasonable answer to certain things like audio. (ie. I want a laptop)

As my ideal may be a once, if at all, design, then I want it to take me as far into the future as possible. That means as much performance and as many current features as possible. My ideal feature set has been stable for years. And I do not think that a southbridge is an unacceptably complicated addition to a PC computer. It's rather standard way of doing things, not an oddball method of gaining features. In fact, the southbridge of choice would be a quite standard AMD PC part, as A-Eon used on the Nemo. If this is unusual, then why does everyone have more than half a PC? (Now that CPU and northbridge have been combined, much like in our PPC SoCs) Also realize that of all the things I've actually worked on toward my own goal, I am actually today more prepared to do things with a southbridge chip than I am with a CPU chip...

Also understand that I would expect that the sb600 drivers would be relatively easy to port onto other AMD southbridge/FCH parts. Certainly easier than if I would do the same with an intel PCH equivalent. So I do not expect this would be a gigantic problem for drivers. Then from there we can move to additional interesting drivers for the sum total.

If you're afraid of new IPs and new drivers, then we should never make anything new at all. Stick with the old, and never make any modifications whatsoever to that. I am not afraid of this driver situation. I think it sounds like fun, and educational. Besides, being afraid of not already having drivers for on-chip SoC peripherals is actually an argument in favor of southbridge peripherals that have been seen already on Nemo...

If I were actually working on any such thing (I wish I had time to do so, but this is a rarity for me), then I would get a dev system from Freescale, and work on a standard PCIe style plugin board with my southbridge choice. The way I think of things, this plugin southbridge should be considered as more than just a development thing for myself and maybe a couple other people, I would approach desining it as a standard product, to be used in Sam460, Cyrus, etc. as well, and then perhaps in A1XE, Peg2, etc. boards via a PCI-PCIe bridge. But for my own end product goal, I wouldn't design a laptop with such pieces missing, as I don't think it makes sense to add so many of them later as addons.

Finally, considering my ideal end use in a laptop form, I would prefer to use an MXM3 slot for graphics. No GFX chip soldered to the motherboard. If there's something built-in to the CPU, like in the 8610 and some others, then maybe provide amux between that and a MXM board, which may or may not be doable within a current AMD or Intel FCH/PCH southbridge art. I think they can do that for a VGA port, but am not sure about the same to an LCD panel. But I do not expect that I would be working with a CPU with internal gfx, so I do not expect to do such muxing myself.

Now, to get me thinking in terms of desktop... My ideal for this is to make a ComExpress module, which is an industry standard thing. There's several with PPC on them. Freescale is workign on their own proposal for an official merging into the ComExpress standard... Think of this like a Classic Amiga motherboard with an accelerator card. But move more of the fun stuff onto the accelerator, as some of the fancier ones like CSPPC or CS-3 did, with SCSI, Graphics slot, etc. In my case, if I was workign with a CPU with internal gfx, then connect that. bu tdo not add a gfx chip to a non-gfx CPU on the module. That's better in a PCIe slot anyway for us. Though I had long ago intended to have a Radeon on the module as well, I have since let that one go. Now, we essentially have the "new Amiga" on the module, and the motherboard becomes little more than wires to connectors. This lets us do the hard part once, and reuse it with relatively simple motherboards in various ATX, Mini-ITX etc. shapes. And also lets us show the module to non-Amiga markets for whatever things they might like it for.

And for me, saving a little power is important, so long as I can source parts and maintain feature set. I want performance. I also want long battery life. What can reasonably be done is somewhere in the middle of those two goals. I won't go without an SD slot to increase battery life, I want a way to power down that slot when not in use. if I;m just looking at webpages or emails, then power down some processor cores and maybe a memory contoroller, if more than one of them. Turn them back on when I load up Quake 7.

Doing things as minimal as possible is one philosophy. Some people share it. I have a somewhat different philosophy. Some other people might share it. If we all agreed on only one way to be correct, then there would only be one product. not an OS4.MOS/AROS split. No arguments of 68k vs PPC vs ARM vs x86/x64. No arguments about QT vs WX Widgets vs GTK vs however many others there are. And for desktops, there is a little more freedom to go your minimalist route, like the SAM and Cyrus boards tend toward. Variety lets various customers choose the winner in our nanoscopic market.

But for what I myself really want, there is not so much freedom to make a likable product that way. I do want to include as much expandability for later things, as laptop standards would let me do, but Olegil still gives me a hard time about all those slots and connections. :) In my opinion, if I'm cloning a laptop motherboard, I want to make real use of as many case holes as posisible, as they were intended by the original manufacturer. No unused USB holes. No unused HDMI holes, eSATA holes, ExpressCard slot holes, SD holes, headphone holes, webcam holes, etc. Give me a laptop shell with an AMCC 460 chip and some minimal graphics, but no sound, and I'll be unhappy. No SD slot, and I'll be unhappy. No webcam, and I'll be unhappy. No ExpressCard slot, and I'll be unhappy. That doesn't mean I won't buy one. I wanted very much to get an even lower powered $500 netbook. But I will certainly wish for a bit more. And as an engineer, I've spent about 10 years now imagining how I'd do these things in my own way.

Now, if only I had a boatload of free time...

Last edited by billt on 14-Feb-2014 at 09:45 PM.
Last edited by billt on 14-Feb-2014 at 09:43 PM.
Last edited by billt on 14-Feb-2014 at 09:27 PM.

_________________
All glory to the Hypnotoad!

Quote:

Signal wrote: Quote: matthey wrote: Hyperion messed up by dropping 68k support instead of creating a unifying API between PPC and 68k. What 68k support?

Hyperion supported 68k Warp3D and games for a while. They could have compiled upgraded 68k versions of Reaction for 68k to make it easier to port to AmigaOS 4 but instead MUI has taken over as the GUI gfx standard and nobody wants to do Reaction interfaces. If they were smart, they might even consider compiling AROS versions of Reaction. P96 and CGFX are compatible enough and AHI is available for all systems so at least those aren't problems.

Quote:

Quote: The slowness of AGA was due to C= not upgrading the speed.

It wouldn't have been that difficult to speed up chip memory even back then. AGA could have had revisions. The blitter speed was not increased but could have been. It was the "good enough" mentality.

Quote:

Quote: It would be faster in some ways on a SoC than through PCIe. Users wanting more performance may still want a gfx card in a PCIe slot.

Putting the gfx on a SoC is faster because of shorter electrical pathways to gfx memory and I/O registers. A faster gfx card will more than make up for this small inefficiency but most use a lot more electricity also.

Quote:

Quote: 2) PPC is under attack and it's losing. General purpose x86_64 processors are more powerful... Is that your opinion? Or just your rhetoric? More powerful? What ?

I think most benchmarks will back up that modern x86_64 processors are more powerful than any PowerPC (excluding Power) processors in the market. They have stronger per core performance, are stronger in caches/memory and are stronger per/MHz. PowerPC probably wins in electrical usage at the same die size. That's not even considering cost.

@matthey

Quote:

They could have compiled upgraded 68k versions of Reaction for 68k to make it easier to port to AmigaOS 4

Many OS4 components were originaly 68k (before 2004 pre-release and even while after that).

Quote:

but instead MUI has taken over as the GUI gfx standard and nobody wants to do Reaction interfaces.

There is greater number of Reaction based than MUI based applications on OS4depot.

Quote:

If they were smart, they might even consider compiling AROS versions of Reaction.

Or MorphOS version.

Quote:

I think most benchmarks will back up that modern x86_64 processors are more powerful than any PowerPC

Right!

PowerPC 7447 1 GHz (as used in Pegasos2 G4) is 7 % as fast as single core Core i5-2500 3.3 GHz (except AltiVec). This x86 CPU is nearly 3 years old...
970MP 2.5 GHz (most powerful single core PowerPC performance) reach only 30 % of single core performance of i5-2500 (again except AltiVec).

@matthey

"Different embedded processors will likely have different weaknesses making it a pain to create one standard OS code for all of them. It wouldn't be so bad if one line of embedded Processors was chosen and didn't disappear but that leads us to problem 2."

Does not seem that problematic on PPC so far. Same SW for G3, G4, PPC440, 460, PA6T,... ?
Current state of PPC looks like they have bottleneck pretty much sorted out for our needs.

"PPC is under attack and it's losing."

It also seems that they are a little bit fighting back. But it might be too late. And perhaps they should have had broader portfolio years ago.

"General purpose x86_64 processors are more powerful ..."

True. If ISA would be changed today for AOS4, then x64 would be only sane option (with custom motherboards, not going to be any cheaper...).

IIRC, in MIPS rating intel is nbr1 in MIPS/Mhz/core, then is PPC(I think), then ARM.
In CoreMark the situation is different, etc...

"... ARM64 looks poised to gain major market share."

Only future will tell. ARM64 is only now starting, first generation will be behind AMP series of freescale PPC.

" PowerPC never lived up to it's promises."

Depends on it's promise... For desktop, right. As a network/embedded chip .... after being nbr1 for a decade or so .... I think it lived.

"There isn't that much difference between ARM64 and PowerPC 64 bit but ARM has been more successful in the market."

32bit ARM. And success via cellphones is what you meant. ARM gained momentum there and now spreading.
PowerPC guys never wanted to spread .... perhaps Irvin Gould and Medhi Ali is there somewhere...

" ARM is automatically associated with low power consumption"

But not as "high performance per watt".

"PowerPC will be around in the embedded space for a few more years but if Freescale starts manufacturing ARM64 then PowerPC is toast."

Freescale has ARM64 and e6500 based layerscape chips coming (e6500 as the higher performance core) and PPC NGCore in the plans.
+they promise 10+ year for a new PPC chip to be in production

We do not need to do hasty decisions. SW work needs to be done first.

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

I just realized/remembered that T1** chips are e5500, not e6500 core. So I now change my tiering that low-end for me would be T2*** to have the e6500 core, and highend woul dbe T4***.

_________________
All glory to the Hypnotoad!

@billt

Quote:

billt wrote: @Boot_WB (billt struggled with several levels of nested quote tags before settlign on this)

(boot_wb snipped and hacked mercilessly instead)

billt wrote:
Quote:

The biggest missing bit is audio, so let's do a Nemo and stick a southbridge on one of them PCIe ports.

Just for audio? It seems like overkill if so, why not just a dedicated audio chip connected to the pcie bus (ie using a pcie card).

Quote:

Understand that what I'm interested in designing, I do not expect to happen several times. Nor do I see later expandability as a reasonable answer to certain things like audio. [...] Also realize that of all the things I've actually worked on toward my own goal, I am actually today more prepared to do things with a southbridge chip than I am with a CPU chip...

Ah, so not just for audio then.
Now I'm not a computer engineer, just a hobbyist in hardware terms, so your experience in almost infinitely more than mine, but to me it sounds like you're proposing more your dream design of an Amigaone rather than discussing a pragmatic approach to designing a reasonably priced board for a small market where sales can't be guaranteed to reach 500 units.

Nothing wrong with that, but we may be talking of a potential market of 1 person.

Quote:

Also understand that I would expect that the sb600 drivers would be relatively easy to port onto other AMD southbridge/FCH parts. Certainly easier than if I would do the same with an intel PCH equivalent. So I do not expect this would be a gigantic problem for drivers. Then from there we can move to additional interesting drivers for the sum total. [...] If you're afraid of new IPs and new drivers, then we should never make anything new at all. Stick with the old, and never make any modifications whatsoever to that. I am not afraid of this driver situation. I think it sounds like fun, and educational.

It's more about realism and available resources. If you want to write your own drivers there's nothing to stop you (you could even write drivers from scratch for hardware which is already supported, or add further support to the envy24 drivers, or scsi to blizzard ppc, or, or, or...
...but would that (byo drivers) be a selling point for any other customers?

Quote:

Besides, being afraid of not already having drivers for on-chip SoC peripherals is actually an argument in favor of southbridge peripherals that have been seen already on Nemo...

Complication=Cost.
The layout for stable pci/e bus is a well-trodden path. Component cost is in the pence. Acreage of pcb isn't really that much additional cost, hence Cyrus is likely to be ATX only, not with an option for miniATX/ITX/whatever-size-Cyrus+-was (size does not drive cost, variation drives cost).

Quote:

I want a laptop

That's a different discussion though. For a desktop (which is what Cyrus is intended) the form factor is fairly irrelevant (atx>=Cyrus>=itx).

Quote:

I do not think that a southbridge is an unacceptably complicated addition to a PC computer. It's rather standard way of doing things, not an oddball method of gaining features. In fact, the southbridge of choice would be a quite standard AMD PC part, as A-Eon used on the Nemo. If this is unusual, then why does everyone have more than half a PC?

Adding any complication adds iterations to the design, prototyping, testing, and debugging cycle (ie the NRE costs). From the example in my previous posts, assuming 500 boardsd produced, the manufacture of a single design revision would add $100 to the cost of each board. Add into that a repeat of testing, debugging, possibly rewriting drivers as a result of changes (see Nemo ethernet), and that most expensive of components: time to market.
Now, granted there aren't a whole lot of other companies competing in this market, but time marches on regardless, customers move on, and technology around supercedes even further the tech you've had on the testbench for a year... two years... three years through revisions.
In the meantime you're probably also paying Hyperion for intial porting to the protoypes to get the ball rolling... and you have your fixed overheads with no income at all in the meantime until sales start.
And if you want to keep a market, you might want to spend some on marketing in the meantime.

Quote:

If I were actually working on any such thing (I wish I had time to do so, but this is a rarity for me), then I would get a dev system from Freescale, and work on a standard PCIe style plugin board with my southbridge choice. The way I think of things, this plugin southbridge should be considered as more than just a development thing for myself and maybe a couple other people, I would approach desining it as a standard product, to be used in Sam460, Cyrus, etc. as well, and then perhaps in A1XE, Peg2, etc. boards via a PCI-PCIe bridge. But for my own end product goal, I wouldn't design a laptop with such pieces missing, as I don't think it makes sense to add so many of them later as addons. [...] Finally, considering my ideal end use in a laptop form, I would prefer to use an MXM3 slot for graphics. No GFX chip soldered to the motherboard. If there's something built-in to the CPU, like in the 8610 and some others, then maybe provide amux between that and a MXM board, which may or may not be doable within a current AMD or Intel FCH/PCH southbridge art. I think they can do that for a VGA port, but am not sure about the same to an LCD panel. But I do not expect that I would be working with a CPU with internal gfx, so I do not expect to do such muxing myself. Now, to get me thinking in terms of desktop... My ideal for this is to make a ComExpress module, which is an industry standard thing. There's several with PPC on them. Freescale is workign on their own proposal for an official merging into the ComExpress standard... Think of this like a Classic Amiga motherboard with an accelerator card. But move more of the fun stuff onto the accelerator, as some of the fancier ones like CSPPC or CS-3 did, with SCSI, Graphics slot, etc. In my case, if I was workign with a CPU with internal gfx, then connect that. bu tdo not add a gfx chip to a non-gfx CPU on the module. That's better in a PCIe slot anyway for us. Though I had long ago intended to have a Radeon on the module as well, I have since let that one go. Now, we essentially have the "new Amiga" on the module, and the motherboard becomes little more than wires to connectors. This lets us do the hard part once, and reuse it with relatively simple motherboards in various ATX, Mini-ITX etc. shapes. And also lets us show the module to non-Amiga markets for whatever things they might like it for.

Cheaper, easier, quicker, with no couple-of-years-development-to-support-new-buses: design a multi-pci(e) card using such components.
1 PCI(e) card with NEC USB, Envy24HT audio, Radeon r200, gigabit ethernet all behind a bridge chip. Basically the common standard of well-supported peripherals - everything but the cpu and the mass storage.
Make it the defacto AmigaOS/MorphOS/AROS expansion card.
Maybe not as exciting, with fewer TLAs, but more achievable and probably more useful.

Quote:

And for me, saving a little power is important, so long as I can source parts and maintain feature set. I want performance. I also want long battery life. What can reasonably be done is somewhere in the middle of those two goals. I won't go without an SD slot to increase battery life, I want a way to power down that slot when not in use.

That happens anyway when not in use. Flash is static memory, it doesn't consume energy sending a strobing refresh signal.

Quote:

if I;m just looking at webpages or emails, then power down some processor cores and maybe a memory contoroller, if more than one of them. Turn them back on when I load up Quake 7.

ACPI, NAP, Frequency-stepping are all fairly standard things. Turning off memory controllers and cpu cores though, is that even possible without reinitializing hardware?

Quote:

Doing things as minimal as possible is one philosophy. Some people share it. I have a somewhat different philosophy. Some other people might share it. If we all agreed on only one way to be correct, then there would only be one product. not an OS4.MOS/AROS split. No arguments of 68k vs PPC vs ARM vs x86/x64. No arguments about QT vs WX Widgets vs GTK vs however many others there are. And for desktops, there is a little more freedom to go your minimalist route, like the SAM and Cyrus boards tend toward.

Imo this is not a philosophy which results in a minimised cost, and is the main reason Engineers need managers (to tell them when they're close enough, as nothing is ever fully optimised or perfectly designed, well nothing with a business plan behind it anyway). I don't like them (managers) either, but it is a valid point: good engineers usually aren't good businessmen.

Nothing wrong with dreaming, or such abstract discussion as this, but unless you have a spare $1.2Million to play with, it will remain abstract discussion.

Quote:

But for what I myself really want, there is not so much freedom to make a likable product that way. I do want to include as much expandability for later things, as laptop standards would let me do, but Olegil still gives me a hard time about all those slots and connections. :) [...] In my opinion, if I'm cloning a laptop motherboard, I want to make real use of as many case holes as posisible, as they were intended by the original manufacturer. No unused USB holes. No unused HDMI holes, eSATA holes, ExpressCard slot holes, SD holes, headphone holes, webcam holes, etc. Give me a laptop shell with an AMCC 460 chip and some minimal graphics, but no sound, and I'll be unhappy. No SD slot, and I'll be unhappy. No webcam, and I'll be unhappy. No ExpressCard slot, and I'll be unhappy. That doesn't mean I won't buy one. I wanted very much to get an even lower powered $500 netbook. But I will certainly wish for a bit more. And as an engineer, I've spent about 10 years now imagining how I'd do these things in my own way.

So you wouldn't allow your design to be limited by support for pre-existing peripherals, in terms of buses/standards used, simplifying design choices... but you would limit your design's layout around the existing physical holes in a re-used laptop shell?

That's crazy talk!

Last edited by Boot_WB on 15-Feb-2014 at 03:17 AM.
Last edited by Boot_WB on 15-Feb-2014 at 02:28 AM.

_________________
Troll - n., A disenfranchised former potential customer who remains interested enough to stay informed and express critical opinions.
opp., the vast majority who voted silently with their feet.

@matthey

Quote:

We need a more standard API between Amiga flavors with PowerPC and 68k

Well here is the problem there are two types of developers how do not want to take the next step.

There are hardware banging Assembler programers.
And then you people who stick to AMOS, who does not want to move on, to learn some thing new.

We can't really do anything about not having hardware registers, if hardware is different, so this is unlikely to happen.

Quote:

Hyperion messed up by dropping 68k support instead of creating a unifying API between PPC and 68k.

Drop it, you don't know what your taking about.

Quote:

Even a cut down AmigaOS 4.x for 68k would have made their APIs more common and development easier.

Its a wast of time and resources, better to investigate what 68k users use of software to be happy, try obtain the source code, upgrade the source whit modern features like long file support and 32bit graphics, and port it over.

Quote:

AGA has kludges but it could be fixed up and used for the top layer of 2D gfx.

AGA is the hardware it can never become top layer of anything its chips, bit and wiring. If put on top anything you might as well call it UAE.

Look what was suggested here was some kind hardware emulation of AGA, enabling the video output to be rendered into a frame buffer, that can be Composited onto Window ow Screen in order to speed up UAE or even possibly be integrated into a motherboard, to enable higher compatibility.

Quote:

AGA is not slow anymore in more modern logic chips and memory.

Don't be a idiot, if AGA was that grate don't you think a big company from China might have copied by now.

AGA lost its momentum in 1993, when planar graphics become outdated technology and the fast PCI bus speed, enabled faster and better graphics on PC at the time, Commodore when bankrupt over lack of being in innovative, the hardware was too limited for most PC games, too little graphic memory, to slow bus speeds, and many other problems.

Last edited by NutsAboutAmiga on 15-Feb-2014 at 07:51 AM.
Last edited by NutsAboutAmiga on 15-Feb-2014 at 07:49 AM.
Last edited by NutsAboutAmiga on 15-Feb-2014 at 07:48 AM.
Last edited by NutsAboutAmiga on 15-Feb-2014 at 07:32 AM.
Last edited by NutsAboutAmiga on 15-Feb-2014 at 07:27 AM.
Last edited by NutsAboutAmiga on 15-Feb-2014 at 07:20 AM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@matthey

Quote:

Hyperion supported 68k Warp3D and games for a while. They could have compiled upgraded 68k versions of Reaction for 68k to make it easier to port to AmigaOS 4 but instead MUI has taken over as the GUI gfx standard and nobody wants to do Reaction interfaces. If they were smart, they might even consider compiling AROS versions of Reaction. P96 and CGFX are compatible enough and AHI is available for all systems so at least those aren't problems.

So way do you want to have Reaction for AROS, if not one uses it? Your trolling, Reaction is being used by a lot of software on AmigaOS4, not all developers care about AROS, its simple as that, if they wanted there software to run on AROS, they write there program in MUI, its that simple.

MUI is different cake, it was never part of AmigaOS until AmigaOS4, and even now you need to pay a license to fully unlock it, way the developers of MUI does not update it for 68K you need to ask them, the core MUI developers are MorphOS developers.

AROS has Zune, its a MUI clone it might not fully be MUI compliant at all.

Hyperion have a lot of other stuff to take of care, back porting, wasting time on hardware from 1990's is just a wast of time, MorphOS did the right thing when they dropped classic Amiga support.

Last edited by NutsAboutAmiga on 15-Feb-2014 at 12:46 PM.
Last edited by NutsAboutAmiga on 15-Feb-2014 at 10:20 AM.
Last edited by NutsAboutAmiga on 15-Feb-2014 at 07:53 AM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@68k support

When Amithlon was being built to be a product, it was said that some AOS4 elements would appear also for Amithlon. ... the rest is history.

But I think it would be cool if 68k could be given NG API compatibility. It would be usable for some 200Mhz+ 68k systems (FPGA or JIT emulated).

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

@KimmoK

What 200Mhz, 68k systems? One or Two hardware hacked motherboards?

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@NutsAboutAmiga

Quote:

Well here is the problem there are two types of developers how do not want to take the next step. There are hardware banging Assembler programers. And then you people who stick to AMOS, who does not want to move on, to learn some thing new.

Eh... there is nothing wrong in banging hardware or using AMOS. Hardware banging is only way to get games to run at decent speed on low end Amigas.

Quote:

Its a wast of time and resources, better to investigate what 68k users use of software to be happy, try obtain the source code, upgrade the source whit modern features like long file support and 32bit graphics, and port it over.

Eh, long file names were supported since 1985. 32-bit graphics on the other hand, unless you redraw graphics it is no use.

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

@itix

I was thinking about > 4Gbytes files not long file names.

Last edited by NutsAboutAmiga on 15-Feb-2014 at 10:18 AM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@pavlor

Hyperion could support 68k if they want

But I have read a thread where someone asked one of the Friedens to compile a 68k version I think of Warp3D. The answer was (in a not very friendly tone) "if you want Warp3D buy AmigaOS"

So I do not think they are very interested to support the 68k platform. They finally harm themselves in my view but that is the situation.

@OlafS25

What company today try to support a 486?
way does not Windows7 run on 486?

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@NutsAboutAmiga

the "Game Master System" included in AROS Vision is written in assembler and supports both AGA and RTG. Common APIs mean we need Libraries that behave the same on all platforms and make it easy to port software. If they are written in assembler (including hardware banging) and on another platform written in C using system libraries is not important.

@NutsAboutAmiga

if 80% of all users still would use 486...

@NutsAboutAmiga

and you underestimate how fast 68k is (depending where it runs). You propably identify 68k with classic hardware but that is not the case anymore. I can watch videos on 68k now.