Clearly a sheet of paper i hardware. Postscript is an example of a language that is used to describe how the print on the paper will be. Same for 3D printing, you program the model. Same for FPGA, you program the layout.
You don't program the layout in FPGA, you configure switches. Which is what software is, a configuration of switches. The switches are volatile therefore the configuration is software, if you put it in an ASIC then it's hardware. There is a convention that if the software is executed from flash memory then it's called firmware (all firmware is software, not all software is firmware).
Postscript is software, it's passed through other layers of software and finally configures the print hardware. The paper is just passing by when this happens, the closest CS description for it would be a "storage medium".
So your argument is that because the Apollo core doesn't currently run those non-Amiga operating systems people shouldn't buy Amiga accelerators that use the Apollo core.
I'm pretty sure I haven't said people shouldn't buy anything. Those are three applications from the top of my head that would really use a faster cpu. There are other applications that require an MMU and FPU, but those in particular would be much harder for any software patching schemes to deal with.
Does your argument also apply to all the other Accelerators using Motorola/Freescale CPUs without MMUs?
If people want to choose to buy an accelerator without an MMU it's up to them, the problem is that Gunnar and all you fanboi's are saying that because you don't see a need for an mmu then I deserve to be mobbed for saying I do. Then after that childish behaviour you all then try to shut down any logical discussion with more mobbing and victim blaming.
Are they all incompatible/inaccurate too?
The incompatible and inaccuracy in Apollo is nothing to do with the MMU, so no.
Do you also go on Unix and BSD forums telling them not to buy x86 hardware as it can't run Amiga OS without emulation or classic Mac forums telling them their machines are useless because they don't have a the custom chips to play Amiga games natively?
Childish strawman.
I'm not sure you're argument will win over many people since the "failings" of the current Apollo core would seem to be insignificant to the vast amount of Amiga users who probably have no interest in running the software you want to.
Sure there are people who will vote for Donald Trump the same way a lot of people voted for Putin in the rigged elections and a lot of people screwed the UK's future by voting for Brexit. But being on the popular side is not the same as being on the right side, but you wouldn't understand as you just want to get in on the mobbing.
Ok lets discuss.
Lets start with a simple question. Which Amiga do you own.
So which specific part of the Apollo-Core do you believe to be defective or feature missing.
I own three A1200's, an a500 and an a500+ (I sold my A1500 a couple of years ago). I've got a phase 5 030/50 card in my A1200 with scsi and 48mb of ram and 2(ish) gb ide hard drive. For the a500 I've got a 28mhz 68000 and trifecta 250mb scsi drive.
Apollo is missing the fpu and mmu & the exception frames/what instructions it supports don't match any existing 680x0 design. While I understand that the mmu and fpu haven't been implemented yet, Gunnar was quite clear at the start that he has no intentions of ever providing a 100% accurate implementation of them. None of the statements made since give any indication that he has changed his mind, only that any issues will be dealt with by software. Which is fine for a lot of people, as you're so desperate to get anything that you'll accept what he says without question. So desperate in fact that any negative discussion is met with pure hostility, in case it upsets Gunnar & he leaves with his closed source core.
I just think it's all a real shame.