Apollo Team announces the Vampire V4

[wawrzon]

Hmm, I don't always agree with kolla, but he and psxphill aren't the only two that Gunnar has rubbed the wrong way.

could you explain how are you getting rubbed the right way, so that people dont make uneccessary mistakes in the near future?

[wawrzon]

@bszili
i understand your considerations. especially its a matter of opinion if the whole apollo approach is (even if closed for most part) a community driven anarchist project or already a business with all its consequences.

however from (potential) users practical point of view it seems that all that is actually a non issue. the vampire boards are sought for, im certain if you putchase and actually got hold of one and then realize that the product is not up to your expectations, you should still be able to sell it, perhaps even with some gain on your part.

i see these all discussions, arguments valid or not, as an attempt on forcing the project in certain direction. alas, for better or worse, its not possible to satisfy everybody, and so the choices that also in this case may be unpopular with some, will always be taken.

i see vampire/apollo project simply as an interesting option. it may not be completely along the lines, what i think is best. but im satisfied with dynamics it wakes. and if i have doubts, i might propose an alternative but would never think of demanding anything from people who do actual work.

[wawrzon]

@bszili
i wasnt talking particularly about you but about the kind of discussions arising around the apollo/vampire project. even if i dont know the internals about the state of the project i personally dont mind waiting till situation clarifies. certainly whoever needs to have fastest amiga ever just immediately, may take some risks, such as the project may develop its mechanisms and concepts as they go and eventually even drop support for previous models.

[wawrzon]

Maybe he is just blind, or a standard troll....

bszili is certainly not a troll. he delivered really a lot of software (ports) for all amigoid platforms and been supportive even though sometimes critical about them all. the more i wonder about his insistence in this case, even though i can sense and partly understand his disappointment. i think a number of people who have been either closely involved in or expectant about this project, became now critical. love lost..

[wawrzon]

Thanks for accepting my point of view, but your point is nonsense. You are basicly saying that if something is popular / succesful it some kind of shield for cristisim.

go on and criticize them all day long, if you ask me.

[wawrzon]

That does fit my observations that apollo isn't meant for amiga, this is just a test ground so people can do his QA and marketing and then he'll be off.

may be. but then what market would that be for?

030 embedded replacement? but then why even 040 and 060 instruction compatibility? why extra features?

060 replacement? but then why no full mmu and fpu compatibility?

coldfire replacement? it isnt compatible at all, in the shape its tested and provided for amiga community? if it was meant as coldfire replacement why even bother for amiga compatibility?

and what about superaga? of what value is it washing mashine or aircaraft industry? why even bother with amiga chipset if you want an embedded chip?

[wawrzon]

https://web.archive.org/web/20150901...o-core.com:80/

yes, buts its not whats being tested with the amiga community, so if they aim at coldfire replacement why dont they test coldfire features and instead test 68k and amiga features with the amiga community.

besides, what you obviously dont know, gunnar was considering coldfire as base for his project, or as cpu solution for natami project at that time. the threads about it are certainly preserved on a1k, in german. he and his associates have been even sponsored wit coldfire evaluation boards by former bplan owners. and they evaluated it and confirmed it inappropropriate for amiga acceleration task, which was exactly the reason to develop an own faster 68k core.

Who knows.

yes, who knows. so simply draw you occam razor, treat it as a black box and stop trying to find some scar on that project, that only exists in your prejudice. its not for you, it may not be for me, fine. but why refrain to poetry to make it look bad for others?

[wawrzon]

We are experimenting with a CPU running an emulation for the 68K processor.

thats what i proposed for years. i hope all the best for your project. however it is completely different approach than gunnar and apollo team. and its a bit misleading, not to mention the obsious differences. also im not sure if mmu can be run with jit, which, if enabled, would certainly cripple the performance below that of apollo core. emulating 040 on a fast pc without jit doesnt reach integer performance of apollo core in current vampire hardware implementation.

[wawrzon]

WinUAE: well, it runs on Windows. I don't have Windows, and I don't pay Microsoft to be able to emuate an Amiga. I already have a perfectly working Os, without spyware, thank you.

i currently mostly use fs-uae under lubuntu under vmware player for immediate testing of changes to aros 68k. it isnt fast, especially that it runs without jit (testing) virtualized on so to say low mem machine (have only 4gb ram to date) but is is usable. you can also run winuae with wine under linux, no problem. you dont need windows.

[wawrzon]

MMU is problematic with jit correct.

thanks for confirmation. thats what also toni says however psxphill has been always challenging that statement.

With jit, and note the ARM is Only emulating the CPU core, it flies. And its a ~$25 BOM cost.

fine, its expectable, as it is the case on a x86 with jit, but what precizely does it mean it files on your arm? are there benchmark results?

We are aiming for different things here, my model is something like an A1200 with a fast accelerator card and RTG plugged in.

sounds pretty close to what an amiga with vampire is supposed to be. curious which approach proves better in a long run

[wawrzon]

I need a cycle accurate 68000. I can extend this to 68020 and faster performance, but I think ARM emulation for the CPU core is the way to go for 68060/MMU/FPU etc.

i think cycle exact 68000 and everything above are two different things.
certainly it doesnt make sense to build a faster porcessor based on pure 68000 design. software that requires it expects it to run precisely at specified speed. all te software that reuires faster 68k cpus are better served with a reimplementation without strictly sticking to 68000 restrictions and design. good deal of binary compatibility is enough.

[wawrzon]

@kolla
ok, we already really well know how gunnar is ignorant and hurting people who know better, how he should proceed with his project. and good to know, that an amiga compatible is not capable of running a linux distro, because it is very important for every true amiga user to run linux on his hardware. thanks for pointing this out constantly.

[wawrzon]

There are a few others who go out of their way with caps-lock and exclamation marks

look, what you describe isnt a pattern with whoever is a vampire interested audience. you described a generic "fanboi" (or whatever it is called) who will hype whatever he sees an opportunity in. the particulars have usually been banned everywhere else, having previously hyped beyond reason and recognition something completely different.

the minions typically don't even have Vampire boards, but wow do they plan to!

yeah, here goes another characteristic of the kind, as if anybody was concerned to know their opinion or whatever they want to buy or not every other post.

do i need to mention, that i dont have a vampire an im still hesitant, if ill be ever getting one;)?

[wawrzon]

Currentely there seems to be more opportunities to run FreeMiNT upon EmuTOS, at native speed, on SAGA screen (already working and demo'ed).

funny enough atari people are far more reasonable in applying their available open options that the majority of amiga crowd. i really envy these people. their practical approach gains my full sympathy.

That said, and not having a MMU on my vamp, i (had to) learn how to do without, and there is many way to debug without a MMU with some good or bad habbits. Also using UAE as companion, helps in this task.

vamp or no vamp, me mostly working on aros68k would be great to find out some way or share experience on debugging stuff, except of having a to put a debug statement every other function in the code. on aros we have gdb, but i have never got it working reliably with 68k.

I might have misunderstood about the minions, sorry then, english is not my native language, thus i agree there are such guys on the apollo forums, yes.

we cannot reduce a wider community to some maybe not so positive examples.