The Vampire II has had a GOLD3 core ALPHA released!

[ronybeck]

So I flashed Gold 3 onto my A600 last night and tried it out a bit. It is really nice to be able to connect my Amiga to a HDMI monitor without a scandoubler and play old games using native Amiga screen modes. Despite being still an alpha version, it is quite far along. AGA + RTG will be a really great combination once the development is that far.

[kolla]

Originally Posted by majsta

1. But what is the plan.

I was talking about the big plan of Gunnar, which started long before Vampire cards, and I guess long before you were involved. Since then, his plan has of course changed a lot, becoming less and less ambitious.

2. Interest for core itself outside of Amiga community we never discussed seriously.

Wayback Machine on the apollo-core.com site, shows that it for a very long time barely even mentioned Amiga, and all performance comparisons were against other CPU softcores of various architectures. It always seemed, to me at least, that it tried hard to attract a much larger audience before it ended up "just" being an Amiga project.

3. Interest in Amiga community, hm numbers says differently. At the beginning I didn't expect that there will be more than 100 cards build, now, like grond said we are at 4 digits, every single day at least 5 card orders.

I am talking about the interest for Apollo Core, not Vampire cards - of all people, I would have thought you could tell the difference. It was not long ago that Gunnar made it very clear that Apollo Core is one thing, and Vampire cards something entirely different.

4. Coders and developers are supporting Apollo, question is how many of them exist anymore.

There are plenty of m68k developers around, you just need to look in the right places, but by focusing on Amiga, you do not really attract the right crowd. This is one of my grudges with AC68080, it should be developed in cooperation with skilled 68k developers, those who maintain and develop compiler backends and toolchains etc.

Potential users are always plenty, what's interesting is number of people who have time and skill to bring the platform forward, developers, developers, developers.

Once standalone version or A1200 version is done this will change for sure. With standalone version others will join, others who are exploring, searching for something new, who are not connected with Amiga.

One can hope, I am not so optimistic. Have you done any kind of survey? Why would people who are not connected with Amiga select V4 standalone? And why would they use Apollo Core instead of using the FPGA with their own cores? Why would they chose V4 over other, cheaper and dare I say, more open FPGA solutions that use the very same FPGA?

5. As for support for compilers, there is support, how do you think people in the team compile stuff for Apollo, search our forums for more info.

They compile for Amiga using compilers that don't support AC68080 whatsoever, and then tweak in 68080 optimisation by hand using assembler. The only thing directly supporting AC68080, that I know of, is VASM. Am I correct?

6. As for support, it should exist the way it works in other communities

Then at least have a bug tracking system, so one can see what bugs and issues are known.

7. About other hardware manufacturers. ... Many don't see how is hard to organize production and each batch needs huge investment.

Exactly, this is what I mean with low interest for Apollo Core. Thank you for supporting my view.

Also for simple test, who is capable of doing what I have send number of blank PCB's. ... That also shows that interest is good.

And this is not about Apollo Core, but about Vampire cards. Again the two are mixed together as if they were one.

8. Price of the hardware was not something Gunnar decided.

And I never said so. What I did say, is that he criticised other Amiga products for being overly expensive, and often ridiculed other amiga hardware developers for wasting time with 030 cards, delays etc. But now, with Vampire cards costing what they do, and there for sure being delays... all that becomes rather ironic.

You blame him for everything, and there are plenty decisions he didn't make inside team.

I don't blame him for everything. I blame him for bad communication, and I blame him for decisions with the CPU core. It is often very hard to draw the line between "the team" and him.

When there is something I talk with him on private he always says, let's include others in this discussion because their opinion could be valid.

I notice that you use the word "could" there, it is typically Gunnar who decides which opinions that are valid or not, right? That was certainly the impression I got.

9. Apollo core as a new baseline is happening, for me, it is next logical step.
Again, numbers tells me differently and mails I m receiving how people are excited after one day of usage.

More interesting are the mails from people who are excited after months of usage. For whatever reason, the people who seem most excited on forums, are the ones who are still waiting.

The fact that there are no new programs is not related to the core. I didn't see lot of software done for 060 also, or for PPC. It is simply lack of developers. They could be motivated to do something if strong user base is established.

I have never, ever, seen developers being motivated by sheer number of potential users. Developers are motivated by good developer tools, great documentation, interesting challenges, good spirited developer communities, open platforms, meetings and conferences etc.

But think it like this, good programmer, if we are talking about the money, could earn more doing PC software then he could earn from user base counting, eventually, 10 000 Apollo core users. This needs to be done differently. Same way you can't motivate me with money to enter some other project than this.

You can never expect to attract good developers with money alone, those you do attract are typically not the ones you may have wished for. Then there are things like... of the 10 000 Apollo Core users, how many are interested in your piece of software, enough to pay for it?

10. Delays, you are talking about. Let's be rude now, people, all of them got inside their cards what they paid for. They got it already.

Again, I am not criticising anyone for delays, I am criticising certain individuals for having ridiculed other projects for having delays, as mentioned, the situation is not without irony. I am sure you did not mean "Let's be rude now".

11. Speaking about core too tied for some FPGA. Porting takes time. As you can see Intel published new Cyclone 10 who is based on Cyclone III/IV and not on Cyclone V. Why is that? If you track some other projects FPGA related, you could easily see that some things Altera did with that series created lot of problems for even more serious projects than we have here. So what, we overcome so far many problems, with this one it will be the same.

Of course, such things are to be expected. And the Apollo Core is not just any FPGA project, it is a CPU architecture. Questions like how tied it is to the Cyclone IV, how much work would it be to build AC68080 and SAGA to other FPGAs, how feasible would it be eventually to make ASIC...

I really hope that I have answered your questions.

I asked about Apollo Core, but was mostly answered about Vampire cards. But fair enough, I appreciate your long answers and the spirit of your posting.

[grond]

Originally Posted by kolla

Since then, his plan has of course changed a lot, becoming less and less ambitious. ... tried hard to attract a much larger audience before it ended up "just" being an Amiga project. ... you do not really attract the right crowd. This is one of my grudges with AC68080, it should be developed in cooperation with skilled 68k developers ... I am not so optimistic. Have you done any kind of survey? Why would people who are not connected with Amiga select V4 standalone? And why would they use Apollo Core instead of using the FPGA with their own cores? Why would they chose V4 over other, cheaper and dare I say, more open FPGA solutions that use the very same FPGA? ... this is what I mean with low interest for Apollo Core.... I blame him for decisions with the CPU core. ... I am criticising certain individuals for having ridiculed other projects

OK, we got it. You are 100% right. Gunnar is an asshole and his project a failure. Why again do you care about it?

[meynaf]

Originally Posted by Don_Adan

No, JIT crashes your port of HOMM2. But this is PPC version (E-UAE).
https://www.ppa.pl/forum/strefa-gier...-2-na-amidze/2
f.e. post 51.

Perhaps E-UAE does not correctly support all 68020+ stuff.
But under Winuae it's fine.

Originally Posted by Don_Adan

And yes, SMC can crash Win-UAE JIT too, but i never tested this, because I dont use WinUAE.

I tested it. It's a real performance killer, but it does not crash (at least on 3.1.0).

Originally Posted by Don_Adan

Only some people asked me to remove SMC code from TFMX 7v replayers, because replays doesnt works OK.

SMC is more a problem for 68040/68060 than for emulators.

Originally Posted by Don_Adan

I dont think that you can create 68040 compatible CPU code, which can crash 68080. But you can try. Good CPU test program will be perhaps very useful, for Apollo Team.

It is very easy. Check availability of 68080, and crash if you find it
It is also possible to implement extra instructions in free opcodes (with emulation like 68060 does for movep & co). Some exist for something else on 68080 and the program will not work or even crash.
Or write some 020+ code that depends on move sr being supervisor only - will not work for 68080.
Etc. As long as there are differences, you can write incompatible code.

Originally Posted by grond

OK, we got it. You are 100% right. Gunnar is an asshole and his project a failure. Why again do you care about it?

Imagine you buy some product for which you have high expectations and get deceived.
Then you see, over and over, new ads for said product. Do you popup and tell your friends and other people, or let it go ?

[grond]

Originally Posted by meynaf

Imagine you buy some product for which you have high expectations and get deceived.
Then you see, over and over, new ads for said product. Do you popup and tell your friends and other people, or let it go ?

I do have friends and I think part of the reason I do is that I don't continue telling them the same thing for years.

[meynaf]

Originally Posted by grond

I do have friends and I think part of the reason I do is that I don't continue telling them the same thing for years.

Even if they don't get it and behave as if you had said nothing ?

[grond]

Originally Posted by meynaf

Even if they don't get it and behave as if you had said nothing ?

I think you are onto something...

[modrobert]

<!-- you're a hot blond girl and do not want to wait anymore for GOLD3 ? just contact me :-) -->

[meynaf]

Originally Posted by grond

I think you are onto something...

Don't jump on conclusions too quickly.

[Don_Adan]

Originally Posted by meynaf

Perhaps E-UAE does not correctly support all 68020+ stuff.
But under Winuae it's fine.



I tested it. It's a real performance killer, but it does not crash (at least on 3.1.0).



SMC is more a problem for 68040/68060 than for emulators.



It is very easy. Check availability of 68080, and crash if you find it
It is also possible to implement extra instructions in free opcodes (with emulation like 68060 does for movep & co). Some exist for something else on 68080 and the program will not work or even crash.
Or write some 020+ code that depends on move sr being supervisor only - will not work for 68080.
Etc. As long as there are differences, you can write incompatible code.




Imagine you buy some product for which you have high expectations and get deceived.
Then you see, over and over, new ads for said product. Do you popup and tell your friends and other people, or let it go ?

Yes, perhaps E-UAE JIT has problems/bugs with handling bitfields instructions.
Sorry, Phil. But you still cant crash 68080 using normal code 68040 CPU code. Move SR works in user mode for 68000, 68020, 68030, doesnt works for 68040 and 68060 only, not tested for 68010 (i dont have access to Amiga with this CPU). I tested this because some Amiga players used move.w SR, -(SP) , and move.w (SP)+,SR in user mode. You want to be, to puristic to me for SR handling. No difference for normal code, maybe if you want to detect 68040 or 68060 only. But noone used SR handling for CPU detection.

[clebin]

Originally Posted by meynaf

Even if they don't get it and behave as if you had said nothing ?

For me, it's more like this:

Friend 1: "You have a crap hairstyle"
Friend 2: "I know, but I don't care."
Friend 1: "Yes, but it looks stupid"
Friend 2: "I know, but it really doesn't bother me."
Friend 1: "Why don't you care?! You should care!! It's very important that you understand this!!!"...and on and on...

eg. Gunnar said it would have a FPU AND AGA!!!! He said it would all fit!!!!11. And he's arrogant!! Well, it didn't. Whoop-de-shit.

[TroyWilkins]

Originally Posted by clebin

For me, it's more like this:

Friend 1: "You have a crap hairstyle"
Friend 2: "I know, but I don't care."
Friend 1: "Yes, but it looks stupid"
Friend 2: "I know, but it really doesn't bother me."
Friend 1: "Why don't you care?! You should care!! It's very important that you understand this!!!"...and on and on...

eg. Gunnar said it would have a FPU AND AGA!!!! He said it would all fit!!!!11. And he's arrogant!! Well, it didn't. Whoop-de-shit.

Yes, that's how it seems to me too.

Or to put it another way, if your friends keep hearing the same thing, over and over from you, and they don't agree or don't care, if you keep going on about it, do you really think they'll stay your friend, I think there is a good chance they may at the very least reduce how often they interact with you...

Like how a lot of people, back in the day, didn't talk to me about computers because I would tell them they needed to get an Amiga. Over and over again. I learnt from that - eventually, hahaha

[clebin]

Originally Posted by TroyWilkins

Like how a lot of people, back in the day, didn't talk to me about computers because I would tell them they needed to get an Amiga. Over and over again. I learnt from that - eventually, hahaha

Hmm, that sounds familiar! A few years ago I ran into my schoolfriend Michael, who told me what a pain in the arse I was incessantly banging on about the Amiga and rubbishing his PC. I felt a bit bad about that...

Sometimes, you have to accept that people understand the message but you won't change their mind. Continuing to repeat the same thing just makes you a fucking irritant. Lesson learnt for me too..

[ronybeck]

So what exactly was the topic of this thread again?

Maybe we should change it to "Vampire Accelerators: Awesome Hardware or Emotional Trigger"

[clebin]

Originally Posted by ronybeck

So what exactly was the topic of this thread again?

Maybe we should change it to "Vampire Accelerators: Awesome Hardware or Emotional Trigger"

Come in, join the circle...

BTW, I must congratulate you on having one of the few on-topic posts at the top of this page!

[Gorf]

Originally Posted by TroyWilkins

Like how a lot of people, back in the day, didn't talk to me about computers because I would tell them they needed to get an Amiga. Over and over again. I learnt from that - eventually, hahaha

Originally Posted by clebin

Hmm, that sounds familiar! A few years ago I ran into my schoolfriend Michael, who told me what a pain in the arse I was incessantly banging on about the Amiga and rubbishing his PC. I felt a bit bad about that...

You both were simply not persistent and enthusiastic enough back that, or otherwise Amiga would rule the world by now!


(And how could you call them "friends", if they own a PC???)

[majsta]

Quote:
Originally Posted by majsta
It can be simple memory expansion using original onboard CPU or advanced device.

Originally Posted by kolla

Do tell how.

By the fact that you actually asking me this says it all so I won't bother to reply on your last post. Write it on your own. It is only 100 lines of code. If you can't do such simple thing then really what's the point of this talk or your accusations. To put down someone's else work you don't even understand. In reality, what's stopping you to do your own core or architecture. Do stuff differently than we do. We can talk seriously only when you show something but this is something that will never happen. True purpose of your posts is just to waste our time and destroy every Apollo-core topic opened by constantly putting words out of context and with silly talk Gunnar said that and did this. Who fucking cares what he said or did. Get a life, not like you were married with him.

I just gotta go off the topic and post this:
[ Show youtube player ]
This is the video everyone should see.
Man said it all and this should be lesson to everyone.
This is something happened number of times for most Amiga projects.
Do we really want this happen again.
Is our desire to chase away anyone who is trying to create something by feeding personal ego and not respecting effort someone puts int projects.
Just watch it.

[Tigerskunk]

I often feel there is a lot of wounded pride at play here and from other sources (there are some dudes on FB who sound the same everytime anything Vampire comes up).

It's always "Gunnar that idiot didn't listen to me, and look how crap everything's now" or some variation of it.
How Gunnar dare not listen to them.

Meanwhile all the other people are playing Cannonball Engine and Games on the Neo Geo Emulator on their Amigas on those crappy Vampire cards with that shitty Apollo core.

[Gorf]

Originally Posted by Steril707

Meanwhile all the other people are playing Cannonball Engine and Games on the Neo Geo Emulator on their Amigas on those crappy Vampire cards with that shitty Apollo core.

so the predominant use-case is to play ports and run emulators, that one could run on any PC?

[Tigerskunk]

Originally Posted by Gorf

so the predominant use-case is to play ports and run emulators, that one could run on any PC?

So what are your predominant use cases for computer hardware?

(and what this have to do with the fact that people enjoy emulators and ports (meaning: more options) on their hardware, be it vampire or 68060 driven?)