PiStorm kills Vampire?

[nikosidis]

@Promilus
I agree 100% JIT will not work on everything. Far from it.
The good thing is that it is not that difficult to choose what Emulator to use with PiStorm.
Either have 2 microSDcards or there will be some kind of boot menu.

Most Amiga freaks that is interested in either Vampire or Pistorm have more that 1 Amiga.
This was mentioned in the thread before.
There is simply no Amiga that does it all.

Some times I wish Amiga was not that complicated regarding different hardware and what software will work or not.

I think that is one of the reasons C64 is more poular.

[meynaf]

Originally Posted by manossg

As for your claim that nobody implements AMMX because nobody wants it, I think you might be projecting your opinion to every amigan. Vampire is admittedly a commercial success, so somebody must find these features desirable.

Not really. The commercial success comes from the fact it faster runs existing apps. The new features have nothing to do with that.

Originally Posted by Promilus

Users might want whatever they like (something you should be familiar at apollo forum ... requests for things Gunnar and team will never make), decision rests in hands of developers and when it comes to WinUAE and Musashi - atm those ppl aren't interested in AMMX and for good reasons.

Toni Wilen said it himself : whole thing is developer hostile.

New features are complex and don't help at all for programming flexibility. Actually, quite the opposite. They're 100% speed oriented and pay no care about ease to code, code density, cleanness of architecture. IOW in another cpu family they might have been 'nice', but on the 68k they completely miss the point.

The problem lies in that the new features need coders to support them, where it's the new features that should instead support coder's needs.
As for switching to PPC, x86 or Arm, that can't fully replace 68k for the same reasons.

[Tigerskunk]

Originally Posted by nikosidis

Some times I wish Amiga was not that complicated regarding different hardware and what software will work or not.

I think that is one of the reasons C64 is more popular.

The Amiga is a horrible machine unfortunately when it comes to these things.
You have to test on a range of weirdo hardware, and if you don't release on adf, WhdLoad and CD32 at least you have will a bunch of people complaining.

Compare that to developing for consoles where there is usually only one config, or the C64 where you are set with the main unit and a 1541 and the only thing that you need to take into account is the SID and some slightly different colours on some very early units.

[Thomas Richter]

Originally Posted by Steril707

The Amiga is a horrible machine unfortunately when it comes to these things.

It's not the Amiga that is horrible, it is that software (and sometimes hardware) authors are horrible - in ignoring specifications and best practise guidelines....

[mschulz]

Originally Posted by Thomas Richter

It's not the Amiga that is horrible, it is that software (and sometimes hardware) authors are horrible - in ignoring specifications and best practise guidelines....

Ignoring specifications, best practise guidelines and, that's the worst, even ignoring standards of the programming language currently used. I have seen that more than enough

BTW. Regarding guidelines - you have a PM from me

[Tigerskunk]

Originally Posted by Thomas Richter

It's not the Amiga that is horrible, it is that software (and sometimes hardware) authors are horrible - in ignoring specifications and best practise guidelines....

Yep, when you are doing OS friendly software and if you see the Amiga as just another PC.
For games that run on vanilla hardware you want all the power you can get, switching the OS off.

Apart from that, the Amiga still has the problem of all these different modes of delivering games.

[Thomas Richter]

Originally Posted by Steril707

Yep, when you are doing OS friendly software and if you see the Amiga as just another PC.

Hardly.

Originally Posted by Steril707

For games that run on vanilla hardware you want all the power you can get, switching the OS off.

Which seems to imply that the Os hides the power from the user, which is not the case.

[manossg]

Originally Posted by meynaf

Not really. The commercial success comes from the fact it faster runs existing apps. The new features have nothing to do with that.

Although I find your point legit and speed truly sells a lot of Vampires, it's not the only factor. Considering Apollo team's marketing campaign, AMMX must have something to do with the Vampire's success, since "Apollo 68080 AMMX with 64-bit support" is the first feature they mention in each one of their products, existing and upcoming. If it had "nothing to do with that", as you claim, and potential buyers were indifferent about it (spoiler-they're not), I guess the Apollo Team wouldn't bother spear-heading their marketing campaign with it.

Source:
http://www.apollo-computer.com/products.html

[tomcat666]

So, which apps and games take advantage of AMMX ? Must be some killer apps to have anything to do with the vamps success ...

[manossg]

Originally Posted by tomcat666

So, which apps and games take advantage of AMMX ? Must be some killer apps to have anything to do with the vamps success ...

I don't know about 'killer apps', but Diablo, Sonic 1 and 2, Jake and Peppy, the Neo Geo emulator are a nice taste of what's to follow.

[Promilus]

@manossg - that's exactly how marketing works. That's how AMD sold their K5 processors introducing PR (performance rating basically understood as pentium rating) and then K6-2 and K6-III with 3DNow!

And truth to be told - at least 3DNow! had fairly wide support - mostly because popular game engines supported it so any game built upon it had a good chance of supporting it as well. And also 3dfx glide driver included 3DNow! support which gave AMD quite few frames per second once new version was installed - not enough to combat Pentium II of course but noticeable difference for all AMD users (there weren't many but most likely more % than vampire users across whole amiga community atm).

So what I'm trying to tell you - it doesn't matter if AMMX and 64bit lured ppl to buy Vampire. It has not been used in great many applications so I find no reason to assume ppl bought it for those features. Many ppl bought vampire because of performance results presented by the team (and those were pure 68k code benchmarks), integrated FPU, IDE, USB and SAGA (beta cores for A500,600/2000 users). Those were selling points. It was compact device giving most of things requiring pretty big sandwich before. And for small Amiga models were basically impossible to have.

Also titles you did mention are quite new (at least to amiga). AMMX is with us for few years already.

[meynaf]

Originally Posted by manossg

Although I find your point legit and speed truly sells a lot of Vampires, it's not the only factor. Considering Apollo team's marketing campaign, AMMX must have something to do with the Vampire's success, since "Apollo 68080 AMMX with 64-bit support" is the first feature they mention in each one of their products, existing and upcoming. If it had "nothing to do with that", as you claim, and potential buyers were indifferent about it (spoiler-they're not), I guess the Apollo Team wouldn't bother spear-heading their marketing campaign with it.

Source:
http://www.apollo-computer.com/products.html

As Promilus said -- it's just how marketing is done.
They badly want their AMMX/64-bit to be used massively, they're proud of these things, but in reality they don't count.
I wouldn't be surprised if this even has pushed potential customers away.

[Tigerskunk]

Originally Posted by Thomas Richter

Hardly. Which seems to imply that the Os hides the power from the user, which is not the case.

On a vanilla machine you usually need every byte you can get for developing quality games, and the OS takes space.

So?

[manossg]

@Promilus

This is the core of our disagreement again, mate, a different viewpoint. You see a 'trap', I see exciting opportunities. You see dead ends, I see things being taken a step forward. You see fragmentation, I feel that if a thing is good and worthwhile, it will survive-if not, it won't. Does any one of us hold the absolute truth? I don't think so, but I am looking forward to see and experience all the exciting developments in the Amiga scene.

The fact is that the PiStorm will not be able to run these apps, even if (when) it surpasses the Vampire on raw speed. I am sure PiStorm will manage to overcome most or all of the obstacles and become another truly kickass option. I also would be happy if the PiStorm team (or CS labs or iComp or whoever with their respective hw) decides to introduce new stuff as well, to take advantage of this power.

We Greeks have a saying: "I wish the neighbour's goat would drop down dead", describing the need to blemish the achievements of others, instead of feeling happy for them and/or trying to improve ourselves. Unfortunately, that's the vibe I am getting from this thread, even from the title.

Originally Posted by Promilus

Also titles you did mention are quite new (at least to amiga). AMMX is with us for few years already.

+@Meynaf

So what? The fact is that titles optimised for AMMX are being released, things are moving forward and there is a growing interest in the Vampire community. You have every right to like it or dislike it, to praise it or criticize it, but that's the way things are, at least for now.

[Chucky]

AMMX must have something to do with the Vampire's success, since "Apollo 68080 AMMX with 64-bit support" is the first feature they mention in each one of their products, existing and upcoming. If it

I must say that AMMX is one of the biggest failures of the boards. they boost of it as something gigantic good etc. still there are no real software using it and interest is to the least say: not high

people buy it for it being a fast board. for extra stuff like AMMX, 64 bit etc it is a utter failure. and I am happy about it being a failure.
they should have scrapped all those ideas and first focus on compability (meaning behaving like the item they try to mimic) FIRST.. then when that it spot on maybe put in shit.

Anyway this is my typical vampire-ranting I know.
the vampire is a success of the meaning of sold cards for sure. but AMMX have nothing to do with it.

[YouKnowWho]

Originally Posted by Chucky

I must say that AMMX is one of the biggest failures of the boards. they boost of it as something gigantic good etc. still there are no real software using it and interest is to the least say: not high

Anyway this is my typical vampire-ranting I know.
the vampire is a success of the meaning of sold cards for sure. but AMMX have nothing to do with it.

So...form before function? It could be argued that a lot of upgrades for the Amiga are just about the SysInfo numbers, regardless if it actually changes the user experience much.

Funny, we were willing to wait for floppies to load back on a 68000, yet today we're not happy with 5s load time to Workbench from CF and 68020 making everything playable and usable.

[Samurai_Crow]

The blitter emulation uses AMMX. I suspect software mixing in AHI drivers to also use it. It is common on all platforms to see vector units used in drivers mostly.

[Bruce Abbott]

Originally Posted by YouKnowWho

So...form before function? It could be argued that a lot of upgrades for the Amiga are just about the SysInfo numbers, regardless if it actually changes the user experience much.

It's a credit to the Amiga's design that the user experience on a stock A500 doesn't suck as much as it would on a similarly specced PC, but even just adding some FastRAM and a hard drive makes a significant difference. I added 2MB RAM and a 20MB hard drive to my A1000 and it transformed the machine from a toy into a useful development platform. SysInfo would say there was a modest improvement, but the 'user experience' was greatly improved.

Funny, we were willing to wait for floppies to load back on a 68000, yet today we're not happy with 5s load time to Workbench from CF and 68020 making everything playable and usable.

It's not just about Workbench (and I bet very few are not happy with 5 second load times). Some applications need way more speed than a 68020 can provide. Stay way from them and sure, a stock A1200 is plenty fast enough. But if you want to play games like Breathless or Alien Breed 3D, work with modern image formats, compile large programs or get onto this site with IBrowse, you will want a lot more speed.

A 50MHz 030 is OK if you don't mind lower resolution and slower operation, But a Vampire definitely 'changes the user experience much' in comparison. Whether you want that change is another matter. I mostly use the A1200 because I prefer the bold look of AGA on my TV, and I am happy to run 8 colors on Workbench. But the Vampired A600 is great for when I need to compile a large C program or run stuff that needs RTG and the fastest possible CPU.

My A500 only gets used for stuff that can't handle the extra speed and power of my other machines. I have a minimal WB1.3 disk to cut down loading time, which is OK because I don't need all the stuff my other machines have to load. Nevertheless I hope to add a lowly 14MHz 68000 with a bit of FastRAM and CF Card or hard drive someday, because some games appreciate it.

[Bruce Abbott]

Originally Posted by Samurai_Crow

The blitter emulation uses AMMX. I suspect software mixing in AHI drivers to also use it. It is common on all platforms to see vector units used in drivers mostly.

Amiga OS was designed to work that way from the start, and we already have programs being compiled for different CPUs etc. so AMMX is no different. If AMMX can make some stuff a bit faster then what's not to like? Plenty of Amiga expansion boards have FPUs that are almost never used, but nobody calls such underuse a failure.

I think some people are just upset because they don't have a Vampire and are worried that they will be left out. But so long as drivers or libraries are used it's no problem - except for possible slower operation when AMMX isn't available. That's no worse than having to put up with emulating instructions on an 060 that even stock A1200s have.

[Chucky]

well I call it a failure as it was a way to lure people into their closed platform and their economical eco-system.
however thankfully people seems not to be that very interested in doing stuff with it. sure some software here and there but no real deal-breaker. and for me : no demos. (and for me Amiga is all about the demos!)

for me the total fiasko is to put shit into a "cpu" that isn't used yet before actually implementing what we got. FPU and MMU (well they DID put in a mediocre FPU after a while that sure does its job but not correct)

however. still: AMMX is not why people buy Vampire.. and this is also why I call it a failure as it is not a "buying point" for people. it is more that it is a board that makes your amiga run faster. NOT the AMMX crap.