@OlafS25


Chris also made a post about the status of the closed source version of poseidon on a.org at the time.

_________________
Support AROS sponsor a developer.

"AROS is prolly illegal ~ Evert Carton" intentionally quoted out of context for dramatic effect

@Barana

If Poseidon is not legal then how is this card available with Cd and software ?
http://amigakit.amiga.store/product_info.php?products_id=193

_________________
A1200 ACA1232 128MB Indivison MkIICr
Elbox empty Power Tower
RPi3 Emulating C64 ZX Atari PS BBC
Wii with Amiga emulation
Vampire v4 SA

@g01df1sh

ability to read brings advances... I answered that already

@Barana

Quote:

All I'm saying, yet again, is that kollas trolling won't be tolerated and we will report him when he does.

Less drama, please.

_________________
"Unix is supposed to fix that." -- Jay Miner

@Snorg

Quote:

Should know better, yes, but recall: Matt 7:1

I read recently that Matthew 7:1 has become the most popular verse in the Bible, having displaced John 3:16 in that repect. It's become a sort of get-out-of-jail free card; people tend to invoke it when they get caught with their pants down. Even famous people do this, which is curious.

Still, it hasn't shown up on (American) football game signs yet, AFAIK.

_________________
"Unix is supposed to fix that." -- Jay Miner

@cdimauro

Quote:

Indeed. Another bad design..

I can understand the OS putting it there for internal use even though a 68K reset vector still looks like bad place.

Quote:

But nowadays you don't really have ROM code. Unless you talk about embedded systems, where it's likely to have EEPROM used to store the firmware.

I would have been talking about 68K library code in ROM since this was a 68k Development discussion.

Quote:

Well, the Bubble sort code looks like 68K assembly. You can try to (completely) rewrite it in a way that it looks like PowerPC code (hence: favoring it for the compiler), in order to favor this architecture. At least it will be as "fair" as what Gunnar did...

Ha. That woud be fun. Well PPC can't be the worse. I can read it so there's a chance I could write it.

I actually needed a sort routine once. I started off with a bubble sort translated from BASIC or something. It was still slow. I was sorting bytes. In the end I just scapped all sort routines and optimised with speed using a histogram table of sorts. (None intended.) I used an array and counted each value by adding it as an index. The end result was a table with a count for each byte. I just used the array to recreate the values in order. Worked really well.

Quote:

Or even less if you should give some address space to peripherals.

Which the Amiga does within the first 1MB and other Zorro space.

Quote:

AROS can run on Vampire, which has already 64-bit support. However this creates problem mixing 64 and 32-bit applications; see above, when I've replied to Mat.

This can give a warning about what would happen to OS4 if and when it goes 64-bit. As it stands it would need to have a split personality three ways. 68K compatibility layer. 32-bit PPC compatibility layer. And new superior 64-bit OS4 layer overall.

Quote:

Hum. This depends on what happens with a 64-bit architecture when it loads a 32-bit data into the register: if it zero or sign-extend it, then the 64-bit pointer which you loaded at the beginning is lost. If it preserve the upper 32 bits, then it's fine, but it hurts performances on superpipelined processors.

Good point. I'd need to look it up. Where I would be thinking is how PPC32 usually loads an address by loading an EA from offsetting another register, for local small data address, by loading hi when needed and lo. But for PPC64 it uses two instructions for a full 32-bit half, then one for a swap and another two for other 32-bit half. So five instructions needed for full 64-bit load. I'm betting on the upper 32-bit being untouched if PPC64 has an extra instruction for swapping 64-bit half words.

If it is affected as you say my idea is moot. But I'm sure they thought of this. Even if the registers are generic.

Last edited by Hypex on 18-Sep-2018 at 04:23 PM.

@bison

Sure, the whole notion of having no consequences for one's own misdeeds is popular. The point is not about turning a blind eye, of course, but rather not being motivated so as to seek harm (condemnation) upon another.

Like Rose, I was going to point out the 'discrepancy', but *perhaps* with a different motive.

Cheers.

@Barana

Quote:

Barana wrote: @cdimauro All I'm saying, yet again, is that kollas trolling won't be tolerated and we will report him when he does. Not allowing trolling to go on is not trolling. It is reporting.

kolla is not trolling: he's reporting facts. Which you don't like, and only for this you cannot classify it as trolling.
Quote:

Threres no need for the personal attack. Please remove it. Or I shall greatly delight in telling the world how well your wife rides my saddle,and how long she moaned for. And how she keeps ringing me in the night and asking for more. ;)

This is really disgusting, and clearly proves what I was talking about blind fanatics which starts the holy war against the ones that have "touched" their totems.
Quote:

I'll remove my personal insult, when you remove yours.

How old are you?

You're putting on the same plane critics about forum fanatism with personal attacks even arriving to my wife.

You're insane. And I've nothing to correct.

Insulting remark removed - _Steve_

Last edited by _Steve_ on 30-Sep-2018 at 06:13 PM.

@cdimauro, @all

Can we all drop our (verbal) stones now, please? People say stupid things all the time - it's part of the human condition, as we all can attest. Give a little grace, watch it multiply, as it were.

Peace!

@Overflow Quote:

Overflow wrote: @cdimauro You keep throwing around words like fanatics and defenders, which isnt really helpful either. Yes, there are a few rabid Vampire/Apollo users, that seems to type everything in CAPS, and seems to bounce off the walls when discussing anything.

See above. This isn't the first time. And not even he's part of small minority.
Quote:

Fine, we got those people in every facet of life, and over the years, Ive seen examples of similar behaviour in pretty much every socalled "camp".

That's what I've already stated. Unfortunately fanatism is homogeneously spread...
Quote:

When you say "dont like critics", thats wrong, to a point. It goes back to precived style of feedback. Several weeks ago kolla and flype had a very good exchange, where kolla tested issues and gave good feedback to flype. Without the "poke in the eye" style. Great stuff.

Because he was on another forum and not being attacked by the usual fanatics, included the Big Gun.

When he talked about some facts on the Apollo forum, he was immediately finger pointed and treated like a troll. With messages deleted and/or edited, and even complete thread which where removed.
Quote:

Same as when Daniel/Daytona makes comments; Hes very direct and to the point, almost to a harsh level, but he points directly to issues and his solution/thoughts around it, and doesnt include negativity to his feedback.

And he was immediately attacked by another idiot...
Quote:

You are not doing anyone any favours putting a whole community into 1 camp, based off the behavour of 1 or 2 persons.

In fact I haven't generalized. And you're wrong: it's not about 1 or 2 persons.
Quote:

Im pretty sure most of us knows a few people that might be good at their jobs (in your own sphere), but are always adding negativity to the workplace. A guy I used to work with was really good at what he did, but in his mind, everything was wrong, and he wasnt shy about talking about it, 24/7. I personally didnt mind, cause I appriciated his performance. But quite a few others at work found it annoying/fatiguing to listen to the same complaints for days, months and years. Take ppcamiga1; I would never dream of attributing his behaviour to the general PPC camp/community. Thats a dishonest and lazy approach.

Sure, it happens, and I don't like it, but you can spot the difference between the above guys and one which just express his opinions and reporting facts, which is attacked because of those things are against the holy word.
Quote:

With regards to AROS development: Thats better for OlafS25 and wawa to comment on. Asking the Apollo team to put even more work on their schedule, in addition to core development, seems unrealistic.

I haven't stated this. Please read carefully what I've written.
Quote:

I havent read that callout on EAB by Gunnar vs Thomas,

It wasn't on EAB, but on the Apollo forum.
Quote:

but if it is as you presented then it sounds like he has done several times on Apollo forum and IRC, which amounts to "People should stop posting all the time, and code instead!". Those comments usually come up when people moan about delays or starts dreaming about new features.

The point is that such statement should apply to Gunnar too, which writes TONs of messages instead of working on his product. That's why I cited it.

Blaming others because they prefer to write on forums while he's doing exactly the same (and in a massive way!) is, at least, a deceiving behavior.
Quote:

Not saying Thomas did any such thing, but it does sound like Gunnars less than optimal way of herding people towards the compilers.

See above: he should shut up and work, since he's the first one which spends A LOT of time writing thousands and messages (I don't on IRC, because I don't frequent the channel).

This is the point, and I hope that it's clear now.

@Snorg

Quote:

Snorg wrote: @cdimauro, @all Can we all drop our (verbal) stones now, please? People say stupid things all the time - it's part of the human condition, as we all can attest. Give a little grace, watch it multiply, as it were. Peace!

Sorry, I've just read it, but I've already written another message.

I'll reply to Olaf in a different manner, and I hope that thread will go back talking of interesting technical stuff, as it was before.

@OlafS25

Quote:

OlafS25 wrote: @cdimauro Where are the 68k programmers helping Aros on 68k? Expecially those doing lots of talk on forums. I have not seen them in recent years.

You're an Apollo forum user, so you should already know many of them.

Anyway, the point was another, as I've explained in another message. See above.
Quote:

Aros runs already on original amiga hardware, it replaces the roms and offers lots of advanced features already but would need help to improve speed (not just for vampires). The truth is they were not interested at all. So you cannot blame lack of support on gunnar, most simply are not interested. That is all, the simple truth. I do not blame anyone, everybody can do with his spare time what he wants but stop all that nonsense...

Indeed, and that's why I've cited that fact.

Thomas, like any other people, has the right to spend his time in whatever way he likes, and nobody should complaint about it, especially if he's doing the same thing.

When I'm back at home I've very little interest on coding again, after a day of work. I prefer to relax with other things, included writing something on forums (if/when I've some spare time). And when I like to do some coding, then I've my own projects which are more interesting.
Quote:

Gunnar is sometimes a little overenthusiastic. The project is his baby and he is proud on it, he has a plan what he wants to reach. People should not look on benchmarks but talk to other users who already own it and buy what is now available without expect wonders in future. Even if the products are not satisfying everyone but in contrast to the moaners he has delivered something, not just talk.

He also contributes to the benchmark drama, as you should know.
Quote:

Regarding Thomas he will never support Aros because he has signed NDA at Hyperion and dislikes open source in general. Also Gunnar has not blamed Thomas but Thomas has blamed Gunnar to violate Copyrights because the Apollo team offered a P96 driver based on Aros sources. Also he requested that Vampire better integrates in existing amiga hardware ignoring the fact that this is not the concept of the project. In this case I clearly know the discussions and can indeed correct you...

I've reported what I've read before, and I'm pretty sure about it. I'm not the one which invents stories, especially in a forum with so many people which are very quick on finger pointing.

BTW, you were reported as an "Apollo Team Member" on the Apollo forum, for long time. Then there was a discussion about the CoffinOS, which you strongly reported as being illegal. After your message, I've seen that the "Apollo Team Member" disappeared. What happened?

@OlafS25

Quote:

OlafS25 wrote: @Overflow There is not much interest in supporting Aros 68k what I personal think is a wasted chance. I personal have no problems with it, it is how it is. People are happy with Amiga OS (or Coffin how it is called). It is not legal and not adaptable and I am already interested to see how they support f.e. USB because Poseidon on 68k is not legal to install on Amiga OS except backporting it from Aros sources. Of course you might again solve the problem by offering drivers and illegal install it in coffin. If that is the right way has everybody to decide for himself. It will certainly not motivate new developments (I have to agree there with Thomas). To me it seems Gunnar has a plan for hardware development but somehow thinks software or OS development magically appears. He has no idea or plan there and it will not happen. For that the market even with a couple of thousand users is much too small.

The reality is very simple. With the Amiga you have two different uses cases: either play the good old games or run the applications.

In the first case you don't need new hardware, albeit some accelerator can help on 3D games. You don't need new hardware because nobody will recompile/rewrite those games for the new platform. What's worse, it's better to have an hardware which as close as possible to the original ones, due to the bad programming which was used at the time. AROS 68K is good because it can provide a free rom which allows to boot the games.

In the second case you want a super-doped hardware platform which let squeeze the most from the applications. Here makes sense to have a much faster processor, as well as a big RTG performer and a decent AHI support. However new hardware features aren't unlikely to be supported, because applications are usually coming in binary form, so not recompilable or modifiable. Here AROS 68K can be used, but it needs some optimizations and drivers.

However there's another point: nowadays you can easily cover both scenarios with WinUAE or Amithlon...

Last edited by cdimauro on 18-Sep-2018 at 09:31 PM.

@cdimauro

Thanks, but no need to apologize - and I don't want to come across as a busybody.

@Hypex Quote:

Hypex wrote: @cdimauro Quote: Indeed. Another bad design.. I can understand the OS putting it there for internal use even though a 68K reset vector still looks like bad place.

Well, an application can receive a pointer to ExecBase on some register, when it starts.

Even bootloaders receive some pointers to some o.s. structure that allows them to load some tracks (with the custom bootloader) and take control of the system in a polite, system-friendly way.
Quote:

Quote: But nowadays you don't really have ROM code. Unless you talk about embedded systems, where it's likely to have EEPROM used to store the firmware. I would have been talking about 68K library code in ROM since this was a 68k Development discussion.

68K != Amiga.
Quote:

Quote: Or even less if you should give some address space to peripherals. Which the Amiga does within the first 1MB and other Zorro space.

It should be in the first 16MB.
Quote:

Quote: AROS can run on Vampire, which has already 64-bit support. However this creates problem mixing 64 and 32-bit applications; see above, when I've replied to Mat. This can give a warning about what would happen to OS4 if and when it goes 64-bit. As it stands it would need to have a split personality three ways. 68K compatibility layer. 32-bit PPC compatibility layer. And new superior 64-bit OS4 layer overall.

The problem is the latter, which cannot share all its resources with the 32-bit apps.

This is also why AROS x64 supports only 64-bit applications: it doesn't make sense to support the 32-bit ones because of the intrinsic Amiga o.s. design.
Quote:

Quote: Hum. This depends on what happens with a 64-bit architecture when it loads a 32-bit data into the register: if it zero or sign-extend it, then the 64-bit pointer which you loaded at the beginning is lost. If it preserve the upper 32 bits, then it's fine, but it hurts performances on superpipelined processors. Good point. I'd need to look it up. Where I would be thinking is how PPC32 usually loads an address by loading an EA from offsetting another register, for local small data address, by loading hi when needed and lo. But for PPC64 it uses two instructions for a full 32-bit half, then one for a swap and another two for other 32-bit half. So five instructions needed for full 64-bit load. I'm betting on the upper 32-bit being untouched if PPC64 has an extra instruction for swapping 64-bit half words. If it is affected as you say my idea is moot. But I'm sure they thought of this. Even if the registers are generic.

Hum. PowerPCs could have instructions which preserve the high part of a register by design, to allow an easy loading of big constants/data.

Dishonest for who? Dishonest for what?
Ten years ago gunnar von boehn announced his 68k wunderwaffe.
Wonderfull natami - Amiga that Commodore made if not bankrupt.
Wonderfull natami will run workbench faster than uae on fast pc (in 2008 it was Core2Duo).
Graphics on wonderfull natami will be better and faster than current amiga AGP solutions.
Other members of natami team state that natami will be faster than 600 MHz G4.
Ten years passed. After ten years gunnar von boehn has cheaper 060 with faster memory and 2D graphics.
No MMU, no 3D, non standard FPU without full support from c compilers.
Some ISA 68k extensions without support from c compilers.
Overall performance and capabilities of Vampire are roughly comparable to fast pc from 1994.
Vampire is not good enough to compete with cheap pc from Windows 95 era.
From developer pov Vampire has not things which are in standard since 1995.
Like MMU, FPU, 3D, accelerated 2D. And of course vampire is too slow compete with cheap pc from Windows 95 era.
From developer pov Vampire has archaic performance and capabilities.
Even people in Commodore know that 060 is too slow and work on something faster.
Where is 68k revolution?

Last edited by ppcamiga1 on 18-Sep-2018 at 09:53 PM.

@ppcamiga1

great to see you united with the usual adversaries to get down whatever you fail to etablish as alternative.

@wawa

Is this guy the same bloke who forgot to put the negative rail in the new Ppc accellerators and blamed his customers for it?

_________________
Never underestimate the bandwidth of a station wagon full of tapes hurtling down the highway.

I serve King Jesus.
What/who do you serve?

@Barana

Some has problems with that there is no real progress in 68k development since Commodore bankruptcy.
Harsh truth, but truth.

@ppcamiga1

Quote:

Some has problems with that there is no real progress in 68k development since Commodore bankruptcy

100% agree : all is done by Hyperion, Cloanto, Phase5, Apollo Team, Jens Shoenfeld, Elbox, Bill McEvil and AmigaKit to make impossible any 68k progress...

They use licences, closed sources, and of course division to block everything...

Last edited by CosmosUnivers on 19-Sep-2018 at 07:09 AM.
Last edited by CosmosUnivers on 19-Sep-2018 at 07:08 AM.