PC still playing Amiga catchup

[GadgetMaster]

As this thread just refuses to die I thought "What the heck! let me just throw some more fuel on this fire".

Anyway does any Bus Arch 'expert' want to explain how these darn things work?

USB to Gameport Converter

I'm just a little confuzzled about all the polling malarky.

:roflmao:

[jkirk]

That's worse for APIs.  If there are hardware errors, they are usually published and kept so you know they will be there.  If you fix them in APIs, then you have to worry about whether the user has the erroneous version or newer version.  Sometimes hardware errors are taken as features like enabling 7 bitplanes in OCS to enable a static mask bpldat.

i see you really don't know what an api is do you? hardware errors are fixed in the driver NOT the api. the only reason an api changes if there is new additions to the api. aka the difference between directx 9 and 10.

Sorry, but application API access is protected by OS but drivers using APIs aren't.  So they are likely to cause OS lockups just as well as one going direct to hardware. But you can also protect applications going direct to hardware using IOPM.

http://www.grahamwideman.com/gw/tech/Delphi/iopm/index.htm

Under NT, "user-mode" code (ie: applications written by mere mortals) is not allowed to access hardware directly, hence when your application attempts to execute an I/O instruction you get an exception.  The idea is, of course, that hardware resources are things that no application should just take over at will, instead it should be up to the operating system (and its drivers) to arbitrate between different apps requests to use those resources.

That's the theory.  Turns out that the NT kernel maintains a map of I/O port addresses that each process is allowed to access, and for your apps that's normally set to "none". But we can tell NT to use a different I/O Permissions Map (IOPM) for our process and thereby gain access to the ports.  This approach is of course very naughty from a disciplined OS standpoint, so not recommended for widely distributed commercial apps. But for those times when you just need to hack on some hardware, who has time to write a proper NT device driver?

Sorry, that's not how IOPM works and there's no API model involved.  If you wanted to protect your hard drive I/O ports, you can protect them via OS and application can go direct to hardware and only OS will get control if application accesses those I/O ports you protected.  Application does not have to use API.  My application (MPDOS Pro) is an example that goes direct to hardware on I/O ports and does not use any API calls and keeps OS stable.

see above.

[ElPolloDiabl]

PC market constantly evolves and newer machines replace older ones, which get used for other purposes or dismissed/trashed.

No it was the cheap add in cards actually. VGA card, SVGA card, sound card etc.

I came close to buying a 2000HD in 89, but splurged on games instead.

PC parts went down in price a lot quicker too. It was AU$1000 for 20MB, then 52MB hard drive. You could get one for the PC for only AU$400.

[ElPolloDiabl]

P.S. Can we drop the hardware calls argument. They are fine for dedicated arcade systems, but have no place on a modern computer. The bitching should be aimed at said operating systems.

I learned PEEK/POKE statements on the Vic 20. But I had no need to hit the hardware on the Amiga. Did I miss out on anything? Probably not.

[alexatkin]

I have been arguing that in certain realtime cases, Amiga hardware is superior.  I never said as a blanket statement that amiga hardware is superior modern PC hardware.  In real-time cases, you need to know best/worst case times.  So using API makes things worse since drivers/OS calls vary from system to system and you don't know the code for all the systems out there.  If it was direct to hardware, you can better estimate best/worst case scenarios.

Yeah, it's true-- floppy disks I tried today are lower quality than older ones.

And we have been arguing back, that its illogical to use that as an argument because the PC is not aiming for that market.

The point being, the fact that the PC cannot do things as real-time as the Amiga is not a failing of the PC so is not "playing catchup".  The PC has evolved to do what the "average user" wants it to do.  If you want to do something specialised, use specialised hardware - thats the point.

Also any argument about digital joysticks is pointless.  Yes most Amiga games used digital, because that was all you needed for the games of the time and it was cheaper to build a digital joystick.  Nobody is going to spend good money mass producing something which is not needed, the reason less used peripherals are always (and were always) more expensive than common ones.  But today, games are in 3D and you need the control of not just an analog joystick but TWO as well as analog triggers.  The argument for analog buttons (something Xbox had but Xbox 360 might not, but Playstation 2/3 does) is less so, its a personal preference and most people do not care.  (I am comparing games consoles as those pads are often used on PC by gamers or PC clones of them) It allows more precise controls for the games of today and is why I actually enjoy modern games more than I ever did Amiga ones.  I did not really become a gamer until analog joypads game out with 3D gaming, those old games just did not hold my interest for the most part.

There are exceptions, I used the love Deluxe Galaga and would play Warblade (its PC reinvention) if I could on a games console with seamless hiscore tables for my friends list.  But even that potentially would play better analog, as you have more precise control over the speed you are moving (although that is part of the skill of the game so it would remain a digital control).

Bottom line, the power of the Amiga joyport is not an advantage for games.  As others have said, there is no logic to reading what the user is doing any faster than the screen redraw rate because you need the user/game feedback.  If you measure what the user was doing with the joystick more than once between refreshes they most likely would end up doing something they did not intend in the game, because you can only react to what you SEE.  What use is measuring the user pushing right even 2 times per refresh when you cant move them accordingly on screen as it may cause them to hit something, because they couldnt SEE that they needed to let go after 1 sample.

[ElPolloDiabl]

Well amigaksi,

Using the philosophy, "If you can't beat em join them." Have you polled the Catweasel Mk IV? (the joystick ports) It should be a much fairer comparison.

[the_leander]

So why are there Linux kernels compiled with different schedulers readily available in most Linux repo's?

Because What works well in one situation (desktops) does not work so well in others (servers). When you support as much hardware, in so many different circumstances as linux, or BSD or even Windows for that matter, it makes sound sense to have a platform flexable enough to allow for such tuning.

[amigaksi]

You need to step back and look at the bigger picture. An algorithm written in asm for 68k will vary across the different 68k processor models just like it does across the various IA32 model processors. There are real time operating systems for both families but they're not called AmigaOS or Windows. Keep in mind that real time os's also make use of APIs to abstract from the hardware. Take a look at QNX for example. Where people demand the most performance, disk and graphics, you can't get away from abstraction. Nobody in their right mind would want apps writing directly to disk interface and people expect applications to play nice with the windowing environment.

These days people expect to be able to play World of Warcraft in a window and be able to see what's going on in yahoo instant messenger and occasionally check out what new messages are floating in on their facebook page.... on the same screen at the same time. You can't get this kind of functionality by allowing software to bang directly on the hardware.

You can still have APIs and go directly to the hardware.  I am not saying everything should be done over again by every application.  I use Amiga's API to set up things and then go directly to hardware where I need it.  I don't get that option for most things on PCs.

[amigaksi]

P.S. Can we drop the hardware calls argument. They are fine for dedicated arcade systems, but have no place on a modern computer. The bitching should be aimed at said operating systems.

I learned PEEK/POKE statements on the Vic 20. But I had no need to hit the hardware on the Amiga. Did I miss out on anything? Probably not.

PEEK/POKE would be slow although you can use them to go directly to hardware.  LDA/STA would be the method for 6502 based machines.  Here's an example for Atari 400:

MyDLI:      PHA
      Lda   #128
      Sta   205
      Sta   53771      ;start 8 pot counters (53760..53767)
NxtScanLine:    Lda   53764
      STA    WSYNC      ;54282 is wait for end of scanline
      STA    COLBK      ;53274 is Background color register
      Dec   205
      Bne   NxtScanLine
      PLA
      RTI

This one uses the analog joystick inputs on paddle #4 on the Atari as a scanline counter to show 128 colors (one per scanline).

[the_leander]

I never even argued point about ASM vs. high level languages with you before so stop lieing.

Actually you have repeatedly gone on about how APIs are so slow without providing any data whatsoever to back it up. You sir, are lying.

You are in a fairyland because you are inept to even comprehend what I am stating.  I can also write insults, but I prefer rationality.

Right... The hardware guys, the software guys, and everyone in between else have effectively stated in various degrees of detail where you are going wrong. You willfully ignore this and continue talking about something that isn't (your makebelief world of total hardware compatability) in order to shore up your argument. Anyone using logic would have long ago started to re-evaluate their position. That you haven't shows you to be nothing more then a fundie.

I already gave citation in this thread.

No, you gave a p-poor example that was shot down soundly by every other person on this thread for a variety of reasons.

You only reply many days later

Oh I'm sorry, some of us have a life outside of evangelising 20 year old hardware.

>You're bitching at someone else dismissing a highly specialised and none too often used tech?

I compared with both USB and Gameport.  Where have you been?

Which brings us back to the point that Gameport has been dead for the better part of 10 years now and unavailable to purchase via soundcards outside of ebay and similar for at least 5.

>Hypocrite much?

You are a biased side-kick.  He seems to understand the subject more than you so why not let him reply.

Hah! Biased? Probably, I have a low tollerance for demonstrably BS posts such as those you have been ejaculating.

That's not the point.

Yes it is, your original point was that the Amiga was better suited on the basis that it could react faster. You have since modified (I'm being kind here, slimed would be more accurate) your argument when it was shown that your original had no legs.


 

I said USB is slower than reading joystick on Amiga.  You are lost.

And you were shot down, you then added the nonsense about the gameport.

How many times are you going to keep repeating the same question and not reply to the responses given?  I don't forget that fast.

I have replied whenever asked and no one else has replied with a more technical answer. And to answer your question: I'll stop when you start backing up your argument with facts instead of baseless statements.

Good for you.  There's about a billion Gameports out there so it's still should be considered in the analysis along with USB.

Err, no. As has been pointed out (repeatedly), Gameport is dead, in the same way that ISA is dead. Or are you now going to go on about other random legacy I/O to prove the supposed superiority?

You can fantasize with your fairy tales all you want.

It was a reasonable point. That you can't see it only goes to back up my assersions that you're nothing but a fundie.

It's a fact that gameports and USB devices are both out there.  As I stated, people still sell joysticks based on gameport.  XP supported gameports as well and most people where I live still use XP.

By that same argument, so is ISA and I suspect if you looked hard enough, MFM and other wonderful ports of yesteryear. So again you've been thoroughly shot down on, well everything you've said so far, got anything else or are you done making a fool of yourself?



TL;DR

NO U

[amigaksi]

I guess you prefer to play flight simulators(eg. Falcon 4.0) on digital joysticks too? It's pointless to compare analog joysticks to digitals. They are meant to be used in totally diffirent purposes.

PS. Analog mouse is just too flimsy for Workbench use. I prefer Amiga+cursor key for everyday use.

My mouse sends digital data on Amiga and PC.  Digital joysticks are also easier to use unlike your keyboard example.  Many PC games I have seen source code to -- sample analog values and convert them to digital values.

[the_leander]

Tl;dr

no u

[ElPolloDiabl]

PMyDLI:      PHA
      Lda   #128
      Sta   205
      Sta   53771      ;start 8 pot counters (53760..53767)
NxtScanLine:    Lda   53764
      STA    WSYNC      ;54282 is wait for end of scanline
      STA    COLBK      ;53274 is Background color register
      Dec   205
      Bne   NxtScanLine
      PLA
      RTI

I'm sorry I actually understood that, I feel dirty.

[amigaksi]

Are you familiar at all with analog circuit design?

Yes.

[amigaksi]

Actually you have repeatedly gone on about how APIs are so slow without providing any data whatsoever to back it up. You sir, are lying.

...

If you think I need data to back up a logical statement that APIs are slower than going direct to hardware, you have a problem.

>Right... The hardware guys, the software guys, and everyone in between else have effectively stated in various degrees of detail where you are going wrong.

Give the link.

>No, you gave a p-poor example that was shot down soundly by every other person on this thread for a variety of reasons.

Wrong.  Some people sided with me.  But you haven't even replied to refutations handed to you so why would you read those.

>Hah! Biased? Probably, I have a low tollerance for demonstrably BS posts such as those you have been ejaculating.

You have no understanding of what I wrote so I think you are better off leaving it to others to reply and take your insults elsewhere.  Because insults don't seem to change the truth or affect me.

>Yes it is, your original point was that the Amiga was better suited on the basis that it could react faster. You have since modified (I'm being kind here, slimed would be more accurate) your argument when it was shown that your original had no legs.

Amiga is FASTER to react to joystick input.  I am sticking to the point.  Amiga can sample at 15Khz not just 1Khz.

>And you were shot down, you then added the nonsense about the gameport.

I originally started with gameport not added it later.

>I have replied whenever asked and no one else has replied with a more technical answer.

Go reply then.  All my arguments have a REALITY basis.

>Err, no. As has been pointed out (repeatedly), Gameport is dead, in the same way that ISA is dead. Or are you now going to go on about other random legacy I/O to prove the supposed superiority?

If you read your own comments in this thread, you will see that you have modified your views.  I can still use Gameport under Vista/XP and I never dismissed USB either which by the way can also be put in Amiga machines.


It was a reasonable point. That you can't see it only goes to back up my assersions that you're nothing but a fundie.



By that same argument, so is ISA and I suspect if you looked hard enough, MFM and other wonderful ports of yesteryear. So again you've been thoroughly shot down on, well everything you've said so far, got anything else or are you done making a fool of yourself?



TL;DR

NO U[/QUOTE]

[Wayne]

Is it about time to close this thread as pointless bickering over a subject which is clearly not true?