Amiga vs PC

[Karlos]

Yeah, but one graphics card hits one register while another one hits another-- inconsistent and non-standard.  You have NO IDEA what you are talking about.  First understand and then reply.

Well, you are a fine one to talk

You're still thinking of graphics cards as they were a decade ago. They have totally and utterly changed. I guarantee you wouldn't know where to begin trying to bang modern GPU hardware if you are still thinking in terms of direct register access, at least if you want to use any modern features.

Drivers exist for a  good reason. My graphics card, for example, has literally thousands of GP registers (as well as "constant" memory) that are shared between 240 stream processors on demand by a hardware many-thread scheduling engine. Even when writing code for the GPU using CUDA, the object code that NVCC produces is not specific to that GPU. Instead, it's a bytecode that is JIT translated at runtime for the hardware.

And, FYI, the VESA compatible registers haven't really changed much, which is why it's almost always possible to use a modern card without any specific drivers. As long as "use" means open a basic frame buffer. If you want to bang those registers, you'll probably get away with it just fine.

[Karlos]

I am not against drivers but it's better to have both options-- being able to go direct to hardware as well as driver or API interface. I would think you already know some of the great stuff written using direct to hardware method on old PCs and Amiga. And it's a superior interface to have the standardized hardware.

It was fun, banging the metal. However, you cannot deny that many 68000 OCS/ECS titles that did it, mysteriously stopped working on 68020/AGA and point blank refused to do so without degraders and so on. Ultimately this is where blindly depending on hardware configuration X gets you. Hardware changes, even on the Amiga.

The driver model exists not just to ensure that applications have a consistent API to hardware but to allow hardware vendors to radically change their internal hardware. This has to be done, if you want to improve performance. If graphics card manufacturers stuck to using a fixed IO/commandset that any old hacker could bang away at, there's no way we'd ever have migrated from old fixed function graphics pipelines to modern fully-programmable stream-processor machines.

[Karlos]

The fact that there are OCS/ECS demos/games/etc. that do work on AGA proves my point.

And the fact there are many that don't equally proves otherwise.

[Karlos]

I haven't found one yet, that without a bit of hacking, or using degraders or WHDload slaves that don't work with AGA.

That's the point though. Without hacking or degrading, such examples just don't work, which is the only proof you need that hacking hardware directly is not a good idea if compatibility is important to you.

[Karlos]

No, it doesn't.  That it's do-able is the point.  You can directly go to the hardware and still have it work across the board on all Amigas.  There's a lot of API-based software that no longer works on modern PCs.  See my other post.

By your own (flawed) reasoning the fact that API-based software still does work on modern PC's demonstrates that it is doable, and thus proves that API based systems are the way forward.

You're trying to have your cake and eat it. Hardware banging, in the modern age, is for embedded projects and the like. it has _no_ place whatsoever on modern desktop machines except for the implementation of hardware drivers.

Recent GPU's have billions of transistors crammed onto their dies dedicated to the job at hand. How far do you think they would actually have gotten if they had to waste precious silicon to appease people that insist on total hardware backwards compatibility with 20-year old designs? Apply the same observation to every other component in your machine.

As for not being able to run old software thanks to hardware changes, I take it you forgot about features like hardware virtualisation or even emulation? After all, there's not much you could do in DOS on a 286 that you couldn't comfortably emulate today if you really wanted to.

Hardware banging and ignoring the OS was never truly encouraged on the Amiga (at least by the people that designed the hardware and software), you are lucky that anything worked from ECS to AGA. And, given that post AGA was set to go in wildly different directions, you'd be even luckier if anything worked beyond that.

You can assume all sorts of utterly ridiculous nonsense if, as an application developer, you view your machine as hardware up rather than software down. For example, the VBR on the 68000 was at a fixed location starting at address 0. From your standpoint, it's t thus perfectly fine when taking over the machine to put your own handlers in that first 1K of memory. And some obviously did, which is why when 68010+ based systems started appearing and the OS got the opportunity to relocate the VBR to somewhere helpful, like fast ram, their code failed miserably. And that's just from "hardware banging" the CPU, let alone assuming other facets of the system architecture would never change.

Frankly I'm glad that not too many people thought your way about development or there's no way we'd ever be using 68060, RTG or AHI. You might find it comfortable in 68000/OCS-only land, but don't assume everybody else did.

[Karlos]

Bullshit.

Is that your qualified opinion? Browsing on my old A1200 with 040/16MB fast (when the 040 card works, but that's a different story) on AGA is not a lot of fun, either. It starts off fine but sooner or later (depending on how many image intensive pages I've visited), I'm down to my last MB of memory and my 040 feels more like an 020 until I either flush the images or turn them off. That's with a minimal installation of OS3.5 and the usual 040 speedup patches installed (RemApollo, 040 ieee math libraries etc).

It's not just a case of misremembering either. This behaviour was observed recently, during the development of the old browser proxy for iBrowse/Aweb.

[Karlos]

They work without hacking or degrading.  You are mixing up OCS/ECS/AGA registers with OS calls and memory differences.  The AGA chipset was purposely made to be backward compatible and register compatible.  Have you narrowed it down to the registers-- no.  You just assumed it.

On software that shuts down the OS and takes over the hardware, you can only blame the OS for instability so far.

The point is that making assumptions about the hardware configuration is not sensible. To claim otherwise is to ignore the advice of the people that actually designed and built it and I'm sure they were better qualified than you to make that call.

[Karlos]

When?  In the Win 3.1 days or today?  



Aah so its today. You want to view the bloated mess thats the web today on a 40 mhz cpu with 32 Mb ram.  Sheesh.

It was 16MB and 25MHz. So I guess you feel even more let off the hook. FYI, the proxied version of amiga.org has considerably less "bloat", especially while it was in early development, so comparable to the good old days. The first version was somewhat overzealous in what it stripped out, the resulting pages could hardly be described as bloated.

FYI  I used the same spec A1200 in 64 colors dblscan(fblit and Ftext) till about 2002 with Aweb/Ibrowse.  Sure over the years, it got slower as web pages became bloated with more and more useless banners and images.  Being on dial up didn't help matters so i switched off images.  I also ran netscape 4 communicator under shapeshifter with Mac SO 7.5.5 and the savage driver in 640x480 256 colors (the screen was in fast ram mapped with the MMU). The Aweb/Ibrowse set up was faster. The speed  advantage of YAM over communicator was even more stark.

Eventually i went to the A4000/68060/CV64 running Ibrowse 2, on dial up till 2007 or so, and never experienced this BULLSHIT about worrying if the next web page would crash the system!

I'm also guessing your 060 had more than 16MB?

With some of posts some "Amigans" here make, you'd think their machines did nothing but crashed at the first single mouse pointer movement after booting.  All the software on aminet must have been created with amiga's in an alternate reality, as according to Leander, the Amiga was one useless, perpetually crashing mess.  For him NOTHING worked.

That's not true, his needs simply exceeded what his machine was capable of. It happens to most people.

[Karlos]

Be that as it may, I was viewing this site's pages under a different user name from about 1998.  I never experienced this event that you and Leander describe.  NEVER.  Sounds like the lack the of protected memory argument: yeah you could crash your amiga, yeah you could lose data, yet gigabytes of software, pics, songs were created by users anyway.

The experience I described was of system slowness, not of system instability. My old 040 machine was unstable for a number of reasons but not because of web browsers using too much memory.

As for the protected memory issue, well, over the years, I had applications crash my entire machine many more times than I care to remember. It's just something you learn to live with, until you get used to a system that doesn't require a reboot because something has corrupted Exec's free memory list resulting in amber alerts every time an application calls AllocMem() of FreeMem() or trashes some other resource that ultimately causes a crash or freeze.

You can keep saying "never" in your experience, but obviously you are exceptionally lucky. I don't know any other users that have "never" experienced any of these problems.

[Karlos]

I love that poster, thanks

At least we'd be comparing reasonably comparable technology.

People talking about IE4 or 5 and comparing it to the amiga are doing the usual "lets take stuff made years later and compare it to a computer from 1992"


For edification, this http://en.wikipedia.org/wiki/File:Internet_Explorer_2.PNG is IE2, which came out in November 1995.
The original version came out a bit earlier that year, according to wikipedia.

http://www.youtube.com/watch?v=TZixhLSooH8
this gentleman has a video of browsing the web on windows 3.1, using a 486 though using a later browser maybe ?

Ahem, iBrowse was released around 1996. If you want to compare things properly (in a timeframe when amiga machines were still relatively new and on sale), then you can compare AMosaic (released around the end of 93) with either Mosaic itself or Netscape which appeared around the last quarter of 94.

[Karlos]

I actually used Shapeshifter to run MSOffice when at university. Those pesky professors and their insistence on MS word documents and the like.

It was actually pretty usable, especially under RTG.

[Karlos]

I've been meaning to get a wireless PCMCIA card for my 1200, and I figured Lynx would be the best way to go

Still a bit daunted at the task of setting up TCP/IP and all that

I seem to recall Patrik and Fx put up a no-nonsense guide for configuring AmiTCP somewhere, but I can't seem to find it atm.

Anyway, the trick is to tell it you have an A2065 network card when installing it. After installation, edit the file AmiTCP:db/interfaces file and change the line that reads "a2065.device" to "cnet.device" (assuming that's what you are using for your NIC).

Job done.

[Karlos]

No, _this_ is an example of "top posting".

First of all, you might have told me what JFGI stands for (had to google it to find out)

Once I found out what that meant I then googled Top Posting and the answer was...

Definitions of Top posting on the Web:

When a message is replied to in e-mail, Internet forums, or Usenet, the original can often be included, or "quoted", in a variety of different posting styles.

http://www.google.co.uk/search?hl=en&defl=en&q=define:Top+posting&sa=X&ei=UZdlTOG-MsfQ4wbu6YzLCg&ved=0CBcQkAE

So by this definition, then using all those quotes makes you a Top Poster...

Glad that's cleared up now... :biglaugh:

(Whereas _this_ is bottom posting)

[Karlos]

^ what the_leander uses is interleaved or inline posting. It's also what I use, unless I'm replying to something very short.

Quoting a long message then shoving an equally long reply above or below it is just tl;dr fodder.

[Karlos]

the PC is still using the same design architecture as the 286.

Hmm, let's see. Current x86-64 architecture and supporting components:

16-bit ALU? Nope.
External FPU? Nope.
Segmented memory? Nope.
ISA slots? Nope.
Mono/CGA/EGA video? Nope.

In my previous job, none of the server machines bought in the 4 years I was there had a floppy drive, nor did most of the already installed ones. USB or CD only. Most had their OS installed via a bootimage on a USB or CD and used a network based install procedure.