1. Go to Dave Haynie's old Amiga docs site: The Dave Haynie Archives

2. Download the single zip file with all the stuff. Direct link

3. Unzip it and go to the devcon93 directory.

There is a document in both pdf and tiff formats.
It's about the upcoming changes to Workbench for the AAA generation.
It's an historical curiosity that I had overlooked and there are some interesting things in there that I want to share:

Windows:
- They planned to revert to 1.x Depth gadgets for windows (two instead of one).
- Substitution of the Zoom gadget for an "Iconify" gadget.
- Resize windows from any corner, no "Resize" gadget needed.
- Opaque window movement.

Most of these things have come reality. They were very logical changes, I guess.

Mouse Pointer:
- Inclusion of standard imagery for different uses: (I-beam for text, crosshair for drawing, etc..).

Gadtools, BOOPSI gadgets:
- Lots of improvements. We had to wait until OS 3.5 or 3.9 to get most of them as default.

Some features have only seen the light of day in OS4.

Menus:
One of the most interesting parts:
- Pop-up style.
- They can be moved.
- They can be "nailed", which means they turn into a window and all its contents are
turned into gadgets adequately. (!)
- Amiga Menu (represented with the Amiga logo). A menu option that was to be shown in all screens and that contained user-configurable general tools.

Preferences:
- Lots of changes, like with GadTools and BOOPSI.

Programmability:
- A series of optimizations and additions to the API.

Tool Port:
- This one looked important, because it was about introducing a new IPC mechanism to unify all communication of an Amiga applications into a single protocol... Or something like that, there are no details in it.

Finally, there are some short remarks about improving on-line help, and a final note about the overhaul of Workbench. "The file requester will become a client of Workbench and will use its code to display file information".

I don't know if any of this is even important, I didn't notice it for years and I thought that for some of you is new. Anyway, is an interesting look into the past.

@xispo

Quote:

- Substitution of the Zoom gadget for an "Iconify" gadget.

I have probably never used the Zoom gadget. You never know what happens to your window when you press that button.

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

I'm using zoom gadget all the time. You can't iconify WB drawer windows without 3rd party patches.

_________________
Rock lobster bit me. My Workbench has always preferences. X1000 + AmigaOS4.1 FE
"Anyone can build a fast CPU. The trick is to build a fast system." -Seymour Cray

@Tomppeli

Quote:

without 3rd party patches

Exactly, a lot of these ideas were finally developed by third parties.
Even years after Commodore's demise, people was trying to follow their steps.

Quote:

I have probably never used the Zoom gadget. You never know what happens to your window when you press that button.

Much like the MacOSX "maximize" gadget... Guess the Amiga was ahead of its time here as well ;)

_________________
http://www.warpdesign.fr/

very interesting!!! wonder if the AAA based amiga would have had some networking tools

_________________

@xispo

I just have discovered here the complete collection of Documents of DevCon 93.

The one about multimedia is quite surprising. There is a lengthy description of a some kind of complicated system that allowed to control all types of capture devices in an unified fashion.

You would have a drawer called "Media Types", like the one called Data Types, but instead of file formats, it would store descriptions of devices like scanners, genlocks, frame grabbers or samplers. And then it allowed to connect these two concepts and even sinchronize their activity.

They really wanted to expand support in the OS for these kinds of things: motion video or audio capture and mixing. So you would need less especial apps for dealing with it and have more flexibility.

It looks like they never managed to code anything of this.

@xispo

Sounds interesting... a real pity that they never managed to pull it out.
But... could be used as a good guide of suggestions for new releases of AOS4.x

_________________
God created Paula so that Allister Brimble and Dave Whittaker could do music

Check my Amiga gameplays (ITA)!

@Tomppeli

OS4 tip: clicking the zoom gadget while holding down the shift key will now cause the window to be maximised to fit the screen nicely just under the screen titlebar.

@ErikBauer

Quote:

could be used as a good guide of suggestions for new releases of AOS4.x

It's bigger than that. They had a clear vision with the audiovisual thing. This was the good part of their strategy. It's sad that we can only remember the flaws of Commodore. I think we lack this focus. They built the entire system on the idea of doing only one thing. But they were excelent at it. It's better than being mediocre at many. It makes you standout.

I think that, in general, the new Amiga systems only appear to exist for the sake of existing. They don't look to have a mission. The motivation should be: "I want this computer to do this thing in a more effective way than any other". In our case "this thing" is audiovisual and multimedia related things.

First the motivation, then the technology. The computer is a tool.

@Tomppeli

Quote:

I'm using zoom gadget all the time. You can't iconify WB drawer windows without 3rd party patches.

It is probably useful for that. But visually you dont get a clue will it minimize or maximize window. Sometimes it does nothing... that is why I find it annoying mostly.

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

@xispo

Quote:

xispo wrote: @ErikBauer Quote: could be used as a good guide of suggestions for new releases of AOS4.x It's bigger than that. They had a clear vision with the audiovisual thing. This was the good part of their strategy. It's sad that we can only remember the flaws of Commodore. I think we lack this focus. They built the entire system on the idea of doing only one thing. But they were excelent at it. It's better than being mediocre at many. It makes you standout. I think that, in general, the new Amiga systems only appear to exist for the sake of existing. They don't look to have a mission. The motivation should be: "I want this computer to do this thing in a more effective way than any other". In our case "this thing" is audiovisual and multimedia related things. First the motivation, then the technology. The computer is a tool.

Exactly

@Bobsonsirjonny

Yep! Words of wistom these are...

_________________
God created Paula so that Allister Brimble and Dave Whittaker could do music

Check my Amiga gameplays (ITA)!

@xispo

Quote:

I think that, in general, the new Amiga systems only appear to exist for the sake of existing. They don't look to have a mission. The motivation should be: "I want this computer to do this thing in a more effective way than any other". In our case "this thing" is audiovisual and multimedia related things.

In case you haven't noticed, the documents quoted were from a time when things were more or less even. Nowadays, the rest of the world has zoomed by and AmigaOS had stayed the same for years. The time for a "mission" hasn't come around yet. For now, it's catching up work. You cannot allow yourself to be different for difference sake, which seems what you are promoting there.

For now, AmigaOS needs to catch up severely to even get remotely into an area where we can start at looking things better than the others. And that moment might never come. We're no longer in a time where Commodore has capital to back up ambitious development plans, and, well, let's say certain events in the last two years and the events that lead up to that haven't exactly helped the situation along.

There is no big money capital behind the current AmigaOS. The system itself must generate the income for its development. If it cannot do that, it will die. Period.

Now, I know what will come up next. "It should be available on x86 to get a wider audience". Yes I know that, but that is missing the point. In case nobody noticed, AmigaOS is a big-endian system, and porting to a little endian CPU will cause problems. LOTS of problems. It will require a LOT of work. Work that cannot be done without any backup capital. It will create a several-years gap.

Next thing that will come up is the usual "you should have done this before". Yeah, I know, looking back it's always oh so easy to say "you should have done xxx". I can only point to the aforementioned events again that lead up the current situation.

The current motivation to do anything is "Let's try to get to a level where competition with Windows, Linux and MacOS starts to get a realistic possibility". Anyone claiming that AmigaOS can do that now needs a serious reality check. I certainly don't. It doesn't matter if you try to do something better than others - they'll ask "do you have a Web browser and Flash" and you'll have to say "er... well... sort of".

Result: Not interested.

You should not forget that Commodore-Amiga ceased to exist. If you think they could have done better, then let me tell you they had their chance. Commodore had every chance to make Workbench 4. Problem is, reality overtook them. As such, any plans remain just that - plans.

Now, don't get me wrong. I have the greatest respect for Dave Haynie and I am not saying that it was his fault, nor that the ideas are bad once. They aren't. But we all know who was at fault for Commodore's demise. The matter of fact is that history has decided against Commodore-Amiga Workbench 4, and whether our current system is good or bad, or does have some other reason for existing other than existence itself, the alternate reality you quote is only alternate and exactly not reality.

Now, future talk. As I pointed out in the past, we do have a clear idea about that. I think I had hinted at Porter/Duff Image compositing as the foundation of future graphics quite early on already, way before 4.1 was even announced. That should give you a hint that we're not just making up features as we go. However, the situation today is radically different than it was in 1993. There's no world-wide cooperation backup up the development process, we don't hold devcons for the simple reason of economical impossibilities, and we'll have to do with two full-time developers and a whole bunch of volunteers. It's a miracle we got anywhere at all, and a testament to the enthusiasm and effort that everybody has invested into this project over the years.

As for the porter/duff stuff, I wasn't even sure it could be incorporated into Intuition. If you have never seen the source code of it, or Workbench, you cannot know what kind of system you are facing. Workbench in particular is extremely fragile. File requester using Workbench to display the contents is a good idea, but I highly doubt that the current workbench could be brought to do this. A rewrite would be required.

And there we are back at the beginning, a rewrite takes time. Would it be worth it? Absolutely. But with the resource limitations, it won't happen.

And I must be absolutely out of my mind to write all this crap... Back to work.

_________________
Seriously, if you want to contact me do not bother sending me a PM here. Write me a mail

It was theoretically possible at one time to have AOS widespread and cheaply available, but A Inc screwed that up through not being forward looking enough.

The main points about AOS were that it was small and efficient.. therefore the LAST place you would aim it for, is the desktop, against the superbloat and overwhelming lead of Windows, Mac etc.

If AOS had been coded for ARM, such as ARM7, ARM 9.. your looking at £20 Mobile phones, £40 DVD, Freeview boxes, and £60 nintendo DS, as well as a lot of other major embeded ARM soft cores in FPGA and other boards.

Instead, because of th family progression, PPC is the best overall version to code for as 68k is dropped, and alternate chip sets are far more complex to code for, and fortunately PPC is being fairly widely used in Non Desktop uses, such as consoles.

The Amiga was created as a Console, that happened to end up being a full computer.
AOS was created to run on a Console, that could deal with all the functions of a computer.

It wouldve been intresting to see Commodore displaying a Sega Megadrive running AOS back in the days, showing the software running faster. 8) (68k based machine)

If Amiga had been taken along a road similar to Microsoft, but supporting only a subset of specialist modular chipsets that came with their own ROM drivers for Zorro class Plug and Play, such as sound cards, graphics cards, IO cards etc, so that it moved to a pure backplane with minimum origional chipset, what then would we have been looking at?

RodTerl

Two neat things on Aminet I found.. DepthGadget displays a list of all open windows when clicking on a window depth gadget, and a list of all screens when clicking on the screen depth gadget. Very useful 8)

DriveIFF .. Bring back first person driving capabilitiy through the computers memory space, lets have screen grabs of the various areas people can find. Why not a 3D version so we can really Tron through the system, and possibly file system and networks. 8)

_________________
The older and more respected a scientist is, the longer it takes to prove him wrong.

Some images from the "Future directions for user interface" document:



Buttons with rounded corners in 1993!!



Pull down menus from an ancient age!



WTF! I remember that program called Magicmenu... They invented nothing!
And look at those three icons at the top.
The first is the "Amiga menu", described in my first post.
The second was a kind of "handle" that allowed you to move the menu.
The third one is the "nail". It turns the menu into a window.

@Rogue Quote:

Now, I know what will come up next. "It should be available on x86 to get a wider audience". Yes I know that, but that is missing the point. In case nobody noticed, AmigaOS is a big-endian system, and porting to a little endian CPU will cause problems. LOTS of problems. It will require a LOT of work. Work that cannot be done without any backup capital. It will create a several-years gap.

There are several reasons why going x86 would be a lot of work (e.g. need for PPC emulator & loss of 68k JIT emulator), but IMHO endian-ness is not one of them: You just have a special version of GCC (which already exists!) that makes x86 act big-endian from the software point-of-view (it should only have a minor impact on speed, and will be vastly outweighed by the faster CPUs & larger caches).

_________________

@ChrisH

Quote:

There are several reasons why going x86 would be a lot of work (e.g. need for PPC emulator & loss of 68k JIT emulator), but IMHO endian-ness is not one of them: You just have a special version of GCC (which already exists!) that makes x86 act big-endian from the software point-of-view (it should only have a minor impact on speed, and will be vastly outweighed by the faster CPUs & larger caches).

I doubt that would work. Yes, you can do that for quite a number of things, but for others, especially when Aliasing is involved, this will not work.

For example,someone might cast an uint32 * to an uint8 * somewhere down the line; that will cause all sorts of issues that cannot be handled at a compiler level.

I am not saying endian is the biggest problem, at least not for the OS itself. There are other issues however. For example (and this is a very generic issue which will also occur with PowerPC 64 bit architectures) all system structures are open and therefore cannot be changed without changing compatibility. Tag items are two longwords.

What I am trying to say is that at one point in the future, there will have to be a new API. One that does not expose system structures, one that is 64 bit clean. Then, and only then, does supporting x86 make sense.

_________________
Seriously, if you want to contact me do not bother sending me a PM here. Write me a mail

@Rogue

Quote:

"It should be available on x86 to get a wider audience"

x86 AOS means more users & we dont have to wait billion years for hardware..

_________________
http://aros-exec.org/

http://serk118.blogspot.com/

@serk118

x86 Sounds all nice and well, one really big problem is getting ALL the hardware available to be supported! Choosing something like SAM440 is far easier as less drivers need to be written and tested. If you'd port it to x86 I guess only a small choice of hardware should be defined as 'AmigaOS Compatible'. As the x86 hardware seems to come and go quicker than the tides you would be in a situation of 'no available hardware' before the OS could be launched.

So let SAM create a little momentum so that development on the current platform can continue.

_________________
Back home...