Top 3 best ideas in amiga history?

[Ilwrath]

Two I can't believe aren't mentioned yet;

ReTargettable Graphics/Audio : Sure, in most OS now you can pick a full screen resolution for a game.  Of course, the Amiga did it first, and took it further, also allowing you to pick which video card you use, and which rendering mode, and also being able to choose differing sound cards.

ARexx : A single scripting language that allows you to create interactions between any major software packages you happen to have.  If only modern PC, Mac or Linux applications got along this well, think of what we could do.......

[dannyp1]

1.  MultiTasking
2.  Custom Chips
3.  DataTypes / Libraries

[spookyx]

The  built in switch on the a3000 that let you change the video...
using SCSI (at least for a while)
AREXX

[Daedalus]

That´s true only in 31kHz. If you are using a 1084 for example, and you have a hires and a lores mode they are running at different pixel clocks. As long as you keep the sync signal at the same rate, you can change resolution without promoting modes

Yup, should've said that, seems only to be the case in native 15kHz modes. I'm pretty sure mode promotion was an option which had the effect described, and could be enabled or disabled using IControl prefs...

[tokyoracer]

1: The A500
2: The A1200
3: Custom Chips

[save2600]

Couldn't go with just three... lol

1) 2-button mouse (the original Amiga mouse is still my favorite)

2) DMA for the already fast SCSI

3) Kickstart in WCS or promoted to 32-bit via MMU and protected

4) Amiga RGB monitors having amplified speakers built in (especially stereo) and composite video input

5) classy case designs (especially A1000 & A3000)

6) "plug and play" architecture for the most part (until you start asking a lot out of your machine anyway)

7) Keyboard garage of A1000  

Innovative software from both Commodore and 3rd parties

9) Hardware and software support for the emulation of several other platforms

10) AmigaOS!   any flavor, take your pic  

[Colani1200]

Well most of the killer features / ideas have already been mentioned, but still...

1.) Turning the Amiga into a full computer (and not a games console as planned in the beginning)
2.) The invention of multimedia
3.) The icon system

[Karlos]

Two I can't believe aren't mentioned yet;

ReTargettable Graphics/Audio : Sure, in most OS now you can pick a full screen resolution for a game.  Of course, the Amiga did it first, and took it further, also allowing you to pick which video card you use, and which rendering mode, and also being able to choose differing sound cards.

These were certainly good but they were basically third party extensions that have come to be a part of most serious user's OS installs. Their very existence, however, exposes a weakness in the OS in that the concept should really have been thought of by the OS designers rather than hardware vendors and the like.

ARexx : A single scripting language that allows you to create interactions between any major software packages you happen to have.  If only modern PC, Mac or Linux applications got along this well, think of what we could do.......

Definitely a +1 for this, too.

Regarding linux, scripting is well served by shell script, Perl and more recently Python. Of course, what none of these have is the basic notion of "application hosts" which is what made (A)Rexx special on the Amiga, rather than the language itself.

I actually always thought Javascript (with importable application interfaces) would make a nice scripting system, but I am a curly brace whore.

[Tenacious]

1   Ram Disk:  and  RAD:
2   Small, extensible, near real-time, multitasking OS
3   The integration of the whole OS and it's very logical organization

[Karlos]

Actually, on my 256MB A1200, I use RAD as a SSD for the OS. On cold boot, a RAD is created of sufficient size and an image (basically just an archive file) is extracted into it, the Install command invoked to make the RAD bootable then the machine resets. The bootable HD partitions are set pretty low priority so that the machine (re)boots from RAD preferrentially.

[Tenacious]

I'm probably being thick.  What's a SSD?

[Karlos]

I'm probably being thick.  What's a SSD?

Solid State Drive.

[Tenacious]

OH!

That's clever.

I wish I could send a note back to myself in 1989 and alter the path I tokk upgrading my A500.  I was years without a HDD.  Getting 4 to 8 Megs of fastram early-on would have allowed a virtual HDD (RAD:) and a much improved experience.

[Karlos]

OH!

That's clever.

I wish I could send a note back to myself in 1989 and alter the path I tokk upgrading my A500.  I was years without a HDD.  Getting 4 to 8 Megs of fastram early-on would have allowed a virtual HDD (RAD:) and a much improved experience.

The main reason I did this was that it speeds up reboots and perhaps reduces some wear on the hard disk. With the best will in the world, there aren't many amiga 3.x apps that will need the 256MB of memory (though my coding experiments have managed to do it), so sacrificing ~32MB (in my case) to create a RAD for a full OS3.9 install with all the various essential additional libraries/drivers etc seemed a reasonable idea. I make a few minor modifications to the RAD bases startup sequence, removing things like Resident and LoadResource since they are pretty pointless for items that are already in RAM and are just going to be copied to another area of RAM.

It also means I can create versioned snapshots of my OS, since at any time I wish I can simply archive RAD to a file on the HD and use that snapshot at any time in the future.

Also, at cold start up I can choose from one of a number of such OS images to be extracted into the RAD, though I rarely use a different version.

[tone007]

That's, ahem, HELLA RAD!

[Boot_WB]

It also means I can create versioned snapshots of my OS, since at any time I wish I can simply archive RAD to a file on the HD and use that snapshot at any time in the future.

Also, at cold start up I can choose from one of a number of such OS images to be extracted into the RAD, though I rarely use a different version.

Nice! Similar in principle to suspend-to-disk, but with a lot more potential (and of course, no need to do messy things like mount/unmount journalling filesystems etc).