Reiter Software developed only one expansion, the WEDGE hard disk controller for Amiga A1000 and A500. A WEDGE for A2000 was planned, but due to the relatively cheap A2090 never released. Parts of the software were written by Bill Henning, though he was not an employee of RSI.

RSI planned to market the Wedge-Tech 'Easy Streamer', a VCR backup solution - but that expansion has also never reached the market.

---

Interview with Jim Brooks, the creator of the WEDGE

How did it come you developed the WEDGE?

To clarify things a bit, I joined the Vancouver Amiga Users Group when the 1000 came out.
After playing with the single floppy for a while, and being shocked at the cost from Commodore, I designed a cable to use standard PC floppy drives.

It was tedious to make the cables so I made a couple and then gave the design to the club. Someone later made a PCB which simplified the implementation.

Next I looked at hard drives. Seems like at the time, the SCSI controller was about $1000 US, and the drive the same totalling about $2000. With a the exchange rate, that seemed bad so I set out to design an alternate/more affordable solution.

I looked to the PC. 20 Megabytes for $500 CDN and a disk controller for about $100, plus a $50 case and the Wedge and you had a working system.

It was an interesting trip as many people said you can't do that, they don't work etc. I remained unconvinced.
The hardware was actually the easy part. Delving into the software internals of the Amiga - now that was a different story.
The Amiga only knew SCSI while the PC only knew well PC SCSI, a bastardized cutdown version of SCSI.
I only knew C while the Amiga knew C at a very high level. I had/have the RKM's but. they were written by the original creators in a way I couldn't fathom.
Somewhere along the way, I ran into a couple of fellows. One was Larry Phillips know as the Amiga Guru, and a second one Bill Henning a dedicated programmer.

Larry at first scoffed at the idea being a strong anti-PC kind of guy. As I worked to figure how to get the system working, I ran across a SCSI driver. This turned out to be my Rosetta Stone.

From it I was able to bridge the gap between the Amiga and the controller.
I remember the first time I was able to get the Amiga to recognize the drive. It was amazing, and at about 3AM. There was no one to tell about it.

I could hardly wait till the morning to talk to someone about it. I brought it to the next Amiga meeting and everyone was impressed. Strangely, later that week, Larry phoned and asked if he could build one. At this point I didn't have PCB's.

This is also where Bill Henning showed up. I had noticed that when reading/writing from the drive, the accesses were very slow.
Turns out the original disk format while very robust was very hard for hardware and software to deal with. A sector contained forward and backward links, and about 488 bytes of user data.

Separating the two took more time than was available when multiple sectors needed to be read. Since there was no interleaving used, the next sector was past, requiring one full revolution per sector as opposed the potential of 48? Sectors. Bill, Larry, and I talked about this and decided to add 16 sector cache. This worked very well at first but within a short time, the disk crashed. This is where a real flaw in the Amiga design showed up. On boot, if the disk was corrupt as happened above, the OS tried to fix the disk. This is real hard when the system caused the problem. Back to floppy booting.

Turned out the cache was not properly updated. The result was we had a disk system faster by far than the $2000 Commodore solution at a fraction of the price.

This is where Dave Allen came in - he was the sales force. The 1000 controller led the A500 controller for the 500 and later to the ROCTEC controller which included IDE, SCSI, and ram.

It was a real ride and if you happen to need a board, I just happen to have a bunch available free to a good home.

The one sour point for me was a company that came out with a clone of my product. Theirs included schematics, a blank PCB, and a boot disk and sold for $50US more than mine did.

The ironic part was ended up getting the bad rap as selling it for too much.

You mentioned a $1000 SCSI controller. Which controller was that? Possibly Tecmar, Xebec or Micro Forge?

At first, when I did the Wedge, there was the Commodore one. Period. That's partly why it was so expensive.
The sad part was that if you ran Deluxe paint, something that used the copper and blitter, and locking out the 68000, the drive would overrun the controller.
This made disk transfers a little slower than with a floppy. The wedge on the other hand had not such problem. It's amazing what three generic IC's and little thought will do compared with a board full of ASIC's. They (Commodore) had no mechanism to stop the SCSI device and so most of what was sent was lost. This caused the controller to have to start the process over and over and over...

The WEDGE is a ST-506 controller - is it (software-wise) similar to a SCSI controller?

Yes the Wedge is PC but the layer between the Amiga and the hardware, dealing with the IO request block data etc was in the SCSI driver.
Turns out all I had to do was take the SCSI block offset as a long, divide it by sectors*heads and out popped the cylinder and sectors/heads as 16bit ints. Another divide by sectors and I had heads, and starting sector. Stuff the values in the PC controller and push the go button.

You also mentioned a connection between the WEDGE and RocTec controller. This is a bit surprising - could you tell more about that?

ROCTEC was fun.
I used to go to COMDEX during the exciting times. Now it's all PC's but at the time, there were many new and different things.
One company making floppies for the Amiga was ROCTEC. At this point I considered the Wedge dead as the sales people had melted down leaving me with about 100 A500 cards they demanded I build. But that's another story. So I'm walking by the ROCTEC booth thinking about the Wedge and thought "wonder if they'd be interested?".

They were sort of. A case of be careful what you wish for. First they wanted a sample so I built up a box and sent it to them. Not exactly what they wanted. They wanted IDE, which I had started on but not finished.
Here's the be careful part. They also wanted SCSI, in fact they wanted two IDE ports, a SCSI port, and a couple of meg of DRAM along with Auto configure and auto-boot.
That Auto-boot is a dirty word. Commodore is a dirty word. I'm trying to figure out how to do it and some women from Commodore is reading from notes by another Commodore employee about patching here and patching there. I asked for something on paper and she replied no way. It's all copyrighted. So much for open-source. We did finally get the ROCTEC module working but Commodore in it's infinite wisdom discontinued the 500 and brought out a new unit with no expansion port. Not to mention ROCTEC was not amused. But I had fun. A guy from ROCTED stayed with my family for a month during development and then I got a free trip to Hong Kong for a month and money to boot to finish the SCSI and autoboot parts. I tried to sell them on the tape backup but they weren't interested and my software guy flaked out. Too bad cause the hardware worked.

In retrospect, I should have done things differently as there was tremendous interest blocked mostly by the small sales force, Dave and money, me.
It was cool coming into the bi-weekly meetings and seeing several Wedge systems busily copying discs (we got two discs full of PD stuff at every meeting thanks to Larry).

What a ride. Way more interesting than the current PC's.

Thanks for the story!