The collected works of big_mark_h

Hi

I'm having a bit of clearout at the moment, and I've got a few bits and pieces to get rid of. Ages ago I bought two ex-Royal Navy deskside Onyxs to rebuild. I managed to get one up and running, but the other is being broken up for spares. It is basically complete except for having no power supply, no hard drives or sleds, no system controller board, and no display system (the display generator is dead and I put the RM4s into the working machine, though the geometry engine still works). It has two R4400 200MHz processors. The skins are in good condition, except for the front, which has a whole in it.

If you want any bits from it, let me know and we can sort out postage costs. If no one wants any of this stuff, it will be going into the bin in a couple of weeks. I am located on the east coast of England so postage to the US could be a bit pricey, but I can look into costs if anyone really wants the bits.

I also have the following parts to get rid of :

6 x 64Mb RAM kits (each kit has 4 x 16Mb SIMMS)

1 x MC3 memory boards (part no. 030-0604-106 rev. B)

1 x Black Box ethernet 10baseT transceiver model LE2040A - converts your 15 pin network port into a standard 10Mbps ethernet port

1 x O2 standard audio module. This is the one without video support

If you want some memory and an MC3 I will assemble and test them before I post them.

I've still got *all* of this stuff. Gues no one else is rebuilding Onyxs these days...

Which bits are you interested in?

Hi Joe

The purple door and skids are fine. The hole is in the black plastic door, through the air vents near the top. Happened in transit when I bought them

If you definitley want them, I'll look into shipping costs.

What version of IO4 do you have? You'll need the latest version (part no. ends in 107, I think) to run R10k procs successfully.

Damn! April fools day gets earlier every year....

This technique sounds like it might work, at least until you switch your brain on and think for a second. For example ...

without knowing too much you can identify (from the exterior) what the microprocessor calculation results are, just by looking at the time it takes to do the computation steps.

For the love of god... Thats a bit like saying "since it took 3 seconds for you to work out the result for (a*b)+c, the answer must be seven" Even if you know how long each step of a calculation took and can use this to somehow infer what those steps are, you still need to know the value of all variables before you can get the correct answer. Which means it will still take years for a supercomputer to crack any half decent crypto algorithms.....