ImageDisk project is canceled
ard at p850ug1.demon.co.uk
Wed Dec 21 18:35:23 CST 2005
> > Without seeing the source code to Imagedisk, how do I know that it
> > actually follows that spec. I will assume Dave is not so malicious as to
> > XOR the sector data with pseudorandom bit stream or anything stupid 
> > but how do I know that there isn't a bug in the progrma that causes it to
> > fail if there are more than 27 sectors/track or something?
> Then you must be a better programmer than I, to be able to tell if a
I don't claim to be a programmer at all. But I will make this claim. It's
a lot easier to discover a bug in software if you have seen the sources.
I know for a fact it's a lot easier to see possible design bugs in
hardware if you have the schematics (because that I have done many times).
> 1. Do you have the source code to every operating system for every
> computer you own? I also know you have PCs, so what in the world
For the machines I _use_, yes I do. Even a couple of my calculators. No,
they are not open-source in the FSF sense, but the source code is available.
> did you do prior to Linux? Refuse to use MS|PC-DOS?
My views changed majorly once I started to run Minix (not open-source,
but source available) and more particularly linux. I found I had many
fewer problems because I could read the source to find the cause of an
error message (rather than just an entry in the manual), I found I could
modify and fix things. Sometimes the problem would be due to my somewhat
unconventional hardware (which I would not expect anyone to support other
than me). I could either fix the problem with the hardware (e.g. a port
bit was not doing the right thing in some odd circumstance) or I could
patch the program.
And having dscovered the advantages of being able to do that, I am not
going to go back, certainly not on a machine that I depend on.
> 2. (might as well ask, as long as I'm asking) I know you dislike
> the whole "board swap" mentality and that you prefer replacing
> blown chips over replacing a whole board. But why not repair those
> non-functioning chips? I mean, one blown gate and you toss the
I am not sure anyone can repair a chip. Are you suggesting I start moving
atoms around to correct the wrong doping levels, oxided metal traces, etc?
> *whole* chip? What's with that? 
> -spc (And you actually trust chips to do what they say?)
>  Not my footnote
>  Or in other words---the "components" in today's computers are bigger
> than yesterday's computers. Right now, the "components" are
> boards/cards and I can certainly see it being whole computers
Maybe to you, but it's not happened here yet.
More information about the cctalk