Baslinux mailing list

[Day Brown <daybrown AT hypertech.net>]

Sindi Keesan wrote:
> 
> James suggests our computer with a hard disk problem was at the curb
> because of the bad hard disk.

> Maybe something is wrong with the timing and DOS and Windows don't care
> but linux does?
I have seen a motherboard which would not run DR-DOS or ROMDOS, but
would run FREEDOS & MS 6.22.
And I have seen it run some Linux distros, but not others. And of
course, there's the new release of Redhat expressly for the VIC C3 CPU,
which doesnt seem to be a problem with other distros.

You might try a lowlevel format on the drive. But put your finger on it
while it is running. If you hear it hum, it is not long for this world,
the bearings are going out. Course, your milage may vary. Nobody can
keep track of all the variations of design, or what internal hidden
corners were cut which dont show up until you try to change the OS or
the hardware configuration.

It really complicates things for developers like Steven. For instance,
he might be able to solve the backspace thing with an alias, or maybe a
bitmask down in the kernel to read the key before something else messes
with the keycode. But we'll never know what is mucking around with it in
the first place, and he prolly aint got the time to track down the
hidden cause of the bug. A situation we have all inherited from all the
programs that have been written. 

Carving code always has markers, break points, inserted in the code to
debug it. But as soon as it actually works, most of the time it gets
shipped, without anyone going back to remove all the useless breakpoints
and error traps. And sometimes, something else stumbles into this code
that should not still be there.