User: Password:
Subscribe / Log in / New account

All bugs are bad

All bugs are bad

Posted May 8, 2008 5:11 UTC (Thu) by jmorris42 (guest, #2203)
Parent article: Time to slow down?

> Others assert that, while it will never be possible to fix every
> reported bug, the bugs that really matter are being addressed.

This is a dangerous attitude.  Bugs are bad.  Reported but unfixed bugs have a nasty tendency
to have security implications.  Unreported bugs are the ones that should really frighten ya.

Eventually a policy needs to be enforced that only very well reviewed code goes in.  Obviously
that won't work for device drivers since they continually churn.

(Log in to post comments)

All bugs are bad

Posted May 9, 2008 7:36 UTC (Fri) by jzbiciak (subscriber, #5246) [Link]

Not all bug reports describe kernel bugs. Sometimes faulty hardware is the cause. So, while it might be fair to say "all bugs are bad," it's not fair to say "all bug reports represent an actual bug in the Linux kernel, and therefore must be fixed by changing the kernel."

Anyone who's owned a computer that's a poster child for the Signal 11 FAQ knows what I'm talking about. So does anyone who insists on buying ECC RAM.

All bugs are bad

Posted May 9, 2008 7:42 UTC (Fri) by jzbiciak (subscriber, #5246) [Link]

I should add, I did have a 486 system that was a poster child for the sig-11 FAQ, and I did
have a filesystem get hosed by a bitflip on a machine I built w/ 512MB RAM that wasn't ECC.
(The machine had nearly a year of uptime and not much load.  A bitflip on a disk buffer wasn't
actually all that unlikely.)

My current boxes all get ECC RAM now, and tons of extra cooling.

Copyright © 2017, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds