Alois Steindl <[log in to unmask]> wrote:
...
>maybe the compiler writers could safe themselves and their users some
>trouble by publishing known problems with their product.
>
>Of course, I know it myself that nobody really wants to disclose his
>own mistakes.
In the old CFT days, Cray used to send out a complete accounting
of all known bugs, their priority, and an estimate of when they'd
be fixed. I used to try to publicize the highest priority bugs (bad
code that generated wrong but plausible results was always the
highest priority).
--
J. Giles
|