Re: Developers dodging bug reports if things aren't Linux.

[Jonathan de Boyne Pollard]

> Yes, that looks like the bug. Surprisingly I did remember my login
> there and added my 2 cents. I don't have much hope for their bug
> system. When I first installed Ubuntu 5 or so years ago, and it
> installed Grub into the MBR I was upset that unlike other
> distributions, it did not add BootManager to the grub/LILO menu even
> though it was the active partition. While I was knowledgeable enough
> to add BootManager to the grub menu, others probably aren't.
>
> They closed the bug as cosmetics.
>

You think that's bad? I recently filed a VirtualBox bug about a
firmware API that's wholly unusable when used as requested.
VirtualBox's firmware erroneously reports the PCI BIOS32 code/data
mappings that it needs in order to operate as zero length. This is a
simple error in one line of assembly language. It's easily fixable, and
it's obviously incorrect on its face, even if one doesn't read the PCI
BIOS Specification to see the interface contract explicitly spelled
out. I even provided a reasonable fix (based, in no small measure, upon
experience of what real firmwares on real machines request) right from
the get-go when opening the bug report. I then had to endure all sorts
of back and forth with a nitwit from Oracle who argued at length that
the specification was vague, quoting something that wasn't the
specification as evidence, tried to pull the "You're the only person to
have actually told us about this, so it isn't a bug." trick, and tried
to pull the "It's not Linux, so VirtualBox developers like me won't take
you seriously." trick.

It's VirtualBox bug #8244. Go and read and despair of the state of
VirtualBox bugfixing at Oracle Corporation.