comp.os.os2.misc,comp.os.os2.beta
RRXuF8ysuwvpMn7w";
er.org"; posting-account="U2FsdGVkX19PS0+OBNxfAfhhTNEXON5u"
Gecko/20101207 Thunderbird/3.1.7
mp.os.os2.programmer.misc:1886 comp.os.os2.misc:2832 comp.os.os2.beta:12
From: Peter Flass
On 2/1/2011 5:51 PM, Jonathan de Boyne Pollard wrote:
>> This is VirtualBox 2.0.4 btw but I don't think it changes - still
>> works on 4.0.0 which is the latest that I've tried (USB is another
>> matter which is why I'm still on 2.0.4).
>>
>
> Versions are important. (-:
>
> I've tracked down the source for the ROM image that VirtualBox uses. It
> doesn't have, as I surmised, Plug-and-Play support. Interestingly,
> there's a common ancestry here with the ROM image used by Bochs, from an
> earlier version of which the VirtualBox source was taken, and which
> (since then) gained skeleton Plug-and-Play support somewhere in between
> version 2.4 and version 2.4.5. VirtualBox hasn't kept track with the
> Bochs improvements, it seems.
>
> What Bochs has isn't enough for real world use, however, so even if
> VirtualBox caught up with Bochs it wouldn't be enough. It only
> implements the version check. All other functions fail with an error.
> This is pretty useless. So I suppose that we're lucky that no-one with
> Bochs 2.4.5 spoke up. (-:
>
> This is exceedingly annoying, because these virtual machines don't
> operate like real machines in this regard. (Real machines have had fully
> implemented Plug-and-Play firmware support as standard since the late
> 1990s.) But its one more datum supporting the point that I've long made
> that x86 virtual machines don't set out to exactly duplicate real hardware.
>
Why would you need P-n-P? It's virtual after all, and you can't attach
a virtual display on the fly while the VM is running.
--- Internet Rex 2.31
* Origin: A noiseless patient Spider (1:261/20.999)
|
