On Fri, Jun 19, 2009 at 5:46 PM, Pavel Roskin
<address@hidden> wrote:
On Wed, 2009-06-17 at 15:27 +0200, Vladimir 'phcoder' Serbinenko wrote:
> Hello. This is a fix for problem in subject line. I believe this fix
> to be correct and legitimate but as I don't know ruby well could
> someone review it? If nobody knows ruby in depth I think we can commit
> this even without deep review
I would prefer that we take a different approach with grub-emu. I think
the USB support should be enabled whenever libusb is enabled.
See my other patch for it
If libusb
is missing, then the USB related code should be omitted from the sources
using preprocessor directives. The build system doesn't need to know
anything about it.
This proposition would go to Marco Gerards - original author of this code. I personaly don't like putting a whole file in #ifdef but it's matter of taset (AFAIK there is no technical reason not to do so). Perhaps we can discuss it on IRC?
Generally, please think how to reduce complexity, not how to add to it.
Complex systems are harder to maintain.
This patch wouldn't increase complexity. It just moves a bit of complexity from ruby to (g)make by defining a variable.
--
Regards,
Pavel Roskin
_______________________________________________
Grub-devel mailing list
address@hidden
http://lists.gnu.org/mailman/listinfo/grub-devel