r2010

r2010

[Yoshinori K. Okuji]

At r2010, Bean added fake_bios_data, and this function calls grub_printf. Why 
not grub_dprintf? Was there any reason to print a message on the screen?

Regards,
Okuji

Re: r2010

[Bean]

On Sat, Apr 4, 2009 at 3:43 AM, Yoshinori K. Okuji <okuji@enbug.org> wrote:
> At r2010, Bean added fake_bios_data, and this function calls grub_printf. Why
> not grub_dprintf? Was there any reason to print a message on the screen?

Hi,

Well, I'm actually planning to move fake_bios_data to a separate
module loadbios, so that linux loader would be cleaner. The message is
for debug purpose, it could be replaced with grub_dprintf.

-- 
Bean

Re: r2010

[Yoshinori K. Okuji]

On Saturday 04 April 2009 04:56:14 Bean wrote:
> On Sat, Apr 4, 2009 at 3:43 AM, Yoshinori K. Okuji <okuji@enbug.org> wrote:
> > At r2010, Bean added fake_bios_data, and this function calls grub_printf.
> > Why not grub_dprintf? Was there any reason to print a message on the
> > screen?
>
> Hi,
>
> Well, I'm actually planning to move fake_bios_data to a separate
> module loadbios, so that linux loader would be cleaner.

I am afraid that you've got a disease of making more modules. This is a bad 
disease. I have seen some projects which failed in refactoring due to this 
disease (e.g. Zope3).

> The message is 
> for debug purpose, it could be replaced with grub_dprintf.

OK.

Regards,
Okuji

Re: r2010

[phcoder]

Yoshinori K. Okuji wrote:
> On Saturday 04 April 2009 04:56:14 Bean wrote:
>> On Sat, Apr 4, 2009 at 3:43 AM, Yoshinori K. Okuji <okuji@enbug.org> wrote:
>>> At r2010, Bean added fake_bios_data, and this function calls grub_printf.
>>> Why not grub_dprintf? Was there any reason to print a message on the
>>> screen?
>> Hi,
>>
>> Well, I'm actually planning to move fake_bios_data to a separate
>> module loadbios, so that linux loader would be cleaner.
> 
> I am afraid that you've got a disease of making more modules. This is a bad 
> disease. I have seen some projects which failed in refactoring due to this 
> disease (e.g. Zope3).

In this case I think that loadbios must be moved to a separate module. 
Not for microkernel or something. Just loadbios will also be used to 
boot multiboot kernels and perhaps bsd too. Also if EFI-related 
functions are moved to a separate module then i386/efi/linux.c can be 
merged to i386/linux.c (I'm also working on this)
-- 

Regards
Vladimir 'phcoder' Serbinenko