Dan Nicholson wrote:
> On Feb 17, 2008 6:16 AM, Alexander E. Patrakov <[EMAIL PROTECTED]> wrote:
>> Jeremy Henty wrote:
>>> On Sun, Feb 17, 2008 at 06:00:12PM +0500, Alexander E. Patrakov wrote:
>>>
>>>> This is because the new kernel defaults to UTF-8 on the console. You
>>>> need a new version of the "console" bootscript to drive it out of
>>>> this mode.
>>> OK, running "unicode_stop" fixes things. I guess just adding a call
>>> to "unicode_stop" to the "console" bootscript will do the job?
>> No, this will fix things only for the first virtual console. You really
>> should
>> get the script from LFS-6.2 or 6.3, and replace ${ECHO} with just "echo"
>> (even
>> Debian demands the ability to understand the "-n" switch in addition to POSIX
>> requirements, so the whole "does echo understand the needed switches" test in
>> the current bootscripts is stupid). For the record, in your case, it does
>> this:
>
> It's not about -n. That's a POSIX requirement. It's about -e and our
> use of color throughout the scripts. So, it's not stupid unless you
> like a lot of "-e" characters showing up on your screen instead of
> colors or other formatting.
Sorry for the delay with the reply, I was at the semiconductor conference.
Indeed, both of the switches are not in POSIX. The point is that Debian thinks
that this bit of POSIX (no requirement for -n even though all shells de-facto
support it) is stupid enough to be overriden by their official "policy"
document. What you say about the "-e" switch is, of course, right even for
Debian (they work around this limitation by having the literal ESC character in
the initscript).
--
Alexander E. Patrakov
--
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page
