mec> I think you just have to call word_eval here.
regis> What is word_eval?
It's my function in new-config that expands $FOO constructions.
> if [ "$CONFIG_MCA" = "y" ]; then
> tristate 'SMC Ultra MCA support' CONFIG_ULTRA
> else
> tristate 'SMC Ultra support' CONFIG_ULTRA
> fi
>
> I see this as a variable redefinition. What is your opinion about it?
I don't have a problem with this. Perhaps I just don't see why it would
be a problem.
mec> I do. A configuration tool is less useful if the user has to install
mec> something else first.
regis> Do you consider that a X-server plus TCL plus TK is not
regis> "something else"?
That's right. xconfig is already part of the 2.1.XX product, so its
requirements are already part of what people need to do gui-based
configuration.
I believe very strongly in compatibility at the user level. Someone who
types 'make xconfig' to configure his kernel today should be able to
download the new mainline kernel, type 'make xconfig', and it still works,
with zero no software. I am not willing to break this lightly.
> Agreed. We can have both of them. But the problem then is to decide
> which back-end must be included into the mainstream distribution.
Both of them, of course, under different names.
> Perhaps Linus will refuse to have too much back-ends.
If it's ok with us, I think it will be ok with Linus. We are talking
about a few hundred lines of code here for an xconfig back-end; code
that doesn't even get compiled unless the user asks for it.
Regards,
Michael Chastain
<mailto:mec@shout.net>
"love without fear"
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu