"J.A. Magallon" <[EMAIL PROTECTED]> writes:

> ...what I do not understand is why you throw away all the kernel
> autoload of modules, that was invented to not have to do what you
> are doing.

show me the patch that throw it away

> > > alias char-major-81 bttv
> >
> > we may do this later. as for now, we fully support quite a number
> > of tv cards. i won't change something that seems "just an easy
> > trick" just before release.
> >
>
> What looks like an 'easy trick' is to preload everything.

read my leaps: we're ~10 week away of final release; qa has frozen
packages uploads and only bless pure BUG FIXES.
we won't do ESTHETIC changes or alter a working mechanism just
because we can do better.
in the long way, we'll do better. i don't say you're wrong: what i
say is this is the WRONG MOMENT.

> If you have the autoloader, you can use it. See
> scsi_hostadapter.

i know it, i've spent enought time to fight against kmod and devfsd
interaction :-(

> You could define a generic module in shipped /etc/modules.conf,
> something like

scsi_hostadapter isn't just a "generic module": it's what kernel try
to modprobe.

> alias radio-tv-card off
> alias char-major-81 radio-tv-card
>
> and then, at install time, detect the card and
>
> s/radio-tv-card off/radio-tv-card the_chosen_one

kernel doesn't try to load radio-tv-card unlike scsi_hostadapter.
it'll be very much simpler to do "alias alias char-major-81 <MODULE>"

> > for now, we only fix bugs. not enhance something that just works
> > smoothly.
> >
> > you see the point ?
> >
>
> Really, I don't. But 8.2 is just there.

i'll repeat me: software development is a cycle. you develop, you
test, you fix, you retest and you upload.
for the distro, it means that we spent several month breaking
it^h^h^h^h^h^h^h^h^h^h^h updating it;
then qa freeze because next release is 2-3 monthes away. we finish
featues currently developed, update carefully packages.
the goal is to stabilize the distro, test all packages and fix all
bugs found.
later, we enter a deeper freeze where no more development or savage
alteration is authorized.
at the very end, we only fixed critical bugs.

we're at the end. this is _NOT_ the moment to change working stuff
for the beauty of the art.


Reply via email to