Thorsten Kranzkowski said:
> On Tue, Feb 01, 2000 at 10:21:47PM +0000, Terry Dawson wrote:
> > I would guess that a large percentage of people on this list are still
> > running Linux 2.0.* based systems. Many of the Amateur Radio Linux
> > fraternity are Amateur Radio first, Linux second and they are less
> > inclined to routinely follow updates in code, especially operating
> 
> I would guess, that exactly these people don't install stuff themselves but
> rather eventually do an update with a new RedHat/Suse/Debian/whatever
> distribution. They will get a new Kernel _along_ with matching application
> software.

As someone who used to maintain the ax25-tools and the ax25-kernel for
one distribution, I cannot fully explain the NIGHTMARE something like 
this does to the distros.

Sure, sometimes you just have to do it, but make sure that this sort
of binding the tools to the kernel sort of thing is really needed and
worth the extra pain that it will cause.

Someone suggested a printk to the old API.  I believe this was done for
some socket code (and why listen used to print that kernel message).

> I think it's absolutely sensible, targetting for 2.4 or later, to intentionally
> cut support for older kernels.
> Look at the new ax25-tools: they _require_ glibc 2.1.x - because supporting
> older libs is ugly and older libs don't provide funcionality the tools need.
> If you already have glibc, then a new kernel is no problem!
Binding a tool to a library version is usually ok because people who 
compile their own libc are quite rare. But most people compile their own
kernel. The distribution's dependency systems can handle libraries.

Quite often the kernel falls out of the dependencies system so there
is no guarantee about what is installed, let alone running. That's when
wierdness starts to happen.

-- 
Craig Small VK2XLZ, PGP: AD 8D D8 63 6E BF C3 C7  47 41 B1 A2 1F 46 EC 90
Eye-Net Consulting http://www.eye-net.com.au/     <[EMAIL PROTECTED]>
MIEEE <[EMAIL PROTECTED]>              Debian developer <[EMAIL PROTECTED]>

Reply via email to