svn commit: r257696 - in head: libexec/rbootd share/man/man9 sys/compat/svr4 sys/net sys/sys

[Gleb Smirnoff]

On Wed, Nov 06, 2013 at 10:54:59AM -0800, Adrian Chadd wrote:
A> Gleb,
A> 
A> I think you are confusing evolving and improving with tearing out legacy
A> APIs.
A> 
A> You can likely do the former without the latter.
A> 
A> So don't think you need to stop fixing things. Just don't conflate removing
A> old APIs with evolving things. They're not the same thing.

When you rewrite some code, that has a very large percentage of compat APIs,
you must rewrite more code, handle all compats, and test them all to comply.
This can double amount of job needed or even more than double.

When it comes to really ancient APIs like COMPAT_43, it is very difficult
to find motivation to continue working on all this stuff.

4.3BSD is already 22 years in the past. I suggest to drop it in 11.0-R,
which would probably happen in 2015, meaning close to 25 years. What
commercial product ever offered such long compatibilty? And we aren't
commercial, we are enthusiast project. Do you know any person who finds
fun in maintaining compatibility layers?

-- 
Totus tuus, Glebius.