Testing status?

[Eric S. Raymond]

Hal Murray via devel <devel at ntpsec.org>:
> Would it help if we made a chart of the status of various features cross 
> OSes/distros?  How many different OSes/distros do we support?  Can it fit on 
> a page?  ...
> 
> How many different levels of testing are there?  I'm thinking of something 
> like 0-3, where 0 is unknown and 3 is in regular use.  (and -1 means known 
> broken :)
> 
> Is a box for individual features enough or do we have lots of interactions 
> such as the recent ntp.d vs early-droproot?

I don't think we have much in the way of OS-dependent features left - that's
an effect of writing to the POSIX/C99 standard.  Looking...OK, there's seccomp,
interface scanning, presence/absence of routing sockets, and whether or not there
is kernel PPS support.  And one strange Solaris error trap.  That looks like it.

I'd say 0 = unknown, 1 = tested, 2 = in regular field use.

I don't think we have a lot of interactions like that, but it's the kind of
thing that tends to be invisible until an error report lands.
-- 
		<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>

My work is funded by the Internet Civil Engineering Institute: https://icei.org
Please visit their site and donate: the civilization you save might be your own.