Daniel Hartmeier wrote:
> Then, people will not agree whether 10:20 or 10><20 is prettier, and
> further variations will be suggested. 10:20 is actually 9><21.
> 10<>20 would be odd at best for redirections. If you replace a:b with
> a><b in the from/to port part of the rdr rule, what syntax do you use
> for the replacement port (the one after the arrow)? 'a><b -> a:*'
> seems inconsistent, too. So, a><*? There's more stuff like that
> around the corner.
>
> The code change within pf is trivial, reaching a strong consensus
> probably not. It's hardly a severe problem, so the energy might be
> better spent elsewhere. ;)

Sorry, it's early, I am on my first cup of coffee and about to drive miles
to go change an in-the-process-of-breaking scsi root disk on a Saturday when
I should be in bed...

Why can't we support both formats ? From the quick (as excused above) glance
at pf.conf(5), the rdr syntax is only "port x:y" while is it block/pass that
has the complex backward compatibility. Or it the problem that you can't put
the "old" ranges into macros ?

OTOH What would be wrong with allowing the "x:y" syntax *additionally* in
the pass/block rules ?

(Yawn...)
--
Peter

Reply via email to