[arch-general] SystemD poll

[Kevin Chadwick]

> My friends at Red Hat inform me there is little marked improvement with SystemD however "It would be jolly nice if we was all the same." so I'm slightly mystified at the vehement determination to adopt it?

It would be very nice but in fact whilst unifying some it's current over
spec'd design can only cause more fragmentation seemingly with debian
attempting to hold it together with a big dollop of permanently wet
cement.

I guess there should be a consensus drawn up on init script location
and format that as many as possible unixes and init systems can agree on
but that will be a very difficult task that bodies like FSF have
struggled to do and have done badly in the past, though they seem to be
getting a little better at hunting out input on mailing lists etc..


If it happens let's just hope it doesn't end up anything like the
DNSSEC mess concocted with millions of pounds by individuals ignoring
the bad press for the reason along the lines of 'we need to get it out'
and then made almost pointless and more dangerous than it needs to be
due to a bad designs implementation problems demanding so.


This latest review was fired across the OpenBSD list recently. 

http://cr.yp.to/talks/2012.06.04/slides.pdf

Would we have DNSCurve without DNSSEC, will DNSSEC actually ever get
fixed having got it out sooner to do so or would it have died and not
been replaced. Would we have DNSSEC with ECC already, solving a large
chunk of the issues. Perhaps pertinent questions for Linux init?

-- 
_______________________________________________________________________

'Write programs that do one thing and do it well. Write programs to work
together. Write programs to handle text streams, because that is a
universal interface'

(Doug McIlroy)
_______________________________________________________________________