On 03/20/2018 08:31 PM, Theodore Y. Ts'o wrote:
There are also ways in which Unix is just simply
deficient.
I get the impression that you seem to think that I think that what Unix
was around the time of BSD 4.3 was sufficient. - I'm not making that
claim or even thinking it.
I'm talking about the spirit, as in "do one thing and do it well". Not
that e.g. syslog shall be these specific facilities and these specific
severrities.
I want people to understand what was done, why it was done, and to the
best of their ability, make an informed decision when changing from
history. - I'd really like people to be able to answer the question
"Why did you do <bla> differently than it was done in <blaBlaBla>. What
was wrong / lacking / needed to be improved from the old way."
As long as people 1) have answers to those questions and 2) can speak to
why they did what they did, then by all means, move forward with
something new to try.
I hear tell of people putting reverse proxies in containers in front of
web server containers so that they can have basic traffic counters,
which they can't get (for some unknown to me reason) from their web
server container. - Where if they had bothered to ask the network
people, there are very likely multiple ways to get said traffic
counters. Further, ways to do it without adding the additional
complexity (read: exposure) / latency of additional containers.
--
Grant. . . .
unix || die