systemd bugs in F20/F21 -> bug against the distribution?
h.reindl at thelounge.net
Thu Apr 3 20:12:17 UTC 2014
Am 03.04.2014 22:04, schrieb Kevin Fenzi:
> Bad behavior in response to bad behavior just feeds a positive feedback
> cycle ( http://en.wikipedia.org/wiki/Positive_feedback ).
> The way to break out of it is for the person you control (namely YOU)
> to behave well. If others don't do so, things can calmly escalate and
> proper measures taken.
> So, be reasonable. Use logical arguments to convince maintainers that
> your bug is reasonable and your proposed solutions or patches are also
> reasonable. If you cannot do this, step back and let others do so.
what logical arguments do you have if the developer is living in his
own world where he is happy about every log message "his baby" writes
while at the same time that burries critical and relevant messages for
anybody else who cares?
you can come up with logical arguments in case of interested
developers but not in case of "we do what we want, break what
we want when we want to break it and everything falling left
and right is buggy and wrong" developers
> If the problems are persistent or highly impact-full, it's likely that
> higher levels of project management may get involved, but only in good
> time and after looking at the behavior of all the parties involved.
writing thousands of log-lines on a completly stripped down Rawhide
VM with only the default cronjobs *is highly impact-full* because
on production workloads and a central machine receiving all that
messages and store them in a database that is a DOS attack
if there would be a prefix or a specific binary thrwoing all that
messages you could filter them out in rsyslog.conf (by keep useless
overhead burried) but not the way it is implemented
"use journalctl and filter out" is pure ignorance
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 246 bytes
Desc: OpenPGP digital signature
More information about the devel