[TUHS] Version 256 of systemd boasts '42% less Unix philosophy' • The Register

[Michael Kjörling]

On 14 Jun 2024 06:26 +1000, from dave at horsfall.org (Dave Horsfall):
>> Good sysadmins live & die by grep and being able to visually detect 
>> departures from the norm by just looking at the “shape” of logs 
>> scrolling down a screen (before), terminal window now.
> 
> Which is exactly what I do: one window with "tail -F /var/log/maillog" and 
> another with "tail -F /var/log/httpd-access.log"; I've spotted lots of 
> attacks that way (followed by a quick update to my firewall).

journalctl -f -u 'postfix*'

or

journalctl -f -u 'exim*'

or

journalctl -f -u 'smtpd'

or whatever else might map to the SMTP server software you're running.
(Sure, it gets slightly more complicated if you don't know what SMTP
server software is in use, but in that case I think a case can be made
for why do you even care about its logs?)

To filter, one can either add -g 'pcre-expression'; or pipe the output
through grep -P for the same effect. Or you can use something like
--output=json (or -o json) and pipe the output of that through, say, jq.

And I'm pretty sure most web servers still log to text files in
typical configurations, so that plain "tail -F" should still work there.

Is systemd perfect? Of course not. I have my gripes with it myself,
but not enough to actively fight it. And nobody is _forcing_ anyone to
use systemd; plenty of examples have already been posted in this
thread, from specially-made Linux distribution derivatives to ones
that have opted to not include systemd to *BSDs to links to
instructions for how to get rid of systemd from more mainstream Linux
distributions that have opted to use it as a default.

Also, the subjectular headline from The Register seems like something
someone has dreamed up; I certainly don't see anything like that in
the actual release announcement at
<https://github.com/systemd/systemd/releases/tag/v256>. Also using
multiple different search engines to try to find it only brought up
the _The Register_ article and a handful of places regurgitating that
quote as a real representation of a statement from the systemd
maintainers. I don't see anything resembling it anywhere on either
systemd.io or github.com. Until I see someone posting a link to
something like that quote posted by a systemd maintainer in
representation of _any_ systemd release, let alone v256, I'm going to
treat that one as hearsay at best, and actively malicious at worst.

As much as I can appreciate the architectural simplicity of early
UNIX, how about not ignoring the fact that today's systems are quite a
bit more complex both at the hardware and the software level than they
were in the late 1960s, and that to some extent, this complexity
_itself_ (unfortunately) drives complexity in other areas. Also that
much of that simplicity was also out of necessity. There's a reason
why most software these days isn't written directly in assembler or
even C. None of which negates the accomplishments of either UNIX or C.

-- 
Michael Kjörling                     🔗 https://michael.kjorling.se
“Remember when, on the Internet, nobody cared that you were a dog?”