>The use of the LZO compression offers 40-74% cold startup improvements over the XZ compression.
I don't feel like going through past entries, but I'm curious what other encoders they've tried. I can't imagine them settling on XZ for any reason other than trying to minimize size on disk, but maybe I'm missing some context(s).
I'm looking for all sorts of REMOTE tech talent, but this week let's focus on Infrastructure + Ops.
Are you an AWS expert? Love writing terraform? Have strong opinions on observability and loging? Well, then we should talk!
I always love connecting with people over the nerd-networks I am on (Mastodon, IRC, Forums, etc.) - so even the chance to just chat with you all would be great!
Grab a slot! https://calendly.com/davidovmoz/moz-chat
Job postings and more info here: https://moz.com/about/jobs
🔄 Boost Me! 🔄
It's not just a new Noise Unit album today, there's the 20th anniversary edition of Assemblage 23's brilliant, heart-rending "Failure."
I have no idea what the rest of my team is going to think, but I'm kinda proud of some patches I pushed today. Now a script can better respond to being arbitrarily renamed and a config value no longer cares about the name at all.
Seems to work fine, but there's probably some tests that need updating.
Unrelated: if you're writing a Makefile and your recipe ends up being a bunch of calls to sub-make ( "$(MAKE) ..."), unless you're passing the -C flag, you're spawning additional processes for nothing and making your build system harder to trace. Declare those targets as pre-reqs and they can even be done in parallel if their recipes allow for it.
Why is every acme client significantly more complex than acme-client(1) from OpenBSD? I spent maybe 30min reading the manual for uacme(1) and it seems to suggest creating/using a wrapper script for it instead of calling the tool directly...
I just want to know what command to run so I can get a generated cert in place, why is that such a difficult use case to suss out?