exozyme/zine
7
1
Fork 0
Code Issues Pull requests 1 Actions Packages Projects Releases Wiki Activity

Added comments about "Floats are Weird" and ersei post about the new server
Some checks failed
ci/woodpecker/pr/woodpecker Pipeline failed
Details
ci/woodpecker/push/woodpecker Pipeline was successful
Details
ci/woodpecker/manual/woodpecker Pipeline was successful
Details
ci/woodpecker/deployment/woodpecker Pipeline was successful
Details

Browse source
This commit is contained in:
Cloudyy 2024-03-03 22:46:36 -03:00
parent 2bf327bae2
commit ce06e1b5cb
Signed by: cloudyy
GPG key ID: 035104A645BAEADD
1 changed files with 6 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

11
src/content/posts/2024/march-2024.md
View file

@ -10,21 +10,22 @@ description: 'The March 2024 issue of exozine'
We have a website, [exofeeds](https://feeds.exozy.me) with our members' RSS feeds merged into one giant megafeed, and it's awesome... except it doesn't work. The megafeed hadn't been updated since the end of September, which is suspiciously around the same time as our migration to a new SSD. Turns out that migration broke the hard link between the output of the feed merger script and the feed file being served by nginx. @ersei made the script output directly to the folder served by nginx, so we'll never run into this hard link problem ever again.
That was January. But something was still off. The feed was now eternally stuck at January, never changing again. This time, after countless minutes of debugging, @ersei finally found the culprit: The `WorkingDirectory` of the systemd service for the feed merger script was wrong! And once that was fixed... exofeeds was *still* broken!
That was January. But something was still off. The feed was now eternally stuck at January, never changing again. This time, after countless minutes of debugging, @ersei finally found the culprit: The `WorkingDirectory` of the systemd service for the feed merger script was wrong! And once that was fixed... exofeeds was _still_ broken!
What else could we blame? The script works, the systemd service works... what about the systemd timer? So systemd has a handy command `systemd-analyze calendar` for checking cron strings, and it told us that the timer's string `*-*-* *:*/30:00` was invalid! There's a typo! It should actually be `*-*-* *:00/30:00` or even shorter, `*:0/30`. And finally, finally, exofeeds is back, and this time it actually works!
## https://a.exozy.me/posts/floats-weird/
## Floats Are Weird
TK: Someone else write this I guess
Floating-point arithmetic is weird, but why? Why does using smaller numbers results in even more inaccurate results?
[In this post](https://a.exozy.me/posts/floats-weird/), @a goes a little bit deeper into explaining why this happens and what you can do to get more accurate results without sacrificing performance.
## Using PulseAudio as a key-value database (yes, seriously)
When you are holding data, everything looks like a database. @a [introduces PulseAudioDB](https://a.exozy.me/posts/pulseaudiodb/), a lost relative to MangoDB.
## https://ersei.net/en/blog/updates-2024-02
## Ersei server updates
TK: Someone else write this I guess
In this post @ersei tells us about their new server (an actual rack-mounted one!) and their long journey through BIOS problems and broken networks.
## Programmatic Caddy