Dusty @ChlorideCull@fuzzy.systems

This would be a once a calendar month ping (controlled client side) with the project name, project version, OS/Distro, and OS/Distro major version

The server, upon receipt, stores those four fields, and discards the rest

i've been meaning to implement telemetry in pretty much all my software, because

* I have more projects than time
* I want to know what platforms I should prioritize testing and supporting on
* Seeing my stuff get used gives me a fuzzy feeling

ah, yes, greed, the leading reason to add telemetry to software

to be fair I have plenty of reasons to be pissed at the FOSS community

things that started pissing me off once I started learning how legal systems work in general:

- ace attorney
- the FOSS community

and a notebook is easier to bring with you than an airgapped system

basically, my journal needs to be more secure and secret than my SSH private key

"Why don't you just write on your phone or something"

a journal is for private venting and reflection, a place without lies or facades, they are poison to your reputation and can ruin friendships because of something you thought in the moment and regret

I've started keeping a journal

like, hastily scrawled in a real lined notebook, like in the old times

today i found out that the person driving the car is part of the deformation model in wreckfest

treat yourself to a hairdryer if you have any hair whatsoever on your body

seriously

omg

this is so nice

Kinda wish Mastodon didn't just rely on sidekiq automatically retrying after a certain time, but instead had per-instance queues it would drain at a decently high rate whenever it wouldn't return errors

The issue with automatic retries after delay is that I will be stuck getting two day old posts mixed with new ones for about two days

it's really odd when your instance has been down for 48 hours

my timeline is currently a mix of posts that are between 2 days and 2 minutes old

toot

sidekiq queue zero is infinitely better than inbox zero