prologic

twtxt.net

Problems are Solved by Method\" 🇦🇺👨‍💻👨‍🦯🏹♔ 🏓⚯ 👨‍👩‍👧‍👧🛥 -- James Mills (operator of twtxt.net / creator of Yarn.social 🧶)

↳ In-reply-to » Me: “Can we increase the quality of ticket descriptions? Like, can we please have 2-3 sentences at the very top that are straight to the point? State your intention/goal/wish clearly?”

At some point the senior engineers, tech lead (me) and engineering manager all agreed that if we’re not going to bother reading the tickets our agents write sup, we may as well not have tickets haha 😂

⤋ Read More
↳ In-reply-to » Me: “Can we increase the quality of ticket descriptions? Like, can we please have 2-3 sentences at the very top that are straight to the point? State your intention/goal/wish clearly?”

@movq@www.uninformativ.de Mate, don’t laugh, but this exact thing happened at work today 🤣 Agent diagnosed a problem, got it right, wrote a ticket, OP went to the team to get help with the problem, that team said to upgrade, OP went ahead with upgrade, obviously didn’t fix the problem, someone asked “did you read the ticket? the solution is right there!” 🤣

⤋ Read More
↳ In-reply-to » Finally got around to publishing some of my Claude Code skills 🥳 interview, validate, incident-postmortem, humanize and a scam-report-au one for us Aussies 😅 https://git.mills.io/prologic/claude-skills

Install is two commands Btw:

/plugin marketplace add https://git.mills.io/prologic/claude-skills.git
/plugin install prologic@prologic

The rest stay private, they know way too much about my infra 🤣

⤋ Read More

👋 Heads up, my infra will likely go offline for a few hours on Wednesday Nov 18, 2026 between the hours of 00:00 and 06:00 AEST due to NBN™ Network upgrades that obviously I can’t avoid. I don’t have any backup/secondary network paths unfortunately. Not sure how long the outage will last. Sorry 😞

⤋ Read More
↳ In-reply-to » ‼️ At 16:32:54 UTC 13 Nov to 01:12 UTC 14 Nov the infamously known Mills DC suffered an ~8h40m long incident that impacted all Ingress to my infra as well as several other services such as cas.run and mbox.blue. The cause is Vultr Cloud deciding between the 8th Nov 2026 and 13th Nov 2026 to break their GET /v2/instances API whereby ?tags= filter no longer works. This API is marked as "deprecated" on their API docs website, but not removed. The trigger was my own software/services performing VM claned using this API call and filter that no longer works all of a sudden. More to come on this soon™ -- I'll have to write this up as a longer blog post I think 🤔 #VultrCloud #API

For those of you that are used/customers of casn.run:

cas.run was down from 13 November 16:32 UTC until 14 November 02:12 UTC, after the server it runs on was deleted by mistake. We restored it from the nightly backup taken on 12 November at 17:00 UTC, so anything you changed between then and the outage (a new token, service or setting) needs doing again. We’re sorry for the disruption. Paid accounts have been credited a month.

⤋ Read More

‼️ At 16:32:54 UTC 13 Nov to 01:12 UTC 14 Nov the infamously known Mills DC suffered an ~8h40m long incident that impacted all Ingress to my infra as well as several other services such as cas.run and mbox.blue. The cause is Vultr Cloud deciding between the 8th Nov 2026 and 13th Nov 2026 to break their GET /v2/instances API whereby ?tags= filter no longer works. This API is marked as “deprecated” on their API docs website, but not removed. The trigger was my own software/services performing VM claned using this API call and filter that no longer works all of a sudden. More to come on this soon™ – I’ll have to write this up as a longer blog post I think 🤔 #VultrCloud #API

⤋ Read More
↳ In-reply-to » This 👇 is fucking insane 😱 Media -- Even though Picket is blocking this stupid fucking traffic from yet-another-botnet, I have to wonder, surely there's a better way to handle this? LIke seriously wut da actual fuq?!

@itsericwoodward@itsericwoodward.com Not much honestly 😅 During the storms Picket sat at ~0.2 of a CPU core (peaks ~0.4), up from ~0.02 at idle, so ~10× but still tiny. Memory didn’t budge: ~650 MiB working set before, during and after. Picket’s doing its job; it’s the git forge behind it that would’ve fallen over.

⤋ Read More
↳ In-reply-to » This 👇 is fucking insane 😱 Media -- Even though Picket is blocking this stupid fucking traffic from yet-another-botnet, I have to wonder, surely there's a better way to handle this? LIke seriously wut da actual fuq?!

I am at this point very tempted to no longer use Gitea or any fancy Git hosting platform. Maybe just go fully static? 🤔 But hmmm CI/CD workflows is kind of nice. Hmmm 🧐

⤋ Read More

This 👇 is fucking insane 😱 – Even though Picket is blocking this stupid fucking traffic from yet-another-botnet, I have to wonder, surely there’s a better way to handle this? LIke seriously wut da actual fuq?!

⤋ Read More
↳ In-reply-to » Since building a new Edge service and capability I've decided to call "Picket", I'm now dropping about ~40% of incoming traffic to the floor in various ways. Whole botnets, search engines, crawlers and all sorts of unwanted traffic are just going splat 💥 -- Turns out measuring cohorts by fingerprinting and behaviour analysis from connection stats and logs alone is "enough"™ to block most unwanted traffic. Things like Anubis, using a PoW, which I tried, don't really work in my expeirence.

@shinyoukai@twtxt.net Yeah, I think you can get pretty far with a few careful lines of config. But what I have here is a bit more sophisticated. Picket fingerprints connecting clients and is table to work out for example what software and stack your client is, if it really is Chrome or you’re pretneding to be Chrome.

⤋ Read More
↳ In-reply-to » Since building a new Edge service and capability I've decided to call "Picket", I'm now dropping about ~40% of incoming traffic to the floor in various ways. Whole botnets, search engines, crawlers and all sorts of unwanted traffic are just going splat 💥 -- Turns out measuring cohorts by fingerprinting and behaviour analysis from connection stats and logs alone is "enough"™ to block most unwanted traffic. Things like Anubis, using a PoW, which I tried, don't really work in my expeirence.

Here’s a pretty graph 📈

⤋ Read More

Since building a new Edge service and capability I’ve decided to call “Picket”, I’m now dropping about ~40% of incoming traffic to the floor in various ways. Whole botnets, search engines, crawlers and all sorts of unwanted traffic are just going splat 💥 – Turns out measuring cohorts by fingerprinting and behaviour analysis from connection stats and logs alone is “enough”™ to block most unwanted traffic. Things like Anubis, using a PoW, which I tried, don’t really work in my expeirence.

⤋ Read More

👋 Just a friendly reminder that if I don’t actually see any “use in anger” of this little container service I call cas.run – Then I can’t actually see how it will behave, perform and whether or not something like this will even remotely work and is worth continuing ot build out 😅 Please 🙏 and thank you 🙇‍♂️

⤋ Read More