You’re going to have to prioritise.

Find changes that:

Save a decent amount of money Are low risk Don’t take too long to do Can be easily backed-out of

Yup. In PieFed you can get an alert based on comments on a post, comments in reply to a comment, posts in a community, posts in a topic, posts in a feed, or posts done by a particular person. We’re really gone to town on it.

Increasing amounts of code running on my computer and in the online services I use will be written by generative AI.

Emphasis added by me.

Thing is, it’s not black and white most of the time - usually a developer is using Gen AI as an assistant in some capacity. There are a wide range of ways to do that with really big differences in how firmly their hand remains on the wheel of where things are going. Only in the most extreme “vibe coding” scenario would it be fair to characterize the code as “written by AI”.

There reaches a point somewhere on the spectrum of dependency on AI where quality would suffer and developer capacity-building would be stunted. Where that point is, is a more productive question than a binary Yes or No to all AI.

I heard you like spreadsheets so I put a spreadsheet in your game so you can spreadsheet while you game

“Do you regularly watch videos by Jordan Peterson?” kinda needs to become one of those before-first-date screening questions.

I saw a snippet of it.

The code was using a function to connect to mysql that was deprecated in PHP 5 and removed in PHP 7. So they must have been running PHP 5.x. It also contained an obvious SQL injection vuln (although that wasn’t used for the hack).

It’s on the radar, yes.

if you turn off “Show posts from child feeds” on the Forumverse feed then https://piefed.social/f/lemmycategories will load much faster…

Elon Musk, the de facto head of DOGE, lowered expectations of the group’s savings from $1 trillion to $150 billion by the end of the fiscal year.

So, a failure even on their own terms.

This idea is quite similar to what Bluesky is doing, with “Labels” https://docs.bsky.app/docs/advanced-guides/moderation.

We’d need some way to crowdsource the verification and validity of the labels so people can’t just put low-quality or abusive labels everywhere.

It could potentially reduce the amount of work moderators need to do because spam would be labelled as such by anyone and if a few others also label it the same then it would reach a threshold where the label becomes active.

Does anyone have experience with this way of moderating content on Bluesky? How well does it work in practice?

That was more interesting than 18 downvotes. Thanks!

Surely the best way to end all wars is to win them all. /s

It is against the Geneva Convention to use POWs for propaganda.

Imagine you want to write a competitor to PostgreSQL and you start out by importing SQLite into your project and building on top of that. To you it seems like a good idea because you’ve never written a DB app before and the only DB you’ve ever seen before is SQLite. You’ll get a prototype real fast but you’ll never build a PostgreSQL equivalent because you never learned the foundational knowledge of how a DB works and because SQLite forecloses all the pathways you need to get there.

Same thing.

Because outsourcing your core business processes is a bad idea. A fediverse app that relies on a library to do all the fediverse stuff is going to have a bad time. Not straight away, but eventually.

Most of the cheap Androids I’ve had got maybe 1 free update and then nothing ever again. Even if I wanted to pay.

Compact mode has been added now, FYI. Screenshots at https://codeberg.org/rimu/pyfedi/issues/540#issuecomment-3788240

Yeah, federation is kinda inherently “wasteful” in that so many copies of everything get made.

There are ways to reign it in a bit, by aggressively pruning old content, leaving original copies of images on their home instance, etc… But fundamentally saving on resources is pretty difficult to do if you want redundancy, decentralisation and resilience.

Not :)

One day it would be cool if the creation and curation and maintenance of feeds more collaborative, like a wiki. Right now one person can squat on a tasty name/url and just do nothing with it, potentially. It’d be great if there was a process to get others involved in each feed, if necessary.

One day…