We’re nearing the end of the season, so as usual, there won’t be much new this and the coming weeks.

A quick disclaimer regarding Backerkit: They have redesigned their website again, which of course broke the crawler. I will try to get it working again as soon as possible, but until then, there won’t be any Backerkit projects included in the lists.

It still looks a bit quirky in Kbin sadly (especially on mobile where it gets very squished), but at least it should work in Lemmy now.

So it seems like the table formatting bug as finally been fixed in Kbin. In other words, the inline lists are back!

This is however not something I think we, as consumers, should necessarily celebrate. This also means that we are very likely nearing the end of the “free” web that we are used to.

No, I’m not saying that selling out one’s personal integrity is preferable, but if it turns out that advertisement as a business model effectively isn’t sustainable, we will just have to accept the reality that we will more and more commonly have to actually pay to access content and services on the web.

Oof, I’m sorry, but perhaps it was a matter of expectations. I feel that the problem that I had with the show was that it started off on the premise of some Fallout-style retro futurism, with what I guess one could call “dangerous techno optimism” of the 50’s and 60’s.

However, beyond the first episode where the woman gets killed by the automated postal car, this never really gets explored much further than some occasional background elements. The story could just as well have been set in the actual 50’s and it wouldn’t have changed much.

Now normally, I don’t generally have issues with character driven stories where the setting comes secondary (e.g. Ted Lasso is a character driven comedy set on the backdrop of football), but in this case it was the setting that actually interested me to begin with.

It’s a shame, because all the individual pieces, from acting, to scenery and atmosphere were great, it’s just that the show never clicked with me in the end, because it wasn’t what I thought it would be.

They are probably reusing a component that happens to sort its entries alphabetically, since that is most commonly the expected behaviour. If the form is configured in a CMS, whoever built it might not even know it’s happening and has entered the data properly, but it gets resorted in the presentation layer. It’s also not impossible that the behaviour of the component has changed at some point and this particular case didn’t have test coverage or wasn’t actually part of the specification.

Apple TV+ started rather slow, but the quality of the shows is just insane, and not just when it comes to scifi. So far, the only show that I didn’t like from TV+ was Hello Tomorrow, which had a lot of style and potential, but just never went anywhere.

Repeat after me: ChatGPT is a language model not a digital librarian.

Filing a patent means little to nothing for a company like Apple regarding future consumer products. All it likely means is that a patent engineer managed to throw something together outside existing prior art that they could file. Maybe they will do something with it, maybe they won’t. If they do, they will have a patent portfolio that will hopefully give them some legal protection from patent trolls and competitors that will attempt to block them.

So first they alienate the people who are working for them for free and then they go ahead and alienate the people who are giving them money for free.

So they have a bunch of users that have been freely paying them money for virtual coins that you can literally only use to display a few pixels of a gif next to a comment.

Their absolute genius move towards profitability is then to forcibly stop making these people give them free money and also erase those virtual coins that they spent money on with absolutely no compensation whatsoever. Not even a shitty award or literally anything at all.

It’s funny, I’m not sure if I should actually be impressed that they are not engaging in any marketing dark pattern whatsoever; they are just straight up alienating the people who were until now been practically giving them money for doing nothing.

I guess it depends on how this “Data Protection Review Court” actually functions in practice. What is written on paper doesn’t seem to really matter much to US agencies so we’ll see how strong these safeguards actually are.

Nevertheless, it’s good to see that the new agreement is finally in place at least. We’ve had this legal vacuum for years now and it was completely unsustainable in the long run. Sans a complete legal overhaul of the nonexistent privacy laws in the US (hah good luck with that), this is probably the best we could hope for now.

So for some reason the inlined tables didn’t format at all in this post. I’m generating them automatically, so I have no idea why they didn’t work this time. Probably some project that contained some invalid character in the name or description.

The Google Docs still works however. Meanwhile, I’m gonna see if I can find what the issue could be.

Edit: It seems like there’s a bug in Kbin that somehow breaks the tables (even when posting to a Lemmy instance). Until that is fixed, I guess the Google Docs will have to do, sadly.

ActivityPub is very likely about complying with the EU DMA and future development of it. The Fediverse community by itself is practically irrelevant.

I know that is a popular narrative in the Fediverse community right now, but I honestly find it unreasonable. Google and Meta didn’t kill XMPP, they abandoned it and without them it went right back to where it was originally: barely used whatsoever. The Fediverse already has a small, but relatively healthy user base. Meta can abandon ActivityPub or twist it into something unacceptable, but all that will do is bring us back to where we are right now.

All that is irrelevant, though, because the difference is that Meta is legally required to incorporate interoperability with other services and that’s why they are going with ActivityPub. Not from the good of their hearts, but because they need to and is in their own self interest to keep alive.

Right now, they are early adopters of ActivityPub and have a very early strong position there. When they federate they will be by several magnitudes the largest instance on the network. Whether we like it or not, they will inevitably be a major player in dictating the future of ActivityPub. Thus, they want to keep ActivityPub alive because they want to make sure that becomes the future EU mandated industry standard for SoMe. Otherwise, some other technology will be chosen, one that might not be lead by Meta, but by Google, Apple or ByteDance.

Note, that I’m not arguing whether this is good or bad, but only what I predict is happening.

Interestingly it seems like the main reason is not the enormous data gathering list that has been posted a few times, but rather the connection to Instagram and whether the two services are allowed to share user data.

Meta is also using Threads as a test bench for the Digital Markets Act, so they are probably holding off the launch until they can get federation going.

Meta has several billion active users across their platforms. 30M is nothing to them.

Also don’t forget that we’re talking about a microblog, so it will inherently generate a large amount of individual posts, much more so than e.g Instagram. The quality is however likely very low initially and a lot of users are probably just trying out the current talk of the day.

I do suspect that Threads will probably grow to a few hundred million users before the end of the year; anything less would probably be regarded as a colossal failure for Meta.

We migrated to Matomo, which has very similar functionality as GA, but can be self hosted and is GDPR compatible. It can even be configured to run without consent since it doesn’t build a third party ad profile, which should actually improve the data coverage a bit.

The tracking API is a little different than GA so we had to redo some things to get all the events to trigger properly (especially for e-commerce), but for basic usage statistics it’s relatively plug and play in the tag manager.

https://matomo.org/

Oh, it seems like there might actually have been a Firefox issue. Or rather, me apparently not knowing how to use fetch() properly.

See if it works better now: https://greasyfork.org/en/scripts/469139-kbin-subscriptions-panel

You can click on the cog icon in the upper right to see if you have the current version installed (2.6)