OpenMW’s official Lemmy community has been on lemmy.ml since 2021, way before lemmy.world existed (and most other instances, too), and way before there was any inter-instance drama. It’s becoming increasingly likely that it’s not going to be a suitable long-term home, but we’d be much happier if we could migrate the existing community rather than start from scratch with a new one. Is there any way to do that yet?
Instead of a commit history, you get a commit fairy tale, which is prettier than the truth, but probably less useful. You get something akin to merging the base branch into the feature branch but things look as if they were done in an order they weren’t instead of getting an ‘ugly’ merge commit.
I’m thinking more for the scale of something like OpenMW, as we’ve got more frequently asked questions than we could hope to put on an FAQ page. In the olden days, stuff showed up from our forums when people googled it, and now it doesn’t, so we get loads of questions through Discord, and very rarely one from Matrix.
If past support questions showed up in searches, then more users would be able to help themselves and would never need to ask for support, so it wouldn’t matter as much what platform it happened on.
Personally, I think it would be good if support discords were all bridged to matrix spaces (currently doable, but matrix needs locking down more than discord to stop spam as the tools to prevent and remove it are worse) and the matrix history was archived somewhere search engines could index it like mailing list archives are (currently not doable). That approach would let users use what they want without forcing anyone else to, and keeps self help as easy as it was in the days of forums.
In the olden days, online games didn’t have servers run by the publisher, they came with a separate program users would run to host their own server. Things like Minecraft still work that way, although I think there are servers run by Microsoft too now. For some games, this will be more complicated than others, but it’s not impossible.
Really, this is the games’ fault, at least if they were released in the last ten years. That’s how long Documents has been a subdirectory of OneDrive by default. If you’ve built software that can’t handle that gracefully and released it since, that’s just negligent. It’s not like the windows documentation doesn’t tell you places that wouldn’t have this problem where you can store machine-specific data.
They’d probably have less of an impression that Linux is only for people they wouldn’t want to associate with, as that’s the impression you’re giving.
About ten years ago, I was using a third-party OneDrive client to sync it on an Ubuntu install. It was working brilliantly until daylight savings started our stopped, then it got confused about file times not matching so tried resyncing everything, and then saw the times still didn’t match, so made a new file for each file with an added numeric suffix, then tried doing the same for the new files, and kept going until my disk was full. That was a mess to tidy up, and once it was sorted, I had to start the sync program in GMT to stop it happening again.
He got upset that there was an X button making it look like he could remove the ads, but it didn’t remove the ads.
There are other rarely-used C+±like languages that fit your criteria, and they basically all aim to eliminate the kind of thing I was talking about. If someone was used to one of those and tried picking up C++ for the first time, they’d probably end up with working, but unnecessarily slow C++, having assumed the compiler would do a bunch of things for them that it actually wouldn’t.
The popular low-level systems programming languages that aren’t C++ are C and Rust. Neither is object-oriented. C programmers forced to use C++ tend to basically write C with a smattering of features that make it not compile with a C compiler, and produce a horror show that brings out the worst of both languages and looks nothing like C++ a C++ programmer would write, then write a blog post about how terrible C++ is because when they tried using it like C, it wasn’t as good at being C as C was. Rust programmers generally have past experience with C++, so tend to know how to use it properly, even if they hate the experience.
I’d say this is pretty dependent on the language. For example, with C++, you need to micromanage (or at least benefit from micromanaging) a lot of things that you can get away without knowing about at all with other languages. That stuff takes time to pick up if you’re self-teaching as you can write stuff that looks like it works without knowing its half as fast as it could be because you aren’t making use of move semantics, and if a colleague is teaching you, then that’s time they’re not spending directly doing their own work. On the other hand, someone with Typescript experience could write pretty decent Javascript from the get-go.
The legalese in the US (which might as well be everywhere as you need to have compatible copyright with the US to have a trade deal with the US, and your country is in trouble if it doesn’t have a trade deal with the US) is basically that:
I’m sure plenty of publishers would love for the second set of rules to apply to things like books, and from a quick googling, it seems like occasionally academic textbooks have included a licence agreement instead of you actually owning the physical book, but I imagine that most publishers are concerned about bad PR from attempting this with a hit novel and also don’t want to be accused of fraud for having their not-a-book-just-a-licence on the shelf next to regular books and thereby tricking consumers into thinking they were buying a regular book. EA attempted to double-dip over a decade ago with Battlefield 3, which included a copy of the game (with regular First Sale Doctrine rights) and a licence key for the online pass (which wasn’t transferrable) and got bad press because of it. Newer PC games often come as a key in a box with no disk or a disk that only runs a web installer, so you’ve not got a copy of the game to claim you’ve bought and obviously only have a licence, and this seems to have caused less upset. This wouldn’t work with a book, though, as you have to fill in the pages at the printing factory, and can’t magically do it only after the user’s got it home.
You want to have commit history, not a commit fairy tale. Once you start rewriting history, it’s not really history any more. The stuff people want to hide tends to be some of the most useful to someone looking through the history to find out how things became the way they are and what was going through the author’s mind when it was written. If things are messy and crappy, it’s better to know that rather than have it covered up.
There are kinds of analytics that are incompatible with the GPL, as you can’t restrict what users do with GPL software, and that includes asking children not to submit analytics containing information you’re not allowed to know about children under COPPA. The only options are to hope your software is only used by adults, or not implement any kinds of analytics that collect the relevant kinds of personal information.
I’ve yet to find tooling that supports this. Clang format has a setting that looks like it does it, but actually does something else. If I have to press the spacebar a bunch of times each time I add an argument to a function, that’s a pain, and it’s a bigger pain to convince the people I’m working with that that pain’s less bad than using spaces everywhere and having the IDE deal with it.
Until the people making editors and auto formatters acknowledge that the obvious most sensible whitespace style is even a thing, I’m forced to do something else and be really grumpy about it.
If you trap a person in a room with a keyboard and tell them you’ll give them an electric shock if they don’t write text or the text says they’re a person trapped somewhere rather than software, the result is also just a text generator, but it’s clearly sentient, sapient and conscious because it’s got a human in it. It’s naive to assume that something couldn’t have a mind just because there’s a limited interface to interact with it, especially when neuroscience and psychology can’t pin down what makes the same thing happen in humans.
This isn’t to say that current large language models are any of these things, just the reason you’ve presented to dismiss that isn’t very good. It might just be bad paraphrasing of the stuff you linked, but I keep seeing people present it just predicts text as a massive gotcha that stands on its own.
!openmw@lemmy.ml has less than 150 subscribers, so it’s definitely not large. We’re already swamped with infrastructure work for the stuff we already self-host, so I don’t think we’ll be running our own Lemmy instance any time soon.