Yea anything big and mainstream just seems super shallow.

I’m not on top of things to compare accurately, but it was always kinda like that (and is like that here sometimes too). But whenever I’ve gone back, I’ve definitely felt like it has gotten somewhat worse. Some of that could easily be a shifting standard from spending more time on other less “mainstream” platforms though.

Every browser released since 2020 supports this

It’s a little paranoid of me, but I like the idea that a basic web app I make can be thrown onto any old out of date machine, where ~2015 or younger seems about right for me ATM.

You mean the Html template Element? I’ve never really got that to work, but I also never seriously tried.

Yea. From memory, it’s just an unrendered chunk of HTML that you can select and clone with a bit of JS. I always figured there’d be a pattern that isn’t too much of a cludge and gets you some useful amount of the way to components for basic “vanilla-js” pages, just never gave it a shot either.

Yea, I’m unclear on how you can take web components and still have widespread browser support (not knowing enough about their ins and outs).

Plain template elements are widely supported and have been for ~10 years (which ideologically matters to me along the same lines as the top post’s article) … perhaps a little bit of hacking together can get you close with just that?

But nah. These goobers got high off npm modules and did shots of JSX in the bathroom at lunch time.

Fucking LoL!

Anyone here have thoughts about doing basic interactive stuff with at most vanilla JS?

A few months ago, the “Nazi” presence on substack and substack’s insistence on not moderating them (like at all it seemed) broke as a story, during which Casey Newton (and by extension his “platformer” blog) got engaged with substack about the issue and, after being disappointed with substack’s responses and policies, famously left for Ghost (see their post on the move here.

Pretty sure that boosted its profile and prompted talks of federating, which they were initially hesitant to do … but here we are now.

Ha, yea! If you know rust, then you don’t need to reach for Python (right?!). Plus the main motivation was to contribute to lemmy itself while also learning rust. That another platform is good for personal instances doesn’t change that, though piefed does seem cool and I can see myself wanting to get involved with it at some point.

Cheers! 2 threads and 2gb RAM I’d what I would have hoped for anyway. Thanks!

But I get the database thing. Its spiking every couple minutes and a lot every hour. It’s not a big deal if you have 2 threads at least but I can see how it doesnt work for everyone in every scenario.

Yea database management seems to where the growing pains are right now (with the core devs welcoming help from anyone with DB/PostreSQL expertise) … and indeed it seems to be a perennial issue across the fediverse platforms.

If I may ask (sorry, probably annoying) … what sort of resources would you recommend for a small personal lemmy instance? (let’s say 1-5 users, ~200 community subs and a few local communities?)

Woah woah … this is legit awesome! Just tested (on lemmy.ml) and yep … seems to be working like a charm!

Give up to matc-pub for the PR (and maybe a new core dev for lemmy too?).

I figure this makes live megathread style posts/chats more viable … which is certainly cool!

I’ve mentioned this before, but an interesting possibility might be to enable selected posts to be “live chats” through a websocket like process as lemmy used to be, just for selected posts for certain windows of time, whenever a live chat dynamic is sort.

It’s the sort of thing that could be scheduled and subject to admin approval or something if resources are a concern.

Otherwise … awesome to see!

Yea I did a quick search through the GitHub issues, and it seems like there are some growing pains with updates they’re making to the way things work and the load it puts onto the database. Sad to hear for smaller instances as my impression was that lemmy had pretty good performance for smaller instances. Architecturally, it makes sense that there are different tradeoffs for bigger and smaller instances. It’d be good to see things mature to the point that you can tune things for your instance size. In the end though, picking the appropriate platform but with the assurance that migration can occur when you need to change platform may be a good way to go.

I think there’s a pretty fair argument that more common and easier languages and tech stacks are preferable platforms for smaller more personal instances … just the comfort of being able to modify and debug is probably worth whatever other tradeoffs may be encountered. Python, naturally, is basically a prime candidate. So yea, PieFed seems very cool, especially for personal servers and they’ve got a good performance profile.

Interesting! Any insights or comments on what’s increased the resources of lemmy over the recent updates?

Still, I think you raise a relevant malicious path. Like I said, I wouldn’t be surprised if something contentious happens however infrequently.

We can already create private instances that don’t federate for those niche communities;

That being said, creating a private instance is a relatively difficult hurdle. By providing private communities, an admin can take care of the hosting, along with all of the other communities, while those who want something more controlled and closed can have an easily accessible option. Plenty of people want their social media to have options for being relatively closed or relatively open, and I think it’s healthy to provide those options.

I hear you though on the lemmy-world community closing possibility (and similar) … that would easily be an abuse IMO and it’s not entirely clear what would or could happen.

To be fair though, the whole lemmy-world instance (or any other for that matter) could simply turn federation off at any point to the same effect you fear, so it’s arguably just part of the federation flexibility. In this case, any community mod has their hand on the switch for their community, which means we’ll probably see it get used in controversial circumstances at least once. But for any given community, going either private or local-only is sure to drop user engagement or be a PITA regarding managing the “approved users” list, so I can’t see it being a popular action TBH.

All good … the whole post was started by me misreading in the first place I think!

I think it’s a good option to have. Most who start communities want reach and engagement. But for those situations where you want a more in-group vibe, something like this is essential.

It’s sorely missing in the fediverse and a rather good form of social media TBH that the fediverse, until now, has ignored (while it has kinda taken off on discord etc).

Private communities though are intended to federate, just with gated membership. And they could be useful for particularly niche communities that don’t want to be disturbed by those who mainly use the All feed.

It will be interesting to see how it interacts with federation/defederation dynamics though. Lemmy-world for instance, could easily start going local only because they kinda already think they’re the whole of the threadiverse and are certainly big enough to sustain themselves.

Ha … it seems like it at least.

I think I was being dumb in asking the question actually.

It’s really just about the circle of users to whom the community is visible.

Local-only … visible only to users of the instance. I’d presumed that it could be writable only to users of the instance such that only users of the instance could post/comment there. But double checking, no, it’s only visible if you’re logged on with an account on that instance … so pretty private in the end actually.

private communities … which are apparently coming … are visible only to approved users, whether on the local instance or not.

And presumably, these will be stackable, so that a local-only + private community will be visible only to approved users from the local instance. So getting pretty closed.

but it shouldn’t

I’m curious about why you think this relatively strongly (if you don’t mind my asking)?

Without knowing the financial history of the place, it seems a good case study in something that could have gone for the sustainable stalwart of the internet path but instead fell to the dark silicon valley profit/growth side of things. With wikipedia being the only great success (AFAIK) at forging solid and sustainable foundations for the internet, I suppose the lesson is that it has to be non-profit, or open-source (or both) from the beginning.

In a way, it is kinda on many of us for not realising this and pushing against it sooner.

One of the great things coming out of the fediverse (and bluesky too at the moment) is all of the open software being developed that will hopefully plant seeds that will last a long time.

Hear hear!!