OpenStars

joined 2 years ago
MODERATOR OF
[–] OpenStars@discuss.online 8 points 3 weeks ago (2 children)

Are you sure? Did you try counting the number of pieces of candy? I bet you got at least 6 to 7 pieces somewhere in there...

[–] OpenStars@discuss.online 13 points 3 weeks ago (6 children)

I don't get what "1234713" means?

[–] OpenStars@discuss.online 8 points 3 weeks ago

Beauty lies in the eye of the beholder.

e.g. I think this comic is wonderful & gorgeous:-)

[–] OpenStars@discuss.online 8 points 3 weeks ago (2 children)

"When you eat at Arby's, there are always (bigger) butts."

- a wise man, probably

[–] OpenStars@discuss.online 7 points 2 months ago

As a distraction from everything else he was doing at the time and would do in the future?

At some point we really should just stop asking why, especially if we don't want to know the answer 🥴

[–] OpenStars@discuss.online 19 points 2 months ago

Even Russia is a "democracy".

[–] OpenStars@discuss.online 2 points 3 months ago

Yeah and as you have pointed out, the PieFed API is very new and not as mature yet as Lemmy's, so there is value in waiting for it to advance while you work on other things like Svelte upgrades.

So long as you enjoy yourself in the doing, it's all good 😊

[–] OpenStars@discuss.online 2 points 3 months ago

I somewhat disagree: I used to use Facebook, then left that and joined Reddit ! whether I joined Reddit or not though, either way I was leaving Facebook) then left Reddit and joined Kbin, then left that when it imploded and joined Lemmy, then left that, and fortunately PieFed was coming up at the time. I view most social media as "bad" - at least for myself - and the interactions I was seeing on Lemmy were not worthwhile for me to remain. But if you enjoy it, that's fine, I am just sharing my experiences.

And you asked "someone can just fork it and add the missing features", which seems like a competition to me since you aren't going to contribute to both Lemmy development (in Rust) and also PieFed (in Python): someone must make a choice where their time & efforts are going to be directed at.

Which if you choose Lemmy again is fine, but I am pointing out that it is in fact a choice being made. Hence I hoped to help inform that choice by pointing out some of the reasons to choose PieFed rather than Lemmy, which either way that ends up getting chosen will lead to increased efficiency and fewer regrets moving forward, with the cost having been counted in advance rather than discovered only much later on in the process.

Further, I would argue that the set of considerations is quite different for a mere user vs. someone thinking about actually contributing to development of a codebase. Even for an instance admin, I would hope that such a person actually would look at what is technically better than something else, before going to all that effort to set something up that will require much maintenance in the future. Of course, to each their own, I was just sharing my own thoughts on the subject.

[–] OpenStars@discuss.online 1 points 3 months ago (2 children)

That seems far too simplistic imho. Instance admins have done this for years and can tell you how quickly things fall behind: if you want to federate with any other instances (the entire point behind the federated model?), then you need to maintain compatibility. Fortunately Lemmy is fairly mature and far less likely to release groundbreaking changes than it did in the past.

But also, you have to learn to code in Rust, which even people who already know C++ seem to find very difficult, for a number of reasons including major lack of support by a standard library (such as C++ itself has in its STL), which in Rust is still fairly primitive iirc, forcing the user to build every tiny little thing from scratch, or use less well-written and tested code, possibly so poor as to negate the advantages of having chosen Rust over some other, more commonly useful language like C++.

And then you'd be doing all of that entirely on your own, and maintaining it in perpetuity. Don't get me wrong, several people have done exactly that (Admiral Patrick, developer of the Tesseract front-end, comes to mind).

But all of that seems like it would be even slower, compared to PieFed releasing new features practically weekly? And also it is Python, which is a much easier language. And also you could work along with others, fixing bugs in your code that you did not spot, and vice versa. I'm not seeing the advantages there to what you are proposing: I mean yes obviously there are "advantages", but relatively speaking I mean, they seem much smaller than if someone put the same amount of effort towards improving PieFed, which would then be shared and maintained world-wide even if you got sick or busy irl or something?

And even if you were right, that doing this with Lemmy would work out well, for how much longer would that remain true - six months? - before PieFed absolutely blows the set of features that Lemmy uses out of the water? Imagine social media that is actually fun to use, and where the computer automates the most common tasks so as to not require menial labor every hour of the day, as Lemmy does (I am speaking of the requirement for manual moderation efforts)? That much has already come to pass, to various degrees, in many ways on PieFed. e.g. in Lemmy you could search for every cross-posting across all instances wherever you can find them, then click on each one, and read through the comments, making sure to get the version of the community that is accessible from the instance you are on rather than follow a link taking you to a different one... but why do all that work, when PieFed provides it ready-made, instantly upon loading the post?

Starting with PieFed is starting ahead of Lemmy, in most ways (not all though: Lemmy's search functionality is still way better, and reportedly about to get even better still by allowing limiting of search terms specifically to post titles separately from message contents).

Unless you just want to learn Rust for other reasons. 😁

[–] OpenStars@discuss.online 3 points 3 months ago (2 children)

It's great to see you still posting on the Threadiverse! Okay so you've been doing it for awhile I guess but I've been sick myself so not staying up with things, anyway it's still great to see!! 😜

You may want to think about it from the ground up: a lot of the need for Tesseract was due to things like the strict authoritian stance of the tankie devs - e.g. not providing a means to truly block all users from an instance, or not showing alternative image text, or not embedding video playbacks - forcing you to find creative solutions to that problem (note PieFed does all of those things mentioned, usually not as comprehensively implemented as well as Tesseract does it but at least to some degree, e.g. Peer tube and YouTube videos can embedd but not Loops ones). Maybe now Tesseract would not have to be an entire alternative UI front-end - especially when the development pace of PieFed is so rapid in comparison to Lemmy that would increase your difficulty of keeping up - but instead rather a "theme", combined with changes to the underlying codebase that would affect all of the users of PieFed instead of only some of them? These devs I believe would be much more friendly and receptive to your ideas:-).

Although I am not a developer like you so too far away from the problem to see it anywhere close to clearly like you will, as you get into it, but wanted to throw out that oddball idea from left field in case it helps jar your thinking along creative lines. Remember to do six impossible things before breakfast each day!

seven unpossible (sic) things

But most important of all, if I can add, would be for you to enjoy it!!!

view more: ‹ prev next ›