Well, I think a whole lot has changed, and we can point to lists and lists worth of stuff (for example: using Haiku as a daily driver on bare-metal is, for a lot of users, actually viable!). But the biggest thing is something indeed that hasn’t changed at all, and that is:
Haiku is still here.
Two decades ago, there were a lot more “alternative OS” or even “hobby OS” projects: Syllable, SkyOS, etc. Most from that era have long since gone, and nowadays, there’s far less, and mostly the “hobby” kind. Haiku, meanwhile, has survived, and continues on over two decades after it began. And as you can see by the fact that this forum thread got 80 replies (most quite long and thoughtful!) in a week, a still very active community.
Is progress slow? Yes. Are we very focused on consensus, and figuring out details of things before we implement them? Yes. Does that require patience? Oh, yes. Does it get frustrating? Oh, most certainly yes.
But clearly something’s working, because we’re still here, and we’re still at it. Other projects that forked Haiku because people were frustrated with the development speed, or thought things would go better with a different kernel, or just wanted the toolkit running on Linux, have come … and then they’ve gone again. (To be fair, some have come back again later, as with Cosmoe making a re-appearance after over a decade of dormancy. But this is rare, and Cosmoe is mostly just a UI toolkit, it’s not even aiming to be a whole desktop environment and much less a “distribution”.)
UI/UX changes, in particular, are hard to do. They affect absolutely everyone in the present and (until we change something again) in the future, so everyone has a stake in them, so everyone has thoughts about them. That’s natural. But we are not Linux: if an idea really is an improvement, then chances are, we will eventually agree, and when we do agree, and change direction in favor of that idea, then the whole ecosystem will (slowly, but eventually) realign around it. That’s a much, much better model than “sure, we’ll add a 15th way to do that” or “go make your own patches if you don’t like it” as Linux tends to follow. It comes with downsides, obviously, as I’ve already noted. But the end result is worth it; or at least, I think so, and I think the long-time “Haikuvians” who’ve stuck around through thick and thin think so, too.
I think Haiku’s relative “conservatism” then isn’t a ‘bug’, but rather a ‘feature’. We are slow, but we’re steady, and two decades later we’re still here. It appears to me that a great many respondents in this thread like your idea, and think it’s a good one. The “ayes have it”, so far at least. There are objections, yes, and there are problems people are throwing up to be solved. But they’re not throwing them up to try and kill the idea completely, at least not for the most part, I think, but rather to try and parse out what exactly the idea will look like, and whether it really is something we want to implement and can live with.
