3 ms·
I don’t think the 3.9 branch is less real than the 3.2 branch. Both branched off from the original Commodore source code, didn’t they?
by bhaak 2y ago
I don’t think the 3.9 branch is less real than the 3.2 branch.
Both branched off from the original Commodore source code, didn’t they?
- actionfromafar 2y agoCan’t remember if either use the sources at all.
- doctor_radium 2y agoI was living Amiga daily during the OS 3.9 days, and have no reason to think it didn't come from the Commodore source. Back then I got the feeling it was called "3.9" because the various powers assumed it would be the last release of 3.x. I've never used Hyperion's 3.x, but that is official, too, if largely a backport of 4.x, which of course also came from official source. I would say Hyperion jumped into 3.x (against their contract?) in search of cash, as nostalgia for these old systems increased and ARM emulator boards (etc.) made it possible. OS4 is still the future, but the cost of PPC motherboards greatly limits the market size.
- snvzz 2y ago>OS4 is still the future AROS is. OS4 being a hard break while also being privative, closed source means it will never see significant adoption. >the cost of PPC motherboards greatly limits the market size. PPC has no future. RISC-V is where it's at. Of course, back in the 90s, there was no way of knowing this.
- kstrauser 2y agoJust my own take here: 3.5 and 3.9 felt like “3.1 plus some bundled, licensed shareware”. Which is fine! It’s nice to have a baseline where you know everyone running that version has a particular TCP stack or dock app or whatever. But 3.1.x and 3.2 to me “feel” like newer versions of 3.1 with actual OS updates (and also plenty of bundled shareware). I wouldn't argue against someone who sees it the the opposite way.