It was a truly sad day when Oracle bought Sun, because we knew right then that ZFS wouldn't improve nearly as many people's lives as it should have, and that Solaris would die.
The ZFS thing still bums me out to this day, but I think I got over the demise of Solaris when OpenSolaris fizzled out.
What was cool about Solaris for old-timers? In particular, are there things that compare well against modern BSDs?
My vague memories of it are that the userland showed signs of legacy (/usr/ucb), and I could never get the tools I needed installed. But that was a consequence of the setting I was working in, not Solaris.
It would be particularly interesting to hear from fans of STREAMS or other kernel features.
I personally don't give a shit about Java or MySQL, so I wouldn't have been sad if those products died, but I know lots of people who would be.
For me, it was ZFS that I was really excited about. Back then, I thought I might have ZFS (or some ZFS-derived or -inspired equivalent bitrot-resistant checksumming filesystem) on all my computers and phones by, say, 2017.
But I don't. My Linux box can finally have ZFS (with some caveats), but my Macs and phones can't and won't; it was another sad day (for me) when we learned Apple's "modern" new filesystem, APFS, does not do checksumming for user data[1][2].
There's a wealth of knowledge getting lost here in the process. I wonder if some enterprises are willing to hire these devs to ensure continuity of systems running on Solaris ?
Either way, I've always had a soft spot for Solaris, and think they're on of the best examples of enterprises being able to produce quality software. It's incredibly sad to see this happen to the Sun legacy, and it makes me wonder why Oracle acquired Sun in the first place.
Is the knowledge really being lost? These people will get jobs elsewhere. With their experience, they might end up working on some other OS, and bringing the benefits of their knowledge to other systems that are now more widely used, so this may actually be beneficial in the long run. I suppose it's a cliche, but markets are supposed to allocate resources where they are best used.
There is a great demand for systems engineers. It's _very_ hard to find low level developers on the market. Don't be intimidated by stupid requirements on job ads like "linux experience" - is easy to adapt to the OS-specific environment. The valuable skill is not OS-specific, it's "reading long piece of C with understanding"!
This is natural and to be expected. Whether we like it or not Solaris went the way of Novell Netware. They have a lot of customers but no new orders. So the staff needs to be kept around simply to serve existing customers out of which lot of them would move tofuture proof systems every quarter. So every quarter they will see loss of revenue and attrition in their customer base. How long can they continue to support such a group?
The ZFS thing still bums me out to this day, but I think I got over the demise of Solaris when OpenSolaris fizzled out.