The End Of An Era: A Look Back At The Most Popular Solaris Milestones & News

[schmidtbag]

Not really. Red Hat is making money. Canonical was supposed to lose money for a number of years while pursuing some specific goals (I think they may be making money on the server/cloud these days, but I'm not sure). Novell, I know nothing about.

I know that. My point was that your reasoning behind why Sun went under should also apply to these other companies I mentioned as well since they function in a very similar way, but obviously these other companies are doing just fine, which suggests there's a different reason why Sun failed.

[rhavenn]

You could argue the same about Red Hat, Novell, Canonical, etc. The only difference is Sun decided to go solo instead of work toward Linux. I think Sun's main problem is they were trying to juggle too many things at once. Creating their own CPU architecture, programming language, virtual machine, operating system, office suite, etc is just too much considering they had at least one competitor who had something better in every one of those categories. Oracle may have killed nearly all of Sun's legacy, but I sometimes wonder how much longer these products would last if they kept going.

No, I don't think "work towards Linux" would have helped any unless you mean the Linux philosophy. Solaris is a rock solid operating system. The problem was you had to be an admin to use it or get anywhere and all the training was tied up behind paywalls and classes that only corporate could afford.

I used Sparc stations in college and they weren't bad, but I only had to run gcc and handle my own account. I didn't have to admin anything. I started tinkering with FreeBSD during my late college years and found an awesome community and a few admins who were willing to take me under their wing. After my first real job I got one of those Solaris Sun Blade 100s as a toy and I just couldn't do anything with it. There were no repos, no real documentation floating around and just not a big community. Sure, there were manuals, but pretty much everything focused around you already knowing Solaris and how those boxes worked. The learning curve was steep.

If Solaris had streamlined the installs, repos and package management, just made it as easy as a FreeBSD or Debian console install and then embraced UGS, cons and the community they would have been able to keep nerds interested, but they hung on way too long to the enterprise and lost any interest from the bottom end.

[schmidtbag]

No, I don't think "work towards Linux" would have helped any unless you mean the Linux philosophy. Solaris is a rock solid operating system. The problem was you had to be an admin to use it or get anywhere and all the training was tied up behind paywalls and classes that only corporate could afford.

Hence why making it Linux-based (or BSD for that matter) instead would have solved the issue...

I used Sparc stations in college and they weren't bad, but I only had to run gcc and handle my own account. I didn't have to admin anything. I started tinkering with FreeBSD during my late college years and found an awesome community and a few admins who were willing to take me under their wing. After my first real job I got one of those Solaris Sun Blade 100s as a toy and I just couldn't do anything with it. There were no repos, no real documentation floating around and just not a big community. Sure, there were manuals, but pretty much everything focused around you already knowing Solaris and how those boxes worked. The learning curve was steep.

I'm not saying Solaris, SPARC, or anything Sun did was bad (well... Java could've been done better) but my point is they had competition that was just 1 step ahead. Case in point - your Blade 100 situation. Sun had a lot of potential. I'm aware there were things they did that really stood out, such as ZFS, but think of it like this: does piece of treasure locked away in an unbreakable container launched into space contain any real value if you can pretty much never obtain it? Obviously, Solaris and other Sun products aren't that extreme but my point is it doesn't matter how good your product is if it's too difficult to really take advantage of it.

[bug77]

I know that. My point was that your reasoning behind why Sun went under should also apply to these other companies I mentioned as well since they function in a very similar way, but obviously these other companies are doing just fine, which suggests there's a different reason why Sun failed.

I don't see the similitude. Sun had a number of products they either failed to monetize or were incurring losses year after year. And little else. RedHat and Canonical aren't even remotely similar. But maybe I'm missing something.

[schmidtbag]

I don't see the similitude. Sun had a number of products they either failed to monetize or were incurring losses year after year. And little else. RedHat and Canonical aren't even remotely similar. But maybe I'm missing something.

Not sure how you don't see the similarities... Yes, Sun did many more things than the other companies, but at their root they all were doing the same thing - they created open-source enterprise products. Much of what their clients pay for is the professional dedicated support. They all share the same target demographic, though Sun's was a bit more niche.

[bug77]

Not sure how you don't see the similarities... Yes, Sun did many more things than the other companies, but at their root they all were doing the same thing - they created open-source enterprise products. Much of what their clients pay for is the professional dedicated support. They all share the same target demographic, though Sun's was a bit more niche.

Root in making open-source enterprise products? You make me laugh. Their roots were in the hardware business. When that went belly-up, so did everything else.
Yes, they sold their hardware with together with lots of software, but the simple fact is when they couldn't sell hardware any more, nobody was in the market for their software either.

[starshipeleven]

You could argue the same about Red Hat, Novell, Canonical, etc. The only difference is Sun decided to go solo instead of work toward Linux. I think Sun's main problem is they were trying to juggle too many things at once. Creating their own CPU architecture, programming language, virtual machine, operating system, office suite, etc is just too much considering they had at least one competitor who had something better in every one of those categories. Oracle may have killed nearly all of Sun's legacy, but I sometimes wonder how much longer these products would last if they kept going.

I think you can state Canonical = Sun safely as they are basically acting the same on so many things that it's not even funny, but.... RedHat and Novell?
They are pretty much focused on their market, work well with overall Linux community outside of their market, and never really missed a beat overall.
They are both profitable enterprises too, while Canonical has yet to make any profit and still exists only because MS is secretly pouring money into them.

[schmidtbag]

Root in making open-source enterprise products? You make me laugh. Their roots were in the hardware business. When that went belly-up, so did everything else.

Perhaps "root" was the wrong word, though it was implied that I was referring to what they were actively focusing on when they were still around; as you have noted, hardware was not their main source of revenue by the time they went "belly-up".

Yes, they sold their hardware with together with lots of software, but the simple fact is when they couldn't sell hardware any more, nobody was in the market for their software either.

And their hardware was useless without that software. Much of the software worked on platforms outside of their own. Sure, literally speaking, the company was founded because of hardware, but that's a completely irrelevant detail. You're getting really fixated on details that don't matter to this discussion.

So, to get you back on track:
You stated that most of the products that Oracle got from Sun was a sink-hole.
As we have both established, Sun's hardware was not their primary product at the time of their sale to Oracle.
Therefore, Sun's main source of revenue and interest to Oracle was their software, which was primarily open-source and targeted toward enterprises.
That would imply the "sink-hole[s]" include the software, which again, was the most important factor to Oracle's purchase.
However, if this open-source software became the primary reason for Oracle's downfall, then that means other companies that focus on open-source enterprise software (like RH or Canonical) would also fail.
My point was simple: it's not that black and white.

Good job dragging this out and making this a bigger deal than it really was...

[ldo17]

In my view, Sun and Oracle deserved each other, in the end.

Seems like Sun were the technical leader in the Unixes for a while (apart from SGI with computer graphics). But they never quite got to grips with open source. They freed Java with OpenJDK (after a too-long delay), and then refused to open-source the TCK to go with it. They invented the CDDL for OpenSolaris, seemingly in a deliberate attempt to stymie compatibility with the GPL.

As is common with large, established companies, the world changed, and they found themselves unable to change with it.

[ldo17]

Sun was nicely profitable until their core market completely collapsed beneath them. After that their options were limited.

That took a while to happen. The whole Unix workstation market was destroyed in the 1990s by Windows NT. But Sun was still able to make a nice living for a few more years selling Solaris-based Internet servers. Seemed quite a few customers continued to believe that this Free thing called “Linux” was just too good to be true...