From the comments
>This individual case is just the result of terminating Solaris 12 development. Part of the old Sun software, operating system, picked up when Oracle bought Sun. Not a big deal.
So so canning the special operating system for your special hardware which is purpose built for your flagship product in 'not a big deal'?
This week I told the rep from Netsuite that there is no way we are buying Oracle products, if we can help it, and he seemed actually shocked. How long before Salesforce comes in house and is ruined too? I would spend a lot of labour on Postgres to avoid Oracle again.
Well, he'd better get ready to keep getting shocked. Companies have been divesting their Oracle infrastructure for years now. Nobody wants to be audited by Oracle and then find they have to purchase useless products to ensure they don't get sued.
From a security perspective, Oracle is pretty awful. They have a history of responding to bug reports with cease and desist letters for violating copyright.
Their support is dreadful - and given they are such a black box then there are many folks who actually need decent support. It's very expensive to tune workloads, and their CBO is such a pain that anytime even a point release comes out you need to carefully reproduce every query rewrite, index and set of statistics to see if it is screwing up somewhere.
In short: there are vanishingly less reasons to use them, be they either technical, monetary or ethical. No one will shed a tear if this behemoth goes under.
I told the rep from Netsuite that there is no way we are buying Oracle products, if we can help it
I would be very careful of statements like this, unless you're the CEO. Standard modus operandi for Oracle sales people is for them to go around technical people and work relationships with C-suite types.
So don't be surprised if your rep is playing golf with your CEO sometime in the near future.
I can't help but wonder what would have been different had IBM been successful in the Sun acquisition. It would have been a strange marriage, and there were a lot more staffing and product redundancies that would have still hurt customers and employees. On the other hand, IBM is often aloof but never callous like Oracle is.
IBM's acquisition of Sequent might be a reasonable starting place to guess. Sequent customers got a fairly abrupt EoL. There was a lot of drama with Itanium hype and even the x450/x455, then IBM fleeing that IA64 sinking ship. And of course Project Monterrey and the SCO lawsuit. But I'm not so sure how sad the world was to lose Dynix/PTX and instead Linux got RCU and a ton of general NUMA scalability know how.
Anyway, hindsight being 20/20, most of us should have probably switched to OpenSolaris back in 2005 and figured out how to wrestle it out of Sun's control for their own good.
Back in the day (2000-2005), Oracle on Solaris was rock solid. There was basically no better platform or database configuration that was better than this, including SQL Server/Windows 2000/2003, etc.
It's a shame that Solaris is basically dead, but Linux killed them swiftly by being "good enough" on commodity hardware, and then Oracle did a coup de grace with their death-embrace by buying them. It's a shame, because so many great things were being done by Sun.
Oracle Solaris might be dying, and Linux sure has a large market share, but illumos as the open source continuation of OpenSolaris is doing very well.
If you enjoyed ZFS, Zones, DTrace and so on in the past there is a good chance that you'd like OmniOS, one of the illumos distributions for classic server environments.
There are more distributions like SmartOS that tkae a more innovative approach as cloud-hypervisor that is just a slim live system.
Most of the original ZFS Developers are now working on OpenZFS in illumos.
Beeing open source also means collaborations with other OS projects like FreeBSD (bootloader and more), NetBSD (pkgsrc) is happening.
What does libraries team entail? I run a FreeBSD kernel team, and am interested in speaking with any Solaris kernel people you may know to fill some additional roles.
I find this whole thing, as in Oracle itself I guess, hard to understand.
In their presentations at Hot Chips for the past couple of years they've repeatedly highlighted hardware features which seemed to be only accessible using next gen Oracle software running on an upcoming version of Oracle's proprietary Solaris.
So if they're giving up on Solaris 12 development how are their customers going to fully leverage this new hardware they're buying from Oracle? If they're unable to use it, why buy/lease hardware from Oracle at significant premium when you can just use commodity x86 compute... and if you're just using commodity x86 compute why step up for the licensing costs of Oracle software if there is anything else on earth which might suffice?
Does anyone track the frequency and size of layoffs from tech companies? It certainly feels that there has been a notable uptick in layoff news over the past 6-12 months.
So going forward for corporate users it's going to be either Linux or AIX considering the way HP-UX, Irix and now Solaris is slowly seemingly heading to.
What I'd really like to see is harnessing of the knowledge, principles and methods of solid system management and stability that have been built and field tested on Unix platforms over the decades. This would benefit whichever platforms remain (and are created/update) going forwards.
From a sustaining Sun's business perspective, it would have been a very poor fit. Sun had hardware product development as a principle effort, a variety of support channels and organizations for hardware and software, professional services, marketing, technical and executive sales etc. Basically many things Google is anemic at. In a way, Google kind of hacked the requirements expected of a business and it sort of works for many of the products they offer, but you can see where they totally flop with the Motorola acquisition and other ventures like Google Fiber. This stuff is hard, and it's really hard to find people that really care and have the ability to execute instead of simply being "qualified"
From a talent and technology acquisition perspective, it would have been absolutely sage. Sun hardware folks could have shifted away from general purpose CPUs to switching and packet processing. OpenSolaris was a tour de force, especially with a legitimate hindsight comparison to where Linux was at the time. If they could have cherry-picked parts of Sun and shed the sustaining business around storage, commercial Solaris, SPARC etc that would have been ideal.
[+] [-] jimnotgym|9 years ago|reply
So so canning the special operating system for your special hardware which is purpose built for your flagship product in 'not a big deal'?
This week I told the rep from Netsuite that there is no way we are buying Oracle products, if we can help it, and he seemed actually shocked. How long before Salesforce comes in house and is ruined too? I would spend a lot of labour on Postgres to avoid Oracle again.
[+] [-] toomuchtodo|9 years ago|reply
The best time to start was years ago. The second best time to start is today.
[+] [-] chris_wot|9 years ago|reply
From a security perspective, Oracle is pretty awful. They have a history of responding to bug reports with cease and desist letters for violating copyright.
Their support is dreadful - and given they are such a black box then there are many folks who actually need decent support. It's very expensive to tune workloads, and their CBO is such a pain that anytime even a point release comes out you need to carefully reproduce every query rewrite, index and set of statistics to see if it is screwing up somewhere.
In short: there are vanishingly less reasons to use them, be they either technical, monetary or ethical. No one will shed a tear if this behemoth goes under.
[+] [-] PhantomGremlin|9 years ago|reply
I would be very careful of statements like this, unless you're the CEO. Standard modus operandi for Oracle sales people is for them to go around technical people and work relationships with C-suite types.
So don't be surprised if your rep is playing golf with your CEO sometime in the near future.
[+] [-] TazeTSchnitzel|9 years ago|reply
[+] [-] kev009|9 years ago|reply
IBM's acquisition of Sequent might be a reasonable starting place to guess. Sequent customers got a fairly abrupt EoL. There was a lot of drama with Itanium hype and even the x450/x455, then IBM fleeing that IA64 sinking ship. And of course Project Monterrey and the SCO lawsuit. But I'm not so sure how sad the world was to lose Dynix/PTX and instead Linux got RCU and a ton of general NUMA scalability know how.
Anyway, hindsight being 20/20, most of us should have probably switched to OpenSolaris back in 2005 and figured out how to wrestle it out of Sun's control for their own good.
[+] [-] joeblau|9 years ago|reply
[1] - http://oracle-team-usa.americascup.com/en/news.html
[+] [-] tlrobinson|9 years ago|reply
Not sure how much (if any) Oracle itself put into it.
[+] [-] jrockway|9 years ago|reply
[+] [-] pfarnsworth|9 years ago|reply
It's a shame that Solaris is basically dead, but Linux killed them swiftly by being "good enough" on commodity hardware, and then Oracle did a coup de grace with their death-embrace by buying them. It's a shame, because so many great things were being done by Sun.
[+] [-] snw|9 years ago|reply
If you enjoyed ZFS, Zones, DTrace and so on in the past there is a good chance that you'd like OmniOS, one of the illumos distributions for classic server environments.
There are more distributions like SmartOS that tkae a more innovative approach as cloud-hypervisor that is just a slim live system.
Most of the original ZFS Developers are now working on OpenZFS in illumos. Beeing open source also means collaborations with other OS projects like FreeBSD (bootloader and more), NetBSD (pkgsrc) is happening.
This is all independent of Oracle.
[+] [-] pjmlp|9 years ago|reply
They do the same by outsourcing their development expenses to the FOSS community.
[+] [-] ariwilson|9 years ago|reply
[+] [-] jclulow|9 years ago|reply
[+] [-] drdeadringer|9 years ago|reply
Only now do I recognize the blue towers at the GooglePlex as resembling those purple SGI computers.
[+] [-] rayban|9 years ago|reply
[+] [-] kev009|9 years ago|reply
[+] [-] noobermin|9 years ago|reply
[+] [-] Auzy|9 years ago|reply
No idea why Oracle hasn't been split apart
[+] [-] detaro|9 years ago|reply
[+] [-] Quequau|9 years ago|reply
In their presentations at Hot Chips for the past couple of years they've repeatedly highlighted hardware features which seemed to be only accessible using next gen Oracle software running on an upcoming version of Oracle's proprietary Solaris.
So if they're giving up on Solaris 12 development how are their customers going to fully leverage this new hardware they're buying from Oracle? If they're unable to use it, why buy/lease hardware from Oracle at significant premium when you can just use commodity x86 compute... and if you're just using commodity x86 compute why step up for the licensing costs of Oracle software if there is anything else on earth which might suffice?
[+] [-] chambo622|9 years ago|reply
[+] [-] frik|9 years ago|reply
Search for tech bubble, unicorn corpse/unicorpse.
Unicorn list to watch: http://fortune.com/unicorns/
Best sites to watch: https://www.cbinsights.com/research-downround-tracker and https://www.cbinsights.com/blog/startup-failure-post-mortem/ and https://ourincrediblejourney.tumblr.com/
[+] [-] dajohnson89|9 years ago|reply
[+] [-] steinerj|9 years ago|reply
[+] [-] sengork|9 years ago|reply
What I'd really like to see is harnessing of the knowledge, principles and methods of solid system management and stability that have been built and field tested on Unix platforms over the decades. This would benefit whichever platforms remain (and are created/update) going forwards.
[+] [-] HillaryBriss|9 years ago|reply
[+] [-] EduardoBautista|9 years ago|reply
[+] [-] rch|9 years ago|reply
[+] [-] kev009|9 years ago|reply
From a talent and technology acquisition perspective, it would have been absolutely sage. Sun hardware folks could have shifted away from general purpose CPUs to switching and packet processing. OpenSolaris was a tour de force, especially with a legitimate hindsight comparison to where Linux was at the time. If they could have cherry-picked parts of Sun and shed the sustaining business around storage, commercial Solaris, SPARC etc that would have been ideal.