Slashdot Mirror
Oracle Scraps Plans For Solaris 12 (theregister.co.uk)
bobthesungeek76036 writes: According to The Register, Solaris 12 has been removed from Oracle roadmaps. This pretty much signals the demise of Solaris (as if we didn't already know that...) From the report: "The new blueprint -- dated January 13, 2017 -- omits any word of Solaris 12 that Oracle included in the same document's 2014 edition, instead mentioning 'Solaris 11.next' as due to debut during this year or the next complete with 'Cloud Deployment and Integration Enhancements.' At the time of writing, search engines produce no results for 'Solaris 11.next.' The Register has asked Oracle for more information. The roadmap also mentions a new generation of SPARC silicon in 2017, dubbed SPARC Next, and then in 2020 SPARC Next+. The speeds and capabilities mentioned in the 2017 document improve slightly on those mentioned in the 2014 roadmap.
To complete the mashing of jargon, in 2024 - "Objective SPARC Next++" (appending "On Rails" for rack systems).
It must have been something you assimilated. . . .
ZFS for Linux and pretty much every other OS (OpenIndiana, Nexenta, ...) except Oracle Solaris is now being developed by OpenZFS, a fork of the Solaris ZFS code and the two are no longer compatible (version numbers and feature sets have diverged quite a bit).
Not sure what they will do with existing customers, probably bill them a heap load of money for future support, if you're lucky, your pool is old enough or you haven't activated Oracle's proprietary features so that it is still compatible.
Custom electronics and digital signage for your business: www.evcircuits.com
I'll have a soft spot for Solaris, but the writing was on the wall long before this. Solaris was stagnating even before Oracle bought Sun. It was obvious that Sun lacked the resources to maintain their compiler suite and operating as well as actively research new technologies. It was sad to watch, but such is life. Solaris was probably the most usable desktop UNIX before OSX took that crown away. Solaris tried to add much-needed things to UNIX like role-based administration, light-weight virtualisation (zones/containers), non-intrusive profiling/caching (dtrace), advanced storage pools (ZFS), and heaps of other cool stuff. Sun SPARC hardware always had cool high-availability features like being able to disable bad RAM in a running system, really wide system buses for pushing around a lot of data, and was built to survive physical abuse.
The trouble is, they lost out to "good enough". Solaris on SPARCstation was better than WinNT on a whitebox PC, but WinNT on a whtebox PC got to the point where it was good enough, and the added expense of a Sun workstation couldn't be justified. On the server side, Linux became good enough, IBM and Dell x86 servers got to the point where they were good enough, and my 13th generation PowerEdges are definitely better build quality than the Sun V245 servers I still have sitting in a rack for nostalgia.
This took away a lot of their revenue so they couldn't throw resources at research and development. In particular, SPARC fell behind in price/performance/power consumption trade-off, first to AMD's 64-bit Athlons, and then to Intel's post-Netburst Xeon. The UltraSPARC T gave them a bit of a reprieve on highly parallel workloads, but cancelling the Rock was the right decision as it was painfully obvious it wasn't going to compete for single-core throughput/latency performance.
They also lost at the extreme high end to IBM who've managed to get insanely high throughput on POWER with a brute-force approach of throwing better and better cooling systems at a design that's arguably incredibly lazy compared to the E5 Xeon.
Yes, I miss Sun, and I'll shed a tear for Solaris. But I don't miss the Sun that Oracle bought - the Sun I miss had already faded half a decade before Oracle bought the dimly glowing remains.
To some extent we're already seeing this happen, although it's not quite like you describe. Systemd, which is now present in pretty much every major Linux distribution, has caused a lot of problems for a lot of users. A lot of serious Linux users, who need systems that are reliable and robust, have had to switch to FreeBSD thanks to systemd.
Running a Linux distro that includes systemd is not an option for these people/organizations, and running an uncommon or niche Linux distro like Slackware, Devuan or Gentoo is even less viable than running a major one that uses systemd. The only thing that's even less viable than a niche distro is trying to hack a distro like Debian into not using systemd, especially when it breaks dependencies and causes other package management or update complications.
So FreeBSD, and to a lesser extent OpenBSD, have already become "pretty much it" for a large set of former Linux users thanks to unwanted and problematic "innovations", like systemd, forced on users by the major Linux distro maintainers.
I switched to NetBSD because of Red Hat 5.0. It was such a disappointment after Red Hat 4.3.
Is OSX a Desktop Unix? I thought they deprecated X11.
Furthermore, if they had an open desktop environment like maybe KDE or even Gnome, it would be different. Opaque binary-only windowing environments don't qualify as modern Unix. That's like NeWS or any of the other old proprietary croft.
Should really read the official line from Oracle for the reasons for the changes (taken from Register post)...
Here is what Oracle is communicating to customers:
The multi-decade record of SPARC and Solaris platform development and delivery continues with new innovations going forward. Engineering focus on SPARC and Solaris is being continuously applied to leadership in security, scalability, and enterprise reliability for mission critical computing for key customer adoption opportunities in the Cloud and on-premises.
Future features and functionality in Solaris will continue to be delivered through dot releases instead of more disruptive major releases. This addresses customer requirements for an agile and smooth transition path between versions, while providing incremental innovation with assured investment protection. We are amending the Support lifespan for Solaris 11, to extend it considerably beyond any reasonable expected lifetime of use, through at least 2031 and 2034 for Premier and Extended Support, respectively.
See page 37: http://www.oracle.com/us/support/library/lifetime-support-hardware-301321.pdf
Linked off of this page: http://www.oracle.com/us/support/lifetime-support/index.html
"Solaris 11 follows a Continuous Delivery model, where new functionality is delivered as updates to the existing release; upgrades are not required to gain access to new features and capabilities. As a result, Support dates are evaluated for update annually, and will be provided through at least the dates above."
If any of Oracle's customers require an email communication from an engineering executive in summary of the above, Oracle are happy to do so.
I'll have a soft spot for Solaris
You must be a sadomasochist. How can you have a soft spot for anything who's default shell is still ksh? It felt like I was stuck in the 80s every time I had to administer Solaris, AIX, and HP-UX.
IMO, the thing that killed VMS was DEC giving more importance to OSF/1 or Digital Unix. Unfortunately for them, NT on Alpha never caught on, and they tried to make up the difference w/ OSF/1. Instead, had they focused on OpenVMS/AXP, they'd have been a lot better off. That, plus had they complemented NT/AXP w/ Linux/AXP and *BSD/AXP, Alpha might have survived, and w/ it, OpenVMS.
Interestingly enough, Linux has killed off all corporate Unixes - AIX, HP/UX and now Solaris. Only ones left standing are the FOSS distros out there - OpenBSD, NetBSD and FreeBSD on the BSD side, and OpenIndiana, Schillix, Nexenta on the System V side. Ironically enough, it was x86 that enabled Linux to pull this off, even if Linux was cross-platform and supported on just about every CPU out there
SIP (ensuring you can't write to any of the system directories, even as root- this includes /usr and a whole bunch of other locations)
you can if you're an admin worth a damn. This is a very good idea in terms of malware protection and I wish Windows would do it.
it's becoming more and more difficult to use the "Unix" side of OS X.
It's actually exactly the same: open terminal.
Soon enough, they'll remove that utility, along with the one you need to disable SIP, and we'll start seeing people having to "jailbreak" their Mac to bring back any semblance of the freedom they once had under the earlier versions of Mac OS X.
yeah, the anti-mac argument has always been that it's just about to become a problem. It's like how we're always 50 years from fusion. Mac becoming a walled garden is always right around a corner that we never get to.
or one of the BSDs - OpenBSD
FTFY. Only OpenBSD supports Sparc64 on modern Sparc64 - ie with T-series processors. And even that does not support the hardware crypto kit (cos Oracle wont let it).
Oracle need to think again FAST! OpenBSD on the new hardware could be a world beater for serving secure websites (something the world actually needs AND wants). However, they are currently engaged in supporting it in the Ellison traditional manner - with multiple stabs in the back!
Sent from my ASR33 using ASCII
Actually, Linux needs competition or it will start to run out of reasons to make it better. In future, it looks like the BSD family will be pretty much it.
Thanks, Sun/Oracle for erecting barriers around DTrace, thus motivating even better tracing in Linux. Thanks also for doing the same to ZFS, thus saving the rest of us from that sprawling abomination.
Actually, Solaris was de facto a single platform OS - namely for SPARCs. Sun did have that experiment w/ OpenSolaris, but once Oracle sabotaged it, and even surviving forks like OpenIndiana were x86 only, it was a lost cause.
I would like to see SPARC survive, though, w/ either Linux or *BSD on it. It would however be nice if it weren't something available only from Oracle
You can still get SPARC systems from Fujitsu.
But frankly, the only thing SPARC ever had going for it was everything around it. SPARC succeeded despite SPARC, not because of it. Consider:
- SUN produced some awesome workstations and servers.
- Everything used to be open standards (covering SPARC, SBUS, OpenFirmware etc.)
- Solaris stabilized into a nice enough UNIX.
- Lots of Open Source implementations available (Linux, *BSD).
But consider the downsides:
- SUN was swallowed by Oracle.
- SPARC is a nasty RISC architecture. Register windows were really a mistake, and most architectures eschew them as a result. Ditto for delay slots.
- SPARC lagged behind all the other major RISC architectures save for perhaps ARM (which was aimed at low pwoer anyway) in performance.
While SPARC lacked in RISC firepower, it still beasted contemporary x86 CPUs until the Pentium II era (christ, that was 20 years ago!). Since then though, it's just sucked SUN resources as they struggled to keep up with other CPU vendors. They only stayed on top while they could scale up to 64 CPUs when other vendors could not. Once Windows and Linux had caught up with that scaling, and x86 could be reasonable scaled to 16 or more CPUs economically, the writing was on the wall.