Oracle Finally Killed Sun

With the Solaris team gutted, it looks like the Sun skeleton has finally been picked clean.

IMG_2076

The news from the ex-Sun community jungle drums is that the January rumours were true and Oracle laid off the core talent of the Solaris and SPARC teams on Friday (perhaps hoping to get the news lost in the Labor Day weekend). With 90% gone according to Bryan Cantrill that surely has to mean either a skeleton-staffed maintenance-only future for the product range, especially with Solaris 12 cancelled, or an attempt to force Solaris workloads onto Oracle’s SPARC Cloud offering. A classic Oracle “silent EOL”, no matter what they claim as they satisfy their contractual commitments to Fujitsu and others. 

On acquisition, Ellison was scathing about Sun’s management and sure he was going to max out the opportunity. So just how good were Oracle’s decisions with Sun’s assets? I’m not really following Oracle’s business day-to-day, but here’s what seems obvious from reports:

Instead of understanding the real failures at Sun – taking too long to open source Solaris and attempting a marketing-led approach in 2000-2002 instead of Sun’s traditional engineering-led approach – Ellison blamed the man who was landed with the task of rescuing whatever he could from the smouldering ruins left by McNealy, Zander, Tolliver and their clan and their complacent failure. Ellison never understood the pioneering approach Schwartz was taking, instead sneering at blogging and calling all the work-in-progress “science projects” while dismantling the partner channels and alienating the open source community.

The contrast with the approach HPE completed this week with its unwanted legacy products, doing a deal with Micro Focus to look after them, could not be more stark. Oracle said it was going to “reinvigorate the Sun brand” but instead has killed it more dead than any Sun executive managed – the “art of the deal” no doubt. Along with many former Sun staff today, that makes me very sad.

(This post made possible by Patreon patrons. You could be one too!)

68 thoughts on “Oracle Finally Killed Sun

  1. In my opinion Oracle has always been very similar to Microsoft, with only difference, that somehow Oracle managed to present its image much better than Microsoft.

      • The refocus on cloud has led to changes of public stance that value community and especially developer reputation more, sure. The fundamentals in the desktop business (not respecting individual control of systems and using patents to shake down innovators for unearned revenue) is still in place, just not advertised.

  2. Pingback: Links 3/9/2017: Linux 4.13 Out Shortly, Manjaro 17.0.3, ReactOS 0.4.6, Oracle Solaris Layoffs | Techrights

  3. @Libre, you beat me to OpenOffice. When MariaDB came out, I moved there. Who in their right mind would deal with Larry when there is a good alternative.

    I also want to give a thumbs up to Judge William Haskell Alsup. He did a great job of educating himself about how software has been developed from epoch. Without his dedication, things could be much worse in the software copyright/patent space.

  4. Pingback: Oracle Murders Solaris – Mark writes

    • It’s so sad that the “about” page doesn’t even tell what it’s about. 😦 How would I know that’s a solaris or illumos fork without a lot more googling.

  5. I agree with Libre, you are missing the OpenOffice.org story, which had to be forked to LibreOffice.org as Oracle kept ignoring the contributions being made by the community, until in the end it just got forked.

  6. As one of the laid-off, it’s a sad end of a long road. Solaris 12 has been in development for over 5 years, with many features backported to Solaris 11.1, 11.2, and 11.3. But technically 12.0 is not cancelled but rebranded as 11.4. The open question: can the remaining skeleton Solaris staff get a quality 11.4 product out the door.

    • What do you think of the theory that Oracle is attempting to move Solaris workloads onto the SPARC Cloud offering? Presumably a skeleton team would be able to maintain just “one” instance like that.

  7. Netbeans lives on. JDeveloper was migrated to Netbeans Platform so if Oracle wants to continue enhancing JDeveloper the Netbeans devs need to continue employed. Virtualbox is used widely as a platform for devs, won’ t go away anytime soon. Yes Solaris 12 was removed from the roadmap, but replaced with Solaris 11.next. So, what is the hurry, what pressing features does solaris need? Everyone and their mother is moving to “the cloud” FAST. Nobody except Google-Amazon-IBM is buying blades to run in-house anymore. The market and sales are hence slowing. It all comes down to MAKING MONEY which is something Ellison understands and Sun did not, as it was hemorraging money during its last few decades.
    Another innacuracy in your article, Oracle did not sue Google for DESKTOP Java but for
    MOBILE JAVA
    Desktop Java is alive and well… in fact Java 9 will be released soon, with OpenJDK made the reference implementation by none other than Oracle.
    Did I mention that Oracle Linux is also doing well? Oracle’s cloud business is doing well too.

    SPARC is a though call. Competing with AMD and Intel on the top and ARM from below is EXPENSIVE. Again, Oracle wants to make money from services, much like IBM which also de-emphasized its PowerPC business.

    • Hi Max,

      Yes, there are current-market justifications for various decisions, and things like Netbeans do live on. What I’m discussing in this article is not that. When Oracle bought Sun, they brutally criticised its current leadership and bragged about how they were going to do really great things, much better things than the awful management team before could do, tremendous things. Having a long memory, I’m thus comparing that bragging with the actual outcomes.

      You’re wrong about the Google lawsuit – it is actually about Java SE.

      S.

  8. Thanks for the thoughtful write-up. It was becoming clearer to me in last few months when Oracle kept on pushing on moving to a Linux IaaS service rather than Solaris based infrastructure. Whatever the game, I believe companies should come forward and clarify their stand and road map instead of pushing customers for their personal gains. Folks like me will always remember the quality time we had with Sun Solaris..

  9. If they buy or takeovers any company that means that company & its employees are going to garbage.

  10. Pingback: Sunset, finally? | Well Red

  11. Reblogged this on dave levy online and commented:

    Simon was close to the politics of Sun’s “Dash to Open”, my feeling is that Sun had failed before Schwartz was appointed, there was no longer room for differentiated hardware company; Oracle’s failure to monetise the SPARC product line may have been caused by management hubris, but the long term economics of microprocessors and the establishment of distributed & collaborative software technology was key. Sun were late to adopt Plan 9 and making Solaris, an SMP/UMA big iron solution was a cul-de-sac it couldn’t escape from.

    I had promised not to write about Sun’s failure, but I am better now.

  12. The decision to kill off x86 Solaris in the early 2000’s and the bad press, bad customer feelings that followed was the beginning of the end. That and not taking Linux seriously. Buying Cobalt was odd. I don’t agree with you about Swartz. He wasn’t serious enough, he really didn’t care. He could always hightail back to Kinsey with his golden parachute and his wine obsession. Most of us thought he was a plant to take the company down & kill it off. . Giving away IP for the POSSIBILITY of a hardware sale was, and still is, just plain nuts! He seemed to go off the rails a bit in his blog, especially toward the end. There is a reason that no one on Earth has those posts archived.

    • That might be overstating things just a tad. I’ve been a programmer since the 80’s (mostly embedded, including bootloaders) and have never used Forth or even come across it. The wiki page on Forth lists certain projects it is used in – something not even attempted with a language such as C, because it is used everywhere.

    • How Forth is related here? It was invented by Chuck Moore who was not working for Sun nor Oracle. Sure, it is used in bootloaders. Maybe Sun engineers did their implementation but Forth has more implemetations than apps.

    • How Forth is related here? It was invented by Chuck Moore who was not working for Sun nor Oracle. Sure, it is used in bootloaders. Maybe Sun engineers did their implementation but Forth has more implemetations than apps.

  13. Thanks for posting Simon. We just celebrated 20 years of Tech Titan luncheons here in Dallas this year and you were one of the thought leaders who came to educate us about Open Source in the early days. Oracle may have killed Sun but the spirit of the people who grew up there will never die. Thought leadership was in our DNA. The company cared about it employees and it’s customers. We had high ethical standards. That’s the culture that makes great companies and Sun was one of those. Oracle doesn’t come close. #sun #java

  14. The sun has finally set. Sad, but also a relief. Nothing left to be destroyed by Oracle, pain is gone now.

  15. Pingback: Oracle Finally Killed Sun | thechrisshort

  16. What about the Unified Communications System (Sun’s iPlanet/Sun One) Carrier Grade eMail, Calendar, Chat product line?

  17. Pingback: Oracle Finally Killed Sun | Meshed Insights Ltd – Larry Ellison should be ashamed of himself. | BiffSocko.com

  18. What is going to be the impact on the illumos ecosystem? How long for stuff like VMware to stop support for Solaris based kernels? What would be the best move for an illumos distro maintainer?
    Your personal thoughts will be very appreciated.

    • Those are all great questions. The open source Solaris community has been cut off for quite some time so I’d not expect them to be negatively affected.

  19. Pingback: New top story on Hacker News: Oracle Finally Killed Sun – The Internet Yard

  20. VirtualBox “seems unscathed”?
    It seems already being left to die off.
    In particular, after 3 years (see https://www.virtualbox.org/ticket/13471), there’s still no support for Wayland.
    As a dev that’s been using Virtualbox for years and currently locked into Fedora, I’m looking for alternatives…

      • I’ve been saying for years that Oracle must not realize they own VirtualBox, other than rebranding it as Oracle VirtualBox, because they hadn’t ruined it yet. Good to know that they even screwed that up. I’ll look into Proxmox, maybe even get them for FLOSS Weekly.

  21. Don’t kid yourself, HP’s is where acquisitions go to die. Just look at Mercury, ArcSight, and others. While it’s too bad about Sun ( disclaimer I worked there from ’87-94 ) as Scott would say, ” It’s lunch or be lunch.”

  22. Pingback: Oracle Finally Killed Sun - https://meshedinsights.com/2017/09/03/oracle-finally-killed-sun/ … - Subdwellerz Beatz

  23. Pingback: Sun set: Oracle closes down last Sun product lines | Fleekist.com

  24. I’m Roger Faulkner’s (RAF) daughter. I don’t understand a lot of this article since I’m not a hacker. I just want to know if this means that PROCFS will be dead in the water. Will his disappear?
    Bethany Faulkner

  25. Bethany – your Dad’s legacy is going NOWHERE! It lives on not only in any Oracle Solaris deployment that still moves bits, but it also lives on in illumos – the still and forever open-source successor to OpenSolaris. Here’s procfs, for the world to see, for example:

    http://src.illumos.org/source/xref/illumos-gate/usr/src/uts/common/fs/proc/

    I had the pleasure of working at least a little with Roger. He was a champion not only of his own technology, but others’ as well, including some things I did.

    Roger’s legacy will never EVER disappear.

  26. Another aspect of Roger’s legacy which is very, very much alive, is the attitude towards software architecture and engineering which he fostered. We all carry a bit of that with us.

    #WWRD

  27. Bethany, Roger’s baby /proc lives on not only in Illimos, but also in Linux. They extended it to do many more things than just expose the process model. Roger hated that on grounds of purity, but it’s a mark of a good idea when it gets expanded far beyond its creator’s original intentions.

  28. Pingback: Oracle Finally Killed Sun | laage65

  29. Thank you so much, gentlemen. It’s been a year of gradually losing little bits of him. I’m relieved his baby lives on. I’m so sorry for all those who’ve lost their jobs. I’m sure there would have been much loud swearing coming from his upstairs office, had Roger lived to see this.

  30. Pingback: Oracle chính thức khai tử Sun | Caphesang.vn

  31. My time at Sun was one of the best learning experiences I ever had. It was definitely a trial by fire. A trial made easier by the great people I had the pleasure of knowing and working with. Sun became great by placing the customer first. Sun’s downfall started when they lost sight of their customers’ needs and refused to see the potential in the average consumer user. There is a lesson that can be learned from Sun’s sunset.

    • After having been in the IT business for 40 years i can tell that this lesson will never be learned. L’histoire se repete, but not the same way.

  32. It took Ellison 9 years to kill Sun, that in my opinion, was the biggest insult, it should have been swift and painless.

  33. It seemed to me that the Sun business was doomed from the beginning, as Oracle exists on high software margins. The low margin systems business was a drag on the business. It was justified on strategic factors, such as developing on Sparc and leveraging the huge Java following. They were always a top down company, both under McNeally and Ellison. They suffrred the same fate as other top down companies such as DEC, IBM, and HP. Sun was always stocked with engineering talent, even at the end. They should find jobs easy enough.

  34. Sun always put HW before SW and especially services. Man cannot live by HW alone it sayeth somewhere. I worked for Sun for 5 years and here are some of the SW backtracks and executions;
    • Genesys, a futuristic view of linked nodes with dynamic changes to almost everything in the way of resources one desired. This was touted widely for a while, and even I went on the road ‘selling’ the vision but it suddenly disappeared
    • JOE (Object Request Broker)
    • NEO (renamed from Distributed Objects Everywhere)
    • Enterprise Manager
    • Domain Manager
    • Site Manager
    • Job Scheduler
    • Java Management API (JMAPI)
    • Javastations 1 and 2
    • A7000 disk subsystem after the Encore acquisition (I project managed the UK side of this)
    • Scott McNealy’s ‘auction’ plans
    • Numerous ‘unproductive acquisitions’ over the years , the latest being Tarantella (old SCO)
    • The demise of the ‘Sparc and Solaris’ is all philosophy and the acceptance of Linux
    • Utility grid computing
    • Senior manager exits, compounding the earlier Ed Zander exit – unannounced and only visible by searching the Internet.
    The old Communist empire has nothing on Sun when it comes to covering things up.
    Having said all that, they were a great company to work for, unlike Oracle, who I also worked for. Oracle were a ‘hire and fire a will’ company and probably still are. People disappeared as quickly as they did in Nazi Germany.

  35. Pingback: tinman alley » Bye bye Sun and Solaris :-(

  36. And what of VDI? Sun Ray? Ellison coined the term “The Network Computer” back in the 90’s and had a brilliant offering, staffed by really knowledgeable, enthusiastic, brilliant people and he killed it. I guess because although it was a profitable business, it just didn’t make enough money.

  37. The beginning of the end for Sun was the dot-com crash. Sun servers powered the dot-coms then when that all came tumbling down, all that equipment went on the used market and Sun could no longer sell new product with all that nearly-new stuff going for a deep discount.

  38. Pingback: Sun set: Oracle closes down last Sun product lines | All About Tech in News

  39. Pingback: Newsletter: September 9, 2017 | Notes from MWhite

  40. Sun was dedicated to and driven by its channel partners. The direct sales force was great to partner with even in the largest of accounts. Oracle has always been a direct sales force company. They claimed to be channel friendly but hot air is hot air. They really lacked top down leadership in sales. Thats what McNealy got and Ellison didnt.

    If you had an Oracle oportunity from 2008 on then 3 different internal Oracle sales groups pounced. None of them knowing or understanding the role of the other.

    leadershipsalesdirection.

  41. Pingback: ITPS251: Hurricane Equifax | Same3Guys

  42. To be fair, Oracle kept Sun going longer than folks expected, and much longer than the life it would have had at IBM, it’s first suitor. Scott probably had trouble sleeping seeing it go to IBM, and better for a silicon valley home with closer culture match. Oracle made a kickbutt DB stack sitting on Sun HW with it’s Exa line, even running on Solaris, that still runs circles around other stacks.

Comments are closed.