Slashdot Mirror


James Gosling Grades Oracle's Handling of Sun's Tech

snydeq writes "With the four-year anniversary of Oracle's Sun Microsystems acquisition looming, InfoWorld reached out to Java founder James Gosling to rate how Oracle has done in shepherding Sun technology. Gosling gives Oracle eyebrow-raising grades, lauding Oracle's handling of Java, despite his past acrimony toward Oracle over Java (remember those T-shirts?), and giving Oracle a flat-out failing grade on what has become of Solaris OS."

223 comments

  1. Oracle's JAVA by Anonymous Coward · · Score: 0, Informative

    Yep, they have done so well with it that everyone turns JAVA off in their browsers. Gotta love the lack of security it gives.

    1. Re:Oracle's JAVA by hlge · · Score: 5, Informative

      Dhu, The main usage of Java is on the server side, where it's fairly popular. Java skils is still the most sought after skill when it comes to developers. There are a few popular desktop applications written in Java, Minecraft comes to mind :) And of course we have the slightly modified version of Java that powers every Android application. So Java is still around and kicking..... As to the big Reds handling of Java, out of the gate it was pretty bad in it's interactions with the Java community. Not surprising as they axed most of the folks that where doing that part back at Sun...... But they actually got better with interacting with the Java community lately, could improve more but still kinda on the right track.

    2. Re:Oracle's JAVA by Anonymous Coward · · Score: 1

      Nope. Andorid is not powered by Java. That's what Oracle would want you to believe. It's like claiming that anything containing flour is a derivative of "bread". Dalvik isn't anything like Java.

    3. Re:Oracle's JAVA by Cramer · · Score: 1

      Even better... Chrome won't even run the java plugin 9 times out of 10. (even then, you have to OK it's running of any applet)

    4. Re:Oracle's JAVA by unixisc · · Score: 1

      If a VM uses a register base instead of a stack base, does that change the entire VM altogether, rather than simply make it more efficient?

    5. Re:Oracle's JAVA by Chrisq · · Score: 1

      If a VM uses a register base instead of a stack base, does that change the entire VM altogether, rather than simply make it more efficient?

      It changes it altogether, op codes, and the lot. Its like a different CPU (e.g. Arm or Intel) programs have to be compiled differently to run on them and some things may be more efficient on one and other on the other.

    6. Re:Oracle's JAVA by allcoolnameswheretak · · Score: 2

      Maybe the guts of it aren't. To a developer though, Dalvik is 95% Java.

    7. Re:Oracle's JAVA by melmut · · Score: 1

      They both use the Java language.

    8. Re:Oracle's JAVA by Chrisq · · Score: 1

      They both use the Java language.

      They both use Java language compiled to different byte codes, just as if you could write a C program and compile it for an arm machine or an intel machine.

    9. Re:Oracle's JAVA by drinkypoo · · Score: 1, Interesting

      Just to be clear, you're upset that Google's attempt to take over the web doesn't play well with competing technologies? I do hope you're not surprised by this. You're helping Google shit on the web.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    10. Re:Oracle's JAVA by gbjbaanb · · Score: 3, Interesting

      as many people do - get a copy of an app inspector - I recommend Addons Detector - and use it to see what dev tools were used for build the apps on your phone. You'll be surprised to see just how many were built with the NDK. All the fast and responsive games are at least.

    11. Re:Oracle's JAVA by Dcnjoe60 · · Score: 1

      Yep, they have done so well with it that everyone turns JAVA off in their browsers. Gotta love the lack of security it gives.

      Are you confusing javascript with JAVA? None of the browsers I use come with built in JAVA. I have to install it if I want to use it. Javascript, on the otherhand, is a totally different story.

    12. Re:Oracle's JAVA by Cramer · · Score: 1

      No, it's more of a "how many times do I let you fool me" kind of thing. Google has learned Java is not something that can NOT be trusted. As a result, chrome will refuse to load some versions entirely, others nags about constantly, and even the current latest versions require user confirmation before running the known-to-be-buggy-as-hell system.

      While I find it annoying, I do agree with them.

    13. Re:Oracle's JAVA by Anonymous Coward · · Score: 0

      Dalvik isn't anything like Java.

      You don't write programs in Dalvik because Dalvik is not a language, it is a virtual machine much like the Java Virtual Machine (JVM). You write programs for Dalvik and the JVM in Java. You cannot compare Dalvik and Java however you could compare Dalvik and the JVM.

      The version of Java that you write Android programs in is slightly modified in that it contains some different Android-specific APIs that are implemented in Dalvik and not in the JVM but by and large the Java code that you write for the JVM will compile and run on Dalvik.

  2. Forgetting OpenOffice.org by jordanjay29 · · Score: 4, Insightful

    Even though it's since transitioned to Apache, Oracle still deserves to be graded on their handling of OO.o.

    1. Re:Forgetting OpenOffice.org by bhcompy · · Score: 2

      Which is what exactly? I never had any problems with their handling of OO.o. It was free, it wrote files in formats that were easily accessible, it didn't have ugly ribbons, it was intuitive to reformed Office users, it loaded faster than Office, and it handled everything school, home, and most of what work threw at me without issue. For a freeware product, what more could you ask for? Sounds like a good grade to me.

    2. Re:Forgetting OpenOffice.org by R.Mo_Robert · · Score: 3, Informative

      Even though it's since transitioned to Apache, Oracle still deserves to be graded on their handling of OO.o.

      Gosling didn't "forget" to grade OpenOffice.org; he was the (co)creator of Java. That's why this article is treating his assessment of Java as special. You wouldn't get that with OO.o.

      --
      R.Mo
    3. Re:Forgetting OpenOffice.org by Anonymous Coward · · Score: 1

      I will give you everything except "it loaded faster than office", as a user of both for work the speed of OO is one of its worst features.

    4. Re:Forgetting OpenOffice.org by Rennt · · Score: 5, Insightful

      You are taking about the product itself, not Oracle's handling of the project.

      Yes, OpenOffice could open your documents fine. It did all that stuff before Oracle came along, alienated the developer base and ran the project into the ground.

    5. Re:Forgetting OpenOffice.org by geminidomino · · Score: 1

      It actually does (did?) load faster than MS office, if one tweaks their system to prevent the pre-emptive loading of MS office, or allows the OO.o pre-emptive loader to run.

      Haven't seen QuickStarter in awhile though. Maybe they did away with that.

    6. Re:Forgetting OpenOffice.org by armanox · · Score: 1, Insightful

      My understanding was the developers simply left because Oracle acquired the product, not because of anything they did.

      --
      I'm starting to think GNU is the problem with "GNU/Linux" these days.
    7. Re:Forgetting OpenOffice.org by Anonymous Coward · · Score: 0

      OO doesn't have the advantage of preloading some of their code into memory like Microsoft does. Unless, of course, you're running the quick launcher at start up.

    8. Re:Forgetting OpenOffice.org by MrBigInThePants · · Score: 2

      Then you would be wrong.

      There was much written about their views at the time.

    9. Re:Forgetting OpenOffice.org by Anonymous Coward · · Score: 0

      Why?
      It wasn't exactly a first class citizen even under Sun, and it never really fit with Sun's product portfolio or strategy. It didn't fit anywhere in Oracle's strategy either. If you insist on grading them on it, I'd give Oracle an A+ as handing it off to Apache was the best thing that could have reasonably been done of OOo.

      Seriously, in the grand scheme of databases, application servers, middleware, and the provision of your entire server stack down to the hardware, where does an office suite fi into the mix, especially one that brings in zero revenue, and why would you (reasonably, as in not because of that warm and fuzzy OSS feeling, or because you want them to) expect them to handle it any differently than they have?

    10. Re:Forgetting OpenOffice.org by fatphil · · Score: 1

      But he is also commenting on now-Oracle-controlled things he didn't create, such as Solaris. Quite why that was included and others weren't doesn't seem obvious.

      --
      Also FatPhil on SoylentNews, id 863
    11. Re:Forgetting OpenOffice.org by bhcompy · · Score: 1

      And what changed? Not much.

    12. Re:Forgetting OpenOffice.org by Anonymous Coward · · Score: 0

      Well, I'd say by passing it to Apache, they deserve a good grade for that.

    13. Re:Forgetting OpenOffice.org by Anonymous Coward · · Score: 1

      Sun paid a large team of developers and marketing dedicated to StarOffice and OpenOffice.org. Oracle laid off the entire team.

      Source: I was on that team.

    14. Re:Forgetting OpenOffice.org by Jonner · · Score: 1

      Everything you laud was there before Oracle got ahold of it and remains in both OpenOffice and LibreOffice. Are you saying that Oracle should be praised for not destroying OpenOffice? That seems like a pretty low bar.

    15. Re:Forgetting OpenOffice.org by bhcompy · · Score: 1

      More like there wasn't that much to do. There is a point where a product is done and enters maintenance mode. OpenOffice didn't get the MS Office treatment of redesigning it for the sake of redesigning it, and there is something to be said for that

    16. Re:Forgetting OpenOffice.org by cheesybagel · · Score: 1

      I have always found that weird considering Larry hates Microsoft more than anything.

    17. Re:Forgetting OpenOffice.org by Jonner · · Score: 1

      Yes, Oracle probably hasn't been as bad a Microsoft with regard to changing things that don't need to be changed. From what I understand, most of the frustration which resulted in LibreOffice forking from OpenOffice was because people didn't think Sun and then Oracle allowed contributions at a sufficient rate.

  3. That James Gosling fella is of no consequence... by bogaboga · · Score: 4, Interesting

    While I applaud James for his contribution to Java, I am afraid he's of no consequence to its direction now.

    It would have been better if he proposed some kind of direction Oracle should have taken with Java.

  4. hard to fault Oracle by Anonymous Coward · · Score: 0

    for ignoring solaris in favor of linux

    I admined solars for years and it was always a confusing blend of bsd and system v, it is hard to imagine how much cruft could have built up over the past three decades. It is not difficult to see how focusing on linux would seem appealing

    Ignoring their x86 servers in favor of selling pricey sparc gear (because the sales reps made bigger commissions) was part of the reason that sun passed away in the first place

    1. Re:hard to fault Oracle by Anonymous Coward · · Score: 1

      Solaris has actually been quite good since Solaris 10. They have a very worthy replacement for init (the service management framework), a very advanced infrastructure for application debugging and probing (dtrace), and a revolutionary filesystem (ZFS) to name a few things. The only thing that (IMO) sucks is the reliance on the sysv style package manager (which they've since replaced with IPS in Solaris 11; not too familiar with it since our company has largely abandoned Solaris thanks to Oracle, but I am of the opinion they should have went with rpm/yum or dpkg/apt).

      The tech is stellar and arguably better than Linux in many respects, but I agree that with the failure of Sun and Oracle to find addressable market with SPARC and no real chance of displacing Linux on x64 that it really is a dead end now. It's too bad Oracle won't invest any time to port some of the better aspects of Solaris to Linux.

    2. Re:hard to fault Oracle by Anonymous Coward · · Score: 0

      I can see how solaris can be appealing in a lock-down telecom environment, their security is probably more mature than we see in a linux distro. One of my tasks, pre-millenium, was to run oracle backed web apps in solaris and I found it to be pretty straight forward. Around the same time, my friends that were admining porn sites were all using freebsd because it was highly secure as a default, not begging you to install every whiz-band add on, with their associated weaknesses. ZFS was pretty exciting back in the day, along with solstice, but I still do not see their advantages over other san and backup solutions

    3. Re:hard to fault Oracle by armanox · · Score: 1

      Solaris is appealing in environments that you don't want to change or break. It's very stable.

      Think of ZFS along the lines of RAID - it's not a backup or disaster recovery solution. It's performance and disaster prevention (like Shadow Copies in Windows).

      --
      I'm starting to think GNU is the problem with "GNU/Linux" these days.
    4. Re:hard to fault Oracle by Anonymous Coward · · Score: 0

      It's too bad Oracle won't invest any time to port some of the better aspects of Solaris to Linux.

      Why the fuck would they want to port Solaris' better aspects to their competitor, when Solaris is still a billion dollar business for Oracle?

      SPARC highend machines are currently the best. Until POWER8 ships, and it's not clear that will be faster than an M5. Five years ago I would have declared SPARC dead, with POWER so far in the lead, SPARC looked like a joke. No more.

      Intel coherency just doesn't scale, there are some "large" x86 machines from SGI (and others?), but the limitations of present coherency protocol in Intel Xeons means they just won't perform anything like POWER or SPARC on contentious workloads like databases.

    5. Re:hard to fault Oracle by Anonymous Coward · · Score: 0

      I was comparing zfs to san solutions like emc, storageworks, clariion and 3par (and solstice to tivoli) which I have used with sparc (and other) servers since the mid 90's. Maybe there is some appeal in getting that functionality out of your os, but I never really saw the appeal in stuffing a server full of disks over fc fabric

    6. Re:hard to fault Oracle by _merlin · · Score: 1

      What? They were doing a pretty good job of pushing you to their AMD servers by making the price/performance ratio of the SPARC gear shit for years.

    7. Re:hard to fault Oracle by thogard · · Score: 4, Informative

      ZFS is on the right path but it still isn't quite where it needs to be. For example I can't tell it not to reallocate blocks on write so I can't force overwrites of sensitive data -- which is required in several industries that Sun used to be strong in. Someone in ZFS land needs to create an ioctl/fctl to fix that. The boot system also needs to be clear if it is trying to mount a ZFS or UFS disk since that is a bit tricky when the disk looks like both. They should also fix the fsck stub so it knows about ZFS and have a /usr/lib/zfs/fsck even if it is just a link to zfs status.

      How is SMF better than init? They even bothered to break init so you can't pull SMF out the system if you don't want it. They now link init and smf to a number of libraries that have horrible security records. Do you want the main process in your system linking in libraries that need security updates on a monthly basis?

      I know how SMF is worse, it is slower to start up, it is indeterminate in its start up state and order, it keeps its data in unauditable binary files an it takes far longer to shut down. It also isn't very good at what init was, which was making sure programs always ran. Solaris 11.1 turns off auditing, then syslog before killing off all user processes which means you have no idea what a rogue process did when it was told the system is shutting down. That appears to be a result of someone at Oracle deciding all the disks need to be mounted before starting syslog, which requires lots of extra crud to be running like NFS, RPC and whatever YP is called this decade and it appears that stuff is all trusted to shut down cleanly without the need of logging. At least with init, you could have two different syslog entries for the different run levels so you could make sure everything was logged and audited.

      The number of bugs in Solaris 10 is far worse than Solaris 9. You can't build a light weight Solaris 10 or 11 system. Under 10, you could build a Solaris 9 container which would only run a bare number of processes but not any more since that feature was pulled out of 11. I have a number of Solaris 9 systems that are running less than a dozen packages but I'm one of the people who feel that if there isn't any unneeded software on a system, hackers can't use it hack the system.

      Solaris 11 also has managed to break decades of sanity of using ifconfig to build network stacks. Now there are other tools that do part of the job and then can allow ificonfig to finish the job.

      At least with Solaris 11.1 they created a tool to create smf xml files which means they are now no longer hand crafted which means a tool can be written to turn them back into rc.X scripts and they can be put back where they belong. Now if I could just remove svc.* without installing a fake to keep the contract open, I would be back up to the integrity level of a Solaris 9 system.

    8. Re:hard to fault Oracle by hlge · · Score: 1

      A few comments.

      Oracle isn't ignoring Solaris for Linux, they are actually putting a fair amount of resources behind Solaris, the issue is licensing cost for non Oracle HW and maybe even more the way the treated the budding OpenSolaris community early on, when they closed the doors on Solaris.

      One more thing is kinda where they are focusing their efforts, rather than going after new developers, the focus is on Oracle on Oracle. So even if Oracle are investing in Solaris, they are primarily doing so to run the Oracle stack.

      If you "admined Solaris for a few years" a few years back, you should take Solaris 11 for a spin, it's surprisingly fresh and the CLI env can either be configured to use GNU tols like most Linux dists does or old school "Solaris", You might like it.

    9. Re:hard to fault Oracle by gweihir · · Score: 1

      I don't agree. I need local sockets and their performance is atrociously bad on Solaris 10 (no, I cannot use shared memory or the other alternatives). Linux performs something like 30 times better. Thread creation, memory management, etc. are also pretty bad. Unfortunately my customer cannot migrate to Linux at the moment, but they are thinking about it pretty intensely.

      --
      Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
    10. Re:hard to fault Oracle by unixisc · · Score: 1

      The tech is stellar and arguably better than Linux in many respects, but I agree that with the failure of Sun and Oracle to find addressable market with SPARC and no real chance of displacing Linux on x64 that it really is a dead end now. It's too bad Oracle won't invest any time to port some of the better aspects of Solaris to Linux.

      The issue he had was w/ Oracle's pricing of Solaris, and the pricing of their support to the OS, which has effectively killed it. Since Gosling himself had to convert to Linux due to that. He wasn't commenting on any technical aspects to what Solaris now is.

    11. Re:hard to fault Oracle by unixisc · · Score: 1

      Solaris is relevant only in environments where one is stuck w/ a legacy SPARC based environment. Particularly since the OSs that support SPARC have dwindled - right now, it's just Debian on the Linux side, but all the BSDs - F/O/N on the BSD end.

      Otherwise, if one wants ZFS, one could go w/ FreeBSD on either Intel (x64 OR Itanic) or POWER or MIPS.

    12. Re:hard to fault Oracle by Cramer · · Score: 1

      IMO (shared with many of my peers), Solaris 10 is when it died. And it's entirely because of SMF. It has too much of a Windows Stink(tm) to it... want to change a service configuration, run the SMF equiv of "regedit". The init system has a database -- it replicates ("rolling backup") at startup -- instead of clearly defined, easy to understand, and trivial to edit configuration files. While SMF does add one or two notable features -- automatic dependency trees, parallel startup, error handling... Most of it has already been done with the existing (shell scripted) init framework. And parallel is one of those things that looks good on paper, and sounds good in the conference room, but when actually done just makes a mess. (one any windows users should be very familiar with... the complete inability to use a system for several minutes post-boot because 87 applications are thrashing the disk all trying to start at once. Yes, it's annoying having to sit and watch the machine do nothing for a minute waiting for sendmail's dns lookup to timeout to finish booting, but the start-almost-everything-at-the-same-time alternative is *worse*)

    13. Re:hard to fault Oracle by dbIII · · Score: 2

      For example I can't tell it not to reallocate blocks on write so I can't force overwrites of sensitive data

      I know that you have such requirements forced on you by others, but with respect that one is quite ridiculous and appears to be a rule from tape or other removable storage or disk disposal that has been badly misapplied to filesystems possibly by accident or teenager writing Quality Assurance rules. If someone gets to your drives at a block level via root on your system or by physical access to the server you are pretty well fucked for a dozen reasons anyway.
      IMHO overwrites are something to be done when media is about to leave a "secure" area so not something a filesystem, even one like ZFS, should be bothered about.

    14. Re:hard to fault Oracle by unixisc · · Score: 1

      for ignoring solaris in favor of linux

      I admined solars for years and it was always a confusing blend of bsd and system v, it is hard to imagine how much cruft could have built up over the past three decades. It is not difficult to see how focusing on linux would seem appealing

      But didn't Sun move away from BSDisms when they moved from BSD based SunOS to SVR4 based Solaris?

      Ignoring their x86 servers in favor of selling pricey sparc gear (because the sales reps made bigger commissions) was part of the reason that sun passed away in the first place

      But had they gone w/ x86 servers, they'd have been trampled by the likes of Dell, HP/Compaq/DEC, IBM - all of who were far more established in that space. Besides, had they gone w/ x86, they'd have been right up against Microsoft Windows Server as well. So it made sense for Sun to stick w/ SPARCS.

      What they could have done - try proliferating the market w/ SPARC boxes at different configurations w/ different price points. Solaris for top end servers, FBSD or Linux for SPARC workstations, Linux for SPARC laptops, OBSD for routers, NBSD for consoles and so on. That would have helped the SPARC become popular. As it is, Alpha, PA-RISC & MIPS had folded, so this was a golden opportunity for Sun, and later Oracle, to try proliferating the SPARC as an alternative to x86. Particularly once Apple had abandoned PowerPC for x86.

      There were some alternate SPARC vendors like Integrix & Tatung. Had they been able to run w/ Linux/SPARC, that could have been quite a viable platform.

    15. Re:hard to fault Oracle by Anonymous Coward · · Score: 0

      There were law enforcement types that wanted a PC to be able to save and archive as much information as possible eg. file modifications and not simply "lose them" with overwrites. Then there was another problem with some blocks ending up being used more than others. So the algorithms were designed to cycle through the disk drive blocks. I wonder how much influence this requirement has had over file system design.

    16. Re:hard to fault Oracle by Anonymous Coward · · Score: 0

      Not sure why you had such a difficult time but smf is easy as pie for me to configure and manage. Easier than init and its mess of hard links in the correct /etc/rc.* directories and getting the priorities set correctly so that it starts at the correct time in the init sequence - no thanks.

      I don't see your complaint about the system being slow for a short time while everything starts up as being a valid concern; just wait a minute or two or get better disks if it can't handle something that trivial.

    17. Re:hard to fault Oracle by Anne+Thwacks · · Score: 1

      If your server handles anything worth actual money, OpenBSD on Sparc64 is the way to go. Of course you might need to know what you are doing, and have confidence that Sparc hardware will survive under Oracle (that is the toughest requirement of all).

      --
      Sent from my ASR33 using ASCII
    18. Re:hard to fault Oracle by thogard · · Score: 1

      Integer priorities mean I have absolute control.

      The current system has no guarantee of any order of anything. This means if you get hacked at a non privileged user level, that process can hang around until it gets the "system is shutting down" signal, then do a quick fork/exec a few times and keep running until the system sends it a kill -9. Meanwhile it has a system without syslog running and without any auditing running. Take advantage of something running a broken xml library that runs setuid, and you own the system until it power off and nothing is logged at all.

    19. Re:hard to fault Oracle by jedidiah · · Score: 2

      > If your server handles anything worth actual money, OpenBSD on Sparc64 is the way to go.

      You're funny.

      --
      A Pirate and a Puritan look the same on a balance sheet.
    20. Re:hard to fault Oracle by jedidiah · · Score: 1

      > Why the fuck would they want to port Solaris' better aspects to their competitor, when Solaris is still a billion dollar business for Oracle?

      Linux is even bigger business for Oracle. Even 10 years ago, companies were trying to figure out how to ditch the large boxes that you are trying to drool over so hard.

      Oracle helped quite a bit in that regard.

      --
      A Pirate and a Puritan look the same on a balance sheet.
    21. Re:hard to fault Oracle by Anonymous Coward · · Score: 0

      Do you know if OpenIndiana falls into any of these pitfalls?

    22. Re:hard to fault Oracle by tsprig · · Score: 0

      You're doing it wrong ... http://xkcd.com/463/

      sensitive data shouldn't be written unencrypted to disk in the first place.

      Init doesn't make sure programs always run, it makes sure things are run once and never cares about it again. SMF makes sure that things always run.

      Solaris 10/11 has a completely rewritten network stack because the old one was getting sorely long in the tooth. As time went on, developers realized that networking has gotten far more complex in enterprise environments than an interface like ifconfig can cope with. As such, there is still an ifconfig but there are many other tools for configuring the network as well.

      I stopped using any kind of Solaris once they killed off the OpenSolaris community. I was an unofficial Nexenta developer and very keen on some features (and not so much on other bugs) but overall, the changes made in Solaris/OpenSolaris were done thoughtfully and carefully. At least until Oracle took over. At that point I have no clue except that their NFS appliances are kinda buggy and we are forced to upgrade constantly.

    23. Re:hard to fault Oracle by Chris+Mattern · · Score: 2

      You're absolutely correct. We are a Solaris shop transitioning out to Linux on VMs. Solaris is wonderfully stable and reliable and Sun supported it well. We liked it a lot. But it was already becoming unaffordable before Oracle took them over and now you simply can't afford it, no matter how pretty it is. We won't be buying any more SPARC hardware, ever.

    24. Re:hard to fault Oracle by TrogL · · Score: 1

      "...and Sun supported it well." Oh you've GOT to be kidding. The only thing worse than Sun's support has been ORACLE's support.

    25. Re:hard to fault Oracle by thogard · · Score: 1

      Have you read "man inittab" on any system V derived? action=respawn means it will ALWAYS run at the listed run levels. Sort of like how it runs the svc daemon does now. Whoever planned the new system just didn't get "init".

      SMF only runs things as long as the contract system works.

      As far as writing sensitive data to disks, do you know about the "real world?" Take a look at any online credit card system in the world. You will find people enter their card number as their email address, shipping address, reference number. You will find admins sending stuff like "can you fix 4111 1111 .... 1111 for me?" SSNs flow like water as well. Some times you must scrub the empty space on the disk or scrub stuff you know was just wrong. ZFS has NO ability to do that and that is a MAJOR FLAW!!!!!

      ifconfig isn't about the stack. It is a tool to tell the stack what to do and has been for more than 3 decades. Inventing new tools to do the same job was pure incompetence.

    26. Re:hard to fault Oracle by mlts · · Score: 2

      Solaris was made in a time when there were almost hundreds of flavors of UNIX. Everyone had their own different variant.

      When SunOS 1.x to 4.x were out, those were extremely BSD. It wasn't until the renaming to Solaris and Solaris 2.x when Sun moved to a SVR4 base.

      I do miss Sun though. SPARC hardware was extremely well built. Now, Oracle hardware just looks like any other machine sitting in a rack. Plus, it was nice having another CPU architecture than x86 that was commonly available.

    27. Re:hard to fault Oracle by mlts · · Score: 1

      Solaris 11 has had a couple security changes that are interesting. For example, root is gone by default until you use rolemod. Instead, UID 0's functionality is a role, where you su for that access, similar to how one uses UAC in Windows to access administrator functionality that your account has.

      ZFS is also one of the most useful features of the OS. No LVM to worry about, and far better bitrot catching than any other filesystem except for MS's Storage Spaces + ReFS.

      I still like Solaris and SPARC, but for the cost involved in the enterprise (where commercial support is necessary), I can get 95% of the features (all except ZFS [1]) for a lot less hardware investment.

      [1]: My only complaint about Linux in general is that it doesn't have an enterprise-grade modern filesystem. btrfs is still not finalized yet (and it can't be booted from), and a true enterprise grade filesystem merges the LVM and the FS to be able to catch bitrot and file corruption via CRCs. Without both layers working at the same time, that protection isn't there. Even MS tossed out the LVM with Storage Spaces.

    28. Re:hard to fault Oracle by mlts · · Score: 1

      ZFS is a decent alternative to having a "smarter" SAN. One can buy a VNX and have real time deduplication, encryption, snapshots [1], async replication, etc.

      Intead, one can buy a "dumb" array, something that might just take disks, run JBOD, or go with RAID5, or RAID6, and let the OS do the work from there with deduplication, encryption, etc. With ZFS, a directory of critical files can be protected with RAID1 while everything else on that filesystem is RAID5 or RAID6. An occasional scrub of the pool will catch any bit rot that might happen, and if one adds SSD, the filesystem will know to autotier it.

      Of course, we have seen this before. Ages ago, there was a push to move to software RAID from hardware controllers back around the turn of the century. However, a few years after that, it was back to hardware RAID controllers and the software just using it as a plain disk.

      [1]: Careful with snapping over 3-8 terabytes, or you will have a very unpleasant experience, unless this was fixed.

    29. Re:hard to fault Oracle by mlts · · Score: 1

      The POWER8 will be an interesting CPU architecture. It is the first since PowerPC that is available for license to anyone other than IBM. May not mean much, but I'd love to have a generic PC or server running Linux and not x86, although in reality, this is a pipe dream.

      I wonder if SPARC is still an open CPU model, although the days of the Tatung or other clones are gone.

    30. Re:hard to fault Oracle by Anonymous Coward · · Score: 0

      Even 10 years ago, companies were trying to figure out how to ditch the large boxes that you are trying to drool over so hard.

      c.2001, IBM was running a new ad campaign for linux on their big iron and for their large enterprise installs. Everyone had high hopes for linux and IBM even stopped selling AIX (though development continued). But not for long... linux fell short, then, of the requirements of most greybeards, who rallied surprisingly well: IBM quickly returned AIX to market. Subsequently, of course, linux has matured and taken over datacenters not unlike a virus. Where linux excelled then was that it extended usefulness of crusty and dusty hardware. But it isn't the OS that makes a system effective and reliable... it is the OS/hardware combination. The insane and still unmatched reliability of Solaris/SPARC is what is missed today, and the OS/server paradigm has more or less been replaced with the same paradigm used for SANs: you have x systems, and can expect z hw failures per t time interval, and must maintain accordingly. Imagine how pointless and ineffective our new ubiquitous paradigm would be if no system ever failed.

      Interestingly, while an early market failure such as linux suffered when initially thrust into the enterprise data center roughly tenish some years ago usually instantly kills the market integrity and industry trust in a product --effectively turning it into a humerous anacdote-- linux sufferred no such shame. Linux is not perfect... it has problems, but you'd never know it because that's not part of its narrative. I believe its original charter was not to kill Windows, but to be a sane and free alternative, to fix the broken parts of Windows environments, to augment and even replace a lot of MS lockin. But at some point Linux started eating its parents, aunts and uncles and siblings, competiting agressively against the BSDs and SYSVs when it offered little if any advantage over other opensource *nixes.

    31. Re:hard to fault Oracle by Cramer · · Score: 1

      a) they're symbolic links. b) if you cannot count (S01 happens before S02, etc.), then you have no place fucking with a *NIX system.

      The claim that SMF does away with all those nasty shell scripts is an absolute and complete LIE. It just has them in different places. After installing Solaris 10 for the first time, I dug deep into the "init replacement"; the shell scripts were still there -- just not in /etc/init.d -- retooled to use the SMF DB instead of simple, everybody-know-about-them everybody-knows-how-to-use-them config files.

      The issue with serialized vs. parallel boot is the lie of efficiency. The push for parallel is the task that sits idle waiting for an outside event. The thing that camp fails to understand is the possible (and often very real) I/O contention from too many processes reading and writing at startup, which can (and does) make boot actually take longer. It's a fine line they don't know how to walk.

    32. Re:hard to fault Oracle by turgid · · Score: 1

      Solaris has actually been quite good since Solaris 10.

      Solaris 10 was developed and released (in 2005) by Sun, five years before Oracle acquired them.

      I used to run Solaris on my UltraSPARC boxes at home until Oracle made that very difficult.

      I'm afraid that SPARC and Solaris are now nothing but historical curiosities to all but the wealthiest who are locked in to Oracle's platform.

      It's too bad Oracle won't invest any time to port some of the better aspects of Solaris to Linux.

      There's no point. While Oracle tries to milk the cash cow and sweat their assets, Linux will gradually overtake Solaris all on its own.

      Oracle will begin the fade into obscurity like the other dinosaurs before it (ICL, Tandem, DEC, Compaq, HP, Microsoft...)

    33. Re:hard to fault Oracle by unixisc · · Score: 1

      Apparently, it still is - Oracle didn't do to them what they did w/ OpenSolaris

    34. Re:hard to fault Oracle by Anonymous Coward · · Score: 0

      ZFS is on the right path but it still isn't quite where it needs to be. For example I can't tell it not to reallocate blocks on write so I can't force overwrites of sensitive data

      ... then you should use UFS.

      Also, I'm pretty sure you can run guest zones as any flavour of Solaris, such as 9 or 10, under a sol11.1 host zone.

      Having said this, I would agree with the F- ... here are some reasons:
      - terrible memory management. We have numerous server crashes / freezes (10 min+) where it pages itself to death, usually after deleting some hundreds of GBs in a ZFS volume. Even if it comes back, the Oracle database will have flagged the background procs as not responding and killed them , and the db service with it.
      - silly bugs. Such as prstat not working properly in some zones.
      - ZFS - is not as performant as, say UFS on HPUX is...
      - Oracle support ... ugh

      AC in deference to our support contracts...

    35. Re:hard to fault Oracle by Anonymous Coward · · Score: 0

      I might like a lot of things, but I am in a different role than I was. It is easy to 'miss the old days', but not very practical. I currently manage software development on 'whatever platform we are given', and our resident unix demi-god would rather eat snot than run sparc/solaris gear (had a bad experience with Sun 20k's back in the day) our last 'unix' platform was Alpha and have since gone mostly suse on x64 for Oracle servers and vmware for hosting win servers

      We do some big data, and I would be interested in applying one of the Oracle Big Memory machines to the work that our mad scientists do...

    36. Re:hard to fault Oracle by kry73n · · Score: 1

      SPARC International Inc. was independent of Sun and is independent of Oracle. The ISA will stay royalty free, no one can/want to change that.

    37. Re:hard to fault Oracle by Anonymous Coward · · Score: 0

      For example I can't tell it not to reallocate blocks on write so I can't force overwrites of sensitive data

      That could work in principle with leaf data blocks, otherwise the Merkle tree design makes it really hard.

      Setting overwrite-leaf-blocks as a property or as an ioctl lets one expose something fairly dangerous to an administrator, along the lines of the sync=disabled property; a crash will mean that there are likely to be inconsistent leaf data blocks within a given vdev which are likely detectable only when they are actually read. (They will be inconsistent with the checksum tree if they leaf block was partially overwritten at the crash; depending on the vdev and its underlying components, there may be no way even in principle to recover the partially-overwritten data block, for example, if both halves of a mirror vdev have partially overwritten a datablock which was last linked to the Merkle tree by a sufficiently old TXG that is no longer reachable via the label).

      Moreover, since many modern underlying devices themselves do COW for write-levelling or scheduling purposes, overwriting an LBA is not guaranteed to destroy the previous data, although some such devices do provide commands which destroy the data in a specfic (range of) LBA(s), but would you really bet that "destruction" is more than simply unlinking?

      If security against forensics in possession of the physical device(s) is a concern, then probably only massive whole-device overwriting is a feasible approach. The TRIM works-in-progress for ZFS could in principle be adapted to do just that, in a way that always retains ZFS consistency and without necessarily wrecking performance.

      If performance is not so big an issue, then put your sensitive data into a zvol or equivalent and do device-level encryption on that, or alternatively use the Oracle flavour of zfs and its dataset encryption.

  5. Re: Proposed direction by Anonymous Coward · · Score: 0

    He did propose a direction—the same direction that Oracle actually did indeed take.

  6. Why bother? by Frosty+Piss · · Score: 0

    Why even read reviews such as this? We know what the pronouncement from the Great Man will be.

    As Gaius Marius said, "You all suck!".

    Now we know...

    --
    If you want news from today, you have to come back tomorrow.
  7. who cares. by Anonymous Coward · · Score: 0

    What I really want to know is, what does Ryan Gosling think about this.

  8. Should've sold out to soneone else by Anonymous Coward · · Score: 0

    If Sun had sold out to Google, those things would have been in much better shape. Hell, even MS would've been better stewards.

    1. Re:Should've sold out to soneone else by Anonymous Coward · · Score: 0

      If they had sold out to google it would have been discontinued due to google not finding a way to make advertising revenue from it. Oracle are bad, Google are worse. MS "may" have done better, or more likely they too would have let it die off. I think perhaps IBM of 5-10 years ago could have done a good job but even they have gone to shit lately.

    2. Re:Should've sold out to soneone else by mark-t · · Score: 2

      Google probably would have made a better offer initially if they had reason to suspect how things were going to play out with oracle. Although I'm sure that everybody realized as soon as Oracle had made an offer for Sun that Google would have been a far better choice than Oracle for Sun's IP, I don't think anyone else expected just how colossally bad Oracle was going to be with it.

      I have mixed feelings about the idea of MS being a better choice.

    3. Re:Should've sold out to soneone else by Anonymous Coward · · Score: 0

      There's only one word that anyone needed to know to foresee Snoracle's fate: Larry Ellison. Google must have known that.

      MS, despite their many failings, knows how to do languages and platforms. More so than Google, I would say.

    4. Re:Should've sold out to soneone else by Anonymous Coward · · Score: 0

      What a hilarious and idiotic reply. MS would have killed it or made it proprietary. That's how they role and that's how they'll meet their demise. IBM would have been the ideal steward since practically all of their middleware is dependent on Java.

    5. Re:Should've sold out to soneone else by yuhong · · Score: 1

      What is funny is that MS had a Java Virtual Machine in the late 1990s that was infamous for its extensions that led to a Sun lawsuit which eventually led to it being discontinued.

    6. Re:Should've sold out to soneone else by unixisc · · Score: 1

      There really were no others. One thing Sun could have done was try getting acquired by Red Hat. That would have made an interesting merger

    7. Re:Should've sold out to soneone else by jbolden · · Score: 1

      Just as an aside it was also very fast compared to the standard (Netscape) one. Arguably the Microsoft Java VM is what made Java viable for clients and led to the success of Java. I understand why Sun did what they did, in that suit but even in retrospect it is hard to evaluate if they made the right choice.

    8. Re:Should've sold out to soneone else by Grishnakh · · Score: 1

      MS only knows how to do one platform: Windows. When they make new languages, they're completely tied to their own platform, and useless on others. No one uses C# outside of Windows.

    9. Re:Should've sold out to soneone else by mark-t · · Score: 1

      To be fair, C# is also used in Unity3D development, which is multi-platform.

    10. Re: Should've sold out to soneone else by Anonymous Coward · · Score: 0

      Microsoft also has the first port of UNIX for the 8086 platform. I still have an Altos box in my collection that runs Microsoft Xenix. Not badly: it's a box with an 8086 processor and 512k of RAM and supports five users on serial terminals. Microsoft later divested their Xenix product, which led to the rise of SCO. Microsoft Xenix was a licensed port of AT&T Unix. They really didn't like paying that license, though at the time they didn't really have any other os but the fledgling MS-DOS. They were really still an overgrown BASIC vendor.

    11. Re:Should've sold out to soneone else by Anonymous Coward · · Score: 0

      I think IBM would have been a good match.

  9. Re:That James Gosling fella is of no consequence.. by cold+fjord · · Score: 1, Flamebait

    That James Gosling fella is of no consequence...

    The views of those that have achieved something great or useful tend to be solicited repeatedly, especially if they had a string of achievements.

    Ken Thompson and Dennis Ritchie? Yep.
    Woz? Yep.
    Gosling? Yep.

    Bogaboga? Unless you are the originator of the "death by booga booga" joke, maybe not. Of course you could be holding out on us.

    --
    much of left-wing thought is a kind of playing with fire by people who don't even know that fire is hot - George Orwell
  10. oracle and aquisitions by PC_THE_GREAT · · Score: 1

    People do not seem to notice, is that, whenever oracle tried to acquire some piece of technology specially software wise, that piece of software/tech tends to perish, notable examples are Openoffice (when oracle acquired it, luckilly now that's apache'd), MySQl (luckilly the community revived it through MariaDB). You don't hear much about these anymore. Agreed they did acquire a lot more of software out there that.. well you never hear much of.

    I wouldn't be surprised if oracle has a magic algorithm to screw up fine projects, and one would have thought having such huge financial backing will just enhance products, such doesn't seem the case.

    1. Re:oracle and aquisitions by TopSpin · · Score: 1

      that piece of software/tech tends to perish

      Except when it doesn't. VirtualBox hasn't perished.

      A colleague of mine speculated that perhaps Oracle had forgotten about VirtualBox and thus it has been spared the obligatory ruining. Perhaps there is a gang of hard core emulator developers quietly slipping in and out of the building each day, carefully avoiding notice.

      Netbeans does actually suck less than Eclipse. That's a low bar, to be sure, but it appears to be acquiring more users than it is repulsing, so there's another counterpoint.

      --
      Lurking at the bottom of the gravity well, getting old
    2. Re:oracle and aquisitions by Sarten-X · · Score: 1

      Perhaps there is a gang of hard core emulator developers quietly slipping in and out of the building each day, carefully avoiding notice.

      Hey, that has worked before.

      --
      You do not have a moral or legal right to do absolutely anything you want.
    3. Re:oracle and aquisitions by PC_THE_GREAT · · Score: 1

      hahaha true!

    4. Re:oracle and aquisitions by Y2K+is+bogus · · Score: 3, Informative

      This comment, and the other 3 that replied to it before me, show a huge lack of knowledge or care. Oracle isn't very transparent, but it only takes a small amount of effort to see that neither MySQL or VirtualBox are in danger of perishing. There are many people who left Oracle/Sun/MySQL for Percona and MariaDB/SkySQL, but most of those people left for their own reasons and *many* left before and Sun or Oracle influence was upon them.

      I get to see it from the inside, and MySQL is growing and has more market share than either of the other competitors. The newest developments are really spectacular improvements. I get to see the walled garden from the inside, and it's anything but dying, it is in better shape as a product than it has ever been. Oracle is anything but stupid and doesn't have a track record of making stupid decisions with their products, which can't be said for some companies. Oracle is putting a lot of resources into MySQL to make it even better.

      VirtualBox is a fairly decent team and they are not just working on VirtualBox, there is a reason it continues to be developed and the technology doesn't have a dead end to it.

      I think that most of the comments I've read are uneducated and purely people spouting off uninformed opinions mixed with conjecture and hyperbole. The people I work with are the brightest group of people I've ever had the privilege of working with, there are some really notable folks that work on MySQL and you wouldn't know it unless you paid attention to the blogosphere.

    5. Re:oracle and aquisitions by stiggle · · Score: 1

      VirtualBox has a commerical revenue stream.

      Basically when Oracle took over they went around all the departments and asked them what their revenue stream was and so justify their existance within Oracle.

    6. Re:oracle and aquisitions by erp_consultant · · Score: 1

      Exactly. Oracle is in business to make money. They are a publicly traded company, after all. I remember back when they bought PeopleSoft and all sorts of rumors were circulating that Oracle was going to kill it off. Until Oracle realized that it was a big money maker for them. Then they invested heavily in it and now it's a far better product than the one they originally bought. OpenOffice is a nice product but you can't make any money off it. Oracle's customer base is the enterprise. Enterprise customers use MS Office, almost without exception. Oracle is going to spend money on integrating their products with MS Office, not OO. Same goes for MySQL. Great product but it competes directly with Oracle's own database. Oracle as a company might be a lot of things but stupid ain't one of them.

  11. To know Oracle... by Anonymous Coward · · Score: 2, Funny

    To know Oracle is to hate Oracle.

    An Oracle Field Engineer shared the secret meaning for the name, "oracle".
    One
    Rich
    Arse
    Called
    Larry
    Ellison

  12. Java in the server, in the client, in the browser by gwolf · · Score: 5, Insightful

    Are three very different things. Java in the server and in the client is alive and very very much healthy. Ugly and slow applets in the browser thankfully are almost dead — Because HTML5 delivered way better. But applets dying off does not in any way mean Java is any less healthy!

  13. Re:That James Gosling fella is of no consequence.. by cffrost · · Score: 3, Funny

    I think it's too early to tell, as James Gosling just lacks the experience most people are used to from those like him — there's still a lot left for him to learn from his father, industry veteran Jim Goose. Once his father retires, though, I think James will get to chance to really spread his wings, and we'll probably see some very good ideas of his take flight. For now, though, I think he's just a bit green around the beak.

    --
    Thank you, Edward Snowden.

    "Arguments from authority are worthless." —Carl Sagan
  14. I'm still sad... by gwolf · · Score: 2

    That MySQL's space was/is being transferred to MariaDB, instead of just dying a relatively quick death.

    Why bother with MySQL if you can just migrate over to PostgreSQL? Yes, of course, some of the weirdest bits won't work, and errors will now (for a change!) actually interrupt your work instead of silently losing information. But it seemed like a good way to kill that ugly beast!

    1. Re:I'm still sad... by drinkypoo · · Score: 1

      If MySQL did not exist, Postgresql might not be as good as it is.

      Competition is good, it permits the market to try out new ideas.

      Lots of us are using MySQL without problems because someone else has handled the details, and we don't care if it's stupid. If postgres were easier to use, maybe more people would use it.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  15. No mention of SPARC? by kry73n · · Score: 5, Insightful

    SPARC has seen more advances in the 4 years under ORACLE then in the previous 15 years under Sun. I actually enjoy reading about their tech every now and then. But unless they open up Solaris again to attract the open source community the only thing that keeps it alive is backwards compatibility of legacy software.

    1. Re:No mention of SPARC? by Anonymous Coward · · Score: 2

      All of the work that you're seeing now on SPARC was started well before ORACLE picked up. What happens in the next 5 will be most telling of ORACLE's influence.

      Hardware, unlike software, has a very long lead time (years).

      From Larry's perspective, source code is the most valuable asset. Thus they want to keep as much of it behind closed doors as possible. That's the lesson he learned from developing the database business that is ORACLE's foundation. And there is only one opinion that counts in the company: Larry's.

    2. Re:No mention of SPARC? by thogard · · Score: 2

      What advances would that be? The ones out of Fujitsu? The T chips are just now catching up with workloads that they can run reasonably. I have work loads that a 15 year old Sparc IIi will out perform a few year old T2. The V100 was a $1000 appliance box yet the base T2 was selling for more than $6,000. If the UltraSparc IIIi was made at 22 nm (unlike its original 130 nm) and it would scream for most web appliance roles. It would even be a nice cpu for the Lights Out Management system and it could even run Solaris unlike their current LOM which is running Linux.

    3. Re:No mention of SPARC? by unixisc · · Score: 3, Interesting

      He did grade SPARC. He said that it was tough to tell, since SPARC was floundering well before Oracle took over.

      SPARC is interesting, but for the OpenSPARC/sparc.org consortium. I don't see how Oracle gains squat by promoting SPARC: the only reason SPARC is alive is Fujitsu SPARC64. Otherwise, SPARC would have been EOLed, just like the SPARCstations.

      I think SPARC has a limited market, since routers are now MIPS and maybe ARM, consoles were MIPS & Power and moving to AMD, servers are x64 and later maybe ARM64.

    4. Re:No mention of SPARC? by Anonymous Coward · · Score: 0

      I don't entirely disagree, but the T1 was quite a revolutionary chip - massively multi-core for the time, and the multi-threaded idea was pretty new too - I saw workloads that struggled on a multi-CPU multi-core intel box fly on a T2000 with an 8-core T1 CPU. Of course, you could flatten the T2000 by running two gzips at the same time. It was all about the workload.

      The Fujitsu chips are really quite nice, too....

    5. Re:No mention of SPARC? by Cramer · · Score: 1

      And yet, it's still a decade behind. Low clocked RISC processors are just too damned slow for many modern applications. Sure, it can run many threads in parallel, but the answer from each one will be 5x (or more) slower than their cheap, commodity Xeon and Opteron competition.

      (@ 20k+ for a SPARC T4 server, I can buy a dozen (or more) x64 servers)

    6. Re:No mention of SPARC? by kthreadd · · Score: 1

      What advances would that be? The ones out of Fujitsu? The T chips are just now catching up with workloads that they can run reasonably. I have work loads that a 15 year old Sparc IIi will out perform a few year old T2. The V100 was a $1000 appliance box yet the base T2 was selling for more than $6,000. If the UltraSparc IIIi was made at 22 nm (unlike its original 130 nm) and it would scream for most web appliance roles. It would even be a nice cpu for the Lights Out Management system and it could even run Solaris unlike their current LOM which is running Linux.

      Your T2 can do up to 64 parallel execution threads. Your IIi, not so much. That's the difference. More cores, not faster ones.

    7. Re:No mention of SPARC? by drinkypoo · · Score: 1

      the T1 was quite a revolutionary chip - massively multi-core for the time, and the multi-threaded idea was pretty new too - I saw workloads that struggled on a multi-CPU multi-core intel box fly on a T2000 with an 8-core T1 CPU. Of course, you could flatten the T2000 by running two gzips at the same time.

      How revolutionary. If I do that on a single-core Linux box, the scheduler still lets me have resources for other tasks, including CPU.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    8. Re:No mention of SPARC? by thogard · · Score: 2

      No, the t2 can preserve the context of 64 threads but it can and will only run no more than 8 execution threads at a time. In most cases, the pipeline is so starved, it won't even manage 8. When it is running 8 at a time, it is doing each at a much slower rate that the older CPUs would be doing if they were made using the same process.

      The II/IIi/IIIii can preserve something like 4 processes executions context at a time. Sometimes that is better. It is better on nearly all of my workloads.

    9. Re:No mention of SPARC? by Anonymous Coward · · Score: 0

      I think SPARC has a limited market, since routers are now MIPS and maybe ARM, consoles were MIPS & Power and moving to AMD, servers are x64 and later maybe ARM64

      You think embedded when you think SPARC?
      Think outside of your comfort zone of web servers and web application. Think about application servers, Oracle databases, and the high-midrange and high-end enterprise tiers which are Sparc's core markets where its main competition is Power and Itanic. Think of those workloads that need that rely on and benefit from having access to 512-104 threads. The saddest thing is people who fuss and muss about how "expensive" Sparc is. Shop around, at this level of throughput, Sparc T4/T5 systems are significantly less expensive than the equivalent x86/x64 cluster, especially when your workload is the kind where a terabyte of RAM isn't exactly overkill.

      People who fuss about the price and how they can buy :"many more" x64 servers, never consider the ram costs, or that to match the throughput, x86 is far more expensive. Pro-tip, the RAM and disks are going into each of you nodes, you're not just buying it once.

      You're right in that Sparc does have a limited market, it isn't and "x86-killer", and it is designed for a rather specific set of tasks where throughput and high vertical scalability are important. It does quite well in those markets.

    10. Re:No mention of SPARC? by kthreadd · · Score: 1

      True, my mistake. At the end of the day it comes down to that it depends.

    11. Re:No mention of SPARC? by Anonymous Coward · · Score: 0

      SPARC limited market? Oracle has 300,000+ customers running Oracle DB and its been estimated that approximately ~10% are running SPARC so theres huge growth potential to up-sell to SPARC. The RISC/UNIX market is also an $11BN market today and with HP abandoning Itanium/HP-UX and IBM floundering, theres huge opportunity for SPARC to regain its crown. The Enterprise/Cloud/Big Data market is huge and these are markets highly addressable by SPARC.

    12. Re:No mention of SPARC? by thogard · · Score: 1

      Early Java was nothing other than a mess of pointers to pointers to pointers to pointers to more pointers all in a multi threaded system. The T1 addressed that problem but the concept of "All problems in computer science can be solved by another level of indirection*" is false and at some point compiler writers fix part of it. When they win, concepts like the T1 fail.

      Sun tried great things with the T1 and it was like a great chess move that failed. The problem is they did a pawn sacrifice of their core business for that attack and it just didn't work out. Up until the T2000, Sun never designed their high end kit, they stayed with the low end and groups like Cray or SGI did their "big iron". The only great boxes sun designed in house where the small pizza boxes. The SS1, SSP20, x1, netra210 were great little servers. Things like the 690 and e10k were outsourced and while they were impressive as well, they didn't have the personality of the pizza boxes.

      *To Quote David Wheeler

    13. Re:No mention of SPARC? by Anonymous Coward · · Score: 0

      Time to get updated. SPARC T4 is now 2 years old and SPARC T5 is ~2.5x faster with same improvement in price/performance. Todays SPARC T5 runs at 3.6GHz, with a full 16-cores and 8-threads per core, higher GHz, more cores and threads than any x86 processor and 2x more cores and threads than any Power7+ from IBM. SPARC T5 with 16-cores outperforms Xeons by 4% to over 3x depending on the benchmark and how much I/O is involved. There is not a single public benchmark where either Xeon or Power beats SPARC T5 on a per CPU basis on throughput, response times or even TCO when comparing systems like for like (same # of CPUs, RAM, etc). You can look here for all the comparisons https://blogs.oracle.com/BestPerf/

    14. Re:No mention of SPARC? by unixisc · · Score: 1

      When I hear embedded, I think either MIPS or Microcontrollers. Who makes SPARC any more? Ross is dead, Cypress is out, and what Fujitsu makes is purely for legacy market servers. Japanese companies value their long time relationships the way the Suns & HPs of yesteryear used to: hence, they still make them.

      For application servers, all I can think of is x64s, simply b'cos both Wintel & Lintel apps run on that platform. Linux on all RISC platforms was just a proof-of-concept, rather than an actual strategy that anyone followed (w/ the exception of Linux on Itanic when SGI ran w/ it on Altix). Aside from that, the workloads you described are what Oracle already focusses on addressing. I happen to think that had Oracle done a market segmentation of SPARC and offered different configurations at different price points, the SPARC would still be a vibrant platform. But Sun workstations had been discontinued a while back, and Oracle by & large does Opteron servers. Not a bad call, but when the SPARC's leading vendor only offers it for top end jobs, its utility is artificially constrained.

    15. Re:No mention of SPARC? by unixisc · · Score: 1

      That would be nice to see, except that Oracle solely focusses on the upper end as far as SPARC goes. Unlike Sun which proliferated it at all levels.

    16. Re:No mention of SPARC? by bill_mcgonigle · · Score: 1

      I don't see how Oracle gains squat by promoting SPARC

      Oracle only bought Sun for two reasons: SPARC and Java. Both are related to Oracle's database business.

      The massively multi-threaded version of SPARC (Niagara, IIRC) is better at scaling Oracle's databases than other architectures.

      Oracle's database architecture does not scale indefinitely. They need architectural changes, but those are covered by Google's patents.

      Hence, Java, as a club to sue Google with, to try to force a patent cross-licensing deal.

      Everything else from Sun is a toy business. I suppose as long as there's profit, they'll be allowed to exist, but watch out if you're in any of those Sun technologies and you fail to turn a profit, because Oracle doesn't need you.

      --
      My God, it's Full of Source!
      OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
    17. Re:No mention of SPARC? by unixisc · · Score: 1

      If SPARC was that important to Oracle, they'd have continued its development, rather than just use whatever Fujitsu makes. If Fujitsu ever stops making it, then SPARC would be dead, since Oracle is not making it, and the companies that used to, such as Cypress, are out of that business.

    18. Re:No mention of SPARC? by bill_mcgonigle · · Score: 1

      Not sure if you're being sarcastic - you know Sun was always fabless, right? The fab contract would often change between generations.

      Oracle just put out a brand new SPARC design less than a year ago. Fujitsu also fabbed the SPARC in the Sun 4.

      --
      My God, it's Full of Source!
      OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
    19. Re:No mention of SPARC? by Anonymous Coward · · Score: 0

      Sun never designed their high end kit,

      No, Sun never sold kits. They sold complete servers and desktop systems. They're not Heathkit or Apple with their Apple I kit.

    20. Re:No mention of SPARC? by Anonymous Coward · · Score: 0

      > kit

      You're full of crap. I still have my old Sun-1 that I used for three years. Sun make complete computers. They never sold kits like you claim. I had several friends that were there from nearly the start, and your kit claim is a bold-face lie.

    21. Re:No mention of SPARC? by kry73n · · Score: 1

      You haven't read any news in the past 3 years have you? Have a look at the SPARC T3, T4, T5, M5, M6 and the upcoming T6 and see which company developed those. They even beat IBM Power with the latter ones.

  16. An F- for the handling of Solaris by vikingpower · · Score: 2

    is totally merited. Solaris was and still is brilliant, one of the best operating systems ever made. The scalability and reliability are legendary. I do not know of any OS that can run on a tiny PC AND on a big-mama cluster with exactly the same code. Solaris is another example of how mergers and corporate acquisitions boil down, most of the time, to sheer destruction of capital. Observed that with tiny companies and start-ups as well as with mega-mergers & acquisitions. Solaris is dead, and I concur with Gosling: I weep.

    --
    Religous speak to God. Insane are spoken to by God. When all shut up, one can finally hear Shostakovich in peace
    1. Re:An F- for the handling of Solaris by Anonymous Coward · · Score: 0, Insightful

      The Linux kernel actually runs unmodified on tiny ARM microprocessors (much smaller than your typical low-end smartphone); right up to the largest single-system-image machines ever made (the 4096 CPU Altix machines); and the world's most powerful supercomputers.

      Now you know of another OS that does it better than Solaris.

    2. Re:An F- for the handling of Solaris by unixisc · · Score: 2

      is totally merited. Solaris was and still is brilliant, one of the best operating systems ever made. The scalability and reliability are legendary. I do not know of any OS that can run on a tiny PC AND on a big-mama cluster with exactly the same code.

      How about the BSDs? NetBSD? FreeBSD? OpenBSD?

    3. Re:An F- for the handling of Solaris by upuv · · Score: 4, Interesting

      I completely agree. Solaris "was" a great OS. With some very notable monster issues. Oracle has effectively killed Solaris. I simple can't use it anymore. The licensing costs of it and the software that runs on it are more than my total IT budget. Despite it's fantastic attributes I can no longer afford to put this in my Datacenter. With on demand virtualisation I can not afford to have to worry about things like. "Am I going to violate my license conditions if I spin up X more?"

      I had an Oracle sale rep try to sell my that ridiculous Oracle stack in a box Exadata/logic. I was almost crying in laughter by the end of the sales presentation. 2/3 of the way through I stood up and wrote on the white board "Tell me how this isn't vendor lock in?". I called time at the 1 hour mark. I ended the meeting with the simple statement. Everything you have shown me is all about "vendor lock in" every word out of your mouths just re-enforced this concept. I had one question for you the entire meeting and you simple could not in any way respond to it.

      So I priced everything I might need on Amazon. Using free and commercial AMI's with the odd vendor SW package tossed in. My first year spend was 1/25th of the Exadata discounted opening price. Nothing on the EC2 list had anything to do with Solaris. This is how you kill something. Make it financially ridiculous.

      Issues with Solaris. That should have been addressed in the Oracle years.
      - Package manager was brain dead. apt, yum are far better. ( Sorry Solaris 11 was too late. Too much legacy out there. )
      - Patching made no sense. You have no idea what packages are patched with a patch. Patches were just binary disk vomit that spewed crud all over the system. Impossible in the real world to build any sort of verification around them. ( Sorry Solaris 11 was too late. Too much legacy out there. )
      - Zones: Are a nightmare of security and privilege. I don't care what any says a zone is just a change root jail. Which means you will only every be as up-to-date as the host system. And it means you must be compatible and tested against the host system. Which is really no different than not having zones. Zones are a horrible horrible mess.
      - No dependable only repository of packages that is robust or up to date. Far to much package hunting still required to locate software for solaris. Most packages are months to years behind there linux counterparts.
      - Java performs better on x64 than Solaris/SPARC. This has boggled me for years. Only recent sparc architectures let java and other highly threaded applications stacks really perform well. Why do I even have to know about processor binding for processes?

    4. Re:An F- for the handling of Solaris by drinkypoo · · Score: 3, Insightful

      No dependable only repository of packages that is robust or up to date. Far to much package hunting still required to locate software for solaris. Most packages are months to years behind there linux counterparts.

      This is something that has boggled my mind for nigh-on twenty years. Eighteen of them, I guess. Linux came with all the latest tools, but in order to get them for Solaris you'd have to download some old tools and use them to build some new tools. Ultimately I think it's really all about selling you the sunspro compilers, or whatever they're called now, two decades on. If it's too easy to just use gcc, nobody will ever buy sunspro, for which they want a massive stack of cash. It's the only compiler that generates very good SPARC code, and it costs a million billion dollars so many people didn't bother to buy it, and went GCC instead. And then they were throwing away performance. If you're going to run those tools, you might as well run them on x86-Linux. And in fact, that's been eroding Solaris steadily for all this time.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    5. Re:An F- for the handling of Solaris by Anonymous Coward · · Score: 0

      FreeBSD and OpenBSD weren't meant to run on umpteen different architectures. NetBSD was.

    6. Re:An F- for the handling of Solaris by Anonymous Coward · · Score: 0

      - Package manager was brain dead. apt, yum are far better. ( Sorry Solaris 11 was too late. Too much legacy out there. )

      The only thing that was missing in the SysV pkg distro method was the concept of a central repo from where the pkgs could be accessed via the pkgadd command. That has been fixed in Solaris 11 - which is an incredible operating system.

      - Patching made no sense. You have no idea what packages are patched with a patch. Patches were just binary disk vomit that spewed crud all over the system. Impossible in the real world to build any sort of verification around them. ( Sorry Solaris 11 was too late. Too much legacy out there. )

      Patching in all traditional unices were formidable. And that has been resolved in Solaris 11 too. I find Solaris 11's IPS far more dynamic and powerful than yum or apt.

      - Zones: Are a nightmare of security and privilege. I don't care what any says a zone is just a change root jail. Which means you will only every be as up-to-date as the host system. And it means you must be compatible and tested against the host system. Which is really no different than not having zones. Zones are a horrible horrible mess.

      Unfamiliar technology always scares us. Zones are a simple and powerful virtualization solution. They have always been. I've successfully ran multi-TB DWH on zones, multiple zones on the M-class hardware, T-class hardware. You need to learn the technology before pontificating on it's short-comings, it seems that you have rudimentary exposure in this field at best. Zones have far lower overhead of virtualization (virtually none) as compared to other type-1 hypervisor-based virtualization solutions.

      - No dependable only repository of packages that is robust or up to date. Far to much package hunting still required to locate software for solaris. Most packages are months to years behind there linux counterparts.

      IPS resolves this (has done so since the times of the Open Solaris project) - about 5 years or so now.

      - Java performs better on x64 than Solaris/SPARC. This has boggled me for years. Only recent sparc architectures let java and other highly threaded applications stacks really perform well. Why do I even have to know about processor binding for processes?

      Try running java based apps on the T4s and T5s now. Also, You don't have to know about processor binding for processes (you might want to hire someone who does, if you don't have the capacity to learn it yourself). You can also use solaris zones to run your apps on, with dedicated vcpus and memory allocated (as opposed to FSS scheduled). Therein lies the power of Solaris.

      Some more points that the naysayers can't counter yet -

      1) DTrace - no single tool has managed to capture it's power. Those who know how to use it, can make performance management seem, well, trivial. If you haven't experienced it's power, you won't know what you are missing (a case of the unknown unknowns)
      2) Multiple options for Virtualization - Zones, LDOMs, Dynamic domains. You can use one or another as the need arises.
      3) A better designed OS, far more efficient and performant than what's out there.

      For all the melodrama around the cost of ownership (of SPARC and Solaris), in my experience I've found that an LDOM-farm with T4-4s or T5-2s will actually cost almost on par with vSphere 5.x licensing on larger x86 boxes. I'm not even considering the T5-4s or T5-8s or the M5-32/M6-32 class since there are not equivalent x86 boxes.

    7. Re:An F- for the handling of Solaris by kry73n · · Score: 1

      I would not be so sure Linux can handle 32TB or more in the same efficient manner as Solaris does? AFAIK the Oracle had to rewrite the memory subsystem to scale efficiently to these large amounts. Also, how well does Linux do on SMP servers?

    8. Re: An F- for the handling of Solaris by Anonymous Coward · · Score: 0

      Linux is a kernel. There are many other components to an OS. The various linux-based OSes run on many platforms, but it's not one OS scaled across all of them.

    9. Re: An F- for the handling of Solaris by Anonymous Coward · · Score: 0

      There are many OSes based on the linux kernel. Which one of them are you claiming is so scalable? Does it's userland build from the same source tree and scale like Solaris?

    10. Re:An F- for the handling of Solaris by tinker_taylor · · Score: 1

      I think there is some sort of censorship going on (thereby forcing me to get my slashdot password reset) so i can repost my comments, this time not as an AC.

      - Package manager was brain dead. apt, yum are far better. ( Sorry Solaris 11 was too late. Too much legacy out there. )

      Only major shortcoming with pkg manager in solaris' 11 was with the fact that there was no directly accessible repository of software that could be used (albeit the dependency checks etc were missing too). That has been fixed in Solaris 11. You have to remember that Solaris 10 was released in 2004 (10 years ago). If you want to use Solaris today, you should switch to an opensolaris variant or Solaris 11.1

      - Patching made no sense. You have no idea what packages are patched with a patch. Patches were just binary disk vomit that spewed crud all over the system. Impossible in the real world to build any sort of verification around them. ( Sorry Solaris 11 was too late. Too much legacy out there. )

      It's an easy enough thing to fix. Use zones to convert physical servers into solaris 10 branded zones. All of a sudden you patching becomes non-sequiter and generally painless. In my shop we automate patching of 100s of solaris (yes 10) boxes at a time and using ZFS root + Iive upgrade we patch them all under 30 minutes. With Solaris 11, this is down to a 2 minute window. We literally patch Solaris 11 LDOMs in under 2 minutes flat, including reboot. And Solaris 11 has gone away from Patch IDs etc...they are all pkgs, just the same as Linux.

      - Zones: Are a nightmare of security and privilege. I don't care what any says a zone is just a change root jail. Which means you will only every be as up-to-date as the host system. And it means you must be compatible and tested against the host system. Which is really no different than not having zones. Zones are a horrible horrible mess.

      Zones are a very very cool technology. You can do a lot of powerful things with zones at zero additional cost. And with the advent of Solaris 11, you can now P2V physical hosts into Solaris 10 branded zones, and circumvent the whole patch/package issue. I have used zones with great success (100s of them running multi-TB DWH on M-class, T-class and x86 hardware). They are not a nightmare of security and privilege. The rules are very very clear and concise. There are things you can do within zones and things you cannot. It calls for a little old thing called RTFM, that's all.

      - No dependable only repository of packages that is robust or up to date. Far to much package hunting still required to locate software for solaris. Most packages are
      months to years behind there linux counterparts.

      Not anymore. Solaris 11 SRUs are released at regular intervals and upgrading is as easy saying 1, 2, 3 ;-)

      - Java performs better on x64 than Solaris/SPARC. This has boggled me for years. Only recent sparc architectures let java and other highly threaded applications stacks really perform well. Why do I even have to know about processor binding for processes?

      Try running you java apps within zones with dedicated vcpu and memory grants. And Solaris runs on x64 as well. Consider running java on the T4s and T5s and see how they run.

    11. Re:An F- for the handling of Solaris by tinker_taylor · · Score: 1

      [[[Only major shortcoming with pkg manager in solaris' 11 was with the fact that there was no directly accessible repository of software that could be used (albeit the dependency checks etc were missing too).]]]
      This should read "only major shortcoming with pkg mgr in solaris less than or equal to 10"

    12. Re:An F- for the handling of Solaris by Anonymous Coward · · Score: 0

      Apparently you haven't heard of an operating system called Linux then. Say you had a tiny little smart phone with a low power chip and wanted an operating system for it. Or say you wanted the fastest supercomputer in the world and wanted an operating system for it. The little smart phone lets call Android and the supercomputer lets call Tianhe-2 . Then you would be informed to know these operating systems are built with exactly the same LINUX code! Its not a different operating system for the phone or the supercomputer, its.the.same.code. Now you know. Oh, and I left the link up so you can see for yourself. Are you ready? If you go to the statistics tab, then "list statistics" and select "operating system family" and click the big 'submit' button, you might get skooled about Linux on supercomputers. Or you can be chicken.

    13. Re:An F- for the handling of Solaris by vikingpower · · Score: 1

      Linux is not an OS. Linux is the kernel for an OS. Which is quite vastly different.

      --
      Religous speak to God. Insane are spoken to by God. When all shut up, one can finally hear Shostakovich in peace
    14. Re:An F- for the handling of Solaris by sparkz · · Score: 1

      Package management was being fixed before "the Oracle years". S11 was in progress when Oracle took over Your other comments don't make any sense, so I shall do you a favour and ignore them until you can state them more clearly.

      --
      Author, Shell Scripting : Expert Re
  17. VirtualBox? by utkonos · · Score: 5, Interesting

    Where is the grade for VirtualBox. As opposed to the others on the list, I would give them an A+ for their stewardship of VirtualBox so far. They have released regular updates and bugfixes. I have run into zero problems running Linux, FreeBSD, and Windows in VMs. The UI has gradually improved. The project is still open source, and they actually provide binaries for every major OS.

    1. Re:VirtualBox? by Anonymous Coward · · Score: 3, Interesting

      I'd agree. I'm a huge fan of Virtual Box and it's kept improving, all the time, no matter which company "owned" it, Innotek, Sun, Oracle. A really great job by everyone involved. I've hardly used VMWare Workstation ever, and as far as I can see, whatever lead that had over Virtual Box years ago, has vanished, in terms of features and compatibility. Virtual Box is certainly smaller than VMWare Workstation.

    2. Re:VirtualBox? by gweihir · · Score: 3, Informative

      Indeed. As VmWare networking is now completely unusable, unless you are fine with not being able to ssh-in etc. in the free version, I have moved to Virtual Box, and there are simply no such stupid issues.

      --
      Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
    3. Re:VirtualBox? by Anonymous Coward · · Score: 1

      I agree too.

      Wow, I can run VM's and I don't need to install a web server to manage them.

    4. Re:VirtualBox? by Virtucon · · Score: 1

      Definitely. It's great software and hopefully it'll stay that way.

      --
      Harrison's Postulate - "For every action there is an equal and opposite criticism"
    5. Re:VirtualBox? by Anonymous Coward · · Score: 1
    6. Re:VirtualBox? by drinkypoo · · Score: 1

      Where is the grade for VirtualBox. As opposed to the others on the list, I would give them an A+ for their stewardship of VirtualBox so far. They have released regular updates and bugfixes. I have run into zero problems running Linux, FreeBSD, and Windows in VMs.

      Too bad it still doesn't work. Problems like not using VT even when it's clearly present and activated, or the D3D driver pretty much always crashing the VM completely. Stuff that works fine in vmware, which also has superior performance. I'd really like to like Virtualbox, but it doesn't work.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    7. Re:VirtualBox? by lwriemen · · Score: 1

      From the Open Watcom wiki, "An Open Source license from Sybase allows free commercial and non-commercial use of Open Watcom." How much more free do you want it?

      If you want a sense of history, compile it on OS/2. ;-)

    8. Re:VirtualBox? by twocows · · Score: 1

      Apparently a bit more free for some people. From the FSF's page:
      This is not a free software license. It requires you to publish the source code publicly whenever you “Deploy” the covered software, and “Deploy” is defined to include many kinds of private use.

    9. Re:VirtualBox? by utkonos · · Score: 1

      I've never run into the problems that you're referring to. Are you using a current stable version?

    10. Re:VirtualBox? by utkonos · · Score: 1

      I'm curious as to whether or not SeaBIOS can be integrated into VirtualBox rather than using the watcom bios. This is the BIOS that is utilized by KVM/QEMU and therefore most OpenStack implementation that I've used have SeaBIOS.

    11. Re:VirtualBox? by drinkypoo · · Score: 1

      I haven't tried in a few months, but I've had these problems through many stable versions, sun and oracle, blah blah blah

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    12. Re:VirtualBox? by utkonos · · Score: 1

      Which host OS are you using?

    13. Re:VirtualBox? by drinkypoo · · Score: 1

      Which host OS are you using?

      Ubuntu on x86_64 with nVidia.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  18. Re:Patent MAD by Anonymous Coward · · Score: 0

    Sued Google!

  19. Solaris name dead, but OSS code lives on by Marsell · · Score: 3, Informative

    While Solaris itself is no longer relevant outside of some enterprise niches, it has an actively-developed OSS fork named "illumos", developed by former-Sun hackers working at several different private companies. There are several distributions -- I use SmartOS in particular, and OmniTI's OmniOS is also excellent.

    1. Re:Solaris name dead, but OSS code lives on by unixisc · · Score: 3, Informative

      Illumos is itself an x64 only OS, which makes it worthless for most Solaris users who run it on SPARC, but there is a derivative of that, called Schillix, which is a SPARC specific open Solaris.

    2. Re:Solaris name dead, but OSS code lives on by Anonymous Coward · · Score: 0

      Solaris is not dead. It is still alive in the enterprise. There are actually companies who have seen through the FUD of "Linux as a panacea" and realized that big workloads need a heavy-lifting OS like Solaris.They have started moving back to Solaris and SPARC.

      Like I always maintain, Linux is great on my laptop or desktop. It sucks if you want to run something more than a LAMP distribution on it. The Linux propaganda was a result of certain hardware vendors realizing that they don't have the wherewithal to survive in a Big Iron market, and thereby surreptitiously creating a marketing hype about Linux. I may be old school, but I fail to see the so-called redeeming qualities of Linux besides the fact that there are too many distributions of it, and all the relevant ones cost as much to run and maintain as do commercial OSes such as Solaris.

    3. Re:Solaris name dead, but OSS code lives on by SIGBUS · · Score: 1

      I took a look at OpenSolaris late in 2009 when I was considering building a new storage server for home. I really liked what I saw (ZFS, COMSTAR, and the built-in CIFS support), and wound up using OpenSolaris for the new build. Then, Oracle decided to spoil the party. I'm very glad that illumos got off the ground; once OpenIndiana came out, I switched over to it.

      As much of a Linux fanboy as I may be, I really like OI or OmniOS for storage server duty, and OI makes a nice virtual machine host with VirtualBox and SMF scripts as well. I would like to see better hardware compatibility, though...

      --
      Oh, no! You have walked into the slavering fangs of a lurking grue!
    4. Re:Solaris name dead, but OSS code lives on by unixisc · · Score: 1

      But even w/ that, there is still the choice of FreeBSD, which can do exactly what Solaris can on Big Iron

  20. Quickstarter still exists by robbak · · Score: 3, Informative

    It is still there as an optional item in the installer, not selected by default (because that is the way it should be).

    --
    Prediction for end of Universe #42: Fencepost error in Quantum_bogosort.cpp
    1. Re:Quickstarter still exists by geminidomino · · Score: 1

      Agreed. And thanks for the confirmation. :)

  21. Re:Java in the server, in the client, in the brows by Anonymous Coward · · Score: 1, Insightful

    The problem is Java in the client has tainted perception of Java on the server. Many execs (ie the ones the write the cheques) see Java as something untrustworthy and dangerous and really I can't blame them, If they make such braindead decisions on the client side what is to stop them doing equally dumb shit on the server end.

  22. Re:That James Gosling fella is of no consequence.. by Anonymous Coward · · Score: 0

    And that line of thought is a fallacy, even recognized in the old Rome. Any idea or argument should be evaluated on its own merits, blindly buying into something because "famous person X" said it have probably lead to more misery than any other human trait in history.

  23. Re:That James Gosling fella is of no consequence.. by Anonymous Coward · · Score: 0

    I think the proper phrase is yellow around the beak.

  24. Re:That James Gosling fella is of no consequence.. by Anonymous Coward · · Score: 0

    > The "right" ideas of computing like data schemas, transactions, checkpoints,

    Haaahahahahaaahahahaha. Gee, you know a couple of buzzwords. Where did you learn those, your first year computer science class?

    "Guys, hey guys, listen! I have this awesome idea which I call 'computing done right'. I'm a genius, guys. We'll have schemas and transactions and checkpoints. Bet you didn't think I could come up with something so brilliant, did you?"

    What a joke. You really are a master comedian.

  25. Re:Java in the server, in the client, in the brows by davester666 · · Score: 1, Troll

    I wish java apps would die, along with the people who wrote them, because they NEVER follow the native platform UI. They are all kind of a native app, only a little more sluggish, and the windows don't look quite right, and the menu's aren't organized properly. All in the name of saving a buck making a cross-platform version or using those highly training Indian programming group that only knows java.

    --
    Sleep your way to a whiter smile...date a dentist!
  26. Gosling's Solaris alternatives? by unixisc · · Score: 2
    From Gosling's Solaris grade of F-

    I have had to convert all my Solaris systems to Linux. I weep'.

    What has he done - gone w/ Debian? Since Red Hat seems to have stopped supporting the SPARC ages ago, and I'm assuming that Gosling's Solaris systems are SPARCstations or similar. Which makes me wonder - couldn't he have gone w/ OpenIndiana or Schillix? Especially since it seems to have been more recent? I'm assuming that the BSDs were not an option, since he probably wants an SVR4 based Unix.

    1. Re: Gosling's Solaris alternatives? by CAIMLAS · · Score: 1

      Probably SusE, its the best choice.

      http://en.opensuse.org/openSUSE:SPARC

      --
      ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
    2. Re:Gosling's Solaris alternatives? by kthreadd · · Score: 1

      What has he done - gone w/ Debian? Since Red Hat seems to have stopped supporting the SPARC ages ago, and I'm assuming that Gosling's Solaris systems are SPARCstations or similar. Which makes me wonder - couldn't he have gone w/ OpenIndiana or Schillix? Especially since it seems to have been more recent? I'm assuming that the BSDs were not an option, since he probably wants an SVR4 based Unix.

      Or he could have regular x86 machines. Solaris is available on x86 as well, not just Sparc.

    3. Re:Gosling's Solaris alternatives? by WuphonsReach · · Score: 1

      Sun started selling x86 boxes back around 2005 or so. We have a pair of SunFire X2100 units using AMD Opteron chips inside.

      So unless he specifically stuck with SPARC hardware, he was probably running Solaris on x86.

      --
      Wolde you bothe eate your cake, and have your cake?
  27. Re:Java in the server, in the client, in the brows by imthesponge · · Score: 2, Interesting

    Sometimes reliability is more important then having a pretty UI.

  28. Re:Java in the server, in the client, in the brows by Anonymous Coward · · Score: 2, Funny

    And that's where Java really shines... ah no, wait...

  29. Re:Java in the server, in the client, in the brows by djdanlib · · Score: 4, Funny

    It reliably dumps a longer stack trace than my scrollback can handle, anyway.

  30. Solaris fees & SPARC by unixisc · · Score: 1

    Are you counting 'Sun' as non-Oracle hardware? B'cos the only vendors who really sold Solaris were SPARC based vendors. HP, Dell or any x64 vendor offers only Linux. But the only people who were left high & dry by Oracle's licensing fees were SPARC owners: others could easily go to Linux or any of the BSDs.

    Right now, isn't Solaris really a SPARC only OS, since Oracle supports x64 w/ OEL?

    1. Re: Solaris fees & SPARC by CAIMLAS · · Score: 1

      I can't speak for oracle Solaris but all the opensolaris community forks that are based on illumos are pretty much exclusively amd64 now.

      --
      ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
    2. Re: Solaris fees & SPARC by unixisc · · Score: 1

      Schillix is the one OpenSolaris fork done purely for SPARCs - the intended replacement. How does one replace Solaris on systems where it doesn't exist? Or is the purpose of OpenSolaris to compete against Linux or the BSDs? A futile exercise, since Linux has the traction, and the BSDs the high end features. There is nothing that OpenSolaris offers x64 users not offered by either of the other 2.

    3. Re:Solaris fees & SPARC by hlge · · Score: 1

      Nope, Sun HW is Oracle HW

      I would not be to surprised if OEL is dropped sometime in a not to far future, but on the other hand that would be true for their x64 HW as well

      I'm seeing Solaris becoming a FW in Oracles appliances model going forward, the market for Oracle customers running Oracle on other HW/OS platforms is fairly large and it's the easiest market to address for them.....

  31. solaris by CAIMLAS · · Score: 1

    Solaris is dead. Long live Solaris!

    (Illumos/illumian/Nexenta/SmartOS, that is...)

    --
    ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
  32. Solaris packaging options by unixisc · · Score: 1

    The only thing that (IMO) sucks is the reliance on the sysv style package manager (which they've since replaced with IPS in Solaris 11; not too familiar with it since our company has largely abandoned Solaris thanks to Oracle, but I am of the opinion they should have went with rpm/yum or dpkg/apt).

    A better option would have been FreeBSD's combination of Portsng and PBIs

  33. What? by Anonymous Coward · · Score: 0

    I am afraid he's of no consequence to its direction now.

    It would have been better if he proposed some kind of direction

    Judging by your replies you meant something but for the life of me I can't work it out. What do you mean by those two statements? They seem to contradict.

  34. and Hudson automated build system by coder111 · · Score: 2

    AFAIK authors had to fork and rename it to "Jenkins" after Oracle filed for trademark for original name "Hudson".

    I hate Oracle with passion, although they seem to be doing OK job with Java.

    --Coder

    1. Re:and Hudson automated build system by Anonymous Coward · · Score: 0

      Until you try to download an old version and
      1. have to rely on google to get to the right page (because a, "Here's our old versions" link that actually links to old versions is to easy)
      2. have to agree to some license agreement (not required for current version)
      2. have and be logged into an oracle account (not required for current version)
      3. Now, we'll let you download that jdk (or jre) you need, don't lose it or you get to repeat this process

      Let's not forget the wonderful "option" that prompts you for Hey you're version is outdated, why not download the new one?! EVERY TIME IT STARTS
      Considering I work for a java and .net dev,shop, I see these problems frequently.

  35. Re:Java in the server, in the client, in the brows by allcoolnameswheretak · · Score: 1

    You can have a native platform UI with Java if you use SWT instead of Swing or JavaFX. There is some benefit to be had from Java applications looking uniform across all operating systems, however.

  36. Re:That James Gosling fella is of no consequence.. by Anonymous Coward · · Score: 0

    And, apropos of nothing at all... Writing that "InfoWorld reached out to... [Gosling]" is just awful. Can we please all agree to do away with this ugly bit of pretentious jargon?

    If anyone at the office says he'll "reach out" to me, I tell him I will defend myself, then call HR.

  37. Re:Java in the server, in the client, in the brows by xelah · · Score: 1

    I'd agree that Java UIs are not as good as native ones. But, in the niches where you typically see Java clients such as enterprise apps, what do you do? You can write web applications, which are often not better if they're anything more than a simple or one-off interaction. You could use Flash/Adobe AIR, which is also not better in UI terms. (My favourite gripe is that neither handle keyboard interaction well, but it's far from the only one). Or you could encourage your IT department to write a native application and get something of similar quality and for Windows only.

    What's happening with mobile devices is leaving an interesting situation here because the OS makers seem to be rather discouraging non-native apps. That might water down the advantage of (standard) Java. But I think there's still always going to be high demand from 'enterprise' software developers for a way of writing one client they can run across many devices. I very much suspect that the alternative will be to shove everything in a web app instead.

  38. Re:Java in the server, in the client, in the brows by melmut · · Score: 1

    Since a few years (java 6u10), it's quite easy to have native look-and-feel for everything. No need for SWT, which is an outdated hack. All java apps that I've been used for years (IDE's, various clients) have a native look-and-feel, you can't tell it's not native.

  39. You forgot Minecraft! by Mr.+McD · · Score: 1

    Minecraft, as ugly as it is, is rather popular and written in Java and leveraged the JVM quite a bit. As long the children have a desire to try and find Hero Brine or build their own little words, Java will be around for a while.

    1. Re:You forgot Minecraft! by egranlund · · Score: 2

      Considering they've ground-up rewrote Minecraft 3 times for iOS, Android, and XBox, I have a feeling that Minecraft may not always be in Java.

  40. Re:Java in the server, in the client, in the brows by Anonymous Coward · · Score: 0

    Well, back in the old days (late '90s IIRC), when Java 1.1 came out, it was recommended to switch apps' UI from AWT to JFC (aka Swing) because it was much lighter, not to mention it also offered a wider variety of widgets.

    Swing had (and still has) it's own "native" L&F, called Metal, but nowadays it's also possible and quite easy to detect the platform the app is running on and switch the L&F to that of the platform...but the devs may need to test and make sure the widgets' sizes do not become a problem with the relative/absolute window and container sizes they're defining, and adjust properly.

    I've had buttons (well, JButton) disappear entirely from the UI due to a resize from the user or a change in L&F. But this was back when the best thing we had was GridBagLayout and no IDE that could be actually helpful in making the GUI. Good thing that's a distant memory.

  41. Re:Java in the server, in the client, in the brows by Anonymous Coward · · Score: 1

    Sometimes reliability is more important then having a pretty UI.

    Having a non-standard and unreliable user interface tends to make your application less usable (and in some cases unusable). It doesn't matter if you think java is more reliable than something else if it's not meeting its UI requirements. BTW, crashing less because less people can use your application does not mecessarily mean it's more reliable overall. It just means you've successfully prevented your users from finding all of your bugs.

  42. Reach out by Alioth · · Score: 1

    No. He didn't "reach out to Oracle", he "contacted Oracle" or "asked Oracle". He didn't "reach out" like some emo teenager to their ex girlfriend. It's not as if the phrase "reach out to" is shorter, it's three words where one word will do.

    1. Re:Reach out by vilms · · Score: 1

      If I still had mod points.

  43. Will I be able to afford them in the future? by Torp · · Score: 1

    The problem with products bought by Oracle is that if you're small and they suddenly decide to ask for a lot of money for what was previously free, you're screwed.
    Better to take preemptive action and switch away from them as soon as possible.

    --
    I apologize for the lack of a signature.
  44. Re:Java in the server, in the client, in the brows by allcoolnameswheretak · · Score: 3, Interesting

    An outdated hack? That sounds mean... SWT was great at the time when it was needed. It is the reason why Eclipse never felt like a bloated, slow memory hog, in comparison to other Java applications of similar scope, like Netbeans. With SWT you had native, memory efficient UI components, whereas AWT/Swing duplicated everything into inefficient Java heap memory with slow Java2D rendering. It is true that now, with all the performance improvements Java and Swing have received, you barely notice a difference, so SWT isn't as essential as it used to be, but I still think it has the nicer API. Today I would probably use JavaFX

  45. Re:Java in the server, in the client, in the brows by gbjbaanb · · Score: 0

    Java tainted perception of Java on the server - all those Enterprise program's we're forced to use, they tole us all how good Java is.

    The only thing Java has going for it is the salesmen for the shitty consultancies who sell such crap to the management.

  46. Short memory, a? Oracle Java licensing debacle by Anonymous Coward · · Score: 0

    Doesn't anybody remember the Oracle Java licensing debacle. We have Oracle to thank for discontinuing the licensing which permitted Ubuntu and other distributions to distribute a non-free version that everybody and I do mean EVERYBODY tested again. While I'm glad Sun released the source code under a free software license it wasn't sufficient. When Oracle discontinued the non-free Java re-distribution license without notice Canonical and others had to quickly react. Because they could not continue to provide security updates from Oracle to its customers they were forced to release an update that removed Oracle's Java. That broke millions of systems and was a huge nightmare for Canonical and other companies supporting GNU/Linux.

    I'm in favor of discontinuing Adobe Flash, Java, the ill-fated Silverlight that never really got off the ground, and all other non-free plug-ins (or free ones for that matter in most cases, with maybe a few exceptions, like firebug), but come on- Oracle did the worst possible thing for its customers. It made it near impossible to get it's own product.

    Oracle gets no applauds from me. They royally screwed up Sun. Oracle is the exact reason I am so hesitant to allow anyone even a single f'ing share of my (sucessfull) startup.

  47. Oracle Killed Sun by Bigbutt · · Score: 2

    Personal opinion of course.

    We have SPARC gear along with Solaris 10. When we wanted to upgrade the hardware from the T2000's, the cost for Oracle licenses went through the roof. So we stuck with T2000's (still have them). It kept us from purchasing new Sun hardware. No new hardware, no new business for Sun.

    After much investigation, we went with Dell hardware and Redhat and have been spinning up Redhat VM's right and left. For the mission critical stuff we're using HP gear and HP-UX. We've been spinning up Informix, MySQL, and PostGreSQL in place of Oracle as well.

    [John]

    --
    Shit better not happen!
    1. Re:Oracle Killed Sun by Anonymous Coward · · Score: 0

      and by doing so your "mission critical stuff" gained more failure possibility, congratulations!

  48. Re:Java in the server, in the client, in the brows by melmut · · Score: 1

    It was a great idea some 10 years ago, probably drove Swing where it is now. I use both Eclipse and Netbeans, and the former is the one looking like the bloated slow memory hog to me ;-)

  49. Re:Java in the server, in the client, in the brows by CubicleZombie · · Score: 1

    I wish java apps would die, along with the people who wrote them, because they NEVER follow the native platform UI.

    Client side Java would be alive and popular if they'd made the Windows Look-and-feel the default. Or just given up on Metal entirely.

    Your other complaints are the fault of the developer, not the platform.

    --
    :wq
  50. Re:Java in the server, in the client, in the brows by jellomizer · · Score: 3, Insightful

    So what can you program in that those execs will give you a green light for? I mean they really don't make good decisions off of their choices. They really just pick what they think they like.

    PHP/Python/Ruby etc... It is those nasty open source freeware programs that may be out of style in a few year, we don't want to use those. (and they don't seem to have those mythical enterprise features that they want, but yet never tell us what they are)
    C/C++ Too cumbersome to code in, doesn't allow for Rapid Development
    C#/VB.NET Well they are fine for little apps, we want something a little more heavy duty. Sometimes you will get a better debate about needing a more scailable servers then what Microsoft can provide.
    COBOL/FORTRAN/FoxPro etc... These old languages.

    Unfortunately Java, even with its security problems is seen as the best enterprise choice, because Companies thinks for some ungodly stupid reason that Enterprise software is some how good.

    --
    If something is so important that you feel the need to post it on the internet... It probably isn't that important.
  51. Re:Java in the server, in the client, in the brows by jellomizer · · Score: 1

    All fine and good, until you go to your next platform.
    Java is nice in although your UI get kinda antiquated, the program will run on Windows, Linux, OS X and some other. So if you have a major upgrade to your OS and it still supports Java, chances are your App will still run.
    Unlike say going from Windows XP 32bit to Windows 7 64bit. Where you old app written back for Windows 98 will no longer run.

    --
    If something is so important that you feel the need to post it on the internet... It probably isn't that important.
  52. Re:Java in the server, in the client, in the brows by Anonymous Coward · · Score: 1

    Or the fact that Java code is married to its JVM. Want to move from 6.x to 7.x? Time to rewrite your code. Want to use a JVM on a Mac? Rewrite your Windows JVM code. Similar with Linux.

    Java had a promise of write once, run anywhere. That has been broken. Now, it seems to be install once, get hacked anywhere, because the only breach OS X has ever had in its existance that was widespread was due to JVM insecurity.

  53. Re:Java in the server, in the client, in the brows by nabsltd · · Score: 1

    All java apps that I've been used for years (IDE's, various clients) have a native look-and-feel, you can't tell it's not native.

    Yes, I can.

    Pretty much every Java app uses non-native File Open/Save dialog boxes, even when the rest of the app does look correct.

  54. BSD potential by unixisc · · Score: 1

    Both were ported to just about every CPU in existence - x64/x86, SPARC, MIPS, PA-RISC, Power/POWER. FreeBSD even exists on Itanic, which NetBSD doesn't support, while support for Alpha has been dropped from FreeBSD. Given that FreeBSD now has a laptop version called PC-BSD, I see little reason why such a solution can't run on a similar range of non-x64 platforms. NetBSD too - given that Minix uses the NetBSD userland, NetBSD could itself run on big servers, while sit on top of Minix for embedded applications.

  55. Re: Java in the server, in the client, in the brow by Anonymous Coward · · Score: 0

    You can go from java 6 to java 7 or 8 without recompiling your code.

  56. java on the server side by Anonymous Coward · · Score: 0

    Java is actually a great server side language. Java application servers have tons of features like XA and object pooling that other server infrastructures lack. Frameworks like javaEE and spring are extremely scalable. It also has one of the best communities for open source api's Source: I was the developer and architect for a large telecoms system...5 million users 10 000 tps 50 ms average response time on COMPLEX transactions. I think people who hate java on the server side either dont understand it or prefer to write code thats unmaintainable and inefficient

  57. Re:Java in the server, in the client, in the brows by Dcnjoe60 · · Score: 1

    Java tainted perception of Java on the server - all those Enterprise program's we're forced to use, they tole us all how good Java is.

    The only thing Java has going for it is the salesmen for the shitty consultancies who sell such crap to the management.

    So your solution on the server side is what, .Net? How is that significantly any better than Java on the server?

  58. Re:Java in the server, in the client, in the brows by Dcnjoe60 · · Score: 1

    Sometimes reliability is more important then having a pretty UI.

    Having a non-standard and unreliable user interface tends to make your application less usable (and in some cases unusable). It doesn't matter if you think java is more reliable than something else if it's not meeting its UI requirements. BTW, crashing less because less people can use your application does not mecessarily mean it's more reliable overall. It just means you've successfully prevented your users from finding all of your bugs.

    In the enterprise, where JAVA is primarily used, I'm pretty sure you will find that reliability and functionality outweigh UI requirements. The prettiest UI in the world is worthless if the application doesn't do what is needed. Hell, all of those VB apps had crappy interfaces, but people used them all the time. Besides, today, JAVAs primary use is server-side and there isn't any UI requirement.

  59. Re:Java in the server, in the client, in the brows by Dcnjoe60 · · Score: 1

    You can have a native platform UI with Java if you use SWT instead of Swing or JavaFX. There is some benefit to be had from Java applications looking uniform across all operating systems, however.

    I would second this (not the SWT part but the uniformity accross operating systems). Look at various apps that run in browsers (regardless of the technology involved). None of them have a native look and feel, but if you are on a PC or Mac or tablet or use IE or Firefox or Chrome, as long as the rendering works, people are happy.

    Native look and feel is nice in theory, but the reality is that with the rise of the internet, it is no longer relevant.

  60. Re:Java in the server, in the client, in the brows by Dcnjoe60 · · Score: 1

    An outdated hack? That sounds mean... SWT was great at the time when it was needed. It is the reason why Eclipse never felt like a bloated, slow memory hog, in comparison to other Java applications of similar scope, like Netbeans. With SWT you had native, memory efficient UI components, whereas AWT/Swing duplicated everything into inefficient Java heap memory with slow Java2D rendering. It is true that now, with all the performance improvements Java and Swing have received, you barely notice a difference, so SWT isn't as essential as it used to be, but I still think it has the nicer API. Today I would probably use JavaFX

    All true, but let's not forget that SWT was never meant to be used for general applications. From the start, it was created to develop the Eclipse IDE. The fact that other people used it for their own purposes doesn't change that.

  61. Re:Java in the server, in the client, in the brows by gbjbaanb · · Score: 1

    wha? .net? me?

    I think you have me confused with someone else.....

    I do serious software, where my skill with the tools mean I don't have such a productivity hit as others who need java or .net to keep up. I prefer C/C++ but I do turn my hand to quite a few different technologies as appropriate.

  62. Re:Java in the server, in the client, in the brows by Dcnjoe60 · · Score: 2

    wha? .net? me?

    I think you have me confused with someone else.....

    I do serious software, where my skill with the tools mean I don't have such a productivity hit as others who need java or .net to keep up. I prefer C/C++ but I do turn my hand to quite a few different technologies as appropriate.

    Server side programming in C/C++ can be done and many do, but I would question why? Since most server side work is to serve up various web pages, they tend to be constrained by IO not memory or cpu. Not only is there the initial development time, which seems not to be an issue for you, given your sill level, but there is also maintenance work, where the next person might not have your level of expertise in C/C++.

    Just like Java can be used for client side programming, although it isn't optimum, it would seem that C/C++ would not be the first choice on server side. As for Anything.Net, I only use it if a client or project requires it. I find that it is good at doing many things, but not great at any of them. The one thing it has going for it is that it is a Microsoft technology, so decision makers who don't necessarily know any better tend to specify it in the requirements. That keeps it alive and kicking. In the 60s and 70s, the old adage was "Nobody ever got fired for buying IBM." Today, the old adage still rings true if you substitute "Microsoft" for "IBM" at least at the enterprise level.

  63. Re:Java in the server, in the client, in the brows by wiredlogic · · Score: 1

    Ugly and slow applets in the browser thankfully are almost dead â" Because HTML5 delivered way better.

    It's pretty sad when your carefully designed platform-neutral architecture intended to supplant native code, in development for 20 years, can be overturned by an SGML derivative and an ad hoc script language.

    --
    I am becoming gerund, destroyer of verbs.
  64. Re:Java in the server, in the client, in the brows by WuphonsReach · · Score: 1

    I wouldn't lump C# in with VB.NET. C# is a lot closer to Java and they both hit about the same market segment. Both are equally capable for large projects.

    The main advantage of Java over C# is cross-platform compatibility for your business logic / business objects. It's the COBOL of the next 20-30 years. A good read is "Java: The Good Parts" (published in 2010 by O'Reilly).

    Use C# and you're stuck running on Microsoft software stacks (or *maybe* Mono, if you are brave).

    We spent 2 years looking at various languages as we moved off VB6/VBScript/ASP. Locking ourselves into C# was not an option. PHP and a lot of the other languages fail to scale past 100k lines of code. Ultimately settling in on Java / Spring for anything that was going to be more then 5k lines of code.

    --
    Wolde you bothe eate your cake, and have your cake?
  65. The Sun is burning out by Darkroom · · Score: 1

    Sun's acquisition quickly went down the drain, when Oracle started killing off things, Open Solaris, OpenOffice, MySQL and on and on.

  66. Re:Java in the server, in the client, in the brows by gbjbaanb · · Score: 1

    Its a common fallacy that server side stuff if just shuffling disk to network, hence the (current, but fading) popularity of node.js

    What I find in all systems that are not trivial is that the middle tier gets a lot of processing bundled into it - read that disk or DB data, and fiddle with it, often combining it with other data sources and then send it down to a client. That fiddling requires quite a bit of processing.

    Its one reason why Microsoft went as native-code as it did a couple of years back, their 'Casablanca' project (now released as C++ REST Services) because someone measured how much electricity their cloud services were sucking up with managed code (and to be fair, .NET is less resource intensive than many script-based languages) and they needed something more efficient. (Either that or someone noticed how seriously faster C WWS web services were compared to .NET WCF services :-) )

    All the places I've worked that do serious stuff, base their distributed services over 3 tiers - the usual web presentation (or sometimes thick desktop) calling a middle tier business logic layer that calls the DB. The middle tier does all the heavy lifting, so its never just IO shuffling.

    I'm sure many websites simply call a web server that calls a DB and does the mapping of data to UI in the client via javascript, but its not the most efficient way of presenting that, especially if there's a lot of data, or it needs processing.

    Maybe there;'s a distinction to be made between the "website" devs and the "professional" devs in the type of systems they develop. I think its a shame the "website" style where everything is placed in the web server (bad security choice that) should be designed with 3 tiers from the start, and for these types of system, a C++ based service layer is not any more difficult than any other language to develop for. .NET, its easy to develop for, which is why everyone seems to be using it. Its not nice when it goes wrong (like the bug I struggle with today - reading event log entries returns null on my colleague's box, for no F*** good reason.. damn you Microsoft) but even Microsoft knows its their RAD tool, not the one that should be used for performance or resource efficient systems. To put it another way, .NET is the new Visual Basic - where VB used to be used, .NET fills that gap. The trouble is, it also attempts to fill every other gap (but I guess VB devs back in the day used to do that anyway)

    Java.. no need for that anywhere IMHO :-)

  67. Re:Java in the server, in the client, in the brows by Dcnjoe60 · · Score: 1

    Maybe there;'s a distinction to be made between the "website" devs and the "professional" devs in the type of systems they develop. I think its a shame the "website" style where everything is placed in the web server (bad security choice that) should be designed with 3 tiers from the start, and for these types of system, a C++ based service layer is not any more difficult than any other language to develop for. .NET, its easy to develop for, which is why everyone seems to be using it. Its not nice when it goes wrong (like the bug I struggle with today - reading event log entries returns null on my colleague's box, for no F*** good reason.. damn you Microsoft) but even Microsoft knows its their RAD tool, not the one that should be used for performance or resource efficient systems. To put it another way, .NET is the new Visual Basic - where VB used to be used, .NET fills that gap. The trouble is, it also attempts to fill every other gap (but I guess VB devs back in the day used to do that anyway)

    Java.. no need for that anywhere IMHO :-)

    With regards to .Net, it provides a very good IDE, but overall, it occupies the same use case as Java, at least C# does. C# does do many things better than Java, but if one has to maintain legacy code or is involved with non-Windows platforms, .Net isn't an answer. Ultimately, I think that was Microsoft's purpose of .Net, not to build a better Java, but to lock people in to their products.

    With Java or C/C++ or just about any other language, it is platform neutral (that doesn't mean what one develops is cross-platform, though). I can run those on anything from a pc to a cluster to a mainframe. That is not the case, at least not out of the box, with .Net applications. They are microsoft-centric.

    Java is still really useful if your middle tier isn't on a Microsoft platform. Java is like COBOL. Everybody wants it to die, but there is a lot of legacy code out there that needs to be maintained and it will be around for a very long time.

  68. OpenIndiana... by Anonymous Coward · · Score: 0

    Illumos is the kernel/userspace, which AFAIK is platform agnostic. Unfortunately, despite still having sparc64 support, none of the major distros have included a sparc64 port but rather stuck to x86/x86_64.

    I was really pissed to find out all the sparc32 hardware/code was taken out right before the opensolaris release. That was the stuff I would've been most interested in (since the old SBUS hardware was really what SPARC was about. Which it's cool to have PCI sparc system support, getting devices that will work with openfirmware during is a huge PITA. The most critical of these components being Ethernet, Video, SATA, and USB controllers, all of which need either boot proms written, or a forth version of the vm86 code from X. Which is probably about to become much messier with 'EFI Only' PCIe devices coming out.)

    1. Re:OpenIndiana... by unixisc · · Score: 1

      That's why for people w/ such old unixstations, a better idea would be to run OpenBSD or NetBSD on them (I'd imagine FreeBSD no longer supports x32). But if you are talking about SBUS era hardware, SATA didn't exist then - all you had was PATA, and since we're talking about reusing such old boxes, rather than making new computers based on SPARC

      Actually, one more option - MINIX v2 - not the current v3 microkernel - too was 32 bit only and would probably support the hardware that you describe. But I think you'd miss some of the Solaris features you're used to if you go by that one.