Slashdot Mirror

I think you already answered your own question. There is no "one size fits all" and if you can live with a 10% decrease in overall GC performance, then you can enable -Xincgc and have less GC related pauses. it's just depending on what you want and what your use case is.

If I understand Performance Options Option and Default Value correctly, then most the options you mentioned are enabled by default already anyway. Except the flags DisableExplicitGC and UseParallelGC, some other you mentioned I didn't found, and ParallelGCThreads is already set to the value for the platform where the JVM is running.

Some of the performance related options are already enabled by default, and there are some you can set if you know your machine. So I don't really understand your complain.

Actually, re-using objects is an excellent strategy if you are using JavaBeans (where you have symmetric getters and setters). Unfortunately, so many programmers hide setters away from you, supposedly to protect you from yourself, but it makes it impossible to re-use objects even where it would otherwise not only be perfectly possible and valid but also very efficient as well. In short, use JavaBeans when you need complex (non-intrinsic) value-objects and make the setters symmetric with the getters, *always*. You don't generally need to work pools of objects if you can re-use a JavaBean directly. Immutable objects have their uses, but they are very expensive in many performance-critical situations. If an object doesn't have to be immutable, don't make it so (keep it a Plain Old Java Object [POJO] where you can).

Although if you do need pooling (and you do from time to time) then Apache Commons Pool has done all the work for you. Real-life example, I worked on an Internet-scale project where the cost of obtaining a java.util.Calendar was too expensive (since it internally gets a timezone) and affected a critical part of very heavily used code. Pooling instances was very helpful in keeping high performance.

As always with performance you ought to be profiling your code before optimizing. Fortunately the Java SDKs come with JVisualVM which is absolutely fantastic - you don't need to compile a special profiling version of your program (like C or C++) you can simply attach to *any* running Java process and see what it is up to. You can even do this to remote processes (although you get slightly less information than if you hook up to a local process). This is an unbelievably good feature that allows you to optimize your Java sensibly - which should result in code at least as fast as C if you know what you are doing.

Once you have done your profiling with JVisualVM you can figure out what you need to do next:
Am I doing a lot of string concatenation (eg. text or XML processing)? then changing one or two Strings to StringBuffers can often give you a factor of 10 speedup in critical code (I've done this myself)
If you are creating a lot of objects then having a fully mutable JavaBean can help (I've done this with a Matrix class - which saved re-initialising the identity matrix a lot).
If you are CPU bound then how about splitting your working into multiple threads? Much easier than doing multithreading in C or C++
Still don't have enough CPU power? cool, you're in Java so move your application from a little Windows box onto a Linux/Unix huge blade server.
are you multi-threaded but lock contention for shared resources holding you back? then get rid of that fancy shared singleton and use non-shared POJO objects in each thread to hold what would otherwise be shared state.
Running out of memory and is the GC thrashing? simple, just give your application more memory with the -Xmx JVM parameter.
Doing a lot of iteration over lists or arrays to find things? then create a java.util.Map to cache lookup data.
These are all simple fixes that can make Java very fast. JVisualVM will help you make the decision about where to start.

Of course, the classic paper discussing how fast the JVM had got even 4 years ago from the French supercomputing research arm of INRIA:
https://blogs.oracle.com/jag/entry/current_state_of_java_for
If anyone still thinks the JVM is slow then they are out-of-date. As long as your developers are competent you can make Java applications very fast.

It seems this kind of volatile deep non-documented black magic might change radically from JVM revision to revision. Although the Oracle "documentation" page seems to contain a lot of "legacy" options, there still seems a risk that this book would be outdated as soon as the next JVM release.

Oh, well, the tech publishing industry seems to be doing pretty well, even if the rate of technology change means that a tech fact is OBE before it's committed to ink.

We went to all virtual tape about a year ago (although there are tapes, it's all STK robots handling them). OS upgrades are probably FTP'd now (I'd have to ask to be sure). The JCL is unchanged since the OS thinks they are real tapes. But they are not the old round reels with 2400 feet of tape, they are cartridges. http://www.oracle.com/us/products/servers-storage/storage/tape-storage/029139.htm

wow, i thought you were joking until I looked it up

http://blogs.oracle.com/templedf/entry/it_s_the_tachyon_signature

Intel was the top contributor to Linux 3.0 (by lines) (source)

IBM is in there, too at #8

Google pushed the Linux kernel and WebKit into an uncountable number of handhelds

Apple deploys Webkit, too, on a smaller number of handhelds

Amazon deploys Android, too (just without Market support), and they use Linux in their cloud offerings.

If you hate Microsoft, give in to your anger and join Oracle (there are a lot of angry JCP and OpenSolaris fans but hey, they made that Linux list, too!)

Remember those handhelds that run the Linux kernel and/or WebKit?

• Broadcom
• Atheros (are they are part of Qualcomm now? You can check out Qualcomm, part of "Qualdroid")
• Marvell

all made the top Linux contributor list, too.

I'll assume that other posters will cover the Red Hat and Novell bases.

Compared to iOS, OS X is a hippie commune all smoking freedom joints in a big freedom love-in; but it has its quirks...

Oh, $DIETY.... Not taken Scientific Programming in College, I assume? Okay, I'm 35, and I have known about this shit since College, and frankly, even young whippersnappers should know about it. But.. Here's some reading for you

What Every Computer Scientist Should Know About Floating-Point Arithmetic If you don't want to read, but I really URGE you to: it's the "non-precision" is intrinsic to the representation of floats and the errors propagate quite quickly. This has absolutely nothing to do with the example you give, because that is pretty much irrelevant. Besides, it's just illustrates your limited vision due to your local currency. Pennies? How quaint... Yens (for example) don't know these fractional units. My currency before the euro didn't either. (That's one thing, I'd like most programmers to have: get involved in a real international project, with multiple locales, multiple currencies, multiple translations... You'll immediately see how badly most programs are on that level).

Money in Java is the article I usually point programmer rookies to if the above article goes above their head.

I assume they're referring to:

http://www.oracle.com/technetwork/topics/security/cpujan2012-366304.html

My mystification is what is the venn diagram intersection of mysql server, virtualbox, and oracle 11G? Without any details I'm guessing a package signing key got owned?

That CPU contains 78 fixes, one of which presumably is the one referred to in the article. So no, the problem isn't common to all the products, it is just lumped with other problems in the monthly patch announcement.

I assume they're referring to:

http://www.oracle.com/technetwork/topics/security/cpujan2012-366304.html

My mystification is what is the venn diagram intersection of mysql server, virtualbox, and oracle 11G? Without any details I'm guessing a package signing key got owned?

Great joke from the 90's.

In case some of the Slashdot readers take the joke (and what used to be true) as the current state of affairs I thought it worth correcting them (otherwise they will have a mistaken view of the *current* performace of the JVM). It turns out today that Java on the Oracle JVM is faster than pretty much every other general-purpose language except for FORTRAN (which is fast 'cause it so simple - which is why FORTRAN programs still dominate much of supercomputing). Don't take my word for it. Take that of James Gosling (a biased source): http://blogs.oracle.com/jag/entry/current_state_of_java_for and the French supercomputing facilties of INRIA (an unbiased source): http://hal.inria.fr/inria-00312039/en

That may be so, but the GUI handling in Java royally sucks. I very rarely see a graphical Java app that can't bring the fastest of PCs crawling painfully slow.

In case some of the Slashdot readers take the joke (and what used to be true) as the current state of affairs I thought it worth correcting them (otherwise they will have a mistaken view of the *current* performace of the JVM). It turns out today that Java on the Oracle JVM is faster than pretty much every other general-purpose language except for FORTRAN (which is fast 'cause it so simple - which is why FORTRAN programs still dominate much of supercomputing). Don't take my word for it. Take that of James Gosling (a biased source):
http://blogs.oracle.com/jag/entry/current_state_of_java_for
and the French supercomputing facilties of INRIA (an unbiased source):
http://hal.inria.fr/inria-00312039/en

Even then, you can download JRockIt 28.2 separately. As far as I can tell, this is synced with Java 6 r29. Since Java6 r30 is the latest Java6, it's not that far out of date.

Now, whether it will be compatible with Java7 is a whole different ballgame.

Side note: java.com does not yet offer Java7 for download, so 6r30 is the latest version in its eyes, despite 7u2 being out.

Actually, JRockit * is * being killed off.

Nonsense, Oracle still distributes it as part of Oracle Fusion.

(Caveat - I work for Oracle too) Anytime you hear that someone is using one of these big Exadata machines, they are using Oracle Linux. See: http://www.oracle.com/us/corporate/features/exadata-196883.html

Oracle is aggressively marketing it's ZFA-SA (Storage Appliance) as a competitor to the likes of Netapp and EMC. http://www.oracle.com/us/products/servers-storage/storage/nas/overview/index.html

Hopefully not too many people depend on hashCode() always returning the same value for identical objects between runs. Fixing this vulnerability, the way TFA recommends, would necessarily break this expectation. I see in the documentation they word the contract in such a way that people should be scared off from doing it this way, but you never know :)

Linux JAVA from oracle site: http://www.oracle.com/technetwork/java/javase/downloads/jdk-7u2-download-1377129.html

While I love to bash on Ubuntu on every (reasonable and merited) opportunity available, and they certainly aren't scarce, this isn't one of them. As others have already pointed out, the packages were removed because Oracle will not license updates, and the latest distributable version has important security vulnerabilities. It would be irresponsible to keep the current packages in the distribution and illegal to update them.

More importantly, this move is exactly what Oracle wants done, and no, it's not any sort of evil move. Dalibor Topic explains in his blog the reasons behind this change in licensing: OpenJDK is (the basis of) the reference implementation for Java 7, and the Sun (now Oracle) JDK implementation is now (going to be) based on OpenJDK; the gratis, non-free licensing for the Sun (now Oracle) JDK was a temporary solution that's reached the end of its applicability:

That non-open-source license was introduced by Sun Microsystems back in 2006, when the open-sourcing of Sun's Java SE implementation was announced at JavaOne, as a stop-gap measure until OpenJDK matured. It was a way to enable Linux distributions to take Sun's JDK 5.0 and provide their own 'native packages' based on Sun's non-open-source bits.

It was always intended to be a temporary solution, and the final solution has always been migrating to OpenJDK. Yeah, it sucks, compatibility is far from complete, and things will break as a result of this move, but it's always been the plan, and it's not Canonical fucking it up this time. For reference, as one of the comments in TFA points out, Debian did it too.

In short: nothing to see here; move along. If this makes you lose sleep, maybe you shouldn't have used Java, and maybe you should migrate to something better.

On Linux, most java developers consider that OpenJDK is the default implementation and that Sun JDK is more or less discontinued.

OpenJDK is a GPL release of Sun's code. It is the official Java (SE) implementation :

http://blogs.oracle.com/henrik/entry/moving_to_openjdk_as_the

You must work for Microsoft TechNet; while your answer is technically correct, it is also useless for the person who asks.

Ah, an ad hominem as the very first thing. It's kind of hard to offer support (which I wasn't trying to do) without knowing which VM and I would suggest a forum dedicated to that VM would be a more appropriate place rather than /. especially as there are conflicting comments about 3D working/not working under OS X depending on the exact VM and OS X. Perhaps some combinations are simply a no-go.

Neither VMware nor VBox support 3D acceleration for remote or shared images.

3D Acceleration has been available in VBox since around version 2.1, they're now on 4.1.6. One does need to have the guest additions installed
https://www.virtualbox.org/manual/ch04.html#guestadd-video

As to VMWare, one of the things that kept me on VMWare Workstation for so long as the fact that it did 3D accelerations whilst VBox did not. It seems that even Server offers 3D acceleration (I have not tried it myself) although server hardware is not usually known for its killer graphics card.
http://www.dedoimedo.com/computers/virtualization-3d-support-vmware.html

You are quite correct on remote/shared images. Whilst the server might be able to do the acceleration and the guest set-up to run the acceleration, it simply will not work/appear on the remote client. There may be some help in getting that to work, but I'm afraid I don't have the time to look into it just now
http://blogs.oracle.com/vizsun/entry/hardware_accelerated_remote_3d_windows

I have run Mint 12 and Fedora 16 (Gnome 3 Shell) and Arch (KDE) amongst others without issue (not as a remote, obviously). All with full transparencies, eye candy etc. One can even run 3D games in these virtualised environments without issue (as a quick test, "Extreme Tux Racer" at 60fps, Saurebraten [high settings] at 20fps in a Mint 12 guest).
XPsp3 (checked via dxdiag) had failures for DirectX 7 and 8, but the test for 9 worked perfectly.

The acceleration offered isn't as good as native hardware acceleration and one doesn't get all the latest features due to the nature of the virtualised hardware exposed to the guest.
Hardware: Dell Dimension 9200, 4GB with a NVidia GT240, Ubuntu 10.10 (proprietary NVidia drivers installed); not cutting edge by any means.

Oracle supports stored procedures in Java as an example. Personally I prefer postgresql.

And they haven't done anything about it for years, either.
http://blogs.oracle.com/malte/entry/evilgrade_and_openoffice_org

It is a commercial agreement to prolong support and development of a component that is vital to HP's line up.

It is? Where? "Such an important contract, if it existed, would obviously be a heavily negotiated, fully documented formal contract, with terms and conditions and payment obligations and all the other characteristics of real-world commercial agreements. But there is no such agreement for porting the Oracle database to Itanium, which is why HP is advancing implied contract and promissory estoppel claims." http://www.oracle.com/us/corporate/features/opposition-to-motion-to-seal-423345.pdf

That's funny. Not to long ago Oracle stated that they have proof that Intel was killing Itanium and that HP was harming their own customers by not admitting it. Now they say that the exact opposite is true; that HP is paying to ensure that Itanium stays alive. Either this change occurred after Oracle dropped their support for Itanium (unlikely), or Oracle just admitted that they have been printing libelous statements about HP, in addition to breaking their contract with them.

I hope the assholes pay for both.

There is a JVM targeted for research named Maxine that is written in—you guessed it—Java. It needs to be bootstrapped by another JVM in order to compile itself with its JIT compiler and after that it runs by itself. I recently attended a talk by one of its developers and it looks very interesting.

Disclosure: I work in backup and recovery for a living. I've been doing TSM professionally for the past six years.

So let's look at this idea that portable HD is cheaper and faster. I've managed backup systems that were backing up tens of terabytes every day. Let's say three 3 TB HDDs for the sake of discussion. Let's say five weeks' worth of data retention. That's 5*7*3 = 105 hard drives that need to be managed - call it 110, to take into account the fact that there are going to be a few going off and onsite at any given point in time.

How reliable are those hard drives? How long until they fail? Bear in mind when considering your answer that hard drives are designed to be stuck inside a chassis that stays put (for the most part) - pulling them out of the system and lugging them halfway across town is not within their design goals. And then you need to worry about the cost of a hot-swap chassis so you can pull the drives out, and shove them back in, every day. Oh, and don't forget that most plugs have a limited life span - probably in the hundreds of swaps, maybe thousands (I don't know, and I'm happy to hear solid data about this.)

And how long does it take to fill up a hard drive, anyway? Take the WD Green 3 TB as an example: 110 MB/s (source), equals about 7.5 hours, best case. Sure, you can throw more hard drives at the problem in parallel, but that just exacerbates the whole question about reliability in transit. The Hitachi 3 TB is faster - 207 MB/s - which takes about four hours to fill up. That's best case scenario, based upon the maximum data transfer rate - guaranteed it's going to slow down as the drive fills.

Now consider tape. Consider wikipedia's information on LTO when reading this. LTO4: 120 MB/s native, up to 300 MB/s compressed (two hours to fill). LTO5: 140 MB/s native, 350 MB/s compressed (three hours to fill). Pretty damn reliable in transit; they're designed to sit happily in their little plastic shells, in a box, and get thrown around (not quite, but they can certainly take more punishment than a hard drive can.) Capacity per cartridge: 800 GB/1.5 TB native (4/5); 1.6 TB/3 TB compressed. If you have money to burn, you could go for Oracle's T10000C drive: 5 TB native capacity, 240 MB/s native throughput (and that's before you get into the whole question of compression.)

Now let's get onto the whole subject of financial data, Sarbanes-Oxley, and WORM media (so you know the data hasn't been altered since it was written out) ...

Sure, tape sucks. It has major issues when data is scattered all over the place; mount time takes a while; and the drives need regular love and attention. But here's the thing - it survives today because it's better than the alternatives in its niches, and trust me, there's plenty of niches where tape fits in far better than the alternatives.

If you only have a couple of TB of data to backup, the cost of setting up tape infrastructure will probably not be worth it. But when you're talking hundreds of TB - or better, petabytes of data (don't laugh, one client I did work for had over 2 PB of data in their tape library) - the cost equation swings over pretty damn fast. Tape is not dead. Far from it. I can't see the likes of IBM investing in developing LTO6 and LTO7 if there was no use for it. And why would Oracle sell a tape library that scales to 100,000 slots if there's no demand for it? It's not about how to get the most bytes for your dollar - it's also about reliability, and that gets down to the usage. If I suggested portable hard drives to the clients I do work for, I'd be out of a job - because they simply won't cut it for their needs.

Again trying to avoid proving what you claimed. You're such a liar, and SO easy to prove you're a liar. PROVE that Oracle or Microsoft bribed someone.

Name names, dates, and the type of gift. Both Microsoft and Oracle have policies about that.

Oracle's Code of Ethics and Business Conduct - see page 13 - Gifts

Oracle employees generally may accept unsolicited gifts or other business courtesies from actual or potential customers, suppliers, or other business partners provided they are not of material value and are not given with the purpose of influencing oneâs judgment. It is never appropriate to solicit gifts or other courtesies directly or indirectly. If you are offered a gift or other business courtesy of material value from an individual, firm, or representative of a firm who has or seeks a business relationship with Oracle, you must demonstrate that the gift could not be construed as an attempt by the offering party to secure favorable treatment. You must obtain written approval from the Senior Vice President of your organization and from your Regional Compliance & Ethics Officer before accepting the gift.

As an example, on page 14 they say that if a vendor sends them a gift certificate at their home "because it's the holidays", they have to return it to the vendor along with a note explaining that Oracle's policy forbids it.

Microsoft's Code of Business Conduct

Microsoft prohibits corruption of government officials and the payments of bribes or kickbacks of any kind, whether in dealings with public officials or individuals in the private sector.

So, again, since Microsoft prohibits all bribes or kickbacks, even in the private sector, or gifts totaling more than $200 per year under any circumstances, put up or shut up.

You keep trying to weasel out of proving what you said. Prove your claim that Oracle or Microsoft engage in bribes via inappropriate gifts, since both companies have policies against bribes, including gifts. Name ONE instance, since you (falsely) claim to have seen this when you (allegedly) "worked for telcos, banks, and insurance."

Also no logic. Are you a chick?

Riiiight, after all, women are not logical .... and yet, you're the one who keeps looking like an idiot, trying to lie his way out of things. Oh, and now we can add misogynist to the list of your character flaws.

Nobody will be supporting mobile apps within 10 years, because all mobile devices except for the very lowest entry-level ones will be able to run desktop software.

The reason is simple - as entry-level devices get smarter, the only way to maintain margins on higher-end devices will be to make them more capable. Microsoft isn't stupid - MetroUI isn't about making a tablet interface - it's a trojan for eventually getting the full-blown Windows OS onto tablets. It's their first smart original idea since ... well, forever ... and it's going to be effective.

This is where going with iOS instead of a cut-down OSX is going to bite Apple. It's also going to be somewhat of a problem for Android (though not as much, since the underlying OS - a heavily modded version of linux - is still going to be able to be modded to expose the underbelly to non-android applications.

The can be worked around by making mobile devices that have enough functionality to replace desktops, which are specifically approved (or laptops, which are mobile devices, but also approved),

"General Purpose Desktop Computers and Servers" means computers, including desktop and laptop computers, or servers, used for general computing functions under end user control (such as but not specifically limited to email, general purpose Internet browsing, and office suite productivity tools).

If it's running a desktop OS, runs desktop programs, can even look like a desktop in terms of the general form factor (screen + optional keyboard + mouse), you're good to go. And of course, now that we can use things like Wiimotes and kinects (and my laptop comes with a remote control of its own), the use of alternate input devices - including touch screens - is irrelevant.

Additionally, since the mobile restrictions only apply to dedicated devices:

The use of Software in systems and solutions that provide dedicated functionality ... or designed for use in embedded or function-specific software applications ... (long list of dedicated devices, including mobile phones) ... are excluded from this definition and not licensed under this Agreement.

... smartphones are already "general enough" that you could get away with installing desktop java on them - you can use them all day without making a single phone call - I'm surprised that Google hasn't thrown that at Oracle.

The real problem is going to be the same as always - once people find that they can run their Windows programs on certain mobile devices, those are the devices that they're going to buy. But that's several years down the road (though closer than you'd think).

Actually this is a repackaged Berkeley DB Java Edition. See http://www.oracle.com/technetwork/database/nosqldb/learnmore/nosql-database-498041.pdf

C# and Java both have atomic operations in the standard library. See Interlocked.CompareExchange, and java.util.concurrent.atomic for examples.

Multi-threaded programming is particularly easy in those languages, because a lot of their internals are inherently thread-safe. For example, strings are read-only, so they can be passed around risk free. Similarly, mark & sweep garbage collection is thread-safe, and doesn't suffer from the rare but complex to debug memory leaks that occur with reference counting. It's also faster -- there's garbage collectors in common use now in the Java world that significantly outperform malloc. Throw in the overheads of atomic increment/decrement required for thread-safe reference counting in the C/C++ world, and suddenly things can tip towards Java in a big way.

I do C# mostly myself, and I've found that because it makes multi-threading so easy and safe (compared to C/C++), that I use it much more often than I would otherwise. Even if it's a tad slower than C, the ability to liberally sprinkle multi threading throughout the code makes the end result a lot more parallel, and hence overall faster or more responsive.

Take a look at the new await and async keywords about to be added to .NET v4.5. They allow traditional serial code to be converted to a thread-pooled asynchronous version with what amounts to about two dozen additonal characters of code!

Meanwhile, C and C++ have poor support for multi-threading, especially if code needs to be portable. There's basically no threading standard library to speak of, or even a threading aware memory model!

Solaris 10 and 11 are supported on x86 and available for free download.
There's even a VirtualBox VM image linked from that page.

The Solaris 11 release candidate is also available to registered developers on OTN.

Java 7 has closures.

Really? I couldn't find any information on it in the release notes here: http://www.oracle.com/technetwork/java/javase/jdk7-relnotes-418459.html I've heard people talking about it but I thought it didn't make it in.

Tell that to Oracle:

http://www.oracle.com/us/technologies/virtualization/061153.html

The btrfs actually has compression (although I think you can't switch it on/off per-directory in single btrfs filesystem). http://blogs.oracle.com/wim/entry/btrfs_compression

We have been using OCFS (Oracle Cluster File System) for some time in production between a few different servers.

Now, I am not a sysadmin so can't comment on that aspect. I'm like a product manager type, so I only really see two sides of it: 1) when it is working normally and everything is fine 2) when it stops working and everything is broken.

Overall from my perspective, I would rate it as "satisfactory". The "working normally" aspect is most of the time; everything is relatively seamless - we add new content to our servers using a variety of techniques (HTTP uploads, FTP uploads, etc) and they are all magically distributed to the nodes.

Unfortunately we have had several problems where something happens to the node and it seems to lose contact with the filesystem or something. At that point the node pretty much becomes worthless and needs to be rebooted, which seems to fix the problem (there might be other less drastic measures but this seems to be all we have at the moment).

So far this has JUST been not annoying enough for us to look at alternatives. Downtime hasn't been too bad overall; now we know what to look for we have alarming and stuff set up so we can catch failures a little bit sooner before things spiral out of control.

I have very briefly looked at the alternatives listed in the OP and look forward to reading what other reader's experiences are like with them.

Since when do "open source" software vendors try to prevent people from using their software in whatever manner they want?

Since a long long time. Apple, Oracle, Microsoft, are all "open source software vendors". And they all "prevent people from using their software in whatever manner they want".

I see a lot of other tools that do this, but since C# mostly started off as a ripped off Java, it's also worth pointing out that since Java 1.6, that language also provided public interfaces to compile code at runtime.

These are nice features. Sometimes, they are even useful (as opposed to just another hammer developers can abuse). But the announcement makes it seem, wrongly, that MSFT is doing something really unique here.

Anyone ever seen Beehive? When I first saw it, it was a freshly steaming turd. No idea what it's like now, but at the time it was largely broken.

OMFG, we actually sell this shit to people!?

At Oracle, we are actually forced to *use* Beehive. There is no way that the version that people buy can suck any worse than what we're stuck with internally.

Instead of being so high and mighty ... oh never mind, whats the point, its not your fault, its someone elses, your code is awesome and everyone will bow down to you guys. I know you guys like to think Linux is ruling the world, but you're still no where near big enough to start trying to pull an Apple/Google/Microsoft and force people to do it your way. You've tried this before and again, you'll lose.

Um, did you even read the article?

Someone released a driver for Virtual Box, said driver causes instability and crashes.

Do you think it's the job of the Linux Kernel devs to re-tool the kernel to work around this, or do you think it's just easier to push it back to the people who wrote the driver?

I mean, seriously, from TFA:

Even though this VirtualBox driver is open-source (it's under the GPL), the quality of the driver is quite poor and continues to cause issues for many users. In particular, kernel developers have become frustrated that this virtualization driver is causing random memory corruption. Specifically cited is "corrupt linked lists, corrupt page tables, and just plain 'weird' crashes."

The code comment for the patch mentions, "vbox is garbage." The VirtualBox kernel driver is needed for providing some features to guests on this Sun/Oracle virtualization platform. While the VirtualBox kernel driver is open-source, it doesn't live within the mainline kernel tree and is distributed separately with the VirtualBox software package.

So, if you start off with a working, stable kernel, apply this patch, and then end up with a broken, flaky kernel ... what is the conclusion other than the driver is crap?

I'm not a Linux kernel developer ... but I have had someone try to write some badly written code on top of some systems I supported, only to have them come back and start filing large amounts of bug reports ... and by the time you waste your own time to realize this has nothing to do with your own code, it's too late. Hell, I even had one occasion where someone ignored the explicit statement that it wasn't thread safe, and definitely didn't implement transactions ... only to submit a bug report whining that the transactions didn't work like he wished them to. Of course it didn't, it said right up front it didn't and never would ... but he figured if he just pretended that it did, he'd be able to force us to make it do so. How was that my fault?

If this module is leading to support issues, I can see why they'd draw the line and say "not our fault or problem".

If I wrote crappy code for a Windows app, do you think Microsoft would be willing to listen to me submitting bug reports in Windows if it was becoming readily apparent that the problem wasn't in their code? Because, that's really what this is about from the sounds of it.

I mean, really, Oracle throws poor code over the fence into production and makes the user be the beta tester ... that's not exactly new. Anyone ever seen Beehive? When I first saw it, it was a freshly steaming turd. No idea what it's like now, but at the time it was largely broken.

I don't see this so much about NIH as "WTF makes this my problem".

These guys probably believe that if they win they will collect royalties for every copy of tzdata distributed for the next 150 years. Instead, a cleanroom version would be rapidly produced based on primary sources. The initial version may be short on historical data but it would eventually be filled out: all the data has been published by governments. This may be done anyway.

Or, alternatively, they may believe that if they win they can shut down this guy's competing software, and have no interest in pissing off some of these lawyered-up companies by asking them for royalties.

> Swing apps perform badly.
Who is the evangelist? Java2D (=> all of Swing) is fully hardware accelerated (DirextX shaders on Windows and OpenGL shaders everywhere else) since Java 1.6u10 (came out years ago). Boy you are out of date, still clinging to very old notions (since according to the French scientific supercompting outfit INRIA even several years ago Java is faster than C++ and C for their purposes, and is approaching FORTRAN, please read this article and the linked paper http://blogs.oracle.com/jag/entry/current_state_of_java_for). Incidentally, Swing works perfectly for my apps on RDP, but if there is a fault rendering then the fault is RDPs not Swings. Being a pro I most often have to use a flavour of VNC and Swing is fine on this. I run Eclipse on MacOS (after using it for years on Linux, since I've used Linux since kernel 0.20) and it is ok (with some bugs), but the so is Netbeans. Your out-of-date picture of Java/Swing performance is laughable.

Well, you could say that about perl/python/ruby either.

On the other hand, Java applets run in a sandbox defined by security policies.
http://download.oracle.com/javase/tutorial/essential/environment/security.html

That musta been before 2005. Java is *very* fast now. If you read the following (2008) article from Gosling's own blog (ironic?) you see that Java easily beats C++, often beats C and nearly beats FORTRAN for speed. Shame your opinions on Java are so badly out of date as to be mis-informed (like so many on Slashdot). You're missing out on the great performance of Java, fast development times, huge amount of existing libraries, and true cross-platform (Windows, Linux, Mac). Here's the link for more information (and read the linked report by INRIA the French scientific supercomputing folks):
http://blogs.oracle.com/jag/entry/current_state_of_java_for

You haven't named any flaws in JavaScript as a language

This question keeps coming up in every Slashdot story mentioning JS. Frankly, I'm sick and tired of having to re-type it over and over again. I'll just keep linking to this post from now on.

there are a lot of things that JavaScript does really well. Try doing what an asynchronous XMLHttpRequest does in Python, or Java, or pretty much any other language.

What, you've never used an asynchronous API in other languages? I guess then that you don't know that some of them have a single standardized API that is shared beween all async operations (rather than some ad-hoc, slapped-together design for one very narrow use case). And some even have extensive support for async operations in the language itself.

Well, Oracle is of course allowed to stop doing business in the EU if they do not like the rules there.

It might make some people (like you) and even some of their shareholders happy. But allow me to show you some numbers.

Oracle's fiscal 2011 income:
Total income: 35,622 million dollars
Americas (North and South): 18,352 million dollars
Europe, Middle East & Africa: 11,497 million dollars

Realistically most of the EMEA income is from the EU, just like most of the Americas income is from the US.

What you're suggesting is that they dump 32% of their revenue from day to day.

Some of that will be offset by laying off 22,394 employees in the EMEA, but that only makes up 20% of their total number of employees. Compare that to the 45,887 employees in the Americas.

The EMEA is a more profitable area for Oracle than the Americas from a pure income/employee point of view (514,000 vs 400,000 dollars)

But if we ignore the financial consequences, the competitive consequences of giving your main rivals 11 billion dollars a year and the sheer idiocy of believing that you shouldn't have to live up to the rules of the countries you operate in, then yeah - you have a really good idea there.

Go for it - I'm sure you'll have a lot of success at Oracle's next shareholders meeting.

I noticed the same thing. According to http://blogs.oracle.com/java/entry/java_7_questions_answers, the fix is only in "current builds of JDK 7 Update 1". I imagine you can download it from the OpenJDK site somehow, but it does not seem that JDK 7 Update 1 has actually been released yet.

1. C#'s 'using' block so I don't have to use try/finally everywhere.

That's in Java 7, in the form of the try-with-resources statement.

I assume the author meant, "... to improved file system access." See here.

here is the index to the stl by SGI :http://www.sgi.com/tech/stl/stl_index_cat.html
there is the same type of index in java : http://download.oracle.com/javase/7/docs/api/allclasses-noframe.html

in the light of the links I posted, my book analogy was flawed it should have been an English book about the combustion engine and a German one about car subsystems including some chapters on the combustion engine.