Slashdot Mirror

While it may save you from a catastrophic drive failure, despite popular belief RAID generally cannot detect or repair corruption*... you really need ZFS (part of Solaris 10 and being integrated into OS X).

* - the short answer: because it reads only one side of a mirror and does not checksum. Drives and controllers do not reliably report errors and nor can RAID do anything about hardware issues such as RAM, cable, controller, or firmware bugs. (ZFS has other features further improving integrity, such as COW.)

I saw a demo at a Sun conference earlier this year that proposed something similar (Java, not X11 based though...)

http://research.sun.com/projects/mc/mpk20.html

Cheers,
JG

I found the device, it was called nucleostop and it filtered Nuclear generated electrons based on Tachyon signature ;-) Here is a Sun blogger's English translation of this beautiful snake-oil product for the anti nuke industry.

Horse sense is what a horse has that keeps him from betting on people. -- WC Fields

I suggested that everyone on the ITU committee should be asked to read David Ewing Duncan's book "Calendar - Humanity's Epic Struggle to Determine a True and Accurate Year." Ponder the fact that it has taken thousands of years of struggles, scientific advancement and setbacks to get human time synchronized with astronomical time. Great rifts developed in societies and wars were fought over the accurate calculation of time. (Check out the Irish/Roman/Orthodox rift over the calculation of Easter). Now with a single vote, the ITU can undo thousands of years of human progress just to avoid mini "y2k errors." Why not fix the code?

http://sunsolve.sun.com/show.do?target=patches/patch-access
http://sunsolve.sun.com/show.do?target=patches/zos-s8
Solaris 8 was introduced in 2000.

I am almost certain that IBM, HP, etc. still support 6 years old OS's.

http://sunsolve.sun.com/show.do?target=patches/patch-access
http://sunsolve.sun.com/show.do?target=patches/zos-s8
Solaris 8 was introduced in 2000.

I am almost certain that IBM, HP, etc. still support 6 years old OS's.

I agree with you about languages like Python being part of the solution. You could pick Ruby too, or any number of other languages, but the basic idea I agree with. Incidentally, in my experience, Python is sort of becoming some sort of quasi-standard for math programming outside of more specialized languages (e.g., SAGE, written by one of the authors of the piece, is scripted in Python).

Having said all that, however, there's a larger problem with math programming that's not being recognized in the opinion piece: there is currently no good language for doing math in.

I do statistics research, and I am repeatedly frustrated by the lack of truly good languages to do programming in. You either have to write in C/C++ or Fortran, which are horribly outdated or overly baroque for math programming, or in languages like R/MATLAB/Maxima, which are elegant for math programming, but are unnecessarily slow.

What's happened is that the difficulty of programming in languages like Fortran or C has driven people, aware or unaware, to closed-source languages like MATLAB, which are much easier to write in.

I don't think much will change in levels of use of open languages for programming until there is a fast, relatively low-level, modern powerful language that is targeted toward math/numerical programming in particular, and other applications secondary to that.

I think something like Fortress might provide that sort of solution, but it's a little ways off yet.

This is not to say at all that things like R, MATLAB, Maxima, Mathematica, etc. aren't useful, just that I think a certain number of closed systems resulted from dissatisfaction with the existing open languages, that were becoming too cumbersome to use for numerical programming.

Python calculated exactly what its documentation says it will do: ((1 minus the IEEE-754 double closest to 1/100) rounded to the nearest IEEE-754 double). It's not Python's fault if you don't know the basics of floating-point arithmetic. Mathematicians who use or write numerical software do.

I recommend reading What Every Computer Scientist Should Know About Floating-Point Arithmetic.

There's no showdown. Sun supports Android and has said so publicly, just read Sun's CEO's response. Personally I thought the author of the story was an idiot, he could have made his article so much more interesting and just as likely and accurate!

At least, that's what Jonathan Schwarz says in his weblog post on the topic.

I'm going to go with 0, since Java is open-source and free. Somebody correct me if I'm wrong on this point.

The companies listed below have licensed Java Platform, Micro Edition (Java ME) configurations and profiles and the associated Technology Compatibility Kits (TCK). Only Java ME technology licensees can claim compatibility with Java ME technology specifications and TCKs.

The list is quite long and of course includes the usual suspects.

RT
--
Your Bookmarks. Anywhere. Anytime.

The current setup that Sun provides has boatloads of backwards compatibility cruft and old API's like AWT

RT
--
Your Bookmarks. Anywhere. Anytime.

There's a reason Java ME has gone nowhere

What about this list of Networks Operatos and Carriers

Or the Java ME Device Table?

Or, for that matter, what about these phones from Nokia, Motorola and Sony Ericcson just to name a few?

Google is trying to succeed where Java has failed

RT
--
Your Bookmarks. Anywhere. Anytime.

There's a reason Java ME has gone nowhere

What about this list of Networks Operatos and Carriers

Or the Java ME Device Table?

Or, for that matter, what about these phones from Nokia, Motorola and Sony Ericcson just to name a few?

Google is trying to succeed where Java has failed

RT
--
Your Bookmarks. Anywhere. Anytime.

"Oracle is going after its piece of the hot virtualization market by introducing an open source Xen-based hypervisor to compete against those from Novell, Red Hat, and VMware.

I'm convinced that transactional memory (TM) is the right model for concurrent programming in most common situations but software transactional memory (STM) suffers from some performance problems. There has been a great deal of research on hardware support for transactional memory and recently Sun announced hardware support for hybrid transactional memory in Rock. Do you think hardware support for TM will catch on and migrate down to commodity hardware? If so would this translate into significant performance improvements for databases like MySQL? If not do you think specific hardware optimizations for DB applications will ever make sense in the mainstream market?

Keep in mind that the Sun lawyers blogged about it as well.

http://blogs.sun.com/dillon/entry/the_netapp_litigation

I've got something like this for Solaris/OpenSolaris users at:
http://blogs.sun.com/timf/entry/zfs_automatic_for_the_people

(pretty screenshots on the blog post for the previous version
http://blogs.sun.com/timf/entry/zfs_automatic_backup_0_1 )

  -- plug a USB disk into your ZFS-based OpenSolaris box, and any filesystems you marked as being interesting enough for backups will get sent to the USB disk. If previous snapshot streams are already on the USB disk, then we do an incremental backup. GUI notification via notify-send(1) Works for me!

This still work, specifically there's no restore GUI right now, and the configuration is a bit clunky, but still...

I've got something like this for Solaris/OpenSolaris users at:
http://blogs.sun.com/timf/entry/zfs_automatic_for_the_people

(pretty screenshots on the blog post for the previous version
http://blogs.sun.com/timf/entry/zfs_automatic_backup_0_1 )

  -- plug a USB disk into your ZFS-based OpenSolaris box, and any filesystems you marked as being interesting enough for backups will get sent to the USB disk. If previous snapshot streams are already on the USB disk, then we do an incremental backup. GUI notification via notify-send(1) Works for me!

This still work, specifically there's no restore GUI right now, and the configuration is a bit clunky, but still...

There is no good reason not to use Java for real time systems as well. The very first JSR, JSR#1 proceeds to specify what is required for real time Java. Later, it has been enhanced in JSR#282, which is implemented in The Sun Java Real-Time System 2.0 (Java RTS).

Sun's Java has been available in binary form for as long I've been using Linux (I started using Linux in 1994).

Quite an amazing feat since Java didn't come out until 1995. When I first started working with Java (1996) there was no official Sun Java under Linux, there was a port from BlackDown (their web site seems to have gone AWOL recently).

I can't remember exactly when Sun added Linux as an officially supported platform, however I know it wasn't from the get go. As a matter of fact, the number one bug in Sun's Java Bug Parade for a long time was a petition to officially support Linux.

My problem with the Sun JRE is that it is HUGE. Why do I need 100MB+ to run a simple Java application?

you don't. a simple stroll over to java.sun.com will show you that the JRE is 14M for windows and 18M for linux.

the "100M+" is if you're also downloading all their development tools and documentation (and possibly netbeans, depending on the link). not atypical in the least.

because i can't find references on the sun & openjdk site.

Mark Reinholt (chief engineer for the java platform) blogs about it. Also the experimental Mercurial repositories are open.

Slightly offtopic, so are the JavaOne 2008 Call for Papers.

Sun used the product from the patent and created a Free version without permission. That makes them a great 'Robin Hood', but it also makes them the 'bad guy' in the eyes of the law.

You're stating as a matter of fact that Sun "used the product from the patent". This is stretching the truth somewhat. The actual facts of the matter are that NetApp claims Sun have violated their patent (WAFL, etc), and filed suit requesting relief. Sun however disagree and believe they do not violate NetApps' patents - indeed Sun claim, in their counter-suits, that NetApp are violating Suns' patents. However, no-one is violating anyone's patents until either both parties agree they are, or a judge says so.

You can read Suns' response to NetApps' complaint (which #include's most, if not all, of NetApps' complaint).

NB: I am a Sun employee. I have tried to keep the above post be 100% fact-based and opinion-free, but I am obviously biased, I also may be wrong and finally IANAL. Lector emptor.

It's not that Sun is depreciating JavaME. But that JavaME and JavaSE will merge. More detail on this from James Golsing's blog entry.

To provide a more general answer to your question, check out Sun Blueprints' Enterprise Data Center Design and Methodology. It's by no means complete, but it's a great starting point. We used it when we planned the deployment of a data center in Mali last year.

What you propose has already been implemented: the Common DOM API for Java applets.

Sun has new language for client-side scripts.. Just-so-coincidentally, Sun has had a variation of this new "language" available in your browser for a decade! It's called applets and the JRE.

Python, Ruby, JavaScript, Groovy. Whatever. Sun has a Java runtime in the browser, a sandboxed one that can only access the DOM and browser. But you still get all the Java benefits, like multithreading and bytecode compilation. And all the Java benefits, like it's implemented for IE AND Firefox on Mac, Windows, Linux and Solaris. Further, it is available under the GPL, so you can port it to any other platform.

See this web page for details of Sun's leaner faster in-browser JVM.

Sun has new language for client-side scripts.. Just-so-coincidentally, Sun has had a variation of this new "language" available in your browser for a decade! It's called applets and the JRE.

Python, Ruby, JavaScript, Groovy. Whatever. Sun has a Java runtime in the browser, a sandboxed one that can only access the DOM and browser. But you still get all the Java benefits, like multithreading and bytecode compilation. And all the Java benefits, like it's implemented for IE AND Firefox on Mac, Windows, Linux and Solaris. Further, it is available under the GPL, so you can port it to any other platform.

See this web page for details of Sun's leaner faster in-browser JVM.

They're working on it.

The crucial factor is having faith in the technology and convincing backend Java EE experts to endorse a rich client interface over markup based solutions such as struts, JSF or whatever 'framework of the month' they read about on TheServerSide.

It can host a variety of scripting languages such as Python, Ruby and, surprise, even JavaScript, as well as a couple home-grown languages such as Groovy and the purpose built JavaFX Script

Now before you shriek in horror at the thought of a JVM running in a web browser, Sun have made a renewed commitment to the browser via the soon to be released Consumer JRE, which aims to relieve some of the bloat and provide an improved experience.

Still no official 64 bit browser plugin but the IcedTea folks are working on a substitute.

It can host a variety of scripting languages such as Python, Ruby and, surprise, even JavaScript, as well as a couple home-grown languages such as Groovy and the purpose built JavaFX Script

Now before you shriek in horror at the thought of a JVM running in a web browser, Sun have made a renewed commitment to the browser via the soon to be released Consumer JRE, which aims to relieve some of the bloat and provide an improved experience.

Still no official 64 bit browser plugin but the IcedTea folks are working on a substitute.

You mean these reasons?

Honestly, I don't know why people care that an environment supports Java at all. Or C++ for that matter. It's completely irrelevant. High level langauges add NOTHING to computing, and absolutely NOTHING that would be important on a computer.

Really, I don't get it. I mean, they should all be coding in machine language anyway. They aren't missing anything, just some dumb libraries.

• JTabbedPane
• JTable
• Unicode normalization
• Heap memory access
• Array reallocation
• Reflection
• More support for IEEE 754 Floating Point
• Dynamic compilation with access to the compiler from within code

• JTabbedPane
• JTable
• Unicode normalization
• Heap memory access
• Array reallocation
• Reflection
• More support for IEEE 754 Floating Point
• Dynamic compilation with access to the compiler from within code

As I'm sure you're already aware, OEM pricing is very different from full retail, plus it's not like he's going to need Office Professional (the one that would cost $460) just to get Powerpoint. Looking at MS's product matrix, he could get away with "Home & Student" and still get Powerpoint - that can be had from Newegg for $129, and includes Word, Office, Powerpoint, and OneNote. Compare to StarOffice, which is $70 - yes, you get most of the same functionality, but it's still not quite the same; the point, however, is that this is less than a third of the price you quoted.

Windows licensing, meanwhile, is not $210 for OEM licensing. A NewEgg search reveals that you can get OEM licensing in packs of three for roughly $410; that works out to under $140 per license. Obviously, mass-manufacturers of PCs get much more favorable licensing pricing than that, but, for the sake of argument, we'll say that the customer is paying $140. This is still $70 less than the number you pulled out.

So, at this point, we've spent no more than $270 in software. Is this $270 you don't have to spend if you get the WalMart Linux PC? Of course, but if the WalMart Linux PC doesn't fit your needs, $270 is a reasonable number, and certainly much more reasonable than the hyperbole-screaming $2500 you came up with on a whim.

DISCLAIMER: I run Ubuntu Linux on everything I have because it meets my needs and does so at a price point that I am quite content with (free!). I do think that, as far as Linux distros go, it's easily the most user-friendly one that I've ever run across, and would happily recommend it to anyone that has some basic technical acumen. That said, I do not run Ubuntu because of it's philosophy, nor do I do it because of any particular dislike of Microsoft's "monopoly practices". From where I'm sitting, Microsoft did precisely what Ubuntu is doing now - they offered a lower priced (compared to the competition of the time), mostly fully featured set of applications that met the needs of a vast majority of people. Think back to the late '80s - if you wanted a GUI, the only way it was going to happen was if you bought new hardware that was incompatible with your existing IBM hardware or if you paid through the nose for OS/2... until Windows 3.0 came out. Need a server operating system? No problem - your choices were Unix (required expensive hardware, had severe vendor lock-in at the time, licensing was atrociously expensive), Netware (a little better on all counts, but still pricey), or Windows NT (same interface as all your workstations and a little cheaper). If you're a 10 person operation, guess which one you're picking? Hey, it's 1994 and you need a small database. Microsoft Access costs $100. How much does everything else cost? Oh... I see. Access it is! How about an Internet browser? Remember when those weren't free-as-in-beer? How did they get free? That's right - Internet Explorer. Were any of those products perfect? Heck no. All of their products were functionally inferior to the competition - but they met the needs of 99% of the world and cost less than their competition. Sound familiar? What people seem to forget in their haste to hate Microsoft is that, for better or worse, they were better behaved than their competition of the time. Now, their time is passing, and look who's sneaking up on them...

Blame Sun for this.

Sounds like a populist position, or maybe troll flamebait. I'll be generous and assume the former, despite the fact your post seems like a digest from an anti-ODF briefing paper. Disclosure: My job includes the task of receiving complaints about Sun and trying to get Sun to fix whatever causes the problem. If you have proof of any of your accusations, let me know. I may have some of my facts wrong below as I'm working from memory; I'd welcome correction.

With a few small additions, ODF could have supported Office formats as well, but Sun would not allow this.

That is indeed the constant assertion that the three guys who comprise the Foundation make. However, I have personally asked members of the ODF working group at OASIS and they tell me its not so.

• The Foundation guys wanted to add structures to ODF to preserve untranslateable tags in translated documents so they could be regenerated on the reverse translation. Sounds OK at first glance, but in practice it results in very brittle software solutions that work well in demos but not in real life.
• The proposal was thus rejected by the whole working group (not just the Sun employees).
• Rejected, that is, in conversation. A complete solution was never proposed for voting.
• To say Sun would not allow it ignores the actual dynamic of the working group (see below).

Their policy is that ODF will support what is needed for StarOffice, and nothing more.

Naturally every member of a standards group in the traditional standards process is looking out for the code base where they implement a standard, and will have serious questions of any feature that they regard as unimplementable. The features actually put to a vote by the guys from the Foundation would have resulted in very brittle implementations, highly dependent on the version of MS Office with which they were coupled. It may have been possible to come up with a solution that reduced this problem, but the discussion was not sustained. The assertion you make is not true in the general case.

They control the ODF technical committee

Untrue. The ODF TC can have no more than three members from any one organisation and is not under the control of any organisation. The Foundation guys actually flaunted that rule at one point and sent many, many more representatives - OASIS had to step in to fix it. That intervention is one of the issues they have with OASIS, in fact. Sun happens to employ the people who act as Chair and Secretary to the TC but the voting remains democratic.

and their patent license allows them to stop the ODF TC if the ODF TC goes in a direction Sun does not like.

I've heard that interpretation of the patent non-assert covenant that Sun has made regarding ODF, but it's untrue. Sun covenants not to enforce any patents against ODF implementations based on any spec it participates in. To the extent that versions of the spec after Sun's departure are based on version in which Sun was involved, that covenant remains in effect even in the unlikely event of Sun leaving the TC. Sun can't stop the TC from continuing its work.

Are you relaying this all as hearsay, or do you actually have data to back up your accusations? If you have, I'd like to see it (genuinely).

So, it's Windows and Linux for me as the only solid choices.

dont forget solaris.

its not for everyone, but if you want a java friendly environment, you want solaris!

they even ship out free multi-platform media if you're prepared to wait a couple weeks.

or just click through here...

Others are set to join him.

Almost 12 months since Java 6 was released on other platforms. Still waiting, Steve.

Ya, because actually addressing the comments brought up by the ISO and resolving them is Evil!!! It may actually get approved! And who needs more than ONE standard?

Right, because a non-MS employee has seen the source code, and MS actively HELPING Mono so that they can sue them later.
I can only imagine the hysteria here when .Net 3.5 comes with the source code.. I guess its not just MS that likes to FUD..

This is not a NetBeans Problem. It is a JDK 6.0 update 2 and update 3 problem. It takes JFileChooser forever to show up if you have zip files on your desktop or in the directory you are trying to navigate to.
JDK 6.0 update 1 and earlier versions don't have this problem.
here's the bug report for this:
http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6578753

I ran across those HOWTO pages. They are not good documentation, they are just are a reasonably categorized collection of quick HOW-TOs aimed at command-line monkeys: "type this in if you want to do x".

That site doesn't document in detail how every option of every LVM tool works. Yeah, there are some conceptual basics on how VGs relate to PVs and LVs (which is obvious to anybody who reads the MAN pages or understands sotarge virtualization on any platform), but the exhaustive details of LMV2 simply aren't documented there, or anywhere else I could find.

I decided your linked HOWTO page was not a good documentation resource to use when I kept running across pages like this. There are no explanations as to what any of those commands mean, nor why they might need to be different in some use cases. The majority of that HOWTO is similarly constructed.

Where is the comprehensive reference that documents every option of LVM2? Why is there only one nearly incomprehensible page about disaster recovery of an LVM system, which should be the most important topic? I didn't find any true documentation back in 2005, and I still haven't run across it. Compare that to Sun's documentation for ZFS, and you'll see why I think LVM documentation sucks.

I know C, and even wrote a simple C compiler as an undergrad back in the early 1990s. But I don't code C regularly anymore, and I am certainly not going to try to decipher C source as documentation. I don't have time, and neither do most other IT folks. Professionals do not consider source code documentation. My developers write design and architecture documentation for all of their code, which is then passed on to technical writers so it becomes actual documentation for end users and system administrators. It seems in most open source projects, coders write the documentation themselves, which leads to documentation being perpetually incomplete. Writing documentation is not what coders are good at, nor is it what they enjoy doing.

Our servers are plenty fast, thank you very much.

Our servers are plenty fast, thank you very much.