Slashdot Mirror

Mc Nealy from Sun has lied us for many months and years.

We want the updated sources of JVM, NOW!!!

Else we would can do many fucked forks!!!

open4free ©

Mc Nealy from Sun has lied us for many months and years.

We want the updated sources of JVM, NOW!!!

Else we would can do many fucked forks!!!

open4free ©

Mc Nealy from Sun has lied us for many months and years.

We want the updated sources of JVM, NOW!!!

Else we would can do many fucked forks!!!

open4free ©

Mc Nealy from Sun has lied us for many months and years.

We want the updated sources of JVM, NOW!!!

Else we would can do many fucked forks!!!

open4free ©

Mc Nealy from Sun has lied us for many months and years.

We want the updated sources of JVM, NOW!!!

Else we would can do many fucked forks!!!

open4free ©

Mc Nealy from Sun has lied us for many months and years.

We want the updated sources of JVM, NOW!!!

Else we would can do many fucked forks!!!

open4free ©

Woo hoo! DLLs! What won't these brilliant folks at Microsoft think of next? Wow! A whole library of executable code! Shared between multiple processes!....

Forgive me if I'm underwhelmed. Can ASP.NET run on one of these or one of these?

So you want fast, and you run on commodity/toy computers? And you're vendor-locked into doing so?

Also seems like everyone is complaining about ASP. ASP and ASP.NET are two completely different beasts. ASP was buggy and a pain in the rear to work with. ASP.NET, however, was amazingly simple to use with an amazing debugger (VS.NET). Please keep on the subject and leave out ASP.

I note with interest that you did not say that ASP.NET is not buggy.... :-)

And before that ChiliSoft ASP - now owned by Sun Systems and refered to Sun Java System Active Server Pages (though it's only classic ASP compatible).

Ask these guys.

You got closer though ;-)

Yes, he got closer. Closer than you that is.

Here it is.

Consumers don't care what language there application is written in as long as it does what is requested.

Well, not to flame but in my experience java sucks, and always has sucked. Forturnately, the marketing people at Sun have convinced them to follow the Solaris versioning and jump from 1.4 to 5.

I guess I have to be more specific with its suckiness. Back when java was newer Netscape on Linux + Java = long lag & loadtime, then crash. This was back when Java was pushed as _the_ GUI app, because you could "Write once, run anywhere". Then the awt toolkit was not as crossplatform as hoped, then came swing, and then people stopped writing GUIs for java (for the most part). Then a little later, I was tasked to install Oracle on an NT box (or some other MS server, don't remember). It was the Java "Universal installer". It randomly hung, and just didn't work. Later I found out that I had to reboot the machine in 16 colors so the java installer would install. Then Sun ships things with their "webinstaller", which is actually the "javainstaller", but the marketing people were on vaction during that naming process. Anyhow, I've seen these guys fail to start if you launched them with the full path (complete with a traceback with linenumbers, etc, w00t!). Of course, it could find the classpaths if I ran it with ./application instead of /full/path/applicaton. Other webinstallers have failed in random places (again with complete tracebacks!). I've had certain versions of matlab that failed to start over remote X sessions because of jvm versions. A student I work with is doing a project in Java and he's still having problems in certain browsers, etc.

These are all off the top of my head.

Again, I'm not flaming, just stating my obvservations. I've programmed and run some java stuff on very small embedded systems like smartcards and iButtons with no real issues, but these were just toy projects that did not go into production.

Now java seems most happy as a middleware language (application services or whatnot) on 3 tier web services, and having competition from .NET.

Also, it has become the marketing puppet for Sun. Java desktop, Java this, java that.

So yeah, as far as _this_ consumer goes, I care, and loathe running anything that is java.

Incorrect. The link in Sun's page does mention the words "mobile" and "wireless" but if you take the time to follow the fscking link you'll notice that Sun calls it "Java 2 Platform, Micro Edition".

Windows

On Windows, Mozilla can be used with Sun's Java Runtime Environment (JRE). It can not be used with the Microsoft Java VM, as that can only be used with Internet Explorer. Also, versions of Sun's JRE older than 1.3.0_01 will not work.

It is possible to install JRE 1.4.2_05 from within Mozilla, using XPInstall technology. This is by far the easiest way to install Java. If you can not use XPInstall, you can download the JRE 1.4.2_05 full installer and use that instead.

Mozilla's handling of XPI files is changing. For more details, see the XPInstall Changes FAQ.

On some systems, JRE 1.4.2 does not work. In this case, you can use JRE 1.4.1_07 until the issue you are having is resolved.

If you are using a zipped build of Mozilla or Mozilla Firefox, you need to add this registry entry. Some users report requiring it even if they have used the Mozilla Firefox installer, so try it first if you have problems with JRE 1.4.2 or later.

Windows

On Windows, Mozilla can be used with Sun's Java Runtime Environment (JRE). It can not be used with the Microsoft Java VM, as that can only be used with Internet Explorer. Also, versions of Sun's JRE older than 1.3.0_01 will not work.

It is possible to install JRE 1.4.2_05 from within Mozilla, using XPInstall technology. This is by far the easiest way to install Java. If you can not use XPInstall, you can download the JRE 1.4.2_05 full installer and use that instead.

Mozilla's handling of XPI files is changing. For more details, see the XPInstall Changes FAQ.

On some systems, JRE 1.4.2 does not work. In this case, you can use JRE 1.4.1_07 until the issue you are having is resolved.

If you are using a zipped build of Mozilla or Mozilla Firefox, you need to add this registry entry. Some users report requiring it even if they have used the Mozilla Firefox installer, so try it first if you have problems with JRE 1.4.2 or later.

Windows

On Windows, Mozilla can be used with Sun's Java Runtime Environment (JRE). It can not be used with the Microsoft Java VM, as that can only be used with Internet Explorer. Also, versions of Sun's JRE older than 1.3.0_01 will not work.

It is possible to install JRE 1.4.2_05 from within Mozilla, using XPInstall technology. This is by far the easiest way to install Java. If you can not use XPInstall, you can download the JRE 1.4.2_05 full installer and use that instead.

Mozilla's handling of XPI files is changing. For more details, see the XPInstall Changes FAQ.

On some systems, JRE 1.4.2 does not work. In this case, you can use JRE 1.4.1_07 until the issue you are having is resolved.

If you are using a zipped build of Mozilla or Mozilla Firefox, you need to add this registry entry. Some users report requiring it even if they have used the Mozilla Firefox installer, so try it first if you have problems with JRE 1.4.2 or later.

J2ME (Java 2 Media Edition) compatable.

[Ironically, the page has "last week's name" for Sun's product, Access Manager. Even groups that Sun founds can't keep up with the continual name changes!]

My big beef with it is the lack of perl and PHP defined APIs. Given the amount of LAMP (along with perl) being used on the web these days, it seems extremely short-sighted not have them defined. Just think, /. and the rest of the OSDN sites could be using Liberty to cross-authenticate rather than requiring each site to do their own auth systems.

No, Sun's multiprocessor architecture is not "slightly NUMA" at all, and is definitely not SMP by definition.

From here, The SunFire memory design has the following characteristics:
Access type, bandwidth, memory request latency (to get load latency, add the relevant transfer latency from the bottom table as well)
CPU local, 9.6GB/s, 180ns
Board local, 6.7GB/s, 193-207ns
Different board, 2.4GB/s, 333-440ns

So, in the 3-4 level memory hierachy in a Sun system, latency is increased by a factor of 3, and bandwidth decreased by a factor of 4.

SGI's Altix systems are a bit harder to classify like this, because they are not a simple crossbar, and have many nodes.

Their node local memory bandwidth is 10.2GB/s, with aggregate interconnect bandwidth per node of 6.4GB/s, node local latency is 145ns, with maximum latency on a 512 CPU system is around 650ns I think (and IIRC the 1024 CPU topology has the same maximum hops distance between nodes).

...is that HP has completely lost sight of its sound roots in the engineering/geek world. HP used to be known as the producer of such geek icons as the HP48 series of calculators, the fantastic old LaserJets (not to be confused with the modern versions) and, of course, the venerable DeskJets. Today, their calculator business is a ghost of its former self, the new calculators are almost uniformly agreed to suck, and their once-vaunted printer business has devolved into the "drug dealer" model of doing business-- hook 'em with cheap printers, then sell them ink at obscene prices. (I remember reading a quote on SlashDot in the recent past saying that ink, ounce for ounce, is worth more than rare old wines now? Or something to that effect...)

Anyhow, HP used to be an engineer's company-- a geek's company. Didn't the Woz used to work there? And he was a geek's geek. Even as recently as my high school education (I'm 25), HP was a touchstone of geek culture.

And now that it's merged with Comcrap, its devolution into yet another mindless "cheap plastic crap computers" business has been completed.

There seem to be only two companies nowadays with solid geek-friendly engineering-- Apple (excepting many of their first-generation products) and IBM (think: ThinkPads... solid engineering and a simple, robust design virtually unchanged in 10 years). HP is now just Compaq wearing a tie. DEC is long gone ("Compaq Tru64 Unix", anyone?), swallowed by the Compaq beast. SGI is going out with a whimper instead of a bang. Sun sold their soul to Redmond and is now producing x86 and x86-64 hardware that are Windows-certified.

And, as usual... no one gives a damn. We're all too damned addicted to ShinyPlasticCrap(TM) to care about the lack of sound engineering.

As far as I'm concerned, Carly Fiorina's head should be on a stake somewhere, the damned sellout. She robbed us all of a good, solid, geeky company in favour of more anticompetitive, mindless, corporate, plastic crap.

I'm sorry. I must not have been clear. Both Darwin 7.4 and OpenDarwin 7.2.1 seem to have (at least some of ) the kernel level functionality needed for BSM, but from neither package can I find any userspace utilities to activate it or read the accounting logs. I'm not so much interested in when it appeared or which versions support it, but rather if it is, or will be usable.

I'm not sure if the problem is that I'm looking for programs with names like bsmconv and auditd and the audit configuration is handled differently in Darwin or that those pieces don't exist yet.

I'm sorry. I must not have been clear. Both Darwin 7.4 and OpenDarwin 7.2.1 seem to have (at least some of ) the kernel level functionality needed for BSM, but from neither package can I find any userspace utilities to activate it or read the accounting logs. I'm not so much interested in when it appeared or which versions support it, but rather if it is, or will be usable.

I'm not sure if the problem is that I'm looking for programs with names like bsmconv and auditd and the audit configuration is handled differently in Darwin or that those pieces don't exist yet.

-- Scott McNeally September 2000 after the dot-con implosion was already in full swing.

He wanted to be able to take the keyboard home with him at weekends to use on his home PC. He was also annoyed that he couldn't fix a web cam onto the Sun servers either.

This sounds far fetched, but in case it's legit:

He ought to be fired for making decisions using those reasons. Instead of basing his decision on careful analyses, the deal breaker is a keyboard?? He is a CIO of a major bank and yet he cannot afford a $100 keyboard?? Also, if one googles for webcam solaris, one may find that there exists webcams for Solaris.

Some day your management will realize that the expensive servers they're buying really aren't worth that much. It shouldn't take that long to realize that you're paying too much for what has now become commodity hardware.

Some day your management will realize that the expensive servers they're buying really aren't worth that much. It shouldn't take that long to realize that you're paying too much for what has now become commodity hardware.

In related news (for real), Sun's COO Jonathon Schwartz has just recently started his blog.

It takes a lot to kill a company, and Sun's not going anywhere anytime soon. They have $7 BILLION in cash in the bank right now, have a strong R&D budget.

They're not going anywhere. Either is McNealy.

And project looking glass looks really awesome for those who haven't seen it, it's a 3D gui that sits on top of Solaris or Linux and adds a lot more functionality. I'm not sure how they will 3.Profit off of it, but it's pretty badass.

It's worth noting that I never said you couldn't. I said it was HARD. If you look farther up the thread, I reiterated this point.

Why is it hard ? Open Visual Studio, make one or more DLL projects, then use these DLLs from the main project according to what you want to do. Making a DLL is nothing more than pressing a few buttons anyway.

10ms res on 2000/XP

Where did you read that? Windows NT provides 1 milisecond resolution. 10 milliseconds is the default timer interrupt granularity. By using the function 'timeBeginPeriod' the resolution can be set to 1 millisecond. Here is an example.

Furthermore, all waitable functions are guarranteed to return at 1-millisecond resolutions (even 'Sleep()'). I have personal experience with this, as I have used the waitable timer functions in many projects.

Finally, Windows have a hidden API ('NtSetTimerResolution') that you can set the timer resolution in nanoseconds. This is kernel stuff, though. And for profiling, you can't beat 'QueryPerformanceCounter' which returns number of clock ticks passed via the RTDSC instruction.

Most Unix systems have a hi-res timer that dives down to the nanosecond range

Which Unix systems are that ? Linux, for example, also starts with a 10 ms granularity, just like Windows (the link above actually saids that any granularity below 10 ms will result in degrading performance on Intel systems). High resolution POSIX timers is a kernel patch, and it is mostly offered in the context or real-time Linux systems.

Solaris also starts with a 10 ms granularity that can be adjusted. Solaris also offers a timer solution based on CPU timer instruction (RTDSC on 80x86).

OS X, not running on the Intel platform, has the benefit of not having the interrupt limitations of Windows, Linux and Solaris on 80x86.

You mean the documents that say "this timer is not actually accurate and is only useful in certain situations", or do you mean the documents that say "the resolution of this timer is dependent on how many processors are in the system".

The "this timer is not actually accurate and is only useful in certain situations" is valid for 'SetTimer' which sets a timer event for a window. This is because timer events are dispatched on the gui message queue.

The "the resolution of this timer is dependent on how many processors are in the system" is valid for all operating systems running on multiprocessor computers. Google it out, if you don't believe me.

Sorry, as far as the industry is concerned, Windows timing blows. Just about every other OS does a better job.

I've shown you that you overreact concerning Windows timers. The 80x86 platform has an archaic interrupt architecture, and that's where the problem is. And this problem concerns all systems, Windows or Unix, running on 80x86. Windows on Alpha have an 1-ms granularity, by default.

There are solutions though to overcome the problems. A nice solution is to use critical sections (that take nano seconds to lock/unlock) and 'QueryPerformanceCounter'). If you search the web, you can even find MFC-based classes for this.

Yes, because everyone though it was stupid for getenv() to be deprecated in the first place. This method didn't even work in 1.4, BTW.

Well, perhaps they've "un"-deprecated it since 1.4.2:

getenv

Autoboxing, String variable format (printf style!), generics -- all are in Java 1.5, now known as Java 5.0. Finally I get to use the old MS vaporware line -- it's coming in the next release!

Autoboxing/Unboxing
This facility eliminates the drudgery of manual conversion between primitive types (such as int) and wrapper types (such as Integer). Refer to JSR 201.

Varargs
This facility eliminates the need for manually boxing up argument lists into an array when invoking methods that accept variable-length argument lists. Refer to JSR 201.

... if you read the article he's actually claiming that the idea of local portable storage will be dead. That everything will be networked and centralised. He makes the point that why would we carry around some fragile copy of the data when we can just have it delivered across the network to whichever device requires it. This is the microsoft vision now, computers+network access in everything.

Am I crazy, or did he just say, "The network is the computer?" Whatever will he think of next?

... if you read the article he's actually claiming that the idea of local portable storage will be dead. That everything will be networked and centralised. He makes the point that why would we carry around some fragile copy of the data when we can just have it delivered across the network to whichever device requires it. This is the microsoft vision now, computers+network access in everything.

Am I crazy, or did he just say, "The network is the computer?" Whatever will he think of next?

Nah, that looks more like a jazzed-up version of Looking Glass.

It's possible to achieve interesting things just by removing a webcam's built-in lens/filter assembly, and replacing them with lenses and filters from 35mm camera. See Lundycam for examples. You can build an extreme telephoto camera in this way for very little money.

You can also change the webcam's behaviour (improving low-light performance, for example) in software by using something like the Java Media Framework.

Part of the balance could be redressed in many interesting ways through computers reading to humans.

What is the state of that art?

Some groups have spent millions to advance text-to-speach technology.

Others have made a free open source text-to-speach engine.

synthesizer.allocate();
synthesizer.resume();
synthesizer.speakPlainText("Hello, world!", null);
synthesizer.waitEngineState(Synthesizer.QUEUE_EMPT Y);
synthesizer.deallocate();

No GUI. (yet)

These functions can get as complex as you'd like, since DTrace is driven by the 'D' programming language.

Here's a few good examples showing the use of speculations (only tell me when a syscall *doesn't* work, not when it does) and aggregations (summarize the data, so you don't have to).

http://blogs.sun.com/roller/page/ahl/20040701#dtra ce_for_developers

This sort of built in functionality is what separate DTrace from tools like truss, not to mention the huge number of OS hooks those other tools can't touch.

Looking at that Apple blurb (and reports I heard from WWDC attendees), Shark is a performance tweaking tool - a profiler. If you had ever bought the heavy-duty Sun compilers, you'd have a functionally similar set of tools (collect, analyzer, and cc with various flags that generates annotated source). Admittedly not nearly as polished as Apple's tools (and not as cheap), but Shark is also several years newer - and having done exactly the optimization that article describes on my own code with the Sun tools, I assure you they have been around for quite a while.

However, DTrace is NOT an application profiler - though it can be used for that. And claiming that it is a profiler would be like claiming C is just an extension to assembly language. It could be used as such, but you'd be ignoring all the power of the tool.

DTrace is far more powerful: it is a whole system profiling tool. It can give debugging data, performance data, or whatever, for any process, thread, driver, kernel module, or other system component. It includes a moderately complex programming language - YOU get to write code that runs in REAL TIME within any code path anywhere in the system. And, you get zero performance loss in any part of the system that isn't actively being traced.

The DTrace manual is extremely informative. Of course, actually trying it on Solaris 10 Beta is even more so :-)

This paragraph is copied from the Related Work section.

Kerninst is a dynamic instrumentation framework that is designed for use on commodity operating system kernels[13]. Kerninst achieves zero probe effect when disabled, and allows instrumentation of virtually any text in the kernel. However, Kerninst is highly aggressive in its instrumentation; users can erroneously induce a fatal error by accidentally instrumenting routines that are not actually safe to instrument.3 Kerninst allows for some coalesence of data, but data may not be aggregated based on arbitrary tuples. Kerninst has some predicate support, but it does not allow for arbitrary predicates and has no support for arbitrary actions.

Sigh, so now the PHB will be running this and telling us what to fix.

I think your point was that wide availability of a tool will increase the number of people using it and thereby increase sales of (in this case) Sun products. While that is true to an extent I think you're hoping for a bit much for Sun to port it to Linux. Sun are a business. They have an OS and boxen to sell. They are generally a community-friendly company, sure, but if dtrace is as useful as they say then I can see them wanting to keep it proprietary and use it to draw people in to buy their OS and their boxen.

I'm not even sure how well other OSes could use dtrace, at least without a fair bit of work. It sounds to me like it's pretty tightly integrated with the OS (for obvious reasons).

And anyway, you can use it. You can download Solaris for x86 for free and get used to the tools that way. So basically they are doing what you want, but with their entire OS instead of just this one tool.

I'm so nice I'll even show you where you can get it. Right here.

Dtrace Usenix Paper

More examples in answerbook from the BigAdmin Dtrace page

Dtrace Usenix Paper

More examples in answerbook from the BigAdmin Dtrace page

Dprobes is discussed (briefly) in the announcement paper. (PDF - sorry.)

Well you CAN get Solaris 10 preview from the Express program. It's free for non-commercial use and available for x86. Downloads available here. Check out the hardware compatibility list for x86 here (don't see one specifically for Solaris 10 Express, probably the Solaris 9 list applies).

Well you CAN get Solaris 10 preview from the Express program. It's free for non-commercial use and available for x86. Downloads available here. Check out the hardware compatibility list for x86 here (don't see one specifically for Solaris 10 Express, probably the Solaris 9 list applies).

Dynamic Instrumentation of Production Systems - this paper was presented at Usenix, and describes how Dtrace is actually implemented.

Dtrace user guide.

A collection of Dtrace scripts

Dynamic Instrumentation of Production Systems - this paper was presented at Usenix, and describes how Dtrace is actually implemented.

Dtrace user guide.

A collection of Dtrace scripts

> Where are the free software projects investigating
> next generation UI concepts?

Check out Sun's Project Looking Glass, a 3d desktop built on X11.