The latest Sun Microsystems Java 1.4.2 release update is out, version 1.4.2_03. The only change in this release deals with a CA certificates update in cacerts file.
The latest Sun Microsystems Java 1.4.2 release update is out, version 1.4.2_03. The only change in this release deals with a CA certificates update in cacerts file.
Java has massive support and a huge library with something like 5,000 classes and 40,000 methods. Not sure of the exact size but it’s somewhere in this neighborhood. It’s also actively developed by a staff of hundreds of developers, and it runs on Linux, not sure if Java3d is for Linux though, but if people ask for it, it will be. It’s just a no hastle Linux solution, but Python is also great, just that Java has a bigger support structure.
Why does everyone get censored here?
Because we were about to start a quite heavy Python vs. Java flamewar I would assume. Correct catch imho, were mostly just inflammatory.
It is a tad hard to comment on a minor patch release which only changes a security certificate though. Count congratulate Sun on the 1.4 series overall though, very good work, first minor release I consider really mature (this is from a guy who think that no complex software younger than 10 years can ever be even remotely mature so my standards are fairly strict
The two things I need must from the upcoming java 1.5 are autoboxing and a shared VM engine, mustly the later, decreasing the loading time of java apps as well as redusing the duplication of resources usage will give a good “feel” concerning speed to the users. Another thing that would be welcome would be a new face for the default swing Metal theme. Something alone the lines of the JDS theme, but lighter and more polished (of course, something better would be nicer). I think java needs badly a shared VM to keep on pair with the .NET CRT
…and a more stable and more complete sound API, with at least one compressed audio file format.
More than anything else…multhreading when handling Swing events…secured threading. It is a pain in the neck that you cannot call on a method on a label (i.e. changing color) while an event started from that label is still active.
despite being a few releases behind regular Python and a few occaisional limitations that mean that some rarely used aspects of Python can’t be used in Jython, Jython seems to solve most of the conflict between the two languages. Personally i’m a fan of both lanugages for different. for Quick and Dirty work i perfer python/jython but i find that or larger projects i NEED the static typing of java to keep everything clean and consistent
Generics will be a most desireable feature, as well as the mentioned shared VM
I hate to disappoint, but Sun has already commented that the Shared VM feature would not make it into 1.5. 🙁
http://java.sun.com/developer/community/chat/JavaLive/2003/jl0729.h…
dynamite: Is virtual machine sharing going to make it into Tiger?
Neal Gafter: Unfortunately, no, because of schedule constraints. Even if it were included, it would be a trivial implementation that doesn’t actually share, but rather forks off a new VM per isolate.
great for GUI work for Java. why not use that?
Hi,
I want know if this java virtual machine wirks weel in linux kernel-2.6-testX.
Now I am running this kernel and eclipse and don’t have any problems. Can I put it in production systems.
“…and a more stable and more complete sound API, with at least one compressed audio file format.”
Ogg Vorbis would be nice.
“I want know if this java virtual machine wirks weel in linux kernel-2.6-testX.
Now I am running this kernel and eclipse and don’t have any problems. Can I put it in production systems.”
Are you sure you want to put a 2.6-testX into production? Doesn’t it feel slightly premature, somewhat like switching to a new version of Windows within the first year of its release (granted, it’s always premature to use Windows).
Regarding usig 2.6 at this stage within a production, you must be kinding me. Not that Linus, et al. haven’t hack this puppy big time, but nothing compares to having a finish version out for everyone to take a stab at. Recommendation: use the test kernel in a test machine and put it through its paces. Let it on for at least the next several weeks and write or get a tool that will bombard it with requests.
> It would be a trivial implementation that doesn’t actually
> share, but rather forks off a new VM per isolate
Actually I had never thought of doing it that way! Always thought of using a shared memory region/mmap for a set of jar libraries, etc, but simply calling clone()/fork() on a parent JVM process /would/ be IMO the simplest. However, that would mean that the whole base jar libraries would need to be loaded FIRST; all of them, not just some on demand thus slowing down the initial first load. Otherwise the fork()/clone() copy-on-write to that inherently shared area would make it a moot workaround.
Come to think of it, there’s no reason I couldn’t try this myself with a custom C launcher/JNI…
It’s a pity 1.5 won’t have VM sharing. So the closest thing we have to that is libgcj. Does anyone knows if Kaffe or any other VM (other than OSX’s) has/plans this feature?
.03 revision.
But if all the processes share the same VM, wouldn’t it basically take down all of the processes if there were any kind of memory corruption in the VM?
That would be a good way to tune the VM and find rare bugs. In today’s VMs, such a memory corruption are sometimes misinterpreted as a problem of the aplication, thus remaining hidden and unsolved. With time, these kind of bugs would be better solved with a shared design.
can’t find a change-doc on the java.sun.com site. Where do yall see the .03 revision even listed?
this version has been in portage for a quite a while.
How has it been in portage for a while? When I updated my JVM on Friday (directly from Sun’s site) it was still 1.4.2_02. Is there some mystical place where you can download sun’s JVM prior to them putting it on thier own site?