Slashdot Mirror
Sun Open Sources Java Under GPL
prostoalex writes "The embargo is off, and Associated Press is reporting on Sun releasing Java under GPL. Sun is hoping that this step will attract more developers, as well as extend the lifespan of Java. The article notes that this is 'one of the largest additions of computer code to the open-source community', and that Java is currently being run on something like 3.8 Billion devices worldwide." From the article: "Rich Green, Sun's executive vice president of software, said the company hopes to turn more developers into Java programmers, who may then create additional software to support Sun products. 'The open-sourcing of this really means more — more richness of offerings, more capability, more applications that consumers will get to use,' Green said. 'The platform itself will become a place for innovation.' All the Java source code is expected to be released by March 2007, Green said. The move covers all Java technology, which includes software that runs on handheld devices, personal computers and servers."
While I'm psyched about the fact that they went whole-hog and went GPL, there are a number of commercial Java Applications out there. The fact that they have this "Classpath Exemption" makes it sound more like the LGPL.But at any rate, it good that Java is now suitable for Linux.
Will they use the GPL for all the components? That is the compiler, the class libraries and the runtime? Then it will still be easier licensewise to use mono.
Well, on a more practical note, this means that within a few months, I should be seeing a real, complete, working JRE sitting in the main repositories for Debian and Ubuntu. Sweet. We no longer have to go and fetch it ourselves or experiment with incomplete toolkits.
For the ideologues, knowing that there's one less piece of non-free software on your system is a real comfort. For me, personally, all that apparently remains are ATI drivers and Flash Player.
Yay!
~ C.
I don't think the Mono folks are rejoicing ;)
With this step, SUN has became the largest commercial contributor to the free and open source software pool. OpenOffice.org, OpenSolaris, now JAVA - well, kudos!
to post a link to a forked java
This will remove, at a stroke, one of the largest issues of licence trouble in the *nix world. Fully Free server offerings can now be broader, and the Free desktop gets access to one of the three major ways of bundling sophisticated apps into web pages(Not that I like the practice much; but if Flash, Java, and ActiveX have to exist, I'd rather that more be Freely supported than less).
I'm really impressed with Sun on this one.
Come on, you dont do the same kind of applications with a script language like Python and a lower level, general purpose language such as C, C++ or Java.
Script languages are good for small applications but when it comes to big and/or commercial software you need something else. Java is probably one of the most used language (with C and C++) in commercial applications. So yes it is important.
I work for a LARGE government contractor and have had a hard time getting management to realize Open Source in general and the GPL in particular aren't bad for business. Open source here has the same connotation as red communism. Can't get many of them to stop calling it freeware. With Sun making Java GPL they won't have the choice of sticking with that attitude anymore. Many of our existing projects use Java already!
"God fights on the side with the best artillery." - Napoleon, Marshal of France - speaking truth to power
Some more articles I have found, with some substance to them:
InfoQ, also mentions Glassfish.
eWeek.
There is also going to be a official webcast about this by Jonathan Schwartz and Rich Green 9.30 a.m. PT.
In related news, apparently Project Looking Glass, the 3d desktop, is likely to be included in the Ubuntu Feisty release.
Being bitter is drinking poison and hoping someone else will die
Now Stallman can drink coffee again
//WR
Linux Devices has a fairly comprehensive list of just what technologies 'GPL Java' encompasses. The std libraries are GPL with the classpath extension.
I love how easily you just shrug off that number.
Java is an embeddable language at the hardware level. So-called "Java chips" provide a very real platform for deployment on handheld devices. It's easier to develop applets in Java than in, say, C, since Java is higher-level. The fact that there is low-level support for such a high-level language makes it popular with cell phone developers.
Your point of Python is a good one. After all, Python is high-level, intelligent, and permits object-oriented development. It's my favorite, and I'm writing a few programs in it right now. However, Java still has a few advantages. First, it's ported more places, the most obvious off the top of my head being that Java's Mac OS X GUIs are far more robust and less buggy than Python's. Second, it's compiled and then byte-interpreted, giving it a fairly good speed compared to Python's interpretation. Python also has structures that, while easier to read, definitely don't execute as fast. (I do concede, however, that Java is no speed demon.)
Also, Java is embeddable as a web applet. Only a few other languages can do that. You can't exactly drag'n'drop a Python application into a web browser, hook it up to a frame, and project it to the world.
Of course, since this is Slashdot, I'll finish up with a low-blow bit of rhetoric. If Python is superior to Java, then why is the leading Bittorrent client, Azureus, written in Java if the original Bittorrent client was written in Python?
~ C.
http://news.zdnet.com/2100-3513_22-6134584.html?ta g=zdfd.newsfeed
Comment removed based on user account deletion
Java is an easy to learn, fully cross platform (and by fully, i don't mean pseduofully like most languages, if you code a GUI in Java, you don't need to install stuff like GTK+ to make it work in windows / linux), with a 'one class file fits all' attitude. You don't need to spend time making sure that your code is compatible with every OS that java runs on, you also don't have to waste time compiling for each system, It's fast, complete and as i say, easy to use. I'm primarily a C++ programmer, but my current occupation requires that I code java most of the time, and to be honest, i don't mind, it's much less of a problem making sure everything doesn't segfault, and with a true OO nature, it's just a joy to use. The only real quibble i have is with the way the Java VM behaves and handles memory. As far as open sourcing goes, i welcome it, although i doubt it's going to effect anyone really at least for years to come, the only real advantage is that it may (as stated) be shipped 'as default' in some linux distros, which is a good start
I'm wondering what is going to happen with free Java compilers/virtual machines like GCJ and Kaffe?
Well, it is not so simple, Java 10 years ago was nothing in comparation with Java today (I wouldn't actually call it usefull). Also, there was all that MS J++ fiasco, then a patent ligitation and deal MS and Sun had two years ago....lots of stuff happened, and it is really hard to tell how would things played out if Sun OSed Java 10 (or even 5) years ago.
I personally think timing is great. Java6 is a great piece of software, lets start from that point and see if we can make it better.
Remember the "Ice Weasel" nonsense? That was because the Debian team did not want to use a copyrighted Firefox logo. Isn't Sun's Java going to have the same problem?
I'm very happy that Sun is going to bring Java -platform to Open Source and under GPL license.
I believe that this move by Sun will increase development speed of Java and more importantly it will make Java more common. By having Java as a GPL, there is no reason left why Linux distributions wouldn't include Java in default installation. By becoming a standard part of Linux installations, it will gather more mind share from developers, both open source and commercial, and make developing of Java based applications more lucrative. It's interesting to see if this move will bring Java over a tipping point in desktop environments and lead to a situation where more and more applications are based on Java, leading to a total commodization of desktop.
When Java will become more standard part of a desktop, I believe that it will change deeply on how we build our applications. I think that the future is for applications that have desktop application component that is integrated to a server application. Already it's quite easy with Java to develop server software that works via multiple interfaces, ie. web and desktop. The only question for now has been, do all clients have Java, maybe in the future there is no need to ask this question.
Survey research tool for commercial and scientific use
... welcome our new Duke overlord.
I have quoted two soundbites on my blog from videos that will be shown at the announcement tomorrow (the quotes are from the information that was (I believe) given to the press.
I've seen the video shorts (well some of them) that will be shown at the announcement. I think some folks will be surprised. RMS also makes reference to the java trap.
Tp.
I wonder what is going to happen to these three projects?
Let's hope now Java integrates all the good features of C#, like true generics.
Open Source Java Web Forum with LDAP authentication
You've been able to do that for years - just not under an Open Source licence. Sun have provided the entire JDK source (including the VM code) under their own Sun Community Source Licence (see http://www.sun.com/software/communitysource/j2se/
You need a bunch of binaries to get it bootstrapped (i.e. it requires Java to build Java) but the result is entirely compiled from the source you can get from the above site.
GPLing it is a change of licence terms, not a change in the actual availability of the source.
Python is also compiled and byte interpreted. It is certainly not interpreted in the same fashion as a shell script. Python even keeps the bytecode versions around and wil use them later to save recompilation. Any speed difference betwen the 2 will not be due to interpreation.
It is likely that Java will be a bit faster because there has been more resources thrown at and therefore more people able to do optimisation.
However, the big reason is that Python is a dynamic language (similar to Perl) and the compiler cannot make the same optimisations that a more staic language, such as Java and C can. It also tends to mean an extra layer of pointer redirection (hidden behind the scenes). When you have the ability to tie ('tie' is the Perl-speak way of talking about this - not sure if they use another word in Python) variable to external resources, you can't take any chances even with consecutive reads of the same variable. This all adds time.
As you say, you last shot is a low blow. Just because something is 'leading' does not make it the best - is Windows the best OS or IE the best browser? Maybe, maybe not - but I htink on Slashdot we'd agree there is more to it than just 'leading'. I've not used either client, but maybe Azureus is simply a better program with a better interface, maybe it is more widely available for different platforms? I have no idea - but it says nothing about the relative merits of the 2 languages as such.
Second, it's compiled and then byte-interpreted, giving it a fairly good speed compared to Python's interpretation.
Python is compiled and then byte-interpreted, just like Java. If you have to bring the speed argument to the table, at least bother to verify WHY it's slower before spouting nonsense.
Also, while Python maybe be slower objectively, especially in number crunching, the significantly faster startup times and generally smaller memory usage often make it seem much snappier, which is at least as important for many if not most applications. Java GUI toolkits other than SWT are also universally regarded as slow, and Python is significantly easier to extend with C so you can speed up the parts that really need it.
Also, Java is embeddable as a web applet.
In theory, in practice Java applets never worked well, and have been going the way of dodo pretty much since their inception, you rarely see one these days - fortunately, painful as they were.
That's like saying if you code C++, Bjarne Stroustrup has indirect control over your code. It's a ridiculous argument and the reason it gets any weight here at Slashdot is because it plays into that "Star Wars" mythology of the battle between good (FLOSS) and evil (Microsoft). And I'm sorry to say this but Star Wars isn't real.
Second, it's compiled and then byte-interpreted, giving it a fairly good speed compared to Python's interpretation.
Python is bytecode-compiled also, although to be fair I believe Python's bytecode is substantially higher level than Java bytecode. There is also a native-code just-in-time compiler (psyco) for x86 platforms that works reasonably well.
IMO, the main problem with Python, performance-wise, is that a variety of operations it supports are defined by the commonly-accepted understanding of the language to be atomic. This includes modifying lists and dictionaries. Because of this, these structures must have all accesses synchronised, which the standard Python implementation achieves by only allowing one thread to execute at once. I don't think I need to point out how horrible this is for scalability to multiprocessor systems.
Just to make it clear: Java is going to have the GPL with the Classpath-exception, which explicitely allows to create closed source applications when using the standard libraries.
See Gosling's blog:
http://blogs.sun.com/jag/
and here you'll find the mentioned exception:l
http://www.gnu.org/software/classpath/license.htm
Richard Stallman discussed this in a Nov 1st interview. I've put a transcript online.
Please help publicise swpat.org - the software patents wiki
I want a GPL'd latte :-(
No need, latte is already distributed under a BSD license, which is GPL compatible.
If Python is superior to Java, then why is the leading Bittorrent client, Azureus, written in Java
I don't know but it may be connected with why it's also a buggy, slow, memory-hogging piece of shit that crashes or just goes ape-shit crazy on a regular basis. I have a theory that this is actually caused by Java's garbage collection agent trying to delete the whole of the program's code on the grounds that it clearly is garbage.
GPL does not prevent forking (and some forks will develop due to $$$)but if the forker abides by the GPL he has to publish his code for the forks, so Sun could have incorporated those changes making Sun java compatible with MS-J.
Not necessarily:
1. GPL doesn't require patent licenses to be granted.
2. Most of the modifications MS made to their Java implementation wouldn't have worked for Sun because they were heavily dependent on features of Windows that aren't present on any other OS. Sun could've easily reimplemented compatible extensions to Java if the only issue was access to MS's code. They didn't because doing so would have violated the spirit of Java as a cross-platform language.
I don't think so. While Java isn't as fast as C, it also doesn't peg my cpu at 100% on a regular basis. If Novell made their deal with the devil^H^H^H^H^HMicrosoft to protect Mono, they screwed up.
This is interesting enough that I'm going to hit the book store for some more O'Reilly - for some Struts, Hibernate and Spring.
Java is NOT strictly interpreted language. HotSpot (http://en.wikipedia.org/wiki/HotSpot) mechanism dynamically compiles bytecode into machine code.
In theory, Python has Psyco that can do JIT-compiling, but in practice dynamic nature of Python prevents most of optimizations.
IBM licensed Java from Sun a number of years with contracts signed and everything. Assuming the Nazgul did a good job on the contracts, IBM can continue doing whatever they like with their Java code tree.
The GPL isn't some black hole that can suck up any and all licenses into it's undeniable gravitational pull. Come to think it, real black holes don't do this either unless their event horizons are crossed. Put that way, IBM Java can't and won't cross the GPL event horizon. At most, it will orbit at a safely removed distance.
> Just think if Sun had done this in the 90s. There would never have been a GTK/Qt appliactions split because all software would have been written in Java.
Oh thank God they didn't open source it in the 90's!
...these aren't my real teeth.
Debain wanted to distribute their own patched version of Firefox, the Mozilla people didn't want to get blamed for all the bugs introduced by Debian, the name change was the way to resolve that conflict.
My own sympathy goes unreservedly to Mozilla, as a software developer I know what a total pain Debian is to deal with.
That it was all written using Visual Basic.
Take the cheese to sickbay, the doctor should see it as soon as possible - B'Elanna Torres, "Learning Curve"
Anytime I startup a Java program I want to run and hide due to fear that the pages being swapped out to make room for it will crush me where I stand.
Maybe as open source software people will be able to look at it and ask *why* does it have to take up so much memory to do such simple jobs? Compare for example Azureus (in Java) to bittorrent (in Python).
I know. I hate navigating the maze of incompatible GCC, Perl and Python implementations we're stuck with.
Dewey, what part of this looks like authorities should be involved?
Tell that to Google. My own company has switched to Python for all future application development. If Python is a scripting language, then so is Java.
Dewey, what part of this looks like authorities should be involved?
I see a couple of benefits of the GPL, namely that my favourite distributions can now supply Java as a standard part of the distribution, and that I should probably now be able to easily get ports of Sun Java to platforms I want to use (whereas I was limited to IBM or an older port from Blackdown before)
Sun can still maintain control over "Sun Java", which is what most people will use. Sure, people could fork and deliver versions that break the tests you mention, but it is unlikely many people will use them. There are forks of most major GPL packages out there, but for the most part people stick with the main tree.
The other advantage the GPL has over the existing Sun licences is that many people are unable to look at or work with the code from Sun because their employer forbids them. GNU Classpath is an example of this in open source - cleanroom implementations only, no peeking at the Sun JDK source, and make sure you don't sign any NDA's on the way.
I meant suitable as a Linux compatible license. Now distributions can include the JDK without conflict. Personally, I find Java's performance, stability, security, and reliability to be just fine.
I'm moving to Hell this summer. I hear the weather is lovely down there lately.
And I want a link that doesn't crash Firefox 3 times in a row ...
I don't know what's on the page that's so evil for Firefox under Linux, but I had to open it in Epiphany to see ... meybe its another hint for me to get SuSE off my hd ASAP.
That should actually read:
Only if C++ is patent encumbered, there are few implementations, and the only useful implementation is owned by a company that is no longer willing to try to work around the patents to produce a truly free implementation, and, in fact, has made a deal with the devil so only their own customers can legally use their implementation.
Sticking feathers up your butt does not make you a chicken - Tyler Durden
This is one of those cases where being specific about the language you're using might help. What he probably meant was "suitable for distributions of GNU/Linux". There are no (serious) licensing issues any more with Java and it can be easily integrated with pretty much all standard distributions. Java can also become a base for further development of projects that make a point about not being tied to proprietary technologies. For example, it can be integrated into GNOME and KDE such that critical components can rely upon Java being there in a way that would be legally dubious today.
If people "hack" Java in order to add new features that make it incompatible with other versions, then Sun would have every right to enforce the Java trademark against them so they can't offer as "Java" and such confusion is reduced. If one of these forks turns out to be successful, then that's great as it presumably would have reasons for being successful, like, say, being better.
If, on the other hand, people start hacking it to:
- Port to alternative platforms
- Integrate it into frameworks it doesn't "natively" integrate with right now (KDE, GNUstep, etc), IIRC the current GNU/Linux distribution uses GTK, though I'm prepared to be corrected on that
- Play with various features of the system to improve performance and memory usage (there are a million ways you can write a garbage collector)
- Do the latter, but optimized for specific instances (should the same GC model be used on your Sharp Zaurus, Playstation 3, and 100GHz Core8Octomegaseptupal with 8T of RAM*
Either scenario, even your "People will corrupt the holy Java" instance, seems like it can't work out negatively, and may work out positively. They're good things!
* Note to people reading this a decade from now. Believe it or not, people in 2006 thought dual-core 3GHz CPUs were "Where it's at". I don't know what Octomegaseptupal is either BTW. The point is that's supposed to be a really insanely high end system.
You are not alone. This is not normal. None of this is normal.
I haven't tried JSF, but I'm the sole maintainer of a Struts application. It works reliably, but changes are a pain in the behind.
s /Stripes+vs.+Struts
I have zero experience with the Stripes web framework, but this comparison between Stripes and Struts has an excellent illustration of the klunkiness of Struts: http://stripes.mc4j.org/confluence/display/stripe
"One of my prime frustrations with Struts is the fact that just to implement a single page/form, I have to write or edit so many files. And I have to keep them in sync, or else things start going horribly wrong. With Struts I have to write my JSP, my Action, my Form, a form-bean stanza in the struts-config.xml, an action stanza in the struts-config.xml, and if I'm going to do it the Struts way, a bunch of forward stanzas. And let's not go into the fact that since this is all stored in one xml file I'm continually facing merge conflicts with my team mates. Yes, there are annotations for Struts, but they are just literal translations of what's in the XML file, and they don't feel natural to me."
You may with to learn Struts anyway, because it's so common. But if you're builing a new Java web app from scratch and no one on your team is used to Struts, I'd investigate alternatives. The Struts project page even lists a few under the 'Similar Projects' heading. Now, extensibility, stability, and other buzzwords matter just as much as ease of initial configuration. So don't use speed of initial development as your sole criteria.
1. GPL doesn't require patent licenses to be granted.
If you redistribute code under the GPL, you give all the recipients rights to all the patents they need to actually use the code, and they can pass those rights on to others.
What the people who believed this will never happen are saying now?
"I think this line is mostly filler"
That's what you get for using Firefox 3. I had no problems using Firefox 2.
that Java on Debian will be as easy as "apt-get install java"
:-)
Close... More like: "apt-get install java java-doc java-common java-dev java-examples java-gnome libjava libjava-doc libjava-dev"
I've seen a few posts suggesting that Sun couldn't do this till now because of the forking threat in the past, reprepresented by MS's J++. I think this is wrong and probably comes from people who blame everything on MS.
.net (and maybe mono on Linux).
Sun didn't open Java before because it didn't have to. The fact that they are opening it now suggests to me that they are starting to fear
Personally, this makes me much more inclined to go to Java as my next language, rather then C# (I'm currently using C++) but, given Sun's past refusal to open Java, this move by Sun suggests to me that Java isn't doing as well now. (If only Sun had done this 10 years ago I could have been using Java all these years instead of C++.)
One other possible factor is the 3rd party open-source Java efforts. I don't know how they were coming along, but if any of them were doing well then they could have increased the pressure on Sun to do this. The great irony of those projects is that they get wiped out if they are successful.
I felt a great disturbance in the slashdot, as if millions of "sun is the next redmond" trolls cried out in terror, and then vanished.
1. GPL doesn't require patent licenses to be granted.
Actually, it does. If Mircosoft didn't want to grant a patent license, they could not distribute it under the GPL. The funny twist is the one Microsoft pulled with Novell - what you need to make sure is the party distributing the code (Novell) isn't the same as the one holding the patent (Microsoft). They can't give Novell a patent license because that'd hold Novell to the GPL - instead they create a covenant saying they won't sue Novell for those patents. It's in essence a patent license without actually giving one. Mircosoft could have pulled exactly the same with a subsidiary "Microsoft Java, Inc." who'd happen to be in a covenant with Microsoft, Inc. to not get sued, but everyone else is.
7. If, as a consequence of a court judgment or allegation of patent
infringement or for any other reason (not limited to patent issues),
conditions are imposed on you (whether by court order, agreement or
otherwise) that contradict the conditions of this License, they do not
excuse you from the conditions of this License. If you cannot
distribute so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you
may not distribute the Program at all. For example, if a patent
license would not permit royalty-free redistribution of the Program by
all those who receive copies directly or indirectly through you, then
the only way you could satisfy both it and this License would be to
refrain entirely from distribution of the Program.
Live today, because you never know what tomorrow brings
That's a flash 7 bug. export XLIB_SKIP_ARGB_VISUALS=1 somewhere.
Reading package lists... Done
Building dependency tree... Done
Package java-examples is not available, but is referred to by another package.
This may mean that the package is missing, has been obsoleted, or
is only available from another source
E: Package java-examples has no installation candidate
Creationist Textbook Stickers Declared Unconstitutional by CowboyNeal
Java being released under the GPL might have an even bigger impact in the long term than it appears (from these comments) many people think. It may represent the beginning of a huge leap forward for open source. The following line of reasoning seems at least somewhat plausible...
Of course, the idea's not perfect, especially due to the large number of existing platform-specific applications and the fact that some applications need to be platform specific to run efficiently. Perhaps ahead of time compilation will be used/developed further? Anyway, even if this happens, it's probably going to take a very long time- maybe decades. But if the end result is that people coalesce around a freer and more open platform, and better (often open!) and cheaper (often free!) software comes about as a result, we'll all (well, except for Microsoft, of course) be better off.
Now, I can just grab the binary (since the GPL allows binary distribution of derived works, as long as the source is available). The real question is how is the trademark licensed? Will the unofficial ports be allowed to be called Java? Or will I have to run 'KenyanPeaberry, base on Sun Java(TM)?'
I am TheRaven on Soylent News