Get rid of java
No unsigned types, what a joke.
Computer languages and software interfaces may fall under copyright protection if Oracle succeeds in its Java lawsuit against Google. Amazingly, "copyfighters" appear to have paid little or no notice to this rare extension of copyright into new realms. But the consequences and costs for the software industry could be enormous …
No unsigned types, what a joke.
Unsigned types are a side-effect of how the underlying hardware works, a high-level language which is abstracted from that has no conceptual need for them.
Go back to /b
No. signed/unsigned is a programmers choice of what bundles of bits mean. I've found describing 8 bit bundles as unsigned values is usually pretty damn useful.
That a high level language imposes arbitrary limitations on how a programmer should be thinking is the Java disease. Simplified for idiots only sounds like a good idea.
If you need unsigned, there are libraries that provide it. I wrote one years ago, and I am pretty sure it is BSD-licensed. I'd be surprised if I was the only person to do so.
Main thing to note is that there is no difference in the bit-pattern behavior between 2s-complement signed + and unsigned + at the word length. The trick comes in describing the operations that are not punned -- comparisons, division, widening multiplication, conversion to/from float and string.
The other half of the trick is to be sure to use idioms (where you can) that will be recognized by an optimizing JIT as equivalent to the unsigned operations. That probably works for comparisons, but not for the others.
I also wrote a library for packing and unpacking bytes and shorts into ints and longs, not sure whether we released that or not.
@"no conceptual need for them."
Imagine trying to tell that to all the mathematicians throughout the past 2000 years and then see if you can't find any "conceptual need for them"!.
Just because Java chooses to omit them, doesn't mean there is no conceptual need for them! ... Negative integers are very useful and fundamental in many areas of work and its a total fuck up design flaw that the only way a language can support something so fundamental is via a library which slows its performance down. But then a lot of Java developers don't believe in thinking about performance either, so in their narrow myopic mind, there is no conceptual need for them. Try opening your mind and find conceptual needs for them, because if you don't others will and they will beat you in a competitive job marketplace.
But if you want to remain close minded and refuse to see what I'm saying, we can at least just console ourselves that there is one less competitor in the job market than we thought. But I find I can't shake off the despair I feel at some people's closed mindedness!
Think you are misreading that, of course java supports negative number. It doesn't however have _UN_signed integer types
Does Java support enumerations now? The last time I played with it (ooo... 10+ years ago?), you had to use some crazy structure to 'simulate' an enumeration. And I agree; lack of unsigned is just plain stupid and indicative of the half-baked lanhuage that Java is. Oh, and while you're at it, no multiple inheritance (I like to throw that one in because it seems I'm the only person in the world that thinks multiple inheritance can be useful). Java is also responsible for breeding a whole generation of third-rate softies who's idea of efficient and small is "it runs ok(ish) on a 3GHz quad core monster", and "it's smaller than 500MB" respectively.
And on a completely different tack, as far as I know (and apparently well illustrated by recent events), Java still isn't completely "free" and many people have been warning for years of the potential for the mess that we're now witnessing in the latest Oracle/Google spat. But, of course, many more people have chosen not to listen
Ok, rant over.
"I like to throw that one in because it seems I'm the only person in the world"
Not the only one, but pretty sure you are in the minority. Try repeating the folloing mantra 10x a day, "composition, not inheritance"
Couldn't get a patent on that process but I own the copyright so I'm still entitled to some monetary relief from your use of it.
It's the ultimate trolling tool they're trying to get written into law.
That Google should have licenced Java.
They wrote a clean room implementation. It's all language or API's. They should not be copyrightable or patentable for reasons given in the article. Therefore there is no need to licence ' Java'.
Anyone sufficiently long in the tooth to have programmed Intel and Zilog 8-bit microprocessors in assembly language will remember that Zilog were forbidden from using the same instruction mnemonics as Intel because Intel had claimed copyright on theirs. By extension, it should be possible to copyright a programming language, but maybe only as long as the reserved words it used were original inventions and not taken from a pre-existing language.
"Claiming copyright" and having a judge make a ruling on a case are quite different.
BUT, IIRC, the Z80 was completely compatible at the *binary* level with the 8085. So much so, that I was able to take a project I wrote for an 8085 at Uni, and load it into a Spectrum (using a self-written Hex editor) and it worked perfectly.
Hex editor? Isn't that something witches use?
"Anyone sufficiently long in the tooth to have programmed Intel and Zilog 8-bit microprocessors in assembly language will remember that Zilog were forbidden from using the same instruction mnemonics as Intel because Intel had claimed copyright on theirs."
And Zilog's made more sense. But I digress.
It's always entertaining and educational to hear the argument of the opposition, but it is no longer fun if the author starts to make unfounded claims.
"without protection for authors the economic incentive to create disappears"
-- not true. There are other economic incentives for creation besides being able to sell the product. One of the incentives disappears, right, but not "the economic incentive"
"In recent years, a strangely zealous campaign against patent protection for software has been successful"
-- not true. In recent years, the law practice with regard to patent-protecting software has been shifting in the direction opposite to what the "campaign against patent protection for software" is advocating.
So lets carve up every idea, every advancement and every concept into money making parcels. Lets pin down any future development under the weight of contracts and limited licenses and royalty fees for usage. Why stop at programming languages when there are so many other abstract ideas you could horde and sell.
But while you're doing it don't look behind the curtain. You might notice the earlier generations of engineers and scientist that built the giant shoulders your product stands on. Who designed the networks and concepts that got you where you are today.
And if that happened, then you might have to feel just a little guilty about lining your pockets off their backs.
[Needless to say this sort of thing just p***es me off.]
Bakunin. Collectivist anarchism. Nobody owns anything, good will prevails.
Remind me again: how well has that worked out?
Andrew. I hate to break it to you, but I’m not *actually* Bakunin. Unless the El Reg forums are known for being haunted by long dead bearded revolutionaries? A few banshees and trolls maybe.
I’ve had lunch now so I’ll take off my grumpy old man hat. But I stand by the sentiment of my previous post.
Well - to be fair to Bakunin it never really got tested anywhere - except maybe some bits of Barcelona - and it's failure there wasn't really down to systemic failure of the theory. On the realistic hand - it was complete bonkers.
It seems to me that the idea of copyrighting a language or any of its derivatives is also bonkers - what is the point of a language where the use of expressive techniques is controlled? And APIs and libraries cannot surely be considered as anything more than expresive techniques - not processes in their own right? I can see that patenting a language might be a valid route in the current legal frameworks but I doubt that anyone would ever use it.
I vaguely remember it worked rather well. It was only really attempted in the Spanish Revolution, wasn't it? It's biggest problem was interfacing with outside influences.
"generations of engineers and scientist that built the giant shoulders your product stands on"
Yep that would be Larry and his mates within the fledgling Oracle Corp ( called Relational Inc or something? ) "nicking" ideas from both Ed Codd and IBM's System R to get their RDBMS product off the ground!
>>"It seems to me that the idea of copyrighting a language or any of its derivatives is also bonkers - what is the point of a language where the use of expressive techniques is controlled? "
Maybe there might be situations where it would make some sense, but at the very least, if there is going to be copyright or patent control exercised, it should be something that's made crystal clear from the start, not something that can bite someone years or decades later.
Remind me again, Andrew: when did we try it?
Some property has to be public. Not having to pay to go out into the street outside your dwelling, because it is a public road and isn't a toll road, is a requirement in order to have markets in things other than road provision.
Some property has to be private. Nobody likes sharing toothbrushes.
Where your preferred society chooses to create the dividing line will determine how much rent seeking activity is permitted, and how much competition you will get.
Generally, with some notable exceptions, monopolies are considered a bad thing and copyright is a monopoly. That's OK to the extent granting this monopoly causes work to be done that otherwise wouldn't be done, but if copyright goes further than that, such law can't enable as much value to be created as it could otherwise.
Another example of where reasonable social policy encourages an artificial monopoly is that a package below a certain price can be posted to any UK address for the same price. Both the postal package delivery monopoly below a defined price and the policy to have a public road monopoly encourages communications without which open competitive markets could not exist.
What pisses me off isn't copyright. If somebody wants to try and sell their new sparkly word processor at £500 a pop, I honestly couldn't give a damn. I might point and laugh at them a bit unless it really is worth every penny of that huge price, but it's hardly worth posting Angry Bloke comments on a web forum about.
What pisses me off is using copyright or, lately, the even more broad patent concept to control just about everything you can about how people use their products that they bought from you. What pisses me off is after shelling out a shitload already on a decent computer, there's a tax to be paid to Microsoft if you want to use any kind of widely-available software. The continuous attempts to destroy rather than encourage interoperability by shackling every possible means to do so behind patent taxes and threats is a disease, and Oracle's behaviour towards Java on Android is yet another example of this. They didn't invent the language, they bought the business and then started trolling everyone.
So remind me, just how well is this working out for us, Andrew?
Spot on: it's a land grab.
And the best, the very best thing about it is, it doesn't involve any actual *land*.
So up-and-coming, hungry young people and businesses can be allowed to control as much of it as they can get away with, and the richer, older people who own the actual land don't have to give up an inch of it. It's a way for the Establishment to co-opt people who might otherwise make trouble for them.
It's no accident that the modern mania for "intellectual property" has coincided with the growth of inequality in our societies. One trend is providing cover for the other.
Someone managed to invent the wheel without charging royalties for the idea.
Plenty of "work" can be done and given away for free because it is inherently useful to the worker. This is true for tools, but also for some consumer goods.
Perhaps your of the opinion that linux is a communist cancer? There it is, given away for free? Does it hurt MS? Certainly. But MS's hurt is another company's gain. Every time you install something you got for free, you gain the value the item for zero cost. When I say "value of the item" I mean "value of the item for you," not the sticker price which is an arbitrary cost, not related to its value. That could be thousands of £s for a server farm in a bank, or it could be a negative value for pirating a really bad movie you wished you hadn't seen.
On the consumer goods side, if I buy a beautiful car, I can give away value by allowing others to look at it and enjoy its beauty. I don't need to give away the car itself and I don't need to charge others to look at the car. It isn't anarchy or communism.
And it works very well, thank-you.
Nice troll though.
Google are using Dalvik because Oracle won't have anything suitable for mobile until Java 9 in 2015 or so.
On the other hand, should Oracle win this, will their lawyers be meeting the IBM Nasgul in court in regard to their derivative version of IBM's SQL?
And does Dennis Ritchie's estate now own the world?
I feel the legal dispute ignores the real issue, which is the legal status of the "ownership" of developers mindshare and developers' personal investment in acquiring specialized skills. In the case of Java, the impressive mindshare was gained by first leveraging the business community's attention with HotJava (?) Sun's java-based browser that introduced responsive client-run applets in the context of the low bandwidth and high latencies of the web of the time. Developers mindshare is the only thing that Google appeared to "steal" with Dalvik. Applying copyright reasoning to the case is for the least confusing intellectual consumption (entertainment) with production (work).
It is surely time that there was some government involvement in clarifying the purpose of IPRs rather than leaving it to courts to interpret. It should be remembered that the granting of monopolies in the form of copyright, patents, trademarks and the like was (or at least should) be done where it is in the greater interests of society and some recognition of natural justice. However, where this is just being used as a tool for extending scope of market control and reduction of competition, this is questionable. Undoubtedly the increase in the use of IPRs on the part of (mostly) larger US corporations is towards the control of markets and the generation of local monopolies through lock-outs.
It's about time a proper commission was put together to clarify these positions with the purpose of protecting the interests of the great majority and not those of major corporations seeking to generate areas immune from real competition.
Of course this introduces huge international issues, but there's a good case for some economic blocs to take a more liberal view of IPR extent than and increasingly litigious US corporate sector. There are, of course, very different interpretations of the scope of patents in the EU for instance.
That's the mental disorder right there.
It does NOT exist to encourage innovation. After all, if someone knew how to encourage innovation, we would be cruising the galaxy right in a Jetson two-seater by now. It exists to put a "NONE SHALL PASS" fence around someone's crap that he deems to fall under "IT'S ALL MINE, MINE!!", completely forgetting that he stands on the shoulders of all that worked before him and worked next to him.
Just looking at how patents came to pass, how they metastasized into a daily dose of cancer and the historical evidence of the damaging effect on innovation tells the whole story.
Copyrights have their use. But these too should be strongly curtailed. In particular, no control on grey imports, please.
There are people who are genuinely afraid to talk about their ideas for fear of having them nicked.
I'm currently working on some language learning software, but I can't release any early public betas because there is no copyright protection on my ideas, and if someone with more time and money than me decided to reimplement my ideas, they would end up with a killer app and a first-to-market advantage. If software was patentable here (the UK) I'd have a prototype in the patent office this summer and a public beta started. It would then be easier for me to get critical mass to push to v1.0 and start earning the cash to support full-time development (and the hiring of a GUI designer -- not my strong point).
That said, I'm against software patents on an ideological level, but to argue flat that IP protection does not encourage innovation is incorrect. It encourages ground-up innovation (such as it would in the case of my software) but in some ways it does discourage *incremental* innovation (eg if someone wanted to add an improved learning task to my software, they wouldn't be allowed to in a patented world).
There's a balance to be struck. Neither "free-for-all" or "screwed down tight" offers the required protection or flexibility.
"I'm currently working on some language learning software, but I can't [patent] my ideas"
Sure you can: patent a method of learning a language. That it can be implemented in software should be incidental. If the method cannot stand on its own merits then so-be-it.
"It encourages ground-up innovation [...] but in some ways it does discourage *incremental* innovation"
Don't flatter yourself. It's *all* incremental innovation (you've just not performed a patent search yet).
Google would have been happy to give money to Sun for a license to use Java the way they wanted, but Sun would have none of it; they'd only license JavaME (think of Mini Me, only not as effective) for use on handsets, not the full-blown JavaSE.
So they came with a workaround where the programmer works in Java, but the handset does not: everything is translated to a different bytecode language prior to deployment, and runs on a non-Java VM.
Actually I'm kind of surprised this point doesn't come up more often: it should be the crux of Google's defense, since it effectively means Java is only used on the desktop, on development time, and no Java IP makes it to the device.
Thanks for that insight, it clears a few things up.
So am I to understand that a company that has no interest in making mobile phones or competing in that market, would not license what is essentially a free platform on desktop computer - but is now suing a company that has used it to create something brilliant that the original creators did not foresee or intend?
JavaME licensing appears to be the crux of Oracle's vexatious litigation. Larry has noticed that ever more powerful smartphones make things like JavaME obsolete. As smartphones catch up and even surpass some desktops in power and functionality; it's only natural evolution that JavaSE would be desired over something that belongs in a museum.
The real question here should be "what is a computer?" The smartphone in my pocket is now more powerful and functional than several previous desktop and laptop computers I have owned in the past.
Is it the ability to make calls and SMS? Both laptops and desktops are now capable of that - though they are not as convenient to carry around.
Actually, no and therein lies the crux of the current issue. It wasn't one company, it is two with a transfer of IP from one to the other, with the second taking a very different view of how to use and leverage the IP.
The first company mostly liked the idea of the open commons, but not quite enough to put it fully in the open commons like Linux or BSD. They kept one part of it to themselves, even though they granted free licenses to use it. The second company looked at it and said "ooooH! We can makes lots of money! Particularly from those damn fools who never got a license from our predecessor and now dominate the mobile phone market."
"Intellectual property exists to encourage innovation"
Isn't that an oxymoron?
so not a lawyer who knew what he was talking about and well after the event.
About as relevant legally as my opinion here.
Java is loosing grounds for years.
Patenting it will just kill it sooner.
Kudos for Oracle, i never liked Java!
There is "losing" and there is "loosening", "loosing" is not a word.
So either "losing" ground.... or "loosening" their grip on...
Actually "loosing" is a word but it's usually only applied to fateful lightning from terrible swift swords.
Loosing is a bit archaic but it's still a word.
Loosening a noose allows one to breath, loosing it removes it completely.
although not likely the intended word. But instead of loosing rivers of blood and bile about this online*, you could have just sat smug in the knowledge that you are slightly better at spelling than your average commentard.
* italics taken directly from dictionary.com's all-too-relevant example on "loosing".
Why sit smug, when you can inform people the details to tell the difference, have a slightly lower chance of encountering a jarring misspelling in the future, and feel good for helping to raise the average level of knowledge in society?
I disagree with AO about it being the anti-IP crowd's fault that this API's may be copyrighted. [For the record I am in favour of reduced copyright durations some other restrictions on the use of copyright to claim additional rights by means of required licenses for non-distributive licenses but am in favour of copyright in principle].
Under copyright law I don't really see a strong reason that APIs should not be copyrightable. They are clearly creative expressions (good API design is challenging and there is definitely more than one way to do similar things).
However if they are copyrighted that will make various anti-competitive lock-ins possible if people cannot create compatible platforms or even shims to translate from one API to another.
Overall the public best interest is almost certainly served by allowing the creation of alternative implementations using the same API otherwise things like WINE would not be possible. It is a possible option that bad behaviour in this area could be prevented by anti-trust law but that generally seems both too slow and ineffective in many cases.
Maybe the compromise could be that the API could be reverse engineered from existing programmes that people want to run in a clean room style without access to the original API/source code/documentation and that as this would be purely functional reimplementation rather than literal copying it should be legitimate where taking the original API documents could still be regarded as copyright infringement.
I think Google clearly believed that they were taking shortcuts and not properly licensing technologies that they wanted to use is a bit of a case of whats ours is ours and whats yours is ours which seems to be a tendency for them. There was also no need to use Java/Dalvik for Android as they were not trying to make a completely compatible alternative to run existing software but they were trying to capitalise on the design/development/documentation and promotion effort of Sun for Java while clearly ignoring the available licensing programmes (partially because they didn't want to use the mobile profile). I hope that any side effects of Google losing this case (I suspect they will lose at least to some extent) won't have too big a negative impact going forwards.
Those of us following this case since the start understood what Oracle are up to the moment claims were filed. Many of us knew this attempt to extend copyright was coming the moment BS&F signed on to lawyer for Oracle - its exactly the same theory they pissed away 7 years and $30mil+ trying to put before a jury in the SCO vs IBM case. I posted about it here yesterday and many times before that.
The other side of it is, we sat through 7 years of analysis in the SCO case and are confident existing precedent rules out Oracle/BS&F's new legal theory, it's why no-one's panicking *yet*. Warning for the clueless but not panicking. Having been assigned a well informed judge there seems little chance Oracle will BS their way to a catastrophic result.
I suppose what really annoys Andrew is all that precedent in software copyright is the result of decades of *CORPORATE* copyright cases, not some shadowy conspiracy of 'copyfighters' fighting to undermine it. After all, FOSS relies on strong copyright to protect itself from both corporate and individual pirates but also understands there have to be limits or development becomes deadlocked.
Might be me but I thought they forced MS with a crowbar to give out certain API details so (I am aware in most cases they are still hidden or the details provided are not actually very useful), does that mean they can sue anyone who uses it (Because I doubt MS wants anyone using any of their API's unless it makes them more money?
**Below is where I am proved wrong and made to look silly**
fscked by SHA-1 collision? Not so fast, says Linus Torvalds