Microsoft selling eBooks?

[pwalker8]

Originally Posted by darryl

The Oracle case is indeed "a bit more" than the SCO one. An unfortunate parallel with SCO is the uncertainty and the threat to third parties, in SCO's case end users, in this case app developers. And, of course, the possibility of selling licences as a form of insurance until the matter is finally decided. Unfortunately whilst SCO's case seems to have had little merit, the same cannot be said of Oracle's. I expect SCOTUS will hear the matter, but there are of course no guarantees. Where I thought the ZDNet article was going overboard is its talk of an imminent mass exodus from Android, particularly one involving a boom in IOS and a resurgence of Windows Phone. It's not that it is impossible, but I am not expecting it to happen, conservatism and potential risk notwithstanding. There simply does not seem to be an acceptable alternative to reach all of those Android users, though there may be a lot of money to be made by anyone who can quickly provide such an alternative. I haven't followed the situation too closely but there were a number of alternatives under development, some of which could actually run Android software, though I would be very surprised if they were able to manage this without the use of Android API's.

There will certainly be a lot of nervous developers around, especially the larger ones, who will no doubt be seeking advice as to their exposure. I'm no expert in US law nor am I am aware of Google's full Android licensing terms. I'm simply not sure what exposure developers may have over and above Google's exposure. The damages situation alone is fascinating. If only Sun had held on!

My understanding is that Sun actually licensed Java as open source via the GNU General Public License. Oracle put out a new licensing agreement in 2010 which gave people and corporations a free license for development, testing and internal use, but production use is only free for general purpose computing. Frankly, java would have never taken off like it did with Oracle's license in place.

Oracle is well known for pushing the envelope with regards to licensing claims. Back in the mid 90's, Oracle claimed that the company that I worked for (large fortune 500 company) rather than pay a license for each server, should pay a license for every customer who connected to that server to access a program that used the Oracle database. They were not successful in pressing their claim.

My guess is that if this Oracle case is ultimately successful, then you will see big companies start to eliminate java. I know that a lot of companies are starting to move away from Oracle to go with some of the Open Source database products.

[pwalker8]

Originally Posted by fjtorres

The basic model is sensible: patents for ideas, copyright for implementation.
There is value in software and investments of time and money in creating them. Software in all its forms is worthy of protection and copyrighting the source code is the proper tool and place to do it as it protects against unathorized derivatives but allows true clean room copies. It's good policy.

In literary spaces you copyright specific stories but not plot ideas or titles.
You can tell the same essential story idea (boy meets girl, boy loses girl, boy meets alien vampire and dies) a hundred different times as long as you tell it in different ways. Lots of writers make a good living telling the same story over and over.

Oracle vs Google is different from most software catfights because there is a smoking gun and the derivative product is, as the appeals court said, a one for one replacement of the original. Google simply didn't want to pay for a license.

Most of the people opposed to Oracle's position are only opposed because of the expected disruption from a definitive Oracle win, not because they don't think Android "plagiarized" Java. Most would be the first to sue if it was "their" software getting ripped off.

The key thing to remember is the Android API started out as an exact copy of Java *and* the interpreter used Java code. In legal terms that makes the entire product (Android) an unlicensed Java derivative. It is not unlike the mess with J++ where Microsoft had to pay Sun for creating a Java derivative, something they did not have a license to do.

Creatives have to be protected or they won't have reason to create.

As "the irascible Harlan Ellison" says: pay the writer!

https://www.youtube.com/watch?v=mj5IV23g-fE

The basic model doesn't understand how programming works in the slightest. Can a baker copyright baking a specific cake? Can a carpeter patent dresser drawers? Many people to all sorts of creative things that have no legal protection at all.

[pwalker8]

Originally Posted by DuckieTigger

How do you suggest to go about it then? There has to be some copy protection system in place for software. Something that prevents a simple disassemble and recompile with different optimization options. Different executable, same features without doing copy and paste.

Why does there have to be some copy protection system in place for software? Or more specifically, why must there be some legal protection for just software?

Much of the protection for binary code could be protected by licensing agreements, while the source code itself is better protected as an industrial secrets. Is it really any different than a chief's recipe? A lot of companies have jumped on the intellectual properties bandwagon, but couldn't the same argument be made for cooking or the layout of an assembly line? What's so different about software?

[DuckieTigger]

Originally Posted by pwalker8

Why does there have to be some copy protection system in place for software? Or more specifically, why must there be some legal protection for just software?

Much of the protection for binary code could be protected by licensing agreements, while the source code itself is better protected as an industrial secrets. Is it really any different than a chief's recipe? A lot of companies have jumped on the intellectual properties bandwagon, but couldn't the same argument be made for cooking or the layout of an assembly line? What's so different about software?

How do you go about protecting binary code? How far do you want to go? Nice analogy about cooking, but that cannot really apply to software the same. Example a simple bread. One bread is indistinguishable from the next one. Now take Photoshop. Make a copy of it so perfect that it is impossible to tell the difference other than the opening splash screen. Should you be able to do that? Keep all the features, but change the GUI around so it doesn't look like Photoshop any more. Should you be able to do that? Where do you want to draw the line? Maybe at rounded corners.

[pwalker8]

Originally Posted by DuckieTigger

How do you go about protecting binary code? How far do you want to go? Nice analogy about cooking, but that cannot really apply to software the same. Example a simple bread. One bread is indistinguishable from the next one. Now take Photoshop. Make a copy of it so perfect that it is impossible to tell the difference other than the opening splash screen. Should you be able to do that? Keep all the features, but change the GUI around so it doesn't look like Photoshop any more. Should you be able to do that? Where do you want to draw the line? Maybe at rounded corners.

Oh, various types of bread and brands of bread tend to be very different from each other.

What about elevators? the UI on elevators tends to be almost identical from brand to brand. Should they be allowed to do that? How about the UI for a car? They are also almost identical. Indeed, driving would be much harder if they were not. Why should software be different? As far as making a perfect copy of Photoshop, why not? As long as you do a clean room creation, why shouldn't you be able to do so? Is that really all that different than screw drivers, power drills and any number of physical devices which tend to have very little difference from manufacturer to manufacturer?

Software pleads for special protection, yet in reality, I'm willing to wager that Photoshop has plenty of code in it that was brought by programmers from other jobs, or from books and examples of how to do a specific task on the web. One of the prime skills for a productive coder is how to use google.

[darryl]

@pwalker8 and @DuckieTigger. Your discussion highlights some of the very real problems with intellectual property law. No longer do such laws require some significant real invention or innovation. The maxim that ideas alone should not be protected seems too often to be honoured in the breech rather than the observance. And, of course, the length of some protections is just ridiculous. The nature of software development is such that it sits very uneasily with intellectual property constructions, particularly when fragments of products qualify for protection. It is of course desirable to have some protection for software. But in my view software patents do significant damage. Protecting a method of doing something as distinct from its implementation is ridiculous. And copyright is very ill-suited for its purpose. It applies to the source code as if it was a book. And the binary is of course derived from that source code. As for the binary itself, I would have thought that copyrighting a series of 0's and 1's lacks originality and is artificial to the extreme. And in the fast moving world of IT intellectual property terms for the life of an individual plus in my view just don't make sense. Particularly when we are talking not about complex programs but sub-routines and modules and basically code designed specifically for re-use. And of course API's. The whole area is long overdue for a comprehensive review. Unfortunately any new system would likely be heavily influenced by the lobbying of the usual suspects.

[DuckieTigger]

@darryl: Yes, current system is not very suitable. Proposed changes of removing basically all rights and allow everything as long as it is "clean-roomed" is the other extreme that is not going to fly ever.

Not sure what the middle ground is, maybe leave as is and remove the API from protection. Maybe the Supreme Court will surprise us and set precedent for Google and against Oracle. That would be a good start.

[pwalker8]

Originally Posted by DuckieTigger

@darryl: Yes, current system is not very suitable. Proposed changes of removing basically all rights and allow everything as long as it is "clean-roomed" is the other extreme that is not going to fly ever.

Not sure what the middle ground is, maybe leave as is and remove the API from protection. Maybe the Supreme Court will surprise us and set precedent for Google and against Oracle. That would be a good start.

One can hope.

A lot of the current problem in the US goes back to the US patent office throwing up it's hands and saying "I give up, we'll grant even the flimsiest patent and leave it to the courts to sort out" back in the 70's if I remember correctly. If I remember correctly, software patents in the US was the camel's nose in the tent, i.e. it started as software in conjunction with a mechanical device and just kept pushing the envelope.

[pwalker8]

Originally Posted by darryl

@pwalker8 and @DuckieTigger. Your discussion highlights some of the very real problems with intellectual property law. No longer do such laws require some significant real invention or innovation. The maxim that ideas alone should not be protected seems too often to be honoured in the breech rather than the observance. And, of course, the length of some protections is just ridiculous. The nature of software development is such that it sits very uneasily with intellectual property constructions, particularly when fragments of products qualify for protection. It is of course desirable to have some protection for software. But in my view software patents do significant damage. Protecting a method of doing something as distinct from its implementation is ridiculous. And copyright is very ill-suited for its purpose. It applies to the source code as if it was a book. And the binary is of course derived from that source code. As for the binary itself, I would have thought that copyrighting a series of 0's and 1's lacks originality and is artificial to the extreme. And in the fast moving world of IT intellectual property terms for the life of an individual plus in my view just don't make sense. Particularly when we are talking not about complex programs but sub-routines and modules and basically code designed specifically for re-use. And of course API's. The whole area is long overdue for a comprehensive review. Unfortunately any new system would likely be heavily influenced by the lobbying of the usual suspects.

Yep, all true.

One of the issues is that when people started going down the IP route, there really isn't a limiting principle. Everything "deserves" protection now and once they start getting protection, they loath to give it up, which keeps pushing the length of protection.

[fjtorres]

Originally Posted by darryl

Protecting a method of doing something as distinct from its implementation is ridiculous. And copyright is very ill-suited for its purpose. It applies to the source code as if it was a book. And the binary is of course derived from that source code.

Do note that the "ideas" implemented in software are the underlying algorithms, not the source code. The source code is the human readable expression of those ideas. Just as a book is an expression of a thesis, theme, or plot.The parallels are solid enough when it comes to copyright.

The "zeroes and ones" of binary aren't what gets protected.

The system as it stands, warts and all, protects new software ideas (algorithms, mostly) through 17-year patents and protects the expression of those ideas, new and old, via copyright. (And yes, copyrights are too long.)

Binary is just a translation of the source code so it is covered by the source copyright. There is nothing in binary that wasn't in the source. It is no different than a copyright of a spanish novel covering its english translation. Or french. Translation rights come from the original creation, not separately. It wouldn't make sense to protect only the translation and leave the original unprotected. (That used to happen before the Berne Convention.)

The Oracle/Google mess started out simple enough: Oracle suspected and later proved that Java Interpreter source code was directly used to create the Android interpreter. Google countered by saying they only used it to duplicate the APIs and that APIs aren't copyrightable. (Which, individually, might be true. Might not, though. At present the courts say they are.)

Lots of arguments, tons of time and money later, the appeals court took a step back and said "wait a minute. They copied all the APIs. And added nothing. And then used the copy as a substitute for the original." (In copyright law, that part "substitution" is critical. It is what allows parodies and critical analysis to exist.)

Heard of this one?

https://en.m.wikipedia.org/wiki/The_Wind_Done_Gone

There is an extensive body of copyright law about what is and isn't allowable under fair use. But for this one particular case (not a clean room copy) the precedents are clear that if the Java source code copyright is valid (and Google never claimed it wasn't: they have their own IP to protect in other areas) and if that source code was directly copied to produce the Google product (a finding of fact at the trial) then the court has to find Google liable or Copyright has no meaning. Not for software and not for anything. The court was unwilling to go there so the case moves on to the Supreme Court.

Right or wrong, that is the law as it exists.
Changes may come via new laws that supersede existing law but until that happens, that is the reality of the day.

It's worth remembering Microsoft lost a similar JAVA lawsuit years ago and in their case they were creating something new, not just a plain copy.

I see little chance of Google winning this one.

It's an interesting mess.
Made a bit more interesting because both players are equally sleazy.

[DuckieTigger]

Originally Posted by fjtorres

It's worth remembering Microsoft lost a similar JAVA lawsuit years ago and in their case they were creating something new, not just a plain copy.

I see little chance of Google winning this one.

It's an interesting mess.
Made a bit more interesting because both players are equally sleazy.

If that was so easy and clear cut, they would have used that already to make a case in the current law suit. Not so easy now? Microsoft may have technically lost, but they still made out in the end.

This time it is different - Oracle is not interested at all to provide support for Java in Android, they want to cash in from its widespread success. If Google gets a similar deal as Microsoft did, they laugh all the way to the bank, are allowed to keep all current versions as is, and only need to worry about changing the API for the next version.

No backwards compatibility in the next Android? No problem, use this newly included virtual machine and install an older version of Android in it. To top it off, put the next API into the public domain. The essential part of Android, the Google Play services, is already separate from the OS itself, so a clone of Android is not hurting Google in any shape or form.

[GlenBarrington]

Interesting how this thread has 'morphed'! I'm not complaining, mind you, I'm just commenting on what is both an online forum's strength and its weaknesses.

[pwalker8]

Originally Posted by GlenBarrington

Interesting how this thread has 'morphed'! I'm not complaining, mind you, I'm just commenting on what is both an online forum's strength and its weaknesses.

Yep, threads go in the direction people want to take them.

[pwalker8]

Originally Posted by DuckieTigger

If that was so easy and clear cut, they would have used that already to make a case in the current law suit. Not so easy now? Microsoft may have technically lost, but they still made out in the end.

This time it is different - Oracle is not interested at all to provide support for Java in Android, they want to cash in from its widespread success. If Google gets a similar deal as Microsoft did, they laugh all the way to the bank, are allowed to keep all current versions as is, and only need to worry about changing the API for the next version.

No backwards compatibility in the next Android? No problem, use this newly included virtual machine and install an older version of Android in it. To top it off, put the next API into the public domain. The essential part of Android, the Google Play services, is already separate from the OS itself, so a clone of Android is not hurting Google in any shape or form.

Oracle's problem is the growth of the open source database programs. Large corporations, which are Oracle's bread and butter, don't see the point of spending millions of dollars a year for something they can more or less get for free. Oracle has had a monopoly in the database business for so long, they got lazy, content to hire cheap labor and just milk the cash cow. They see Java as another cash cow where they can simply milk the licensing for as much as they can get.

Google first released Android in 2008, before Oracle bought Sun in 2009, and then released the new licensing in 2010 and immediately sued Google. At the time they developed and released it, Java was a good bet with it's GNU license. Now it appears that Apple made a wise choice in developing their own programming language.

Frankly, Larry Ellison is about as close to the late 19th century robber barons as we have seen. The downside is that the excesses of many of that class is what triggered the anti-trust laws and commerce regulation acts of the late 1880s. While everyone remembers the Standard Oil case, the Sherman Anti-Trust act was really aimed at the railroads, who were using their monopolies to manipulate the market.

[DuckieTigger]

Originally Posted by pwalker8

Google first released Android in 2008, before Oracle bought Sun in 2009, and then released the new licensing in 2010 and immediately sued Google. At the time they developed and released it, Java was a good bet with it's GNU license. Now it appears that Apple made a wise choice in developing their own programming language.

That is what happened? How is that possible or even legal. If Google used a version of Java that was under a GNU license, then copyright should not even come into question. As far as I understand, the GNU license allows to break copyright in the one case where the modified software also have a GNU license.