 Mae'r gweithio, yn fawr, a fawr iawn i'r gyflawni Emanuel Côlau, i'r lechau Llywodraeth Cymru. Mae'n fawr i'n fawr i'r fawr i'r profesi panliad Samuelson yma yma. Mae'r cyflawni Llywodraeth Llywodraeth. Mae'n gyflawni Richard M Sherman, profesi Llywodraeth Llywodraeth Llywodraeth, ac ymddirionedd Llywodraeth Llywodraeth Llywodraeth, ond mae'n gweithio'r llyfriddol yma ar y Prydysgrifennu i gyflawni Llywodraeth Llywodraeth, o gyflawni Llywodraeth Llywodraeth, o unrhyw ychydig, o'r yn Llywodraeth, o'r Cornell, o'r Toronto, o'r Llywodraeth Llywodraeth, ac oherwydd angen i'n ymgyrchonnig wedi'u gwneud. Ond mae'n hoffi'n ei bod yn ysgolod, yn gwneud hynny ar gyfer y Llywodraeth Llywodraeth, ac oherwydd iawn i'n gwneud ymddirionedd Llywodraeth. Mae'n mewn gweld fawr i'w cyfrindwyr, ond o'r cyflawn i'r cyflawn, a mae'n cyd-afodol, yng Nghymru, i gynfodol o'r cyflawn yn ddarparu i gael ei bwysig, yn gweithio i gweithio'r cefnogaeth yn y cyflawn, yn gweithio'r cyflawn i gweld yn y Y Llyw Llywodraeth, y Llywodraeth Cymru, y Llywodraeth Y Llywodraeth y Dynor, ac i gael'r cyflawn o'r cyflawn i'r cyflawni. Ond oedd byddwn yn y ddweud unrhyw o'n ddweud rhaid gael ei gweithio eich ysgol yn y Llyfrgellai Llyfrgell. Mae'r bwysig o'r amser o ran ysgol yma. Mae'r cyhoedd ymlaen yn y Llyfrgell Llyfrgell o'r arfer? Ymlaen i gael eich cyhoedd yn rhaid i'u cyhoedd ymlaen i. Mae'r ymlaen i'r ysgol yn llwyddon yn gystyried arfer. Mae'r llwyth yn llwyth, ac mae'n amser o'r gymryd yn ysgol ymlaen i'r ysgol. Mae'n rhaid i'r llwyth i'r cyhoedd ymlaen i'r ysgol. Felly, rydyn ni'n meddwl am gael y ffremio am y dyfodol yma, ychwanegwch a'r cymdeithasol i'w pryd-gwylio ar y rhagleniol. Rhaid i'n meddwl. Felly, rydyn ni'n meddwl am gweinwch Llywodraeth Llywodraeth ford gan gweinio ffordd yn gweithio ar gwaith yma. Mae'n gweld i'n meddwl am gweithio arill o'r reyfiadau. Mae'n ysgrifennu ar gweithio ar gyfer y gwaith a'r hynny o'r fath er mwyn i gael 30 yma yn llyfrneidol yn amser a fyddwch yn neb i gael hwn yn senol. Mae'r prydleidd hon yn dda i'n credu'n mynd i dda i fudio gyda chi o'r Lyfynol a'i hefyd oherwydd rwy'n meddwl y bobl oedd ymdegol yn Brad Sherman o arferwys Unedig y Llyfridol Ploedеннаяidd Llyfridol felly maeth bod yma yn yhe bwysig mewn Llyfridol. Efallai mae'n cael ei edrych yn Llyfridol. that you for that, and part of the why I mention that is not only is it a great book, so if you haven't read it, you should, but also because it was in the course of reading that book that I began to realize just how old the problem I'm interested in is and how difficult it has been to resolve it over decade after decade after decade. So we'll be revisiting some English history as well as old American history as well as new stuff. Now my initial motivation actually to pick this topic is that the issue about the patentability or copyright ability of aspects of computer programs is back up in the air in the United States. So, there's a lawsuit making its way through the courts right now that Oracle brought against Google. The reason that Oracle sued Google is because the Android platform software that many of you may have on your mobile devices is uses 37 of the 166 packages of the Java APIs. And the question about whether or not there's infringement is a question about whether the Java API packages are protectable by copyright law. And that question was decided by a district court judge saying, no, that's not copyright protectable. The Federal Circuit reversed that decision. And now the Supreme Court has the case pending before it, and the Supreme Court has asked the Solicitor General of the United States to consider whether to take the case. That makes it actually 46 times more likely that the court will take the case. And so, it is something to be watched for. The SG will probably weigh in sometime in April or May. So the case won't be on the Supreme Court's calendar until next fall. But one of the reasons why this was a good topic for exploring this old question about functionality and expression, copyright or patent is because one of the arguments that Google has been making is that Java APIs should be protectable by patent law, not by copyright law. And the Federal Circuit has said, well, just because it's copyrightable doesn't mean it's not patentable. And just because it's patentable doesn't mean it's not copyrightable, so it doesn't make any difference. And some of us are not entirely satisfied with that Federal Circuit point of view. But the court viewed the argument that the API should be patented rather than copyrighted as an effort of Google trying to rip the copyright protection out from under oracle for the APIs that Sun created years ago. Now, application program interfaces are intellectual creations that software developers carefully engineered to set the terms on which information is going to be exchanged across the boundary of one program to another in order to enable them to inter-operate effectively. I found it helpful to analogize interfaces of computer programs to the precise configuration that a plug has to have in order to fit into a socket in order to inter-operate with the electrical grid. You want the electricity to flow. You want the boundary to be broken. You basically have to conform to that precise configuration. APIs are sometimes printed in a specification document and sometimes admitted in software code. But APIs really are a fundamental part of what computer programs are about. Now, Sun, before it was acquired, its assets were acquired by Oracle a couple of years ago, so Sun had created these Java technologies, including a set of APIs. And the idea behind Java was that you'd be able to write a program that would basically, you could write it once and it would run everywhere, right? That's at least at the time when we were looking at PCs and servers, there was this idea that we want to break out from the kind of Microsoft proprietary technology and be able to essentially not have to rewrite programs every time. And the APIs that were created for that were actually something that was an important component of the aspiration to be able to do this kind of write once, run anywhere. And Sun, it must be said, was one of the foremost proponents of all the companies in the United States and indeed the world, saying that people should be free to re-implement an independently written code APIs that are developed by whoever. So the Java APIs, as I said, consisted of 166 packages. Each of the packages is hierarchically organized so that there are a set of classes below that and a set of methods that are below that in the hierarchical structure. And I thought it would be helpful to give you just an example of what kind of Java declaration or method header is involved in the case. So there are roughly 6,000 to 7,000 method headers or declarations that are specifications for Java programming. And this is the example where the package is Java lang, the class is math, and the method is basically to compare two numbers and decide which one is bigger than the other. So this is the specification that Java prescribes for achieving that. So the Java command here calls for carrying out this mathematical operation and the line of code here essentially is a more concise expression of that mathematical method. But the method is carried out by code that implements this more precisely. Now the question is something like the method header we just looked at, the command we just looked at. Is that copyright subject matter, patent subject matter, both, or neither? And it is a bit of a puzzle because if you look at the selection and arrangement of words and symbols, it looks more like copyright subject matter than it looks like patent subject matter. But the truth is that there are quite a few API innovations that have been patented. And so it's not entirely outside the bounds of the patent system seemingly if patent offices are issuing patents on these kinds of designs. And the trial court found that Google had written its own implementation code for carrying out these methods. And so there was, at least at the trial court's view of this, no literal infringement. The question was whether or not re-implementing the structure sequence and organization of the APIs was copyright infringement or not. So that's the kind of deep question and now I'll kind of go into a little bit of background. I'm not going to talk about the copyright and patent paradigms. We both know, we all know that the patent system requires you to apply it. You have to have novelty and some sort of invention to qualify for it. And if you get protection, you don't get protection forever. You only get for relatively limited time by comparison with copyright. Of course copyright provides automatic protection for authors of literary and artistic works that meet the originality standard, but it lasts a really, really, really long time. So the concern about patent and copyright overlapping is an old problem. But of course it's been a kind of convention that there's kind of like patent subject matter over here and copyright subject matter over here and they don't have any blurry lines. And in fact they do have blurry lines. But that paradigm of separability in some sense of inventions and artistic and literary works is embedded in the US Constitution, which I show you here. So this clause is often understood in the United States as essentially being two clauses in one, right? One granting Congress the power to promote the progress of science by granting to authors exclusive rights for limited times in their writings. And the second clause is the one that says that Congress has the power to promote the progress of useful arts by granting to inventors exclusive rights for limited times in their discoveries in the useful arts. So I've bolded and italicized the words to just try to emphasize it. But notice that word, respective, right? The word respective basically says, oh it's over here and oh it's over there and what do you do when something has a little blurry stuff in the middle? So the question first came up in the United States in the Baker v. Selden case back in 1880. So Selden had developed a new bookkeeping method and he published a little pamphlet about it. It had a little bit of explanatory text at the beginning. But mostly the book, such as it was, consisted of a series of forms and he kind of put in a few kind of sample entries to give you kind of an idea about how to enter data in it. Baker came along later and published a book and had forms in his book that were substantially similar to the forms in the Selden book. And Selden's widow then sued Baker for copyright infringement. And interestingly enough, I know this because I read the entire Brimgort record, that Selden claimed copyright protection not only in the book, but also separately copyright protection in the bookkeeping system. So the court, the Supreme Court, was faced with the question about whether or not the scope of copyright protection in the Selden book extended to the system or whether there was separate copyright protection in the system and the Supreme Court said no. So the Supreme Court came upon this, I think partly because in the preface to Selden's book, he actually mentions that he sought a patent for the bookkeeping method. Apparently he hadn't gotten one, but I think that ended up being very influential on what the Supreme Court had to say in this case. So the court was very clear that the copyright in Selden's book protected his explanation of the system, but did not extend to the system itself. The bookkeeping system itself was in the court's view a useful art, which was protectable, if at all, by the patent system, not by the copyright system. And the court used a number of examples to illustrate why nobody in their right mind would think that if you wrote a book about how to treat a particular disease with a medicine, or if you wrote a book about how to write and make a new plow that will cause the farming to be more efficient, nobody would think that you get copyright protection for the plow or for the medicine. And so if you can't get copyright protection for the plow or the medicine, you can't get it for the useful art of bookkeeping either. And so that's one of the important things that the court had to say about this. And the Supreme Court has some pretty strong language in there that it would be a surprise and a fraud on the public if you allowed someone like Selden to get through copyright a longer period of protection than they would get if, in fact, he'd been able to get the patent that he'd tried to get. But notice that the way that this is framed is really this kind of what I call a kind of categorical exclusivity approach, which is like something is either protectable by copyright or is protectable by patent, but never the train shall meet. Those, the subject matters are entirely separate. And we find that very similar things showed up around the same time in England. So the Davis versus Committee case was a case involving the face of a barometer. So even though there was a registration with the station or company, the court went faced with an infringement claim said now it's an integral part of the machine. It's not really it's not really artistic or literary work. And it kind of too adopted this kind of like inventions of inventors and literary works and authorship or kind of distinct types of regimes. There is also an old English case that cites to Baker v Seldon and refers to it in part as saying that this kind of the dress pattern was a mechanical contrivance that was ineligible for copyright protection. Now categorical exclusivity was kind of like part of the Baker v Seldon tradition on tell a case called Maser versus Stein. And this is a place where the American story has its own little flavor. So Stein was a guy who created a statuette of a Balinese dancer and used it as kind of the base of a lamp and then put the lampshade, kind of strung wire up the center, put a little thing on the top and then had a lampshade on it. But the Balinese dancer was initially a statuette. And so he registered the statuette with the copyright office. But Maser was like so oh man that Balinese dancer lamp is really selling well. I want to copy it. So he went ahead and copied it and said look there are lots of lamp based designs that have been protected by design patent law. So under the kind of Baker v Seldon exclusivity approach it shouldn't be copyright subject matter. And I think actually at the time the sense of many American commentators was yeah that's probably right. And part of the reason for that is again to get a design patent. The design had to be novel and non-obvious and you had to apply for a patent just like the utility patents and since Stein hadn't applied the design was arguably in the public domain and so what to do? Well the copyright office as it happens had been issuing registration certificates to works of artistic craftsmanship for quite a few years and it was considered that the statuette that Stein registered was one of the things that fit that category. So it was okay and here the Supreme Court says just because something's patentable doesn't mean it's copyrightable that being kind of true for something where you could sort of say there is a bit of subject matter overlap between design patents and copyrights in respect of things like this statuette. But I think the kind of core ruling of Baker v Seldon was still intact because the court wasn't talking about copyright and utility patent. There's never been something in the United States that's been protected by both of those things at the same time. So categorical exclusivity has lived on in the minds of many Americans since then. Now part of what we know is that the laws in England and also in the United States provide some tools with which to manage the boundaries between patents and copyrights in the US and between copyright and design rights in the UK. And so I wanted to kind of just mention a couple of them. So unlike the UK, the United States basically says that if there is any functionality of a pictorial sculpture on graphic work that cannot be separated from the kind of the way in which it either displays impurence or conveys information, right? If there's any integration of functionality and expression in a pictorial sculptural graphic work, the whole thing is unprotectable by copyright law. So furniture, all kinds of designs that actually might be eligible for copyright protection in the UK, those things are ineligible for completely ineligible for protection under US copyright law. And we have a separate statutory provision that basically says that the copyright and a drawing of a functional design doesn't mean you get copyright protection for the functional design. The drawing is protected as a drawing and the functional design can be re-implemented. So there's an old case in which somebody drew a design for a parachute. The United States government started making parachutes like that. The guy sued for copyright infringement and the court said no. It's basically under the kind of the penumbra of the Baker v. Selden case, it was ineligible for protection. In 1976, Congress added a particular provision intending, too, for that provision to help to manage the boundary between copyrights and patents in that section 102B, which reads that in no case does copyright protection for an original work of authorship extend to any idea, procedure, process system, method of operation, concept, principle, or discovery, regardless of how it's embodied in the work. Now, I've highlighted, I bolded here, the words that I'm interested in because I don't know what those words mean in relation to computer programs, although I do know that Congress added those words to the statute for the specific purpose of making sure that the scope of copyright protection in computer programs would not be too broad. So that's what they said at the time that they adopted it. There's also a concept, again, not when I think that you find in UK law that it's called the merger doctrine, where the notion is that if an idea can be expressed in essentially only one or a very small number of ways, it's not eligible for copyright protection because the concern about not protecting ideas is so strong that even if there's some expression that's intermingled with it, you don't protect it at all. And there's a question, a very big debate, and it's core to what's going on in the Oracle v. Google case about whether or not merger means only that when you're first trying to create something, is there any other way to do it? If there's no other way to do it, then there's merger. Or is it the case that something could be creative at the time that you made it, but constrain the design decisions of subsequent people in a way that makes it capable of merger in the long run? I'll give you examples of that as we move through this, but that's one of the big debate issues. Now, again, you know these rules better than I do, but I think that it's interesting that the UK, too, had a kind of categorical exclusivity between designs and copyrights for a period of time. Then it shifted to creating a kind of defense for certain kinds of designs, but as it turned out, the unregistered functional designs ended up being eligible for full copyright protection for a time and so the screws and exhaust pipes were protected by copyright law and that just didn't seem like an artistic or literary work that we wanted to use copyright for. So there were a set of reforms and over lunch today, Lionel tells me that Section 52 is slated at some point for extinction, but too bad for that. But I'm just noticing that there are some similarities between the tools that are part of the UK tradition and the tools in the US tradition. I think Section 51, for example, is pretty similar to Section 113b that I talked about a little while ago. Okay, so let's get back to computer programs for a minute or three. So, yeah, source code, listings of computer programs, literary works, no problem. That's one of the reasons why they initially seemed to be eligible for copyright, although back in 1964 the US Copyright Office initially issued registration certificates under what it called its rule of doubt, like, I'm not really sure that computer programs are really copyrightable, but I'll give you a certificate and if you want to sue somebody for infringement, you can, but we're not saying it's really copyrightable. That's kind of weird, but that's one of the things that they thought. And why did they think that? It was because in executable form, computer programs are machine processes, they're virtual machines, they're machines that are constructed in text, and so it seemed like that's not eligible for copyright protection, and it's notwithstanding that reservation eventually Congress decided to treat the executable forms of programs the same way as the printed source code versions, and of course, as you all know, the Software Directive and the Trips Agreement embody the same norm that copyright is available to computer programs as literary works. But again, the question about what these words of exclusion mean is kind of up for debate, just like the provisions in the Software Directive about to the extent that interfaces, logic, and algorithms are ideas or principles are not protected by copyright law either. Well, what does that exactly mean? So it's a similar kind of puzzle. And the question about whether or not there's a rule for pads to play in protecting software is kind of up in the air too. Now, pads have had a really complicated pattern in the United States, and the time, the cases that I'm going to be talking about in a minute were being decided from the 60s to the kind of mid to late 80s, and the general view was that you couldn't get patent protection for computer programs or elements of computer programs such as algorithms that might be innovative but were ineligible on one or another grounds, and a guy named Benson came up with an algorithm for transforming binary coded decimals into a binary form. Supreme Court in 1972 said that's not protectable by patent law. That's basically a mental process or an abstract idea. And so there was this kind of like no patents available for a very long time. But from the early 19 or late 1980s through the mid 2000s, around the Federal Circuit essentially granting ever broader scope of protection for computer programs, essentially everything under the sun made by humans was considered patentable subject matter for that period of time. Hundreds of thousands of software patents were issued during that period of time. And one of the things that's happened since the rise of patents in that period of time is that courts in the second wave of decisions said, oh copyright doesn't have to be that broad because there's a role for patent to play in protecting certain aspects of software. And so here are the main cases. So again in the era before the rise of software patents, you had three main cases that took a pretty broad perspective about copyright protection for computer programs. So in the Apple versus Franklin case, the court decided that exact copying of operating system program code was an infringement. That was actually not so bad, but one of the things that the court said in that case is that you may want to become compatible with another machine or with another operating system or with another program, but that's just a competitive objective. It doesn't completely irrelevant to the scope of copyright protection and computer programs. And so seem to be signalling that if you copy the interface of a program to inter-operate that that might be copyright infringement too. The Whalen versus Jaslow case was the kind of high watermark of protection, at least until the Oracle v. Google decision by the Federal Circuit, where the guy who was a dental laboratory self-taught programmer re-implemented a program that he'd hired somebody else to do because he didn't like her, I think, anymore. And he wanted to write it for an IBM PC. Anyway, he copied some of the structure sequence and organization of the internal design of the program and some subroutines. And the court decided that copying of SSO, wether it's the way subroutines operate or whether it's the internal design, what file structures, data structures, and the like, that that was copyright infringement, unless there's only one way to perform a function, if there's only one way to perform it, then that's unprotectable. That's a case of merger of ID and expression. But Whalen took the view that the only unprotectable ID in a program was its general purpose or function. So all of a sudden computer programs are going to have way more copyright protection than any other kind of work in the whole universe, and that sort of seemed a little wrong to some of us who were commenting at the time. The Lotus who picked back decision was, Paperback was a company that made a kind of, what they called a work-alike program to Lotus's 1, 2, 3 spreadsheet program and copied the command hierarchy of the spreadsheet program was held as an infringer, and all three of these cases, the courts basically said, look, this provision about no copyright protection for process procedure system method of operation, it's just nothing more than a restatement of the ID expression distinction, and we can't take it literally because if we do, then computer programs wouldn't be protectable at all, and so let's just say that it doesn't mean anything. Now there was a case in 1992 that turned the tide called Computer Associates v Alltie by the Second Circuit Court of Appeals, and it turned the tide in a number of ways. So one was it recognized the functionality of computer programs and recognized the functionality of computer programs meant that there should be a limited scope of protection for the internal design elements of computer programs. It recognized that Section 102b was about trying to stop copyright protection for functionality of programs and said that, you know, if somebody had adopted, for example, a kind of a highly efficient way of doing something in a computer program, we shouldn't use copyright protection to allow that person to fence off that most efficient way to do something. So on a kind of merger-like analysis, the efficient design then is available for re-implementation by other companies and court recognized that there were constraints that might be imposed on subsequent programmers either by the hardware with which they were trying to work or the software with which they had to interoperate and so compatibility considerations under computer associates of the Altai meant that there was not as much copy... not as broad a scope of protection and the court basically threw out this notion of SSO, a structure, sequence and organization saying it's just not helpful. Yes, some non-literal elements of computer programs may be protectable by copyright law but SSO is basically not a very useful way of trying to do it. The court also recognized that copyright for software was kind of weird, it was really hard to apply and so one of the things that it also said is maybe some of the aspects of programs should be protected by patents, not by copyrights or if there was a gap in protection maybe Congress should do something about it rather than trying to stretch copyright to reach something that it wasn't intended to do. The Seigabee Athletic case was a Ninth Circuit Court of Appeals case out in California that was the first case to recognize that reverse engineering of a computer program for a purpose such as getting access to the information necessary to write a program that would interoperate with an existing platform was not copyright infringement. We call it fair use. You have a decompilation privilege. It allows you to do something similar but importantly the court of appeals in that decision talked about the SEGA interface that Accolade essentially extracted from the code that it reverse engineered as an unprotectable procedure under section 102B and moreover said, you are trying to get patent-like protection through copyright and we're not going to let you do that. Congress chose not to protect these kind of aspects of programs and so there. Lotus v. Borland is the last in the series of cases that are part of the second wave. Borland-like paperback made a program that allowed people who had constructed macros, little programs of common sequences of functions that they did on a regular basis. You could like do one macro stroke and have that series of things essentially executed and Borland had both what they called a native interface so if you had never used Lotus 123 we think this is a better user interface you should use that but if you've been using Lotus and you've constructed macros and it will help you decide to come over to the Borland better program if you can re-execute your macros so the emulation interface then presented exactly the same commands in exactly the same order in Lotus 123 and the reason for that was because you can't execute the macros unless the thing is in exactly the same order. Now Lotus basically said, hey, the command hierarchy is part of the structure sequence and organization of the program and under Whalen v. Jazzlow in cases like that it's a protectable part of the program and there's no merger of idea and expression because hey, when we set out Lotus initially we could organize the commands whatever way we wanted so the fact that Borland had its own native interface shows you don't have to use the same stuff Borland's argument was that the commands had to be in exactly that order for the macros to execute and therefore there was a kind of merger of idea and expression or there was a kind of system the macros were a system so that things had to be in this particular order so the fact that it might have had a choice initially is irrelevant because the people who have invested in creating macros want to be able to use those macros you shouldn't have to learn how to write macros all over again part of the macros system and therefore ineligible for protection under section 102b First Circuit Court of Appeals ended up deciding that the command hierarchy of Lotus 123 was a method of operating a spreadsheet program to fulfill some functions and so it was unprotectable under section 102b kind of likened it to the buttons on a VCR machine just more buttons but essentially the same kind of idea and so Lotus appealed that case to the Supreme Court after hearing oral argument on the case the court split 4-4 and so kind of left the question about whether a command structure of a computer program is part of what's protectable by copyright law or not one of the things that Borland argued during that case was a command hierarchy of macros systems are patent subject matter they found patents at an issue for some of these things and so made the argument that it was ineligible under Baker v Selden for copyright protection now a very similar case actually came up in the UK High Court got referred to the Court of Justice of the European Union obviously the legal context is a little bit different but one of the main defenses here was that users of the SAAS statistical analysis program had written these little mini programs like the macros these were called Scripps and they were like common sequences of statistical methods that you went to do on a regular basis and so people wanted to be able to use their Scripps if they were going to port them if they were going to be able to use another program they wanted to like take the investment that they made here and be able to run them over here so WPL essentially re-implemented the functionality of the SAAS statistical analysis program and the court decided that the functional behavior of a program that is to say the fact that you're emulating a functionality of another program that's not copyright infringement the Scripps had a little language just like the macro language and the Scripps language was not protectable by copyright law in the EU and neither were the data formats all of them were unprotectable ideas under the software directive and Mr Justice Arnold referred these questions this and other questions to the CJEU which basically agreed with him on those issues and I have the honor of having co-authored a paper with Professor Cornish on this particular case the result actually comes out to be almost the same the legal issues are very similar to the Lotusby Borland ruling and that was actually one of the arguments we were making let's try to have a compatible legal idea one thing that's actually worth mentioning is that the Court of Justice of the European Union actually invoked the TRIPS agreement which talked about the unprotectability of methods of operation and procedures as kind of a part of the basis for this and so there is a bit of harmonization there at least in terms of TRIPS kind of giving us more of a connection to each other back to the Oracle v. Google case so you find the trial court having relying on the ALTI and Sega Viacoli decisions for the propositions that trying to promote interoperability is a good thing and Google said one of the reasons that we adopted these 37 Java API packages was so that people could interoperate the programs that have been written in Java could run on the Android platform and things that have been written on the Android platform could run on other Java systems and the Lotusby Borland case was important because the Court said that hey, these are commands just like in the Lotusby Borland case and they're either a method of operation or a system and therefore unprotectable by copyright law alternatively the Court said look, you have to use exactly the same names for these functions in order for the Java APIs to work and so that's actually an alternative basis for the decision the Court rejected the idea that this SSO structure sequence and organization was a way to conceptualize what computer program elements might be protected and said there's a danger of giving too broad a protection to computer programs because that may make it a little bit too much like pass now the Federal Circuit in its decision reversing the trial judge said hey, at the time that Sun was first designing the Java API packages they had lots of choices they were really creative and so there's no merger of idea and expression because at the time that they set out to do this in the first place they were not under any kind of design constraints based on pre-existing systems the Court rejected out of hand the notion that there's a compatibility exception to copyright and kind of puts a little dig in also because instead of using all 166 Java API packages Google only used 37 of them now those 37 were selected because the testimony says because these would be appropriate to the use of Java on a mobile device one thing to know is that Java was originally kind of created with the idea that it was going to run on PCs and on servers not on mobile devices and so there was kind of an adaptation going on because of that but the Court of Appeals Google intentionally designed this to be incompatible with Java and what they were really trying to do was take a free ride on the investment that Sun and Oracle had made in getting Java adopted there are now 6.5 million Java programmers who use these API packages on a regular basis and they know the commands they want to be able to use them and this is free riding by the bad guys is Oracle's position on that and as in the earlier cases the Court saw 102B as being nothing but a restatement of the idea expression distinction so Google has asked for Supreme Court review of this case and it's a case of exceptional importance for the software industry there's a conflict among the circuits about what section 102B means there's a split in the circuits about whether there's a compatibility exception to copyright there's a difference also among the circuits on what the merger doctrine means and they also make this argument that APIs are pat and not subject copyright subject matter and one of the reasons the Court so when the Supreme Court takes cases they often take them when there's a circuit split here you have a situation where 20 years ago the Supreme Court split 4-4 on exactly the same legal issue and guess what the circuit splits have gotten deeper because the Court wasn't able to resolve it back then so the question is what to do so Oracle of course says no, don't take it and is very happy with the Court of Appeals decision I think the patent subject matter argument is the weakest part of Google's argument up to date partly because I showed you one of the Java API specifications it just doesn't look patentable it looks copyright predictable of anything in an article I wrote with a couple of other people back in 1994 a manifesto on legal protection of computer programs we identified that a lot of the value in computer programs really comes from them being industrial compilations of applied know-how and I see something like an API or an API specification as one example of that and we made an argument that you know you're going to either under or over protect these kinds of industrial compilations of applied know-how if you use copyright so if Suey Jenner's form of protection would probably be a better idea but that's water under the bridge but more importantly the Supreme Court has not found categorical exclusivity to be appealing since the Baker v. Selden case the patents might have issued an error they may be at a different level of abstraction and guess what? it turns out the patents are eroding now so last June in the Alice v. CLS bank case the Supreme Court ruled that a computer implemented method of settling financial transaction risk was unpatentable subject matter it was too abstract an idea to be eligible for protection since then hundreds of software patents have been thrown into question and dozens of them have actually been repudiated as ineligible for patent protection altogether now one of the things that's kind of interesting to me is that a lot of these things are part of the structure sequence and organization programs so wouldn't it be like completely weird if they're too abstract for patenting and yet because they're SSO under the interpretation given in the federal circuit they'd be eligible for copyright protection even though not for patent protection this drives me crazy I don't think that's what Congress had in mind but you never know part of what may have motivated the federal circuit and the ruling that they made in the Oracle v. Google case is Judge O'Malley who wrote the decision for the federal circuit actually was one of the judges of the federal circuit who would have upheld all three types of Alice's software patent claims and I think she's worried that if the Supreme Court is kind of pulling the rug out from under the patents for software by gum she's not going to let the copyright get gutted and so she's basically protecting copyright for all she's worth now in this particular decision but of course one of the questions that's interesting here is whether if in fact patents recede should the scope of copyright expand because now they're less protected okay so I promised you actually some pragmatic approaches so let me give you two one is a legal one and one is a kind of realistic one so the legal one is basically we've been able in a lot in America anyway been able in a number of kinds of circumstances to separate functionality and expression through various doctrinal and statutory mechanisms and we should be able to do that with respect to computer programs too so Supreme Court did it with Seldin's forms and the explanation, Stein's lamp base as a statuette there are a bunch of ways in which it's like architectural works are protected but the scope of copyright protection and architectural work doesn't extend to the wiring and the plumbing and the things that are basically conventionally functional design elements of those and I think in same measure we should be able to say that program code is clearly expression but things like algorithms interfaces should be unprotectable by copyright land part of what's happened here is that that literary work metaphor that we use to describe computer programs it's sent us off in some weird directions we recognize that functionality was present as in the design area that UK has been dealing with some overlap then we wouldn't be quite as confused so what's the pragmatic approach number two so it's to recognize that copyright is doing a lot of good work in protecting computer programs so mostly it's the code and those expressive parts of a program such as video game graphics and sound that may be just like a regular audio visual work and we should keep our minds open to the possibility that some non-literal elements of programs should be eligible for copyright protection but let's scrutinize them a little more carefully than we've done so far because many of them perhaps most are excludable under section 102B and languages, behavior, interfaces, logic and algorithms really should be beyond the scope of copyright protection in programs but more importantly most of the internal design elements of computer programs don't need copyright protection because they are adequately protected through trade secrecy law and through licensing for the most part it's really difficult and expensive to reverse engineer a program to extract non-literal elements of programs you don't do it unless you really have to and more importantly if you ask software developers what's really important to you is the competitive advantage in the marketplace and I've done a survey of software entrepreneurs and I know the answer to this question first mover, advantage and complementary assets are way more important than copyright, trademark, trade secrecy let alone patents very few software companies in the United States actually get or even apply for patents they don't really think that they're useful to them in attaining competitive advantage let's not worry so much about the receding of patents that's actually not a bad thing for the software and copyright is important by comparison with patents but it's not really that important as these things go except for the protection of code so in conclusion we sort of see that patents and copyrights have a long history of trying to like essentially navigate ways in which to protect appropriate things not get in each other's way there are doctrinal ways to get out of this I think that APIs and command structures are examples of things that probably shouldn't be protected by either copyright or by patent although they can often be protected as trade secrets but we can get lost in the worst possible but we can get lost in this metaphysical model if we try to separate functionality and expression in programs it will make you just completely crazy if you try to do that so let's not drive each other crazy let's just develop a couple of simple rules and as a pragmatic matter just basically relax a little bit software industry is doing real good but Google should win this case www.google.com www.google.com