|
APIs are to functional, working code as the index of a cookbook is to the recipes in said cookbook, in case that helps anyone.lamentable dustman posted:The scummy part on Google's part was to advertise this new language, which I'm not sure even has a name, as Java to developers. It worked well for them as they got a huge programmer base from the get go and was one of the main reasons it became the dominate mobile OS. In the end though this fork means Android is stuck in the Java 1.6 era and hasn't got any of the new features in Java 7 or 8. It is the Java language (if not the runtime and full compliment of libraries) unless it can't handle some subset of the language but that's not the case as far as I can tell. Condiv posted:That being said, google's wouldn't because they copied only enough of the apis to be able to allow dalvik to benefit from preexisting java programmers and libraries without really allowing the inverse to be true. They did vastly increase the user base of and interest in (that one crappy old version of) the language. Kalman posted:"The IP system is outdated" is a stupid opinion that's completely ignorant of historical context, though. "The IP system has no applicability to software out in the real world" Better?
|
#
?
May 29, 2015 07:24
|
|
|
|
| # ? May 9, 2024 19:20 |
|
|
Munkeymon posted:"The IP system has no applicability to software out in the real world" Better? I mean, it at least makes an actual assertion, but it's still loving dumb. Nothing about software is special; IP is applicable there (and should be). There is innovation in software that should be protected, and there's old poo poo that shouldn't be. So, actually, IP seems pretty applicable to me.
|
|
#
?
May 29, 2015 08:00
|
|
|
Munkeymon posted:APIs are to functional, working code as the index of a cookbook is to the recipes in said cookbook, in case that helps anyone. Vastly is probably a stretch, java was and still is extremely popular when it comes to business software. According to TIOBE, java's popularity has never fallen below #2.
|
|
#
?
May 29, 2015 09:25
|
|
|
Designing a good API is very hard, and in my experience, in fact it's usually much harder to design a good, clean, and performant API than it is to implement it. I don't like some of the really nasty implications that fall out of the fact APIs are copyrightable -- but I find it very hard to argue against the fact that if this two sentence creative work hosted on somethingawfuldotcom is worthy of copyright, an API I spend considerably more time designing is not. EDIT : Had a response to Munkeymon here, but really, if your position is so far off the radar that Richard Stallman look like a radical conservative it's not really worth arguing about. Chuu fucked around with this message at 10:18 on May 29, 2015 |
|
#
?
May 29, 2015 10:02
|
|
|
Chuu posted:Designing a good API is very hard, and in my experience, in fact it's usually much harder to design a good, clean, and performant API than it is to implement it. I don't like some of the really nasty implications that fall out of the fact APIs are copyrightable -- but I find it very hard to argue against the fact that if this two sentence creative work hosted on somethingawfuldotcom is worthy of copyright, an API I spend considerably more time designing is not. Copyright cares nothing for the quality or length of the work in question, so "it's harder to make good, so it should be protected" is a bit besides the point. My hamster/vampire fanfic is two pages and terrible by any reasonable standard and was dead easy to write, but in terms of copyright its worth sits side by side with Lord of the Rings.
|
|
#
?
May 29, 2015 10:16
|
|
|
GhostBoy posted:Copyright cares nothing for the quality or length of the work in question, so "it's harder to make good, so it should be protected" is a bit besides the point. My hamster/vampire fanfic is two pages and terrible by any reasonable standard and was dead easy to write, but in terms of copyright its worth sits side by side with Lord of the Rings. Yeah, I think I was veering off the point there. In the end, at least the way I understand it, the only real test for copyright is "is it an original work?" I can't find any line of argument where I could answer "no" for an API.
|
|
#
?
May 29, 2015 10:24
|
|
|
Chuu posted:Yeah, I think I was veering off the point there. In the end, at least the way I understand it, the only real test for copyright is "is it an original work?" I can't find any line of argument where I could answer "no" for an API. Software and IP always becomes muddled, because when it comes to the actual code, the line where an algorithm stops and the specific implementation of said algorithm begins is tough to draw. APIs are less complex, leaving out the exact legalese of it that I know nothing of. Describing a thing is not making a thing. It may be a very good and detailed description of a thing, but I should not get to prevent others from making a thing that fits my description of the thing, just because I made that description. Putting copyright on an API gives the describer power to control who makes things that they describe, because they can claim they used their description without permission. From what I can tell, and realising that copyright is not patent, it follows some of the same logic that dictates that you can't patent an idea, even a very original idea. At issue is ultimately that copyright law lacks the nuance to handle this area well. You could maybe do something that works, where innovators of great APIs can retain their copyright so that others cannot claim they made it, without also granting them a 70+ year embargo power on anyone using that API, but it requires rewriting the law. As it is now, copyright is a blunt sledgehammer of a tool. GhostBoy fucked around with this message at 11:20 on May 29, 2015 |
|
#
?
May 29, 2015 10:43
|
|
|
GhostBoy posted:Describing a thing is not making a thing. I can't really offer a proper response in the time I have, but I will say one of the reasons some love programming is describing a thing *is* making a thing.
|
|
#
?
May 29, 2015 12:57
|
|
|
That API that you spent considerable time designing should be patentable not copyrightable which is the core of the problem
|
|
#
?
May 29, 2015 13:50
|
|
|
Deceptive Thinker posted:That API that you spent considerable time designing should be patentable not copyrightable which is the core of the problem It absolutely should not be patentable
|
|
#
?
May 29, 2015 13:51
|
|
|
Chuu posted:I can't really offer a proper response in the time I have, but I will say one of the reasons some love programming is describing a thing *is* making a thing. In fact, there are entire programming languages designed around this idea. The concept is called declarative programming, and programs written in such languages are still copyrightable. Deceptive Thinker posted:That API that you spent considerable time designing should be patentable not copyrightable which is the core of the problem TBH, patentable apis would be 10x worse than copyrightable apis, because with copyright you at least have fair-use, and with the current patent system it is far far far easier to get a patent to cover a wide range of implementations.
|
|
#
?
May 29, 2015 13:54
|
|
|
I have no idea what you people are taking about anymore. Maybe this is why SCOTUs didn't take the case 
|
|
#
?
May 29, 2015 15:07
|
|
|
WhiskeyJuvenile posted:It absolutely should not be patentable Condiv posted:TBH, patentable apis would be 10x worse than copyrightable apis, because with copyright you at least have fair-use, and with the current patent system it is far far far easier to get a patent to cover a wide range of implementations. You both missed the point of my post Functional elements of something should be patentable, and absolutely NOT copyrightable The problem is the system for both copyright and patent (especially for software) are so broken that poo poo like this happens
|
|
#
?
May 29, 2015 15:13
|
|
|
I'm OK with APIs being patentable in the sense that one would only patent the means rather than the ends, but the cynical side of me knows they'd be as broad as design patents and we'd see a decade of billion-dollar litigation over who had the rights to "accessing a remote or local program through custom software designed to retrieve and return specific information on request"
|
|
#
?
May 29, 2015 15:37
|
|
|
The basic problem with copyrighting code is that functionality is not copyrightable, expression is. This is a tough line to draw in other circumstances about when something is too functional to be copyrightable, and if it is copyrightable what part of it is copyrightable. The amount of effort you put in is irrelevant: architectural drawings may involve shittons of effort but be pure uncopyrightable function while literal poo poo smeared on a canvas is copyrightable. We've essentially decided that normal code is copyrightable (as a 'literary work', hilariously) but the issue with APIs is if by their very nature they are too functional to be copyrightable. They definitely shouldn't be patentable.
|
|
#
?
May 29, 2015 15:43
|
|
|
Kalman posted:I mean, it at least makes an actual assertion, but it's still loving dumb. Nothing about software is special; IP is applicable there (and should be). There is innovation in software that should be protected, and there's old poo poo that shouldn't be. So, actually, IP seems pretty applicable to me. That's fair but I wasn't expressing myself accurately. What I meant to say is "The patent system has no applicability to software as it is created and used out in the real world" although I also think that there appear to be some hilariously outdated assumptions underpinning IP law that need to be addressed. See the outcome of Capitol Records, LLC v. ReDigi, Inc. for an example.
|
|
#
?
May 29, 2015 16:09
|
|
|
evilweasel posted:They definitely shouldn't be patentable. The Java API, and most APIs out there wouldn't past a novelty or obviousness test anyway But if a specific API has a novel element to it, why not?
|
|
#
?
May 29, 2015 16:15
|
|
|
Munkeymon posted:That's fair but I wasn't expressing myself accurately. What I meant to say is "The patent system has no applicability to software as it is created and used out in the real world" although I also think that there appear to be some hilariously outdated assumptions underpinning IP law that need to be addressed. See the outcome of Capitol Records, LLC v. ReDigi, Inc. for an example. Ah, so you were just making a completely false statement. The patent system absolutely has applicability to software as it's created and used in the real world. Lots of software isn't novel, but some is, and that software shouldn't be treated differently because coders are idiots about how IP works.
|
|
#
?
May 29, 2015 16:20
|
|
|
I think the best solution is to come up with a new class of protection for software and code. Patents and copyright don't really apply and instead of jamming a square peg in a round hole a new solution would make sense. Exactly what that new class should be is a huge mess.
|
|
#
?
May 29, 2015 16:26
|
|
|
evilweasel posted:The basic problem with copyrighting code is that functionality is not copyrightable, expression is. This is a tough line to draw in other circumstances about when something is too functional to be copyrightable, and if it is copyrightable what part of it is copyrightable. The amount of effort you put in is irrelevant: architectural drawings may involve shittons of effort but be pure uncopyrightable function while literal poo poo smeared on a canvas is copyrightable. the particular arrangement of java packages is not inherently functional: Java would work just as well were all classes placed in different packages
|
|
#
?
May 29, 2015 16:27
|
|
|
Kalman posted:Ah, so you were just making a completely false statement. Incorrect on both counts. Nobody looks up a patent when trying to figure out how to write software because they're never sufficient technical documentation and generally written in impenetrable legalese (E: the system is also full of obvious garbage which is painful to sift through because of the preceding issues), so patents as a means to communicate ideas between developers who are supposed to need them are useless. That's a fixable problem, but I haven't heard of any proposals that would actually fix it. Algorithms are novel. Software that implements an algorithm is not. That's not a fixable problem unless you start to let people patent algorithms. Munkeymon fucked around with this message at 17:12 on May 29, 2015 |
|
#
?
May 29, 2015 17:01
|
|
|
WhiskeyJuvenile posted:the particular arrangement of java packages is not inherently functional: Java would work just as well were all classes placed in different packages With this line of argument as long as Google rearranged the API; it would no longer fall under copyright. If functionality isn't a factor. Just as a bad movie knock off may use a plot line charactor etc from a blockbuster and not get sued into oblivion. Patent is functionality, copyright is expression. Oracale could I suppose copyright every single function in Java.... I don't k now what I'm talking about but I have learned a lot in the last few pages.
|
|
#
?
May 29, 2015 17:21
|
|
|
Munkeymon posted:Incorrect on both counts. Not saying you are wrong, but that's the exact opposite of what this has: http://www.nolo.com/legal-encyclopedia/qualifying-patent-faq-29120-3.html quote:Even though you can't get a patent on a mathematical formula per se, you may be able to get protection for a specific application of a formula. Thus, software may qualify for a patent if the patent application produces a useful, concrete, and tangible result. Honestly like someone else said we are trying to shoehorn ancient laws to deal with modern technology. Kalman posted:The patent system absolutely has applicability to software as it's created and used in the real world. Lots of software isn't novel, but some is, and that software shouldn't be treated differently because coders are idiots about how IP works. Is a new first person shooter novel? How about the first one to implement a bullet-time style gameplay? The whole thing is just incredibly handwavey.
|
|
#
?
May 29, 2015 17:21
|
|
|
Alice made that a lot harder
|
|
#
?
May 29, 2015 17:25
|
|
|
BlueBlazer posted:With this line of argument as long as Google rearranged the API; it would no longer fall under copyright. If functionality isn't a factor. no, that makes it a derivative work quote:I don't k now what I'm talking about but I have learned a lot in the last few pages. no kidding
|
|
#
?
May 29, 2015 17:29
|
|
|
honestly patentable subject matter when it comes to software should have a lot to do with semiotics, but good luck getting the supreme court to understand that.
|
|
#
?
May 29, 2015 17:30
|
|
|
Xae posted:I think the best solution is to come up with a new class of protection for software and code. I see this a lot. Why do you think patents don't apply? Are there no new applications of ideas in software? As to copyright, a piece of software is a sequence of written words that exhibit some modicum of creativity, that exhibit characteristics of most copyrighted works (hard to create 1, easy to create 2+). Why is copyright the wrong structure for protecting it? WhiskeyJuvenile posted:honestly patentable subject matter when it comes to software should have a lot to do with semiotics, but good luck getting the supreme court to understand that. Software should be per se patentable but should all be treated as mpf and limited to algorithm disclosed in the specification and structural equivalents thereof. tsa posted:Is a new first person shooter novel? How about the first one to implement a bullet-time style gameplay? The whole thing is just incredibly handwavey. If they aren't novel, they aren't patentable. I fail to see the problem.
|
|
#
?
May 29, 2015 17:39
|
|
|
It seems like the real question is "should companies be able to make their own implementations of another company's API?" If you care at all about systems being interoperable, the answer is clearly "yes". The whole purpose of an API is to define a set of common names and data types that you program your software against. If the names of standard functions and their return types aren't known (and constant), programmers can't write code and know that it will work everywhere. What kinda confounds this case, though, is that Android doesn't actually need to be interoperable with other platforms Java runs on. Not only can you not run Android apps on, say, Windows (outside an emulator) because the Android libraries aren't implemented on that platform, who would want to run an android app on a desktop computer? Google reimplemented Java on android simply to take advantage of the large existing programmer base, and possibly because they couldn't get any of their own attempts at a language ready for prime time (some people like Go, but I think it's a pain to use). You could argue that using Java allows them to keep naughty programmers' fingers out of the system and let them manage resources better, but there's no reason they couldn't have defined their own, novel API for, say, C that did all that without exposing too much system control to application programmers. Of course, we can't actually know what google's motivations were, but I think it's all fuzzy enough that I can see why the court doesn't want to deal with it.
|
|
#
?
May 29, 2015 17:48
|
|
|
Kalman posted:Why do you think patents don't apply? Are there no new applications of ideas in software? So is this software API idea a useful process, machine, manufacture or composition of matter or a novel and useful improvement thereof?
|
|
#
?
May 29, 2015 17:49
|
|
|
Kalman posted:
a lot of the poo poo I see recites the algorithm in the claims which doesn't get you very far anyway, stuff like "aa crm storing software to write data in a certain order to make flash memory more reliable" should be patentable, and stuff like "a crm storing software to receive an origin point representing a first geographic location on a map, a destination point representing a second geographic location on said map, and calculating a shortest distance between the two based on a weighted graph" shouldn't
|
|
#
?
May 29, 2015 17:51
|
|
|
Kalman posted:Why do you think patents don't apply? Are there no new applications of ideas in software? Software is the expression of an idea in a particular language and as such only copyright should apply. Why should it be covered by both patent and copyright law? Isn't that a logical, if not legal, contradiction?
|
|
#
?
May 29, 2015 18:00
|
|
|
Munkeymon posted:Software is the expression of an idea in a particular language and as such only copyright should apply. Why should it be covered by both patent and copyright law? Isn't that a logical, if not legal, contradiction? This is not at all a contradiction, logical, legal, or otherwise. The patent is on the successful implementation of the idea. The copyright is on the code used to implement it. Few programs have anything patentable in them. Those that do, the patent applies even to different expressions of that method.
|
|
#
?
May 29, 2015 18:05
|
|
|
LeftistMuslimObama posted:What kinda confounds this case, though, is that Android doesn't actually need to be interoperable with other platforms Java runs on. Not only can you not run Android apps on, say, Windows (outside an emulator) because the Android libraries aren't implemented on that platform, who would want to run an android app on a desktop computer? Google reimplemented Java on android simply to take advantage of the large existing programmer base, and possibly because they couldn't get any of their own attempts at a language ready for prime time (some people like Go, but I think it's a pain to use). There is some interoperability, though: libraries can be written (and sold by ISVs!) that work on both standard Java platforms and Android.
|
|
#
?
May 29, 2015 18:09
|
|
|
WhiskeyJuvenile posted:a lot of the poo poo I see recites the algorithm in the claims which doesn't get you very far Algorithm in claims doesn't get past 101 but that's because everything about modern 101 should be taken out back and shot. Also, "Ntcrm". It's like you don't read your own office's memos. (#2 should be patentable subject matter to the extent the specific technique for doing it is the scope of the patent, and certainly is patentable in the context of using it to route a delivery truck.)
|
|
#
?
May 29, 2015 18:43
|
|
|
OddObserver posted:There is some interoperability, though: libraries can be written (and sold by ISVs!) that work on both standard Java platforms and Android. True. My primary point was that Google's reasons for reimplementing Java rather than just building a runtime around another language or making a language of their own are opaque enough that the court probably doesn't want to attempt to rule from the middle of a fog. It would be nice if they at least took the case to say "this ruling applies only to the specific case at hand and is not binding for future cases". I'd rather not have a ruling be definitive when much of the conflict stems from shady licensing behavior from both sides.
|
|
#
?
May 29, 2015 18:44
|
|
|
Kalman posted:Algorithm in claims doesn't get past 101 but that's because everything about modern 101 should be taken out back and shot. Also, "Ntcrm". It's like you don't read your own office's memos. here's an arbitrary patent: quote:1. A computer-implemented method for blurring location information for an entity, the method comprising: that's an algorithm quote:(#2 should be patentable subject matter to the extent the specific technique for doing it is the scope of the patent, and certainly is patentable in the context of using it to route a delivery truck.) shortest route is an abstract idea in graph theory and "do it on a computer" doesn't add significantly more; that the origin and endpoints represent geographic locations is nonfunctional descriptive matter
|
|
#
?
May 29, 2015 18:49
|
|
|
my get rich quick scheme is "A System and Method for Compiling and Executing Code from Printed Material" where the source code for an implementation is in the spec, then I sue google, freepatentsonline, etc., for infringement when they host a copy of the patent
|
|
#
?
May 29, 2015 18:50
|
|
|
WhiskeyJuvenile posted:here's an arbitrary patent: "Find the shortest path" is an abstract idea; "do it with a weighted graph" is a non-novel algorithm to implement the abstract idea. Don't try and make 101 handle 102's job, that's why we are where we are.
|
|
#
?
May 29, 2015 19:25
|
|
|
Kalman posted:"Find the shortest path" is an abstract idea; "do it with a weighted graph" is a non-novel algorithm to implement the abstract idea. Don't try and make 101 handle 102's job, that's why we are where we are. a weighted graph is literally math; it's like saying "do it with addition" e: not that Bilski was clear about how to handle a novel abstract idea, given the way they said "well this has been known for a long time, and it's a computer, so non-statutory"
|
|
#
?
May 29, 2015 19:25
|
|
|
|
| # ? May 9, 2024 19:20 |
|
|
There should be a separate court only for patents...some kind of different circuit. But here's the catch: nothing gets sent up to scotus.
|
|
#
?
May 29, 2015 19:45
|
|
