Slashdot Mirror
Morfik Patents AJAX Compiler
MikeyTheK writes "It appears that under the radar, the USPTO granted Morfik a patent for the "System and method for synthesizing object-oriented high-level code into browser-side javascript". Reading further, it appears that they have patented the compiling of high-level languages into AJAX apps. The high-level languages include "Ada, C, C++, C#, COBOL, ColdFusion, Common Lisp, Delphi, Fortran, Java, Object Pascal, SmallTalk, Visual Basic, and Visual Basic.NET". It would appear that the application date is September, 2005."
My first thought was, "Is Google Web Toolkit prior art or infringement?" After a bit of looking around, it seems this patent was filed on September 5, 2006 while GWT 1.0 was released in May 2006. Sorry Morfik, but your patent is invalid. (Thank God, too. This patent appears to be overreaching and far too broad. It could prevent an entire industry from developing.)
:)
All I can say is: where was your due diligence, Morfik? It doesn't make a whole lot of sense to spend time and money on filing a patent that will be useless to you after it's granted. The best they could do is scare a few Open Source projects into submission. Anyone with a vested interest in the technology is going to do the due diligence that Morfik didn't, and take the matter to court.
The only "out" they have available is to show evidence that they disclosed the inner workings of their JST product prior to GWT being released. In which case they might have protection from the "one year to file" rule. Maybe. Or maybe they're just trying to carry out this threat in a laughably oversimplified fashion. (They're lawyers must be telling them it won't work?) Go figure.
For those who are unaware of what GWT is, it's basically a toolkit that takes Java programs and converts them down to Javascript. By coding Java to the GWT toolkit*, you gain all the benefits of the Java compiler and type checking without sacrificing the ability to deploy on browsers that do not have Java installed. I'd rather code in Javascript myself, but it has its place.
Javascript + Nintendo DSi = DSiCade
Then maybe the patents aren't such a bad thing.
For all intensive purposes, "whom" is no longer a word. That begs the question, "who cares"?
I swear, I will set fire to the first bastard to reply to this article with "dur dur... I'm going to patent blank"!
April Fools Day is over.
Shouldn't any technology like this be immediately unpatentable? All this does is translate code in one Turing-complete language to another. Since this transformation is mathematically proven to be possible for all Turing-complete languages, this is merely an algorithm and should be unpatentable...
(I know, they'd approve a patent on cheese if you worded it as "a method for transforming milk and bacteria into edible food product.")
GWT does it and very well indeed. I've cross-compiled MD5 hash code to javascript and it works fine. I'd rather not port that manually since I already have perfectly good Java code for it.
-Dave
There is not a patent here. This is a published application. Filed in 2006, not 2005, with priority to 2005 through a provisional application, which almost gives it the same level of priority as if it were file in 2005. Someone wake me up WHEN this gets issued and not when it is just an application that quite possibly hasn't even been viewed yet. (USPTO has almost a 3 yr backlog in some arts.)
"Some days you just can't get rid of a bomb."
Yes. Next question?
Javascript + Nintendo DSi = DSiCade
Since when is ColdFusion a high-level language? Just because they use it in MySpace?
The definition of a compiler seems to differ from person to person, but the best one is something that reads a stream of input, converts it into a richer internal representation (usually a tree but it doesn't have to be), and writes out a different stream based on this internal representation.
Even here in 2007, some people still seem to think the only thing that can be called a compiler is something that takes source code and emits binary code, but that's just one specific special case. The same basic principles that GCC uses will be used by PovRAY to compile its scene language into an image, ignoring the raytracing part. (That is, setting up the internal representation of the scene is just like a compiler.) Compiling C# into IL uses the same basic techniques. Defining anything that uses standard compiler techniques as a compiler is the motivation for my preferred definition.
Given the long history of compilers, and the sheer profusion of them, I really don't think that compilers ought to be patentable anymore. Compiling Java into Javascript isn't a novel idea, it's "just" some engineering by somebody who understands compilers. (Which the recent "Wasabi" uproar over Joel on Software's posting proved is not all that many people, but still, it's simple once you see the tricks.) The only even remotely tricky part of such a compilation is if there's no easy way to get the syntax tree directly from the language parser, and that's still just engineering. There's definitely plenty of copyrightable stuff in such a compiler, but it'd take something very, very novel for it to be patentable.
(Note I'm writing this message as if I weren't entirely against software patents, which I am, at length. This is written from the putative point of view of the patent system; even then, compilers generally aren't that novel an idea. Saying "with a compiler!" is up there with "on the internet!" for novelty.)
The MSFT technology that the poster is referring to used to be marketed as Atlas but no longer. It is not in violation of this patent because it does not compile server side code into client side java script. Instead, it provides a lot of web controls that use AJAX instead of the traditional ASP.NET form post to interact with the server.
For all the handwaving and buzzwords in the application, it converts code written in language foo into language bar. That's a compiler. Now if they want to patent (copyright's possibly more appropriate?) their specific individual implementation of a compiler, then let them get on with it. If they want to patent compilers as a concept they can stuff it and I don't think they have a hope.
Only three things are certain; death, taxes, and apocryphal quotations - Ben Franklin.
Yeah, haha, Hemos, you almost had me! OMG PONIES!!!1!! Very clever April Fool's joke! And for the USPTO to be in on it... Wow. Best one yet!
For as smart as the people on Slashdot are, it continually surprises me to see just how bad people are missing the boat in terms of what a patent application such as this means. Seriously, discussion here might be elevated from ridiculous to meaningful if everybody would just keep in mind the following things:
1. The title of a patent does not describe the scope of what the patent will cover. In this example, they are not trying to patent every single "System and method for synthesizing object-oriented high-level code into browser-side javascript." They are trying to patent *one particular, specific system and method* for doing so. That is a big friggin difference. If you read the patent claims, you will see what the patent is actually covering. Summary: People on Slashdot inappropriately freak out when they see a patent title that looks broad and they really ought to know better by now.
2. This is a patent APPLICATION!!! Not an issued patent. When people try to get a patent, they make the claims incredibly broad at first. And then the examiner will make them narrow the claims as they negotiate what the patent actually will cover. In this example, I am going to hazard a guess that the examiner will probably make them throw out claims 1 through 4. And the examiner might let their broadest claim be claim 5. The people trying to get this patent KNOW that this is what will happen during this process--it is how things are done! I am astounded that the people on Slashdot have not picked up on this yet. Summary: Just because a patent application has incredibly broad claims, that does not mean 1) that the patent will issue that way, or 2) that the people applying for the application realize that what they are in their application claiming is too broad.
Get with it, people! I hope that reading this may take the discussion here from "Patents suxkzorz!!" to "Well, claim 1 looks way to broad because of $somePriorArt . . . but the method described in claim 5 . . . hmmm, I don't know that I've ever heard of anything that does it that way exactly."
Here's to hoping, anyway!
In this case, the application was filed September 5, 2006. Thus, the application filing date is September 5, 2006. However, the application claims priority to a provisional application filed September 6, 2005. Thus, the application has a priority date of September 6, 2005. This difference is important for 35 U.S.C. 102(e) purposes.
Under 35 U.S.C. 102(e), a prior filed U.S. patent or application can be used as "prior art" against a currently filed application. As an example, suppose Application A was filed on March 1, 2006 that discloses each and every limitation of claim 1 of the current application. Application A could then be used under 35 U.S.C. 102(e) as "prior art" to reject claim 1 because it has an earlier filing date (application date) than the current application. However, the Applicant of the current application could then rely on the priority date of the provisional application, which is September 6, 2005, to overcome the rejection. In this case, the Examiner would then have to go back to the provisional application to which the current application claims priority, and determine, whether, in fact, the provisional application discloses the subject-matter claimed by claim 1. As an additional comment to 35 U.S.C. 102(e), I briefly add that it is common for an Examiner to reject a claim in view of a reference that qualifies as "prior art" under 35 U.S.C. 102(e), where the reference is used with other art under 35 U.S.C. 103(a).
Hence, there is a difference between filing dates and priority dates, and these two dates need not always be the same. So, to say that these application has a filing date of September 5, 2006 would be correct, but then to say that this application has a priority date of September 6, 2005 would also be correct.
The opinion expressed herein does not represent the views of any government or private organization.
I guess now I am forced to to port all of my FORTRAN AJAX apps over to Perl
Yes. Next question?
Javascript + Nintendo DSi = DSiCade
This article is so 2006. :-)
http://blogs.zdnet.com/web2explorer/?p=196 explains the patent back in 2006. No, Morfik didn't copy it. They even hypothetically showed it to Google before GWT was released.
-- I'm the root of all that's evil, but you can call me cookie..
I worked for b-process.com in 2001 and we implemented exactly this in Java. I also have done similar stuff in PHP since then. It's nothing new.
All I have to think of is the current progress of the pypy interpreter. They have the ability to take something that is written in RPython and translate it into any backend to the annotator. This currently supports JS among other things. Check out http://codespeak.net/pypy/dist/pypy/doc/news.html
It doesn't matter whether they (or anybody) used it before then. The patent system does not reward invention, it rewards disclosure.
If you invent something and you don't disclose it fully and publicly, you lose your right to use your invention if someone else patents it. That's what the patent system is intended to accomplish.
Here's the process:
Requiring a Java VM is the new "You need VBRUN300.DLL". i.e. a big fecking "Stay Away!" sign.
Compiling to JavaScript means the process is this:
I read through the whole thing, that this is just a very bad patent. Not bad for people in general. It can't possibly stand up to scrutiny. Its just a bad patent at every level I can think of up to and including "badly written".
Its a bad patent because it is so "obvious". All they've done is to define JavaScript as as a p-code machine and created a (set of) compiler(s) for translating arbitrary programming languages into JavaScript "p-code". They have, in effect, attempted to patent compilers, albeit in a very limited way. Doing this was extrordinarily obvious, so much so that it wouldn't surprise me if there were hundreds of instances of prior art.
This may be one reason why the patent is so badly written. A truly general patent would make claims against all possible programming languages, but this one explicitly does not. Almost every claim is tacked down to a specific list of languages, and that list varies from claim to claim. Worse, none of the claims address some of the most widely used web programming langauges, including Perl, RUBY, Python, and PHP. My guess is that either (1) the patent was written by an amateur or (2) that these list variations reflect what they were able to find in their search of prior art. The mere existence of variations in the claims is probably evidence of the obviousness of this patent.
As for prior art, we've already seen claims of prior art in IBM (1996) and Microsoft (1998) products. That doesn't surprise me at all. I can recall discussing use of a JavaScript translator for an Ajax-like project I was in involved with (at IBM Research) in 1996. We didn't actually do it while I was on the project, but it was an option, and certainly not one that anyone would have believed was in any sense patentable. Appearance of such code in VisualAge during that same time frame would be anything but surprising. That is, for instance, the same time frame in which Mike Cowleshaw is translating REXX into Java p-code while retaining the interpretability of the REXX.
There are so many other examples of this kind of machine code translation, going all the way back to the original Fortran. I don't see a chance that this patent will hold up to scrutiny.
Davis http://davis.foulger.net
1.) This process is generally regarded as 'generating' rather than compiling. Compiling implies that something is transferred into a lower level language for speed and better runtime integration. Here it's the opposite. Thus: Generating. The servlet (or whatever) does it is generically refered to as 'generator'. Compiling is the wrong term.
2.) Prior Art. Tons of it. Laszlo and a bunch of other generators have been doing this for years. This patent won't even last a month. To many big players involved in RIA to let it pass. It's about as long lasting (and as silly) as the famous Gary Larson 'Chicken hung by a helium balloon floating into a pub full of Samurai'. Nothing new here, move on.
We suffer more in our imagination than in reality. - Seneca