decoration decoration
Stories

GROKLAW
When you want to know more...
decoration
For layout only
Home
Archives
Site Map
Search
About Groklaw
Awards
Legal Research
Timelines
ApplevSamsung
ApplevSamsung p.2
ArchiveExplorer
Autozone
Bilski
Cases
Cast: Lawyers
Comes v. MS
Contracts/Documents
Courts
DRM
Gordon v MS
GPL
Grokdoc
HTML How To
IPI v RH
IV v. Google
Legal Docs
Lodsys
MS Litigations
MSvB&N
News Picks
Novell v. MS
Novell-MS Deal
ODF/OOXML
OOXML Appeals
OraclevGoogle
Patents
ProjectMonterey
Psystar
Quote Database
Red Hat v SCO
Salus Book
SCEA v Hotz
SCO Appeals
SCO Bankruptcy
SCO Financials
SCO Overview
SCO v IBM
SCO v Novell
SCO:Soup2Nuts
SCOsource
Sean Daly
Software Patents
Switch to Linux
Transcripts
Unix Books
Your contributions keep Groklaw going.
To donate to Groklaw 2.0:

Groklaw Gear

Click here to send an email to the editor of this weblog.


Contact PJ

Click here to email PJ. You won't find me on Facebook Donate Paypal


User Functions

Username:

Password:

Don't have an account yet? Sign up as a New User

No Legal Advice

The information on Groklaw is not intended to constitute legal advice. While Mark is a lawyer and he has asked other lawyers and law students to contribute articles, all of these articles are offered to help educate, not to provide specific legal advice. They are not your lawyers.

Here's Groklaw's comments policy.


What's New

STORIES
No new stories

COMMENTS last 48 hrs
No new comments


Sponsors

Hosting:
hosted by ibiblio

On servers donated to ibiblio by AMD.

Webmaster
MS Gets 2 New Patents
Tuesday, January 04 2005 @ 09:11 AM EST

Here's a charming start to the day, an article on Microsoft getting two new patents, one on a compiler and the other on an "improved" system and method for editing software. Innovation marches on.

Obviously, if you are a programmer, you probably don't want to be reading patents, but for the rest of us, the compiler patent, U.S. Patent number 6,836,883, titled "Method and system for compiling multiple languages", is here and the editing patent, US Patent 6,836,884 titled "Method and system for editing software programs" is here.

The first is described as a method or "process involving the parsing and analyzing of more than one source language to produce a common language file that may then be read by the same or another front end system."

Well. The world is crying out for a compiler that can do that. The patent cites the Free Software Foundation's GCC in the prior art section. I'll say.

It also references: Gosling, J. et al, "The Java Language Environment. A White Paper", Sun Microsystems Computer Company, Oct. 1, 1995, pp. 1, 4-85, XP 002042922. The second patent appears to be a debugger/editor, which permits line-by-line "execution capabilities", whereby you can stop and edit and then continue where execution left off, and it provides "support for common language runtime environments".

The part that caught my eye in the compiler patent is this language:

"Although the invention has been described in language specific to structural features and/or methodological steps, it is to be understood that the invention defined in the appended claims is not necessarily limited to the specific features or steps described. As an example, other source languages may be included in the front end portion in combination with the first source and the common languages. Therefore, the specific features and steps are disclosed as preferred forms of implementing the claimed invention."

Way cool. We get to guess at all it can do. A patent on a secret. They used the same language in their patent application on .NET. (Mono Project's reaction.) Next they can apply for a patent on their patent, "a method or process of keeping the threat of patent infringement lawsuits against GNU/Linux alive, whereby one obtains patents that could be easily debunked by multitudinous prior art if all the facts were placed on the table, but by means of partial disclosure of what the method or process does, you can hint that there is just enough there to require a court to debunk your claims, all of which costs money, which, happily, the FOSS community doesn't have. This methodology creates anxiety in CEOs about using GPL software, without actually having to sue anyone, thus minimizing costs."

The article says that Microsoft was "unable to comment" on the patents. After all, I thought as I read that sentence, they might get deposed one day. Nothing more annoying than having to explain the why of a thing in a hostile deposition.

They were happy to tell the journalist, however, why they need to get patents:

"'The protection and licensing of intellectual property allows companies and individuals to obtain a return on investment, sustaining business and encouraging future rounds of research and investment in the IT industry,' the Microsoft spokesperson explained to internetnews.com."

Is that English? I recognize English words, but what does it mean? Let me parse that sentence out the way my beloved granny taught me as a child. Let's see. They need patents so they can get money from all of us to fund all the innovating they are doing. Oh. Patents ensure the freedom to innovate. Wait. Wasn't that the reason for integrating the browser into the operating system? And Windows Media Player? Freedom to innovate? I'm thinking they need to come up with a better excu^H^H^H^H reason. That one didn't play too well. They lost two antitrust lawsuits with that tagline. Besides, "you have to cross our palm with silver or we won't innovate" sounds crass, especially coming from such wealthy company. So, how about it? I bet you can come up with a better quotation than that one. If you were Microsoft on truth serum, how would you explain their need for these patents?

Here's my entry:

"We need these patents so we can kill off Free and Open Source software."

No? Too plain-spoken? Too direct? OK, let's see. I'll put it in my own editing/debugging tool, my brain, and after execution and some stop and go, out comes:

"The protection and licensing of intellectual property allows Microsoft to get buckets of dough to sustain our bully business without having to actually innovate, because it permits us, by cutting off the oxygen supply of our competition, to overprice our software, because you'll have no other practical choice but to buy and use our software, and all that money you give us will pay us for continued, ahem, research and investment and oppression in the IT industry. Patents enable us to kill off the babies in the innovative software industry, so to speak, particularly pesky Free and Open Source competition, which is threatening our monopoly business model (think Firefox). We'd like to remain still the one and only overbearing monopoly player, the only real game in town, thanks to the wonderful USPTO's patent system. We'll license our patented software under 'reasonable' terms that block all GPL software because of incompatible language, and presto! They are cut out of the loop, and money will flow in to us like a tsunami from all you suckers who didn't pay attention to Groklaw's early warning system that notified you that we'd use software patents to destroy free and open source software, ha ha ha ha ha. "

Perhaps you can tell that I've been reading transcripts of the US v. Microsoft antitrust trial for a a week? Of course, my debugging/editing tool is old and I neglected to file a patent on it, so it could be missing something innovative here. If so, I'm sure you will let me know. It's so old-fashioned, it could have sworn there was invalidating prior art all over the place. Do the examiners at the USPTO need a crash computer course? I'm sure we have volunteers galore willing to give them a helping hand.

Of course, to solve a problem, you need to recognize you have one. In an interview, the head of the USPTO says he thinks things are going along swimmingly there. The problem is, in his view, critics aren't experts in the law, and so they imagine problems that don't exist:

"The bottom line is there is a lot of misunderstandings about patents and that may lead to some of this rhetoric, but the bottom line is it's a pretty specialized area of the law and it takes experts to understand the patentability -- what the process is, how we determine whether to issue a patent or not, how the prior art defines the scope of coverage that an applicant is entitled to, and so on and so forth.

"And what you get is a lot of people that look at the title of the invention or some abstract or whatever and just come to these conclusions with respect to, 'how could the USPTO issue a patent' on such and such? But they haven't actually looked at the legal scope of coverage; they're more or less looking at a broad concept that's described in the patent as opposed to the claims that define the legal scope of the coverage."

Really? With all due respect, may I point out that Microsoft's patent on the FAT file system ended up being rejected, thanks to PubPat.org's requesting a reexamination procedure and presenting prior art? There are patent lawyers who see a problem with patents being issued that are not legitimate. And they do know how to read a patent claim and figure out the legal scope of coverage. I asked Dan Ravicher of Pubpat.org if he would care to comment on the new patents, and he said this: "These two patents specifically have effective filing dates in 2000 and 2001, respectively. So, anything that was publicly available a year before those dates would be prior art to them." A master of economical speech.

The FOSS community would appreciate it very much if the USPTO didn't issue patents so broadly that it is forced to go the trouble and expense of finding prior art after a patent is issued, although we will certainly be willing to continue to do so, if needed. But is there no way we could set up a process to help you find prior art *before* the invalid patent issues? Wouldn't that be more efficient? We're happy to give you a helping hand, if only you'd let us.

Speaking of patents, I wondered how it is that software patents are issued in Europe when supposedly patents on software are not permitted. I have found the answer, and it goes like this:

Europe already has uniform rules about what is patentable and what not.

They are laid down in the European Patent Convention of 1973. In Article 52, the Convention states that mathematical methods, intellectual methods, business methods, computer programs, presentation of information etc are not inventions in the sense of patent law. There is a systematic reason for that: in the legal tradition patents have been for concrete applications of natural science (technical inventions), whereas patents on software cover abstract ideas. When patents are applied to software, the result is such that instead of patenting a specific mousetrap, you patent any "means of trapping mammals" (or, for an actual example, any means of trapping data in an emulated environment).

In 1986 the European Patent Office (EPO) started granting patents that were directed to computer programs but presented in the guise of process claims, typically phrased as follows:

1. process for [using general-purpose computing equipment], characterised by ...

The patents granted on this basis were considered as hypothetical, because the program as such, when distributed on a disk or via the Internet, did not constitute a process and was not an invention. To resolve this ambiguity, the European Patent Office took the final step toward patentability of pure software in 1998 by authorising program claims, i.e. claims of the following form:

2. computer program, characterised by that [with its help a process according to claim 1 can be executed].

Hmmm... "computer program characterised by ..." Microsoft's two new patents, with a minor tweak, would seem to qualify, no? They both begin, "a method or process for". Call them instead "a computer program characterised by", and they are off and running. Of course, they are too clever to apply now in Europe. It might tip off everybody there as to the big picture. First they get software patents permitted, and *then* they apply for patents on a compiler and a debugger/editor.

EU? Are you paying attention? I hope so. Do you want GNU/Linux to be available to you or not? If you do, please read these patents, think about your backdoor patent arrangement and the whole software patents question, and connect the dots, before it is too late. We who love GNU/Linux software are counting on you.


  


MS Gets 2 New Patents | 396 comments | Create New Account
Comments belong to whoever posts them. Please notify us of inappropriate comments.
Corrections here...
Authored by: Anonymous on Tuesday, January 04 2005 @ 09:37 AM EST
...for ease of PJ's finding them

[ Reply to This | # ]

Compiling two languages into one executable, circa 1975
Authored by: OldPro on Tuesday, January 04 2005 @ 09:40 AM EST
Burroughs medium systems (b4x00) Cobol 68 compiler allowed command ENTER
SYMBOLIC, followed by Assembler language code, terminated by END SYMBOLIC.
Resulting compile produced common object code form COBOL language statements and
Assembler (SYMBOLIC) language statements which could be loaded by the Linker and
used to produce executable code.

[ Reply to This | # ]

OT here
Authored by: Anonymous on Tuesday, January 04 2005 @ 09:44 AM EST
cause sometimes the coolest things start out OT

[ Reply to This | # ]

On the EPO
Authored by: dyfet on Tuesday, January 04 2005 @ 09:45 AM EST
The EPO has been granting what are in effect pure software patents contrary to stated European law for a considerable time now. Those patents are issued to date by the EPO, and I recall there is now a backlog of 20,000 such patents issued, are currently unenforcable. Some who are more cynical might even say the EPO was acting in a knowningly fraudulent manner. The purpose of the EU council directive on patents seems to be to "harmonize" the existing EU legal code in such a way that those already issued EPO patents, and other forms of software patenting would suddenly become legally enforceable in Europe.

[ Reply to This | # ]

  • On the EPO - Authored by: MathFox on Tuesday, January 04 2005 @ 10:04 AM EST
  • On the EPO - Authored by: Paul Shirley on Tuesday, January 04 2005 @ 11:07 AM EST
    • On the EPO - Authored by: Anonymous on Tuesday, January 04 2005 @ 01:40 PM EST
  • On the EPO - Authored by: Anonymous on Friday, January 07 2005 @ 03:53 AM EST
You obviously can't see the bigger picture
Authored by: lifewish on Tuesday, January 04 2005 @ 09:50 AM EST
"We need these patents so we can kill off Free and Open Source
software."

Pah! Shame on you for suggesting such a thing. They're merely selectively
leveraging existing assets and contextual infrastructure to innovate dated
business paradigms, thus proactively maximising consumer value and shareholder
ROI.

By which I mean that they're exploiting a buggy patent system to kill off FOSS
software. Their customers and shareholders may reap some benefit; no-one else
will.

On second thoughts, that was too legible. I'll go read some sales literature and
try again later.

---
To err is human but, to really screw up, you need a computer.

[ Reply to This | # ]

The real problem
Authored by: Anonymous on Tuesday, January 04 2005 @ 09:51 AM EST

"The bottom line is there is a lot of misunderstandings about patents and that may lead to some of this rhetoric, but the bottom line is it's a pretty specialized area of the law and it takes experts to understand the patentability -- what the process is, how we determine whether to issue a patent or not, how the prior art defines the scope of coverage that an applicant is entitled to, and so on and so forth.

"And what you get is a lot of people that look at the title of the invention or some abstract or whatever and just come to these conclusions with respect to, 'how could the USPTO issue a patent' on such and such? But they haven't actually looked at the legal scope of coverage; they're more or less looking at a broad concept that's described in the patent as opposed to the claims that define the legal scope of the coverage."

That's right. The problem isn't with the patent system, it's just that we're all stupid.

[ Reply to This | # ]

Nasty claims
Authored by: MathFox on Tuesday, January 04 2005 @ 09:55 AM EST
The Microsoft patents contain some claim language like:
A computer readable medium having stored thereon a data structure ...
and
A computer-readable medium comprising computer-executable instructions ...
which, could even prohibit discussion of the topic matter of these patents via a computer network. Patents have strayed a long way from publicly disclosing inventions...

---
When people start to comment on the form of the message, it is a sign that they have problems to accept the truth of the message.

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 09:56 AM EST
I entered the computer industry in 1980 at (long forgotten) Wang Labs. We had a
debugger on our mini-computers that could handle all of our languages and our
assembler at the same time and do single stepping by high-level language
statement. Of course, ALL of our competitors (primarily DEC, DG, Prime) had
debuggers with similar capabilities.

Twenty five years later this is patentable?

Absolutely amazing!

[ Reply to This | # ]

Prior Art
Authored by: Anonymous on Tuesday, January 04 2005 @ 09:56 AM EST
There are many examples of this. In fact, compiling to an intermediate, internal representation of the source langauge is common compiler theory. They're patenting an abstract syntax tree (AST?) That is why compilers have what they call "front ends" and "back ends". The front end reads the source langauge code, and converts it into an intermediate format, which contains the same meaning as the original program, but in a generic representation that the back end knows how to traverse. The back-end is given this intermediate representation, and is used to generate the output of the compiler. The beauty is you can have multiple front ends utilizing the same optimizer and code-generating back end, or you can have one front-end and multiple back-ends for cross-compiling (producing programs that run on different platforms.) And the power of GCC is that you can have multiple front ends mixed with multiple back ends, for many combinations of languages on different platforms. You can even build a windows executable on linux.

Borland has done nearly the same thing with their C++ and Delphi compilers. They use the same front end to parse the source language, but use the identical back-end to produce the actual executable programs. Does this qualify as prior art? ( link )

There are many, many code generators that take a custom or niche language and translate it into a common language like C. You then take the C output, run it through a C compiler, and have your executable. In that regard, you are using the C compiler as a back-end, and your different translators as your front end. Heck, in college I built one of these in my compiler theory class in a single semester.

Currently, there is work being done on Perl, Python, and Ruby, to create a single representation for byte-compiled output of all three langauges, called parrot. In theory, this means that these scripting langauges could inter-operate, and they become just different ways to say the same thing... using your favorite language they can all compile to the same bytecode(s). ( Link )

Same with Java. There are other languages that can be compiled into Java compatible bytecodes, that are fully capable of operating directly with java classes.

Chris

[ Reply to This | # ]

  • Prior Art - Authored by: Anonymous on Tuesday, January 04 2005 @ 10:19 AM EST
    • Prior Art - Authored by: Anonymous on Tuesday, January 04 2005 @ 10:34 AM EST
  • Prior Art - Authored by: Anonymous on Tuesday, January 04 2005 @ 10:44 AM EST
    • Prior Art - Authored by: Anonymous on Tuesday, January 04 2005 @ 11:42 AM EST
      • Prior Art - Authored by: Anonymous on Tuesday, January 04 2005 @ 12:02 PM EST
        • Prior Art - Authored by: Anonymous on Tuesday, January 04 2005 @ 06:12 PM EST
  • JPI (later Clarion) 1990 - Authored by: Anonymous on Tuesday, January 04 2005 @ 12:03 PM EST
  • Prior Art - Authored by: Anonymous on Tuesday, January 04 2005 @ 12:05 PM EST
  • Prior Art - Authored by: Anonymous on Tuesday, January 04 2005 @ 12:19 PM EST
MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 09:59 AM EST
I still think that patent law should include compensation by the patent holder
for challenges to the patent should the patent be struck down in court. I say
let patent hunters be let loose on the pile of worthless patents that have been
issued.

[ Reply to This | # ]

Sure sounds like the PL/I Checker
Authored by: Anonymous on Tuesday, January 04 2005 @ 10:00 AM EST
that ran under TSO on IBM MVS (mainframe) systems in the early 1970s, re U.S.
Patent number 6,836,883.

[ Reply to This | # ]

Tsk, tsk....
Authored by: Anonymous on Tuesday, January 04 2005 @ 10:02 AM EST
MS: The protection and licensing of intellectual property allows companies and individuals to obtain a return on investment, sustaining business and encouraging future rounds of research and investment in the IT industry.
...
PJ: Oh. Patents ensure the freedom to innovate. Wait.

I'd have expected better of you, PJ. Far more accurate to paraphrase it as "Patents ensure the funding to innovate" -- if less rhetorically effective.

It is, of course, the prospect of a return on investment that patent law should provide, that innovation may be appropriately rewarded. Alas, the degree of innovation of IP varies. If the Microsoft quote was modified to use the phrase "innovative intellectual property," I'd be in enthusiastic agreement. However, far too many patents are awarded these days with far too little innovation behind them, leading to market rewarding with returns on investment in patents, not on investment in innovation.

At this point, the invisible hand of economics needs a thwack on the wrist.

[ Reply to This | # ]

  • Tsk, tsk.... - Authored by: Anonymous on Thursday, January 06 2005 @ 08:29 AM EST
MS Gets 2 New Patents
Authored by: gtall on Tuesday, January 04 2005 @ 10:18 AM EST
The second patent appears to be a debugger/editor, which permits line-by-line "execution capabilities", whereby you can stop and edit and then continue where execution left off, and it provides "support for common language runtime environments".

I once used a beautiful object oriented data flow language that had a debugger/editor that did just this, it was called Prograph. I doubt it exists anymore but I still hope for it. It was made by Pictorus, they got burned in up in the .com inflation.

Gerry

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: jkw on Tuesday, January 04 2005 @ 10:18 AM EST
P-code, anyone? I had FORTRAN and Pascal on an Apple IIe back
in 1986 or so.

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 10:19 AM EST
"'The protection and licensing of intellectual property allows companies
and individuals to obtain a return on investment, sustaining business and
encouraging future rounds of research and investment in the IT industry,' the
Microsoft spokesperson explained to internetnews.com."

Please read as:

We need patents so we can kill our competition and bleed money out of you so we
can get more patents so we can kill our competition and bleed more money out of
you so we can get more patents so we can...

Paul C.

[ Reply to This | # ]

Legalized Mine Mine Mine
Authored by: Anonymous on Tuesday, January 04 2005 @ 10:22 AM EST
Whenever I hear about another rediculous set of patents I always remember Don
Lancasters advicea "That patenting a million dollar Idea dosn't pay because
you cant afford to enforce it". Do We really need another example of how
badly screwed up the patent system is ?

The big question is what will it take to get what amounts to one more shakedown
racket reformed.

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 10:22 AM EST
Doesn't those patents remind you of the UCSD P-System and UCSD Pascal that was
available for microcomputers back in the 70's and 80's. Its purpose was to
translate different languages in to psuedo-code that was then compiled or
interpreted, can't remember which.

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 10:24 AM EST
Anyone looking for compilers accepting more than one language and generating a common format can go back to the 70's and UCSD Pascal, which later accepted FORTRAN and Ada. They generated P-code which was either interpreted, compiled, or executed directly by a Motorola 68000 with hacked microcode.

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: ExcludedMiddle on Tuesday, January 04 2005 @ 10:37 AM EST
So I suppose that this only becomes an issue if they decide to sue GNU for
making prior art. Man, I hate when people infringe my patents by making prior
art!

If you are a programmer, like I am, the best legal advice that I've heard is
that you should NOT read them. I just love that. Doing nothing about this is
tough, but I'll try to manage.

As a strategy, these are reasonable infantry in MS's fight against FOSS, but
hardly a threat singly. It's a waste of their time to go after particular little
projects that might use this: they may not have money, but they also aren't
influential enough. And the major projects not only have money, they are so
important to the computer communities that major companies would shell out to
help with the legal defense. For example, let's say that MS sued GNU for GCC.
That is the foundation of the entire GNU/Linux. Without it, it would not be able
to be compiled. Assuming that GNU didn't even have money or legal defenses
(which of course they do), the donations for their legal defense would be
staggering. And the amount of free researchers to find prior art and for legal
help, grok-style, would be even more staggering.

The patents that I worry more about are the ones that protect interoperability.
For example, patents on the MS file system access that could kill SAMBA. And as
our recent research has been showing, you can't directly use patents to do that,
especially if you are a monopoly, so luckily, it's not a given that these will
be effective.

I would suggest that the best thing to do about these particular patents is not
to directly challenge these, but to instead collect prior art for it now, in
case this does come up in court. Perhaps teaming up with an organization like
PubPat so that in the future, if this comes up, there's a ready made defense.

[ Reply to This | # ]

Okay, people know of prior art.
Authored by: Anonymous on Tuesday, January 04 2005 @ 10:44 AM EST
So what's the next step?

How do we make the USPTO aware of the fact that they've made a couple of fairly
serious mistakes here?

[ Reply to This | # ]

P-Code dates to BEFORE M$
Authored by: Eeyore on Tuesday, January 04 2005 @ 10:44 AM EST
I remember using P-Code (which was a pre-compiler plus an intrepreter initally) on a TI 994A back in the late 70's. If I remember correctly it was PRIMARILY Pascal but there were other front ends (BASIC and Fortran I think - and maybe others). Later it seems like I remember seeing software that would convert the output from the pre-compiler to a native executable. It SURE sounds similar....

Wikipedia article with links

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 10:48 AM EST

I was wondering if someone could help me understand some of the language in this patent. I'm not too knowledge in the areas of compilers, though I've written plenty of code.

For example:

What is claimed is:

1. A front end compiler system for generating code to be used by an execution environment, said front end system comprising:

a metadata module that compiles information to produce metadata information;

a code module that compiles information to produces executable instructions; and

wherein the metadata information and executable instructions are the result of compiling a source file in a first language and the front end compiler consumes metadata information produced by a different front end compiler as a result of compiling a source file in a second language.

That first part I take to mean it takes some code and compiles it an executable file, not necessarily for that system, but for some system. Fine.

When the patent mentiones a metadata module and a code module, I start getting lost. A code module sounds like the source code, but what is this metadata information?

The next part baffles me a little. There are two compilers here, one for the metadata information and another for the source. The source compiler feeds the information to the metadata compiler. Is this akin to creating object code and linking objects? To nitpick, I think consume is a poor choice in words unless they are implying the metadata compiler destroys the data being feed to it from the source compiler. Granted it could be getting a stream of data but couldn't that be writing a little more clearly?

What I need is the patent language on one side and some sort of laymans speak on the other. Or am I the only one having trouble comprehending the patent?

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 10:52 AM EST
It was a long time ago but I remember writing CBasic source then compiling it to
a .hex file. This could then be linked to other .hex files, produced from OTHER
languages (assembler and Fortran spring to mind) to produce an executable file.
It wasn't exactly intuitive (something about the symbol tables needing to be
modified) but it worked.

We're talking early 1980's here, on CP/M machines. Prior art anybody?

davcefai, away from my password.

[ Reply to This | # ]

MS Gets 2 "New" Patents?
Authored by: jim Reiter on Tuesday, January 04 2005 @ 10:57 AM EST

It looks like the new M$# strategy is to recreate the
monoply behind a wall of patents.

It would seem that the appropriate counter strategy would
be to avoid M$# "innovation" at all cost. Using M$#
"innovation" puts you under M$# control and subject to M$#
"over pricing".

This should start with development tools.

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: fredex on Tuesday, January 04 2005 @ 11:02 AM EST
Back in the late 1970s I worked for a small startup that used its own
proprietary programming language (QPL). QPL was compiled to a byte-code language
named QTL. QTL was (originally) interpreted by a module in the (proprietary)
operating system (which was written in assembler for the machine in use). Later
on, various levels of a hardware implementation of the QTL interpreter were
created, giving enormous performance boosts.

Anyway,... though this was only a single source language, it probably wouldn't
have been an enormous deal to target some other (modest) language to the same
QTL back end.

This company began doing this in '75 or '76 and sold hundreds of systems
throughout the US and Canada up into the early-to-mid '80s.

The designer of the language had done the same sort of things before at other
companies, so this was not a new concept even then.

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Bas Burger on Tuesday, January 04 2005 @ 11:17 AM EST
Translations of the outspoken opinion from the head of the USPTO.
This is what I read and think is really ment but not said because of legal
issues.

"The bottom line is there is a lot of misunderstandings about patents and
that may lead to some of this rhetoric, but the bottom line is it's a pretty
specialized area of the law and it takes experts to understand the patentability
-- what the process is, how we determine whether to issue a patent or not, how
the prior art defines the scope of coverage that an applicant is entitled to,
and so on and so forth.

Translation: "Since you are not a Expert in our eyes and never will be
because we say so, Thus SHUT UP!!!"


"And what you get is a lot of people that look at the title of the
invention or some abstract or whatever and just come to these conclusions with
respect to, 'how could the USPTO issue a patent' on such and such? But they
haven't actually looked at the legal scope of coverage; they're more or less
looking at a broad concept that's described in the patent as opposed to the
claims that define the legal scope of the coverage."

Translation:"People that stand in the way of us making large sums of money
by selling air castles should SHUT UP and BEAT IT, otherwise we will give them
trouble"

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 11:17 AM EST
I apologize for not reading all the posts before posting this. A quick search
did not seem to turn up any hard dates. I remember a book Anatomy of a
Compiler. Would have to dig through boxes to find my copy, but I enjoyed the
book when I read it. I have not read the patents, I am a programmer. It seems
to me, and I may be mistaken, that the author refered to P Code. A pseudo
machine code, that a compiler could compile to. A search of the web brought up
the following results:

The first site I found referenced a copyright date of 1967. The copy at Western
Michigan University has a copyright date of 1974. Perhaps if John A.N. Lee by
chance reads this site, he might be able to shed some light on what in MS's
patent is not prior art.

The anatomy of a compiler
By:
John A N Lee
Type:
English : Book : Non-fiction
Publisher:
New York, Reinhold Pub. Corp. [1967]
Subjects:
Compilers (Computer programs)
Programmation (Informatique)
Compilation (Ordinateurs)
Find Related:
Title/Author Search



The anatomy of a compiler [by] John A. N. Lee.

Main Author:
Lee, John A. N.
Title:
The anatomy of a compiler [by] John A. N. Lee.
Subject(s):
Compiling (Electronic computers)
Publisher:
New York, Van Nostrand Reinhold Co. [1974]
Description:
Book

The anatomy of a compiler [by] John A. N. Lee.

Author:
Lee, John A. N.
Title:
The anatomy of a compiler [by] John A. N. Lee.
Edition:
2d ed.
Publisher:

Author: Lee, John A. N.

Title: Computer semantics; studies of algorithms, processors, and
languages [by] John A. N. Lee.

Primary Material: Book

Subject(s): Electronic digital computers--Programming.
Algorithms.
Programming languages (Electronic computers)--Semantics.

Publisher: New York, Van Nostrand Reinhold [1972]

Description: xvi, 397 p. illus. 24 cm.

Series: Computer science series

Notes: Bibliography: p. 386-387.


Call Number: QA76.6 .L38
Location: Waldo Library, General Stacks

================================================================================



++++++++++++++++++++++++++++++++++++++++++
Institution Name
Institution Address
Institution Phone Number
Institution E-mail Address

[ Reply to This | # ]

USPTO is lying
Authored by: Anonymous on Tuesday, January 04 2005 @ 11:26 AM EST
The bottom line is there is a lot of misunderstandings about patents and that may lead to some of this rhetoric, but the bottom line is it's a pretty specialized area of the law and it takes experts to understand the patentability -- what the process is, how we determine whether to issue a patent or not, how the prior art defines the scope of coverage that an applicant is entitled to, and so on and so forth.

The USPTO has granted a patent on faster than light communication. You can read about it here and here

Th e USPTO will grant you anything, even the patently impossible. You just have to write it down. The quotes are misinformation, a patent has nothing to do with invention. It is the supposedly infringer who has to do all the work and bear all the cost of disproving the patent.

[ Reply to This | # ]

Striking back?
Authored by: Anonymous on Tuesday, January 04 2005 @ 11:29 AM EST
Maybe somebody in a patent suit about such an absurd patent should sue the USPTO
for negligence in issuing it. I'm sure that such a suit could get help from
patent-reform groups.

It might be interesting to include language about the USPTO getting into
software patenting against caselaw at the time they made the decision.
Unfortunately, I doubt this part would succeed, as I believe the Supreme Court
has already refused to review a software patent case, but it seems worth
trying.

[ Reply to This | # ]

Patent 6,836,883 -- COBAL?
Authored by: alextangent on Tuesday, January 04 2005 @ 11:36 AM EST
"In order to modify an existing front end compiler for an existing source
language, such as Pascal, COBAL, C++, etc..."

COBAL?

[ Reply to This | # ]

This is really sad
Authored by: Anonymous on Tuesday, January 04 2005 @ 11:38 AM EST
It took me less than 15 seconds to reach around to my bookshelf, consult THE
book on Compiler Construction, and come up with the following blurbs:

----------------------
Intermediate Code Generation

In the analysis-synthesis model of a compiler, the front end translates a source
program into an intermediate representation from which the back end generates
target code. Details of the target language are confined to the back end, as far
as possible.
-----------------------
BIBLIOGRAPHICAL NOTES

UNCOL (for Universal Compiler Oriented Language) is a mythical universal
intermediate language, sought since the mid 1950's. Given an UNCOL, the committe
report by Strong et al. [1958] showed how compiler could be constructed by
hooking a front end for a given source language with a back end for a given
target language.
-----------------------

...1958 is before 1999, right?

The book I'm looking at is "Compilers Principles, Techniques, and
Tools" (aka "The Dragon Book") , by Aho, Sethi and Ullman; (c)
1986 by Bell Telephone Laboratories, Inc.; Reprinted with corrections March,
1988, Reproduced by Addison-Wesley as ISBN 0-201-10088-6 and stored on my
bookshelf since I took Compiler Construction at Queen's University in 1996.

Library of Congress Cataloging in Publication Data

Aho, Alfred V.
Compilers, principles, techniques and tools.

Bibliography: p.
Includes index.
1. Compiling (Electronic compuiters) I. Sethi,
Ravi. II. Ullman, Jeffrey D., 1942- . III. Title.
QA76.75.C65A37 1985 05.4'53 85-15647
ISBN 0-201-10088-6

Excerpts were from Chapter 8, "Intermediate Languages", pages 463
(intro to chapter) and 511 (chapter epilogue). I didn't have time to read
anything else, but at least I had the 15 seconds to do the basic research a
patent examiner should have.

I suppose there are other real-world examples worth mentioning: p2c and c-front.
The former is a Pascal compiler; the later is the original C++ compiler. They
both use C as an intermediate language, and C as an intermediate language is a
VERY common paradigm for people writing quickie compilers (I've done it myself
for a couple of custom VHLLs).

/. 163409 (multipart/mixed)

--
Do daemons dream of electric sleep()?

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 11:39 AM EST
Don't forget things like Informix's 4GL. This was embedded SQL with C code...
-Mikebmw

[ Reply to This | # ]

Confused patent purpose
Authored by: Anonymous on Tuesday, January 04 2005 @ 11:44 AM EST

This confuses me:

"And what you get is a lot of people that look at the title of the invention or some abstract or whatever and just come to these conclusions with respect to, 'how could the USPTO issue a patent' on such and such? But they haven't actually looked at the legal scope of coverage; they're more or less looking at a broad concept that's described in the patent as opposed to the claims that define the legal scope of the coverage."

So, he is saying that what is described in the patent is not really what is covered by the patent? How then, do I, the lowly inventor or small business man, determine if my idea infringes on someone's patent if the patent does not accurately describe what a patent's "legal scope" is?

I read that he is admitting to what the real purpose of patents has become. Because the published patent does not describe the "legal scope" of the patent, one must pay large legal fees for analysis of all patents before doing anything innovative. Therefore, the purpose of patents is to block any possible competition with huge legal fee barriers before they can even start. Just get a broadly worded but narrow "legal scope" patent granted and any potential competition has to hire attorneys instead of simply reading the patent.

Sucks.

[ Reply to This | # ]

Yep, no problem at all...
Authored by: stuart_b on Tuesday, January 04 2005 @ 11:45 AM EST
"And what you get is a lot of people that look at the title of the
invention or some abstract or whatever and just come to these conclusions with
respect to, 'how could the USPTO issue a patent' on such and such? But they
haven't actually looked at the legal scope of coverage; they're more or less
looking at a broad concept that's described in the patent as opposed to the
claims that define the legal scope of the coverage."

I guess he's sort of right: Obviously any prior art shown in the patent is
excluded from coverage. And the same for any demonstrated by third parties
afterward--but of course that requires a trial (he-he).

And Microsoft's totally right about "innovation". Here their
innovation has been to make their claims vague and inclusive enough as to cover
ANYTHING similar that isn't included in prior art. Elsewhere they have
innovated by forcing system vendors to do things (that the system vendors can
legally do) on their behalf that would have been illegal for them to do
themselves--shielding themselves against the letter of anti-trust laws. (And of
course further protecting themselves by making those arrangements subject to
protection as trade secrets.) See, they ARE great innovators.

[ Reply to This | # ]

A Truly Recursive Solution...
Authored by: muswell100 on Tuesday, January 04 2005 @ 11:48 AM EST
So how about patenting a "system of creating wealth through the purchase of
software patents"? Every time someone tries to use a software patent to
extract money from someone else, they have to pay the patent holder a fee. The
patent holder may then choose to pay the money back to the company(ies) using
the original patent to extort money from the target company in the first place.
Since this income is derived from the application of that same "system of
creating wealth through the purchase of software patents", the patent
holders will have to pay a further fee to the patent holders. The patent
holders, however, pay back this sum to the same company and the whole cycle
starts again...

PS: This IS a joke...

[ Reply to This | # ]

MS patents == first step
Authored by: Nick_UK on Tuesday, January 04 2005 @ 11:48 AM EST
This is frightening really.

We all know MS have been applying for patents by the truck
load, whether those patents are prior art of just
downright ludicrous (the to-do et al).

BUT who has to contend the patent? The original owner?
The world at large on a known prior art?

This is where the trouble starts. Now, SCO are in a
similar boat, but as we know with MS money they have
managed to sustain a lengthy court case and at every
opportunity are drawing it out.

Who could take on MS and it's money mountain in the
courts? OSS? I doubt it. MS along with the patents can
produce a legal battle that will cost millions before it
even reaches the courts. Nobody or any one Company can
stop them in that area.

What needs to be addressed is the law (patent law). Just
because it's law doesn't mean it is right or correct.

Nick

[ Reply to This | # ]

EU is not paying attention
Authored by: Anonymous on Tuesday, January 04 2005 @ 12:17 PM EST
The new Luxemburg presidency has made one of their top priorities to introduce software patents in the EU. Since the directive may not pass if it comes to an honest voting, the patent mafia will continue to do everything they can to push this through with dirty tricks instead.

Once again, EU citizens, please do what you can. The German and Dutch governments should not be allowed to continue ignoring their own parliaments will in this question!

[ Reply to This | # ]

A way of making obvious patents irrelevant?
Authored by: vegard on Tuesday, January 04 2005 @ 12:19 PM EST
I'm just thinking....what if, instead of taking the trouble of going through
court for every patent, you just validated their invalidness, and put them on a
list of "patents we're willing to defend law suits against" ? A
defense fund, so to say?

Would it be likely that you were granted your legal costs if you're sued by a
patent holder, and the patent is found invalid?

Of course, you'd have to be pretty sure before you put things on such a list.
Whether or not you could trust that the court system would come to the same
conclusion, is entirely different...

The point would, of course, be to make it expensive to try to enforce bad
patents. And easy/safe to ignore them.

A secondary point would of course be to point out the obviousness of a lot of
patents.

[ Reply to This | # ]

Quickest and best solution to software patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 12:29 PM EST
Regular patents are required to be accompanied with diagrams and specifications
of the given invention. You have to be able to build the invention from the
patent application. The entire shebang is available to the public - meaning that
anyone can take a patent application and reproduce the invention. This would be
illegal in most cases, but there you are.

I say we subject software inventions to the same requirements that physical
inventions must meet - all source code and build scripts must be made available
to the public to qualify for a patent. The same for copyright.

How quickly would companies like MSFT and SCOX back off using patents and
copyright against F/LOSS ??

Feel free to petition your congressional representatives if you like the idea.

[ Reply to This | # ]

Don't get carried away
Authored by: Anonymous on Tuesday, January 04 2005 @ 12:39 PM EST
A couple points:

Microsoft is not patenting "secrets" as PJ suggests. Patents are
<b>always</b> written with expansive language so that they cover
other instantiations of the claims beyond the example that is disclosed. This
does not extend the reach of the patent to cover additional claims that have not
been disclosed; that would be silly. True, the language is sometimes made
overly inclusive, because it generally hurts more if you err on the side of
making a too narrow claim, than if you make a broader claim that later is more
narrowly interpreted by a court.

When you read a patent, you have to be <b>very</b> careful to see
<b>exactly</b> what is claimed. In the compiler patent, they are
not patenting a system for compiling multiple languages with a common back-end.
They are patenting a compiler system where the front-end processes source files
as well as intermediate representations created from other source files. This
is different from a traditional compiler heirarchy with compiler and linker
steps. Specifically, GCC does <b>NOT</b> do this. It could be that
Mono's mcs compiler does.

[ Reply to This | # ]

Pj's Patents Given to All For Innovation List..
Authored by: shareme on Tuesday, January 04 2005 @ 12:47 PM EST
Okay folks its obvious that we need to gatehr soem friends of FOSS..

PJ, I would liek to suggest that we start a List doucemnting all Patents given
rolaty free to all tech players to encourage innovation and their woerns..

In this way we will know which companies and groups to band together with to
fight the MS patent club and eventually kill it..

Suggest Areas to start:

On Computer Langauge Compiling:

Xeorox PARC
IBM
Apple
Red Hat
HP
SUN


I realize that I am asking for everyone to do some work/research..

I can only say that as a community united we STANd divided we Fall under
MS..Wheels of destruction..

The part I am grabbing for this weekend is the Xeorx Patents used in the
compiler byte code manipution feature od Eclipse IDe...


PJ, what details do we need? I know we need the patent number, the summary,
Orginal Patent owner, org that patent was royalty free contributed to...but what
else?



---
Sharing and thinking is only a crime in those societies where freedom doesn't
exist.

[ Reply to This | # ]

MS Gets 2 New Patents -- USPTO 'Harmonizing Patent Laws'
Authored by: Anonymous on Tuesday, January 04 2005 @ 12:48 PM EST
This is would be very silly if I did not think it was serious.
I would just like to be able to run my computer, my compiler, editor, etc. without worry. I want to be able to do the same things I did before 2000 after 2000. (I really want to be able to do this for the rest of my life).
I've allready got ideas on how to improve compilers/editors while I have not talked to anyone yet. Anyone know how I can protect myself?
Also I just checked out the US Patent and Trademark Office
It looks like they are getting ready to update US law to be more compatible with 'international' law. On Feb 3-4, 2005 they are having a meeting on it.Any one have any good ideas of what the best way for a US citizen to provide input at this meeting. (I'm not sure making patent law more like everyone else is the right thing.Unless it is to make it more like Europe, where software is not patentable. I am still thinking this through).
Does anyone know if there is a research and development exception so I can develop new ideas then only check them out when they look promising, and I ready to publish, or commercialize, or submit under GPL etc?
I'm sorry about the 'shortness' of this message but I'm home for lunch and read this and wanted to reply. I go back to work shortly.

A florida guy.

[ Reply to This | # ]

Are these patents really mistakes?
Authored by: Anonymous on Tuesday, January 04 2005 @ 12:59 PM EST
Just to play devils advocate, and to try to get clear ansswers of the mistakes,
are the patents mistakes, and if so how? To get it out of the way, no I have
not read the patents and will not because I am I a programmer.

From what PJ said, it sounds like the editor/debugger might be a legitimate
claim. It sounds like it's a case were you are stepping the the code in a
debugger, get to (oh say) line 500, and go into the source listing in the
debugger and change the code at and after line 500, and then continue to step
through the new code. So the debugger compiles the new version of the code and
inserts it into the running image in the debugger. I can't honestly say I've
see that before, but I know I don't know all that has been done in the computing
world. Is that what this patent is for, and if so, has anyone seen something
like?

I know off the top of my head how I'd implement it and it wouldn't be too hard,
but that doesn't mean it's been done before. I guess the implementation seems
trivial to me, but the need for it was a bit obscure. I know in dbx, I can
evaluate code that is already in the program and skip code I don't want to run.
That's not quite the same, but usually all I've ever wanted.

Second, the compiler patent. I saw a number of people mention that compilers
use an intermediate language between front ends and back end. That doesn't
sound like what PJ described. It sounds like the trick was that the multiple
languages were in the same file. That and analysing the languages together. If
that is right, the first thing that comes to mind for me is perl's Inline
modules. Inline a mess on non-perl code in you perl scripts. Yehaw. But I'm
not sure it pre-dates the patent. Second, I know perl is working on the ability
to deal with multiple grammers (languages) in a single file. The perl6 grammers
would be powerful enough to implement multiple languages in the same file and
even the same code block. But I'm not sure when they first discussed using
grammers to modify the compile time parsing.

I know several compilers for various languages that let you inline the language
and assembler, but is there any that allow mixing two or more non-assembler
languages if the same source file? (I figure they could tap-dance the assembler
situation easier than if you could show the mixing of Pascal, Fortran, LISP and
COBOL all in one file.

[ Reply to This | # ]

Ok, so we don't fully understand patent system.
Authored by: Anonymous on Tuesday, January 04 2005 @ 01:09 PM EST
...meaning that we supposedly aren't qualified to judge them for their decisions on patents. But on the other hand, what evidence has the USPTO given to show us that they have the competence to judge the innovativeness of software?

---
m(_ _)m

[ Reply to This | # ]

It's not just software patents...
Authored by: Philip Stephens on Tuesday, January 04 2005 @ 01:30 PM EST
Recently I had reason to look up some patents awarded to Apple, Nintento and
Commodore on circuits for producing a computer video display compatible with
NTSC and PAL TV's. This was purely to satisfy my own curiousity about how some
of the original personal computers and game consoles achieved such a
technological feat (I'm a computer history buff).

Now, I have to say that it's pretty neat that I could learn about how these
companies solved a particular technological challenge by reading their patents.
But by the time I'd read through all 5 patents (3 by Apple, one each by Nintendo
and Commodore), I realised two things: (1) the patent problem isn't just
restricted to software, and (2) the problem has existed for a long time, 25
years at least, maybe longer.

Let me explain. The 5 patents I printed off from the USPTO web site (4,136,359;
4,217,604; 4,278,972; 4,569,019; and 4,8234,106) are all essentially describing
the same invention! And yet, the USPTO decided to award a patent on the same
invention to 3 different companies!

Okay, so the details of each invention differ. The Apple patents describe how
the video for the Apple ][ computer was generated, the Nintendo patent describes
the video generated by the NES game console, and the Commodore patent describes
an IC used to generate video similiar to that found in the Commodore 64.
However, the fundemental approaches used to produce a video signal compatible
with NTSC and PAL were identical in all the patents.

Furthermore, and this is what really made me realise what a crock most patents
are, there is no way that a hardware engineer could actually reproduce these 5
inventions by following the "blueprints" in these patents. For you
see, the patents only provide a high-level overview of how the inventions work,
with simple block diagrams showing the layout of the components. There are no
detailed circuit diagrams, no electrical signal timing diagrams; in short,
nothing resembling complete blueprints at all.

So, just imagine that you're a company who wants to do the right thing and
license one of these patents. Will Apple, Nintendo or Commodore then give you
actual blueprints for these inventions once you've forked over the cash? Or
will you have to do all the R&D necessary to fill in the gaps, spending
thousands or maybe millions of dollars in the process? For what do you get from
licensing one of these patents? Just a vague description of how your circuit
SHOULD work once it's finished! That's it!

Because these patents are essentially useless as a way of reproducing the
inventions, the patent holders are effectively patenting the IDEAS behind the
inventions, not the inventions themselves. The blueprints for the real
inventions are still being kept secret by Apple, Nintendo and Commodore (the
latter a defunct company), perhaps never to see the light of day.

So is this how the patent system was supposed to work? Somehow I don't think
so. Yet it is clear that companies have been exploiting problems with the
USPTO's procedures for decades, allowing them to bypass the requirement that a
patented invention be fully specified in return for 20 years of protection. For
decades, companies have been able to patent inventions without ever giving up
their actual blueprints! They've been getting the best of both worlds: the
ability to get their competitors to cough up money for the privilege of using
their invention, while not actually telling their competitors how to build it!

If you ask me, either the USPTO needs to be reformed so that it won't accept
patent applications unless they include complete blueprints of an invention; or
we need to dismantle the patent system altogether and force companies to rely on
trade secrets (which they're doing anyway with one hand while asking licensing
fees for their patents with the other).

[ Reply to This | # ]

  • Thanks - Authored by: Anonymous on Tuesday, January 04 2005 @ 07:55 PM EST
Prior Art - JOVIAL Compiler, 1969
Authored by: Anonymous on Tuesday, January 04 2005 @ 01:31 PM EST
My copy of the "Univac 1100 Exec 8 JOVIAL Programmer Reference" (c)
1969,1973 by the Sperry Rand Corporation describes in chapter 10 what it calls
"Direct Code", where you can insert assembly language in your JOVIAL
source code and compile it into one application.

[ Reply to This | # ]

  • Yep... - Authored by: jbeadle on Tuesday, January 04 2005 @ 03:44 PM EST
MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 01:38 PM EST
I'm a bit surprised that Microsoft took this long to patent the compiler "innovation". They've been doing something like this forever. Going way back to the Days of DOS, the first pass of Microsoft's multipass compilers were translating source code into an intermediate "language" that the second and third passes acted on. I'm told that the later passes were the same regardless of the source language (C, Pascal, FORTRAN, etc.). I never tried verifying that by using the second/third pass programs from, say, their C compiler on the output of the first pass of their FORTRAN compiler. It made a lot of sense at the time. I'm not all that surprised that they might still be doing that. (Heck, I understand that the deprecation of old GW-BASIC features only recently was left out of Microsoft's BASIC language.) It would have made it easier to crank out compilers for additional languages. Sort of clever, I guess. But worthy of a patent? Hmm... I'm not so sure about that.

[ Reply to This | # ]

On-topic Humor: M$ Patents 0 and 1
Authored by: eckenheimer on Tuesday, January 04 2005 @ 02:04 PM EST
Here's a link to an old (but funny) humor piece:

"REDMOND, WA -- In what CEO Bill Gates called 'an unfortunate but necessary step to protect our intellectual property from theft and exploitation by competitors,' the Microsoft Corporation patented the numbers one and zero..."

---
In a world without walls or fences, who needs windows or gates?

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 02:37 PM EST
Is sarcasm impolite?

[ Reply to This | # ]

What about MIPS?
Authored by: Anonymous on Tuesday, January 04 2005 @ 03:25 PM EST
If people are going to look for prior art they should be looking at MIPS
innovations in the 80's and 90's. MIPS commercialized u-code compilers
(invented at Stanford I think in the late 70's) which permitted global
optimization across Fortran, Pascal, C++ and Ada. MIPS ucode was really at the
forefront of optimization technology in the 80's. And yes, I could compile my
C++ code, inlining my ucode library and generate a single object if that is what
this patent is about. And no I dont think MIPS was first, but they did do it
particularly well.

As far as the " Method and system for editing software programs"
patent, it's called Fix+Continue and has been available in the MIPS Pro IDE
since the mid-90's. The only difference is that I dont know if the current
(>MIPS III) compilers use an IL (intermediate language) - I think
Fix+Continue was specifically C/C++. Unless the Microsoft patent achieves a
"Fix+Continue" with global inlining (and why bother) the existence of
an IL seems totally irrelevant, but I am not knowledgable about patent logic...
Also, I never thought this was an SGI invention. Doesn't this feature also
exist in HPUX/Solaris/AIX/OSF/whatever IDE?

Karsten

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 03:31 PM EST
I think that this is a very interesting topic, being a lawyer myself (but not a
patent lawyer).

I propose that one or more patent lawyers who read this site analyze the claims
of the patents in some detail and then see if prior art that actually relates to
the claimed invention can be identified by Groklaw readers.

If what is claimed as the real invention can be fleshed out by a patent attorney
or two, this might be a good place to identify, and document, prior art.

[ Reply to This | # ]

FOSS needs one volunteer martyr
Authored by: Anonymous on Tuesday, January 04 2005 @ 03:43 PM EST
One to test MS patents validity!
The community would pay the money for it for the trial itself.
Let's calculate 5 million FOSS users each pay 10$ = 10 M $


[ Reply to This | # ]

Reasons for obfuscation
Authored by: Anonymous on Tuesday, January 04 2005 @ 04:14 PM EST
PJ,

Having two software patent applications in my name, I can tell you a bit about
why patent language is as obfuscated and vague as you found it to be.

There are three reasons:

1. You want to claim as much ground as possible. The description of the
technology serves as an example, the claims are the meat of the patent.

2. You don't want the competition to easily figure out what you're actually
doing while still protecting what you're doing in case of need. Obfuscation as
well as embedding what you're doing in dozens of additional claims serves that
purpose.

3. You want to discourage the patent office from examining your patent to
closely. Obfuscation serves this purpose well.

Item two is, of course, in crass violation of the spirit of patents and three is
just nasty lawyer tactics.

IANAL, but IAASE.

[ Reply to This | # ]

We want MORE patents!!!
Authored by: Anonymous on Tuesday, January 04 2005 @ 05:08 PM EST
The more outrageous patents that appear, the easier it is to show people that
the system isn't working.

I'm throwing down a gauntlet:
Who can file (and obtain) the most proposterous patent?

My money is on the people at Redmond.

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Anonymous on Tuesday, January 04 2005 @ 05:30 PM EST
I have now read about half of the first patent, and I am in need of some
explanation of patent law. This patent is absolutely full of prior art by any
common sense definition. Almost every paragraph describes something that I have
seen before, sometimes in many different systems. There may, however, be some
original (I didn't say earth shattering) suggestions. For example, the reading
of type information from the intermediate rep by all the various front ends as a
requirement (though similar to header pre-compilation) may be novel (please
correct me).

So the question is, do I violate a patent
1) if I implent one or more of the ***novel*** ideas in a patent
2) Only if I implement the whole patent including all the ideas?

If 1), how am I supposed to know which of the ideas is contended to be novel
when they are not clearly identified as such in the patent. Also, if this is
the case, why dont I just take the shotgun approach to patent filing. How about
a million page compendium of random thoughts. If some of them turn out to be
original my patent on those will be valid and I only have to file once.

If 2), who cares. It would be easy to circumvent a patent as long and full of
minutiae as this one, simply by doing something slightly differently.

This patent is clearly not a single invention. At best, it is one or two
middling ideas mixed in with a large amount of completely generic software
engineering.

Karsten

[ Reply to This | # ]

Mono is a huge waste of time
Authored by: Anonymous on Tuesday, January 04 2005 @ 05:52 PM EST
I said this before and now it seems even more relevant. Why would anyone in
their right mind implement a technology that is:

1. Created by a software monopoly, known for their "trade mark"
business practices that border on criminal.

2. Patent encumbered by that same monopoly.

Do we really want to go down the path of polluting free and open source software
with stuff like that? There is heaps of other stuff we could focus our
development effort on, not MS clones.

I know I'm going to get flamed for this, but I had to get it off my chest.

[ Reply to This | # ]

One last point
Authored by: rweiler on Tuesday, January 04 2005 @ 05:59 PM EST
People shouldn't get their knickers in too much of a knot over the 'fix and
continue' patent because while it seems like a nice idea, in practice, it is of
relatively little value other than a marketing bullet point. The problem is
that most fix and continue schemes (including this one) can't back up the input
data to what it was before your fix, so as often as not, you have to restart
anyway (note, before Microsoft gets any funny ideas, there are ways to implement
this if it was really useful) In addition, you should be debugging modules in
isolation before they ever get into a huge long running program (where this
feature is most useful). Finally, modern computers are so fast that it just
doesn't take that long to restart even big programs so FOSS doesn't lose all
that much by not having this feature. At most, it can save a few minutes a day.

---
Sometimes the measured use of force is the only thing that keeps the world from
being ruled by force. -- G. W. Bush

[ Reply to This | # ]

  • One last point - Authored by: Anonymous on Tuesday, January 04 2005 @ 06:08 PM EST
The purpose of the compiler patent
Authored by: Anonymous on Tuesday, January 04 2005 @ 06:43 PM EST
It seems obvious to me. In the .NET framework, multiple languages are supported.
The compiler claim describes a method for re-using interface declarations of
libraries in various languages.

Thus, if someone writes a .NET library, they only need to write *one* interface
declaration, which can then be used with any .NET language. (I think these are
C#, J#, VB.NET at the moment).

GCC certainly does not allow you to use java header files in your C programs for
example. My take on the patent's validity, is the following:

1) To my knowledge, no-one ever needed to do this before, thus no-one has done
it. Library vendors have generally been providing separate header files for
various languages to interface with their libraries.

2) No matter what the exact method used are, there must be a translation from
one type of source to another. Source code translation has been done before.

So, we have commonly used methods to perform a novel task. The task's novelty
cannot make the invention patentable by itself. If the way they have used known
methods to perform the task is not *obvious* then the patent is valid.

However, in that case, I think that if one does the same task using some other,
perhaps more obvious, method, then one does not infringe on this patent: i.e.
the patent's scope is extremely limited.

[ Reply to This | # ]

What patents were REALLY supposed to be?
Authored by: Dr.Dubious DDQ on Tuesday, January 04 2005 @ 06:53 PM EST

The more I've thought about it, the more I've come to the (possibly erroneous) conclusion that most people misunderstand what patents were originally supposed to be for.

Obviously, wealthy corporations and some wealthy patent lawyers like to say that patents are for "making sure inventors get money", though most of us can see that's not accurate.

On the other hand, I don't think it's quite accurate to even say that patents were intended to "protect inventors", either.

The formal language in the Constitution just says "promote the progress of science and the useful arts", nothing about "protecting inventors" (or even "making sure corporations get paid")[1].

From my (again, admittedly possibly flawed) perspective, patents were effectively a mechanism for "the public", through its government, to "buy" trade secrets. The monopoly protection for patent holders seems to me not to have been the purpose of patents, but merely the metaphorical "payment" that holders of trade secrets got in exchange for making that trade secret available for the rest of the public to build from.

If my perspective is reasonable and accurate, it seems to me that the USPTO shoudln't even merely be examining whether or not there is real prior art (though some of THAT would be nice, too), but also evaluating, in general, whether or not the "patented" information will be valuable to the public for future innovation.

Never mind whether or not Dropping to one knee to hit a tennis ball (US Pat#5,993,336) or Swinging sideways on a swingset (US Pat#6,368,227) has prior art. Based on my perspective, it's more important to ask whether or not the public would actually be losing anything if either of those had been kept "trade secrets" instead. Exclusive rights are valuable, and The Public should not be spending them on just anything. (I think the "obvious to someone in the field" criteria is supposed to address this, but obviously doesn't seem too. Have any of us NOT pulled alternately on each chain on a swingset to make the swing go sideways in our youth?...)

Am I monumentally naive, or has the patent process been perverted even further than they cynical among us (myself included) have previously noticed?...

[1] I'm saying all of this from memory. If I've gotten the details wrong, I suppose my only defense is that I attended public school. In California. Corrections welcome, but please speak slowly and use small words, for the aforementioned reason :-)

[ Reply to This | # ]

gcc is cleared
Authored by: argee on Tuesday, January 04 2005 @ 08:06 PM EST
It seems to me that by mentioning GCC as prior art, that
MS has declared gcc as a valid, non-infringing program at
least as of the date of the patent application.


---
--
argee

[ Reply to This | # ]

IP law suits
Authored by: jim Reiter on Tuesday, January 04 2005 @ 10:01 PM EST
Remember, Caldera/TSG has maintained a Ip law suit against IBM for almost two
years without any proof of ownership of Patents and/or Copyrights (IP).

What does M$# think they can do with a questionable Patent? Sue, sue, sue.

There has to be significant risk for a company filing bogus lawsuits like this.
TSG need to be punished for this, put out of business if that is possible.

BTW, having a Patent should not be exculpatory. If a company sues on what
latter turns out to be an invalid Patent, the suing company should be
responsible for full cost, damages and penalities.



[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Anonymous on Wednesday, January 05 2005 @ 12:05 AM EST
not sure how this helps:

<snip>
ANALYSIS AND LESSONS

First and foremost, diagnosing this problem as a black box would have been
impossible. Only detailed traces of actual system behavior enabled the
faulty execution sequence to be captured and identified.

Secondly, leaving the "debugging" facilities in the system saved the
day.
Without the ability to modify the system in the field, the problem could not
have been corrected.
</snip>

from:
http://groups.google.co.in/groups?q=Space+shuttle+inversion+of+priority&hl=e
n&lr=&client=firefox-a&selm=CMM.0.90.1.881704091.risko%40chiron.csl.
sri.com&rnum=2

[ Reply to This | # ]

Welcome Asia, take our jobs, please.
Authored by: jim Reiter on Wednesday, January 05 2005 @ 12:12 AM EST
If Linux/Open Source becomes outlawed in the United States and Europe, Data
Processing can move to Asian countries where there would be no restrictions on
what software companies use.

"When Linux is outlawed, only outlaws will have Linux."

[ Reply to This | # ]

How to utterly destroy software patents *without* changing the law
Authored by: AdultSupervision on Wednesday, January 05 2005 @ 01:10 AM EST
Create a purely mathematical language - I mean one that uses math as its source
code. No procedures, no objects, no nothing but math as you see it in your
hated college calculus textbooks.
Programs are stored as formulas.
Math can't be patented.
*every* computer program ever written is actually math, but expressed in a
language easier on the brains of mere mortals.
Then start teaching jr and sr high school kids *real* math.
Once you start writing widgets, etc in math, then software patents start to fall
apart.


---
All my witty sigs are made up by me, excluding this one, which I made up but
isn't witty.

[ Reply to This | # ]

I think "snoball" did this too
Authored by: Anonymous on Wednesday, January 05 2005 @ 02:43 AM EST
back in the 80s on the early PCs I worked with a language called
"snoball" which took what was pretty much plain english and converted
it into byte code, which then was compiled into binary. you could hook other
things from other languages into that byte code, and generate one final binary.
that seems to be pretty much the same thing.

heck, I guess you could claim the old util called "automenu" did the
same thing, turned a plain english language set of text files into byte-code
which then got "compiled" into a binary.

couldnt a whole bunch of things from this period apply? like comal, forth, and
what of ADA, etc? I remember all the talk that things like ADA would
revolutionize the world... uh, sure it will ;) and then there is always really
weird stuff like Prolog. and how about lisp? these are still current...

but i doubt anyone remembers this stuff....

[ Reply to This | # ]

are there really any "new" inventions?
Authored by: Anonymous on Wednesday, January 05 2005 @ 03:06 AM EST
given it is "late" in the information age, and there are so many
people and so much prior art, I am wondering if, in the general sense, there
could atually be any "new" software inventions.

Again, I am talking generally. certainly there are wide areas of medical science
and research that have lots of room left. but when talking about general use of
computers, internet, and such, is there really any room left? I mean, havent we
"been there and done that" to death by this point?

i mean there are only so many different ways you can peel an apple... or
whatever.

I personally would see an "internet" based patent as frivolous at this
point. I mean, a patent based on how you click a button on a webpage?
puh-leeze!

I remember, back in the day when you couldn't use the "/" in a
computer program to bring up a menu, because Lotus had a strangle-hold on this
and would sue anyone or any company to death who even attempted to use this
feature. Lotus 1-2-3 (spreadsheet) was the defacto standard and no one could
compete, because of a stupid thing as to how you brought up the menu.

no one would buy anything else because the menu was too different. companies
tried, but got sued if they used the "/" - I remember this happened a
few times.

in this particular case, does anyone know if the "/" is now safe,
because Lotus is now IBM, and so forth, history has moved on.

silly things like this are just stupid.

[ Reply to This | # ]

A Larger Leagal Defense Fund
Authored by: Anonymous on Wednesday, January 05 2005 @ 11:52 AM EST
I know that OSDL has a leagal defense fund, but we need lots of people from
everywhere around the world to donate to it. Then when Some 800lb gorrilla wants
to sue a tiny company or the fsf there will be a container of money contributed
by everyone that's large enough to hold out. Meaning it's just going to be very
expensive for the 800lb gorrilla and in the end they won't win.

[ Reply to This | # ]

MS Gets 2 New Patents
Authored by: Cathail on Wednesday, January 05 2005 @ 01:31 PM EST
Had to come out of my long time lurk mode and add another prior art example on
the debugger patent.

In the early 90s there were products for a number of Unix platforms from
Centerline Software. The products were called CodeCenter, ObjectCenter, and
CenterLine C/C++ and among other things included a compiler and what they called
a source-level debugger.

The debugger allowed specifying that one or more modules used in building a
binary that it was debugging to be interpreted from the original C/C++ during
execution. While being a bit slow, this allowed you to modify the source and see
the effects of the changes while stepping through the code without re-building
or re-running.

It also allowed rebuilding a single object file and reloading just that one
object file without re-starting.

[ Reply to This | # ]

Ops! "tsunami" reference BAD Idea.
Authored by: Anonymous on Thursday, January 06 2005 @ 08:06 PM EST
Ops! "tsunami" reference BAD Idea.

[ Reply to This | # ]

Small businesses too
Authored by: Anonymous on Thursday, January 06 2005 @ 09:25 PM EST
<i>Here's my entry:

"We need these patents so we can kill off Free and Open Source
software."</i>

Actually, its also a disadvantage for small businesses who normally cannot
easily afford buying a patent or a lawyer. Even when they do, the big
corporations (e.g. IBM, Microsoft -- Yes IBM has a huge patent portfolio!) can
just say 'oh but your application infringes patent X we own' then that costs
them a lot of money to research and in the end its just better to cross license
elseway their software project dies. Its just evil. Quite frankly i almost find
it criminal behaviour from a moral standpoint.

This was also coverted at RMS and 2 FFII guys' speech at ITCollege Estonia. Here
it is:
RMS: http://www.itcollege.ee/dl/OGG/avaloeng10_1.ogg
FFII: http://www.itcollege.ee/dl/OGG/avaloeng10_2.ogg

[ Reply to This | # ]

About europe and patents..
Authored by: Anonymous on Friday, January 07 2005 @ 10:45 PM EST

And maybee a tip for the US.

The law, both U.S and european, through international treatys as TRIPS, WIPO and the Berneconvention says that software should be protected as litterary works, logical, it is something you write. Now, what they want europeans to swallow is basicly this, start with the law:

[X] shall be protected as [X] (copyright)

then take the argument that the EPO and a lot of lawyers want us to:

[X] is not an [X] as such if [condition that is always true]

But hey... the same three international treaties also says that we are not allowed to legislate in a way which conflicts with a normal exploitation of the work and do not unreasonably prejudice the legitimate interests of the right holder.

Everytime a computerprogram infringes a patent, or the use of one in the way the autor intended, it is an unreasonable prejudice on the authors legitimate interests. A lawmaking that TRIPS, WIPO and the Berneconvention forbids us to do, and it does not matter if the patent is about such konceptual objects that copyright is not considered to protect (ideas). The patent prejudices the author unfairly, because he is denyed his exclusivety by the reason that he wrote it (the patented effect). In perfect analogy to that music in audible form is the reason that the musician wrote it, and to divorce him from the performance of the work on a machine is to deny him the reason that the work is authored (wonder what RIAA has to say about that.. The data of CDs not fully protected under copyright, since it can be technical method for producing soundwawes with a microchip and other technical equipment).

If countries really where allowed to do this, reinterpret laws and redefine software to be a technical field, countries could as well do it to TRIPS article 27:

Patents shall be available for any [A], whether products or processes in all fields of technology.

Now suppose Europe legisaltes that:
[A] is not an [A] as such if [condition that is always true]

I'm quite sure we would see the U.S take E.U straight to the WTO if we tried this....

If lawyers can redefine what a computer program is to make it fit their purpose by their preferential right of interpretation, then they could just as well say that the earth is flat to fit another purpose, would politicians also be fooled by this?

Is it reasonable that ones right to make money through a granted monopoly outwights another mans freedom of expression? - If someone says yes.. i hope we not meet IRL. Voltaire: "I may disagree with what you say, but I will defend to the death your right to say it"

It all boils down to the word "unreasonably". Is it reasonable to allow someone to prejudice authors legitimate interests by allowing patents on the reason that the author wrote his work?

One could argue that copyright isn't all that exclusive, there are other laws too. - But of course authors have to live under the law, but nations isn't allowed unfairly prejudice authors as they please. This logic fails when taken to extremes: Do authors have all the rights to publish their works even if they are illegal according to, say censorship (copyright totally exclusive to other laws)? No of course not - this would mean that authors could do whatever they like illegally. Can Governments do whatever they like without respect of how these affect the exclusive rights of authors? No of course not- the treaties would be worthless. If you take either to the extreme it would fail because it would mean either that the treaty was worthless or that people can write illegal content. The answer lies in between and the unreasonable prejudice standard is key.

Luckily, lawyers do not exist in a vacuum away from the laws of physics and economics anymore than programmers live independently from the law. What lawyers accepts or not isn't law in an absolute sense, there is no agreed starting point. There is a basic reason to why we have both prosecuting and defending lawyers, but it will be a huge problem for the justice if lawyers on both sides, with their preferential right of interpretation in court, doesent accept truth and consequence and treats their interpretations as absolute. I think we have a little of that problem today, especially when politicians put too much weight on this kind of orthodox legal thinking rather than the law and it's consequences, especially when you clearly can see what that way of thinking has led to in the US.

[ Reply to This | # ]

Groklaw © Copyright 2003-2013 Pamela Jones.
All trademarks and copyrights on this page are owned by their respective owners.
Comments are owned by the individual posters.

PJ's articles are licensed under a Creative Commons License. ( Details )