Fog Creek Software
Discussion Board




Software patents

Have you ever applied for a software patent?  If so, did you do it with a lawyer or handle it yourself?  Which is the better way to handle it?

Lee
Monday, December 01, 2003

Yes.  Corporate lawyer handles it.

hoser
Monday, December 01, 2003

Well in my case since I used to work for a small company we had to hire a outside patent lawyer (at $300/hr!) to do it and I was amazed to see how lacking in technical knowledge the patent lawyer was (or may be that was a tactic to spend more hours!) but in the end we got our two patents so I would say it was worth it to the company!

Code Monkey
Monday, December 01, 2003

Nuts.  I need a new screen name.

Used to be Lee
Monday, December 01, 2003

The patenting process is byzantine, and it's not worthwhile to apply for a patent unless you can hire a patent attorney to walk you through the legal minefields.  (Patent attorneys have law degrees plus specialized training in patent prosecution, and usually have an engineering background.)  If you make any false step during the patent application process, you'll either get no patent or a patent that's unenforceable in court if someone contests it.

I applied for a software patent several years ago.  Even though I'm a (non-patent) attorney, I knew my limitations and hired an experienced patent attorney.  It was expensive but absolutely necessary.

Of course, a software patent isn't necessarily the best solution for protecting your intellectual property.  Copyright and trade secret law can provide different types of protection, and software patents are (in theory, not always in practice) difficult to obtain.

Now let the "software patents are evil" debate commence.... <g>

Robert Jacobson
Monday, December 01, 2003

Definately get an attorney.

Engineers are too honest for patent applications. Have you ever read those things? The goal is to bury the application in so much gobbledygook that the inspector goes "Ok, I guess" and rubber stamps it. Laywers are a *lot* better at that kind of stuff than engineers.

Chris Tavares
Monday, December 01, 2003

While you may not be able to perform the whole process yourself, you can probably save yourself some money if you follow these steps:

1. Write down what your invention is, when you invented it, why it's unlike the stuff currently available, and how it works. Print this out.

2. Take your written description to a patent lawyer or two for a consultation. Be prepared to pay, but perhaps some may offer a free consultation. Also if you decide to choose a lawyer, talk about the total costs up front so there are no surprises.  Ask about ways you can help out to reduce the cost, if appropriate. You may wish to finish the fee negotiations by letter or email so there's some kind of "paper" trail.

3. Visit a local patent library or search online (uspto.gov) for related patents. Refine your written document. Try to write the patent yourself.

4. Take your description to the patent lawyer (ideally in machine-readable format) and have them massage and refine it into a form suitable for filing. You will probably go through a couple of drafts. Read them very carefully and mark them up. The attorney will not understand the nuances of your invention as well as you do.

5. Be aware that there is a very strict time limit of one year (I think) from the time you first demonstrate (i.e. at a trade show) or sell the invention and when you file your application. I think there are limits to the way you can sell the invention before you apply for the patent.

By the way these suggestions apply to US patents and is based on what I did to get a patent in 1993. 1993 was a long time ago, however, and many things have changed.

IANAL
Monday, December 01, 2003

"Now let the "software patents are evil" debate commence"

Ok!

It's not that they are "evil", it's just that they are so routinely abused and frequently take advantage of our relatively clueless Patent Office.

Amazon's One Click (or whatever it's called) is a shining example of absurdity in software patents. Another example was the company that claimed to have a patent on e-commerce that occurred on a computer with a CRT. For the love of God, stop the madness...

Another favorite is F5's patent on "Cookie Persistence Technology" (oh geesh) that basically allows a browser to reconnect to the same web server it visited previously within a cluster or web farm. According to F5, they invested "millions" in designing this technology. Millions you say? Then you are obviously better at abusing patent law than writing code...

The key problem is that if you can convince the not-so-clever clerks at the Patent Office to grant a patent, then it becomes enormously expensive for someone to say "Hey! There was a ton of prior art for that!" because now they have to hire attorneys and it becomes a litigation nightmare.

Mark Hoffman
Monday, December 01, 2003

"It's not that they are "evil"  "

<G> I beg to differ.

software patents _are_ evil.

_and_ they are genuinely stupid, its the equivalent of allowing writers to patent their stories.

When I sit down to code, there are certain ideas im _not_ allowed to have.  certain ideas that, should I come up with them, i am _not_ allowed to turn into (released) code.

wtf?  its my idea, if I turn it into code the design is mine, and the code is mine.
_but_ im not allowed to turn it into code.

thats the first problem with software patents.

The second is more insidious:

What are those ideas?  I literally have not a clue.  Im a programmer, not a patent lawyer and I spend my life writing code, not reading software patents.

but soon,  reading patents is something I am going to have to start doing.

Im not going to be able to write commercial code unless I know that its patent free. (well..I _can_ write it, but after doing so I will have to totally rewrite it again before I can release it, a couple of mistakes like that and my small business will be down the gurgler).
but wait, it gets worse....
none of my programmers are able to write that code either.  So I dont just have to learn about software patents myself, but also I have to teach my employees.

Patented software is going to cripple the future of this industry.

Its stupid and unnecessary.

FullNameRequired
Tuesday, December 02, 2003

if you have a brilliant idea for a software algorithm, write a program that uses it, publish it in a magazine somewhere, take all the credit and the profits of being first in the market and then let someone else build on your work as you have built on the work of those programmers, mathematicians, scientists, philosophers and thinkers who have come before you.

FullNameRequired
Tuesday, December 02, 2003

> Patented software is going to cripple the future of this industry. <

Rubbish.  This debate is nothing new; it's probably been existing since the first patent was issued in 1790.  ("Damn farmer Joe for patenting that new plow!  It's going to cripple the wheat industry!")

Patents exist to provide financial incentives for inventors to create novel and useful new inventions.  Sometimes they cover truly novel ideas, like Thomas Edison's patent on the first telephone.  Sometimes they cover more mundane, but useful, ideas, like the patent for intermittent windshield wipers.  Sometimes, as Mark notes, they are issued for completely obvious ideas (the Amazon patent may be "Exhibit A") -- but in those cases, the patents are unlikely to hold up in court.

There's plenty of room for debate over what is sufficiently deserving to qualify for patent protection, and whether the U.S. Patent and Trademark Office's method for examining patent applications should be improved.  However, there's no reason to throw the baby out with the bathwater.  If an inventor comes up with a truly unique idea, then he should be able to apply for a patent.  There's no reason that the software industry should be treated any differently than any other profession.

Robert Jacobson
Tuesday, December 02, 2003

Just applied for a patent myself. What surprised me most is that patent attourneys are basically scribes - all they do is take a description of the invention (which you must write) and re-write it in legalese. So the quality of the patent will depend heavily on the description you write. (the claims can be amended later, but the initial spec MUST describe the invention completely)

It is up to you whether to do a patent search or not. A thorough search can help the prosecution process, but on the other hand, it will expose you to related patents (and thus liability for treble damages if you infringe them). For this reason most inventors deliberately stay AWAY from looking at others' patents. (the USPTO is considering new guidelines on treble damages in order to avoid this catch-22)

Dan Maas
Tuesday, December 02, 2003

I think I'm going to patent religious wars online, you know ill get the patent if I have the money, and then ill start lots of c vs java, emacs vs vi and windows vs linux arguments on here.

I'll be rich.

fw
Tuesday, December 02, 2003

This thread is going to kick up so much bollocks I better get everything in at once.

*History Lesson*
To start let's recognise patents what they are - a legally granted monopoly.  Second, the first patent was not issued in 1790 but wayyyy back -  to  quote the UKPO website:

"The earliest known English patent for invention was granted by Henry VI to Flemish-born John of Utynam in 1449."

Until 1624 , patents were granted more or less however the reigning monarch felt like granting them.  At that point  Parliament outlawed monopolies except (again from  www.patent.gov.uk ) "those "for the term of 14 years or under hereafter to be made of the sole working or making of any manner of new manufactures within this Realm to the true and first inventor" providing they were not "contrary to the law nor mischievous to the State by raising prices of commodities at home or hurt of trade".  This is the origin of US patent law which is enshrined in their constitution.

The deal in essence is that you get a time-limited monoploy on your invention providing you cough up how it works (so other people can have a go afterwards).  Since development time has increased with mechanical and *especially* chemical inventions so the time-limits have been increased.

*My problems with US Software Patents*
The principle argument against is that software is a branch of maths and you don't invent maths you discover it.  I'm not entirely convinced - some software patents could fall into this catagory (a new algorithm say) whereas some relate to software as a product. 

More important in my view is that the length of a patent is disproportionate given development time in the industry and that software patents don't seem to document methods sufficiently.  Mechanical patents require drawings - software patents should require code.

*My advice*
Document you invention, and the state of the art. 

See a patent attorney and listen to his advice. 

If you decide that a patent isn't financially viable publish your method so you have a prior art defence when some other bugger decides that law is more profitable than development.

A cynic writes
Tuesday, December 02, 2003

As IANAL wrote above there is still one year period between you publishing your invention and applying a patent. But it onyl applies to US. If you reveal your invention beforehand to public you can not apply a ptent in europian countries anymore (I don't know about asia).

I have experience only on Finnish patent attorneys (plus couple of attorneys in US handling the US patenting). We have selected those attorneys who have knowledge about the techonolgy we are applying the patent for. I have two friends working in patent offices (both have MSc. in Tech) and they are specialised in certain technologies.

The biggest help we have had, is to get their (good) opinion what should be the minimum coverage of a certain patent. Of course they need also help from inventor, but they can guide us. They also take care of all the paperwork, which is quite demanding e.g. for dates. If you miss one date, your patent is gone...

Otto
Tuesday, December 02, 2003

Yep - in the UK prior public disclosure invalidates a patent.  Wierdly so does threatening to sue for infringement (suing is fine, it's the threat that's not allowed). 

Go on  - see a patent attorney. 

A cynic writes
Tuesday, December 02, 2003

They're only evil if they're not mine!

Fred
Tuesday, December 02, 2003

"Patents exist to provide financial incentives for inventors to create novel and useful new inventions.  "

thats rather the point ism't it?

Im _not_ against patents in general, Im against _software patents_

Software algorithms are _not_ invented, they are discovered.

FullNameRequired
Tuesday, December 02, 2003

"Patents exist to provide financial incentives for inventors to create novel and useful new inventions.  "

Yes, they do, but that is not normally the reason people obtain patents on software nowadays. Many companies patent obvious ideas for the sole purpose of fleecing other companies of their revenue.

Want to argue that you had the idea first? Sure, go hire an army of lawyers to defend yourself and go bankrupt in the process...or...just pay us this little tiny royalty and nobody gets hurt. Kapeesh?

Ever notice that many of these extortionist only target small companies and oddly enough never go after the large companies that appear to be infringing their patent? Hmmm..I wonder why that is..Perhaps it's because the extortionists know that they would be squashed in court. So they target the little guys who can't afford a protracted legal battle.

I'm not against software patents. They have their place. My beef is with the rampant abuse of software patents that are stifling creativity and productivity. The Patent Office needs to do a little more than rubber stamp every application that attempts to patent ideas that already been in use for years.

Mark Hoffman
Tuesday, December 02, 2003

Agree 100%, Mark.  Unfortunately, these abuses happen in every industry, not just software.  Some companies go out of their way to get "blocking" patents just to frustrate competitors, without any intent of actually using those patents themselves.  The patent system is far from perfect.

Robert Jacobson
Tuesday, December 02, 2003

"Software algorithms are _not_ invented, they are discovered."

Whether they are discovered or invented, they should not be eligible for patents because patents are not necessary to promote progress in software.  Most who develop software would have done it anyway without patents.  Most of today's software is built on the algorithms that were put forth by Knuth, Dijkstra, and others in the days before software was deemed patentable.

The multitude of accidental patent infringers is also evidence that software patents are unnecessary.  If they infringed unintentially, they obviously did not need ownership of a patent on the concept in order to have incentive to implement the concept.

If there is anyone who would have decided not to develop something because software patents don't exist, before long there will be someone else willing to develop an equivalent thing.

Software already has trade secret and copyright protection; adding patents as a third form of protection is redundant and hinders progress.

T. Norman
Tuesday, December 02, 2003

*  Recent Topics

*  Fog Creek Home