Hi Cathy,
My direct experience is with the pre- and post-prosecution research of software patents. Funny thing with software patents, there tends to be a negative correlation between the amount of research done pre-prosecution (patentability, market research, etc.) and the amount of prior art searching required post-prosecution (invalidity, infringement litigation). If the client files blind, either the examiner or a well financed legal team will find the prior art at the worst possible time. Software patents due require a good deal of thought and consulting pre-filing on whether patent protection is even the way to go.

The 5 year pendency for software is a problem given its speed of innovation. A strategy several of the startups and attorneys I work with use is to file a provisional application or PCT to get the early filing date. This allows them the time to continue working on their system and business model and it will also give them patent pending status which may provide them with much needed negotiating leverage. Then, when the 12 month deadline approaches, the team can evaluate where they are and see if it's wise to proceed with a full application.

Let's say they do proceed, they don't need to wait until the patent issues to sell or partner with another company. Once the application becomes public, if their invention has merit, the marketplace will seek it out or, like 95% of the patents issued, ignore it.

posted 8 months ago

Cathy,

John alluded to his answer to another recent software patent question, one I thought you answered well, and I'll refer you to my answer in the same. Instead of repeating any of those comments, I'm going to assume the pulpit on different issues you raised.

If I've got my patent history right, Lemelson used to file patent applications and try to get the prosecution on them buried for 10 years or more. Then, if they issued when the technology was being practiced, he would have ready-made licensees that he'd bully into licenses. These were called submarine patents (hiding below the surface). The USPTO changed their rules to combat this practice, changing the term of a patent from 17 years from the issue date to 20 years from filing (with caveats to protect inventors from patent office delays). The patent office delays are forcing most software patents to operate as Lemelson's submarine patents did (although those who allow their applications to publish are easier to spot). It is interesting that when Lemelson did it, it was a benefit to the inventor, but now it is a hardship. I think software patents get less love outside the US than within, so I regularly encourage clients to request nonpublication and follow Lemelson's lead (when you have lemons ...).

I don't think the USPTO's present administration likes software patents, but they don't know what to do about it. Games are played with the definitions of 'tangible, useful results' and 'technological arts' to marginalize the software patents. I had an accounting method patent that was required to be claimed as a software patent a few years ago under a 'technological arts' interpretation, although the invention had nothing whatsoever to do with software (although it could have been calculated using a computer). Similarly, I think they understaff some more complicated technologies because finding qualified people is difficult, while I think they understaff the software patent groups to slow the process while they continue to muck around.

Finally, with regards to the lifespan issue, I would discourage anyone from obtaining a patent on any invention in any field if they think its useful lifecycle is five years or less. If that's the case, the money to be made is being first to market, so focus your resources there and by the time anyone catches up, the technology will be too obsolete to copy. To that end, I would agree with John - focus on 'what it does' as that will have a longer life cycle than a specific software or device. Also, bear in mind that the patent law seems cyclical, so in five years software patents may be all the rage again, but you have to file one now to profit from it.

Regards,
Todd

posted 8 months ago

Cathy:

I think there's a larger question here that I've not seen addressed in any other posts... and that's the "why?". Why would someone want a software patent opposed to any of the other IP protections currently available to them.

Generally speaking, as someone who handles these things all the time, I would say that patents are the wrong IP protection to use with software. They take forever to get, you have to give up your IP to get them... and there doesn't seem to be any real sense of security that the USPTO won't later invalidate.

Rather, I ALWAYS suggest that software be protected as a trade secret... you don't have to give up any of the code, you protect yourself with a solid license, and you can keep the secret forever. For example, do a patent search for Microsoft. You'll find all sorts of hardware devices... and a random piece of software combined with hardware. But look for Windows or Office. You just won't find it.

On the other hand, if what your client is trying to protect is a process (of which software enables), then patent the process, but leave the software to be secret. That way, by the time the process patent expires, there's still a piece of secrecy protecting your client's competitive advantage.

posted 7 months ago

Let me answer this in purely a theoretical sense. I say this as someone with many years of software development experience prior to law school.

First of all, I repeat what I just wrote in another answer. The PTO routinely grants software patents for nonpatentable subject matter and those clearly within prior art. I even read a software patent involved in an infringement suit where the patent claims had no relation to the description of the invention. In some cases these patents have been validated by the Federal Circuit.

1. Given this environment a client may wish to patent for purely defensive purposes. In some computing areas (dictionary-based compression for example) there are so many conflicting patents that it is probably impossible to write an application in the area that is not infringing. They are many software patents that are no more than "Do X {Where X is some long standing business practice} on a computer." In certain situations, having your own patents may help in defense of an infringement suit.

2. Algorithms tend to be in use longer than a particular software implementation. While mathematics is not supposed to be patentable, describe the process as machine and it is treated as such.

3. You don't want to patent a particular software implementation but rather what it does. Take for example Amazon's infamous "One-click" patent. The "what it does" is also going to last longer than the particular implementation.

4. Many software patents have a list of specific claims (often prior are) with one catch all claim. See State Street Bank & Trust Co. v. Signature Financial Group, Inc. 149 F.3d 1368 for a good laughable software patent along this model that is now frequently-cited precedent.It appears from the case law that you can get a lot of coverage out of a few patents.

Among the few down sides to software patents are the cost and the revelation of information.

Some thoughts that I can type into this 6 line window.

posted 8 months ago

I have been preparing and prosecuting software patents for over a decade now. I have found that it can be very valuable for clients who:
- are developing software that was difficult to conceive and/or time-consuming and costly to develop, yet easy to copy once you see it running or even read an explanation of what it does; and
- who have not yet created a working version of the software but who need to attract investors, employees, and/or partners.

The slow processing of patents is not as much of a problem in practice as it might seem at first. Even having a patent application on file provides some benefits. For example, it provides you with a certain degree of protection against someone else obtaining a patent on the same invention as you. Investors will often take you more seriously and have higher confidence in the viability of your business if you have performed a thorough prior art search and then prepared and filed a solid patent application. The patent application itself can also be a valuable asset in a merger or acquisition.

Links:

• http://www.rplotkin.com/SpecialConsiderationsforPatentingSoftware.pdf

Robert P also suggests this expert on this topic:

• Norm G

posted 7 months ago

Dear Cathy:

The timing of your question is quite good. Last week I gave a webinar on the current state of the law with patenting machine and computer implemented inventions, and on February 26-27, 2008, World Research Group is hosting a whole conference on the topic of Patenting Business Methods for Internet-Based Businesses in New York City. I have attached the link below for your reference.

While patenting "software as software" is currently out of fashion at the PTO today, at least according to several recent BPAI decisions, patent protection is available for software when claimed properly. Experienced patent prosecutors can help with this issue, as we often do for our clients.

The point of patents is to protect as a tangible asset, what would otherwise be a intangible idea. When advising clients, I would suggest that patent protection, if appropriate, should definitely be considered, within the confines of the client's budget. Many internet businesses that are no longer functioning have sold their patent portfolios (or patent pending portfolios) as one of the major assets of the company.

Of course, patent protection is only one piece of the internet puzzle. I have also attached as a resource, an article we prepared about how intellectual property can be used to protect an internet website.

While you are right that the long period of time before the PTO will first consider the patentable merits, and the reluctance on the part of the PTO (at least at the present time) to even want to consider the patentable merits in many instances, can make patent protection both costly and perhaps too late to be of maximum use, it is still worth pursuing nonetheless in many instances.

Best regards,

Charley Macedo
Amster, Rothstein & Ebenstein LLP

Links:

• http://www.worldrg.com/showConference.cfm?confCode=FW08016&field=summary
• http://www.arelaw.com/downloads/ARElaw_IPProtectYourSite.pdf

posted 7 months ago

NOT LEGAL ADVISE NOR CLIENT RELATIONSHIP

A patent is one of the few ways of effectively protecting software. If it is a good idea and can make money, it will be copied and stolen. Period.I have had it happen to me with over 55 patents in my name. Software is easy to build and few other vehicles will help protect the investment made. Trade secret is really only effective if the software is never released to the public.

As far as delays are concerned, its the government, and every thing takes time. I think if it is a good invention, it will have a lifespan in todays market and technology of over 5 years. Rights are established at the filing date, but few IP attorneys generally want to start to try to enforce them - it is much better to wait until the grant,

Personally, I think you might want to find another party to refer the work out towards.If your clients are not establishing these rights, I am sure their competitors will try to, which will make it more difficult on your clients later on.

Anthony D also suggests this expert on this topic:

• Mike R

posted 7 months ago