Archive for the ‘Intellectual Property’ Category

Open Source Community Gains Patents, Support

Monday, January 31st, 2005

Earlier this month, IBM pledged 500 patents to the open source community to support development efforts in a wide range of fields, ranging from databases and networking to e-commerce and compression. IBM’s pledge of non-assertion will cover any open source project being developing under a license approved by the Open Source Initiative (OSI) (

During the analyst briefing, Jim Stallings, IBM’s Vice President of Intellectual Property and Standards in the IBM Technology and Intellectual Property Group, explained that IBM had taken the initiative to create a so-called patent commons to help reduce the fear and concern developers had about patents being asserted against technology and products they were developing, and hoped that other companies would follow suit.

The 500 U.S. patents represent just a mere fraction of IBM’s portfolio of over 41,000 U.S. and nearly 30,000 European patents – a portfolio which is growing at over 3,000 patents a year in the U.S. alone.

Stallings indicated that IBM wishes to foster collaborative innovation and enable open standards and that the initial pledge of 500 patents was only the start, and that more patents would be added to the pledge over time.

In a blatant and self-admitted attempt to grandstand and eclipse IBM’s pledge, Sun Microsystems last week upped the ante by granting a license to 1,672 patents, but only under the OSI-approved Common Development and Distribution License (CDDL) – a license that Sun derived from the Mozilla Public License. Sun’s announcement went hand-in-hand with the release of its Solaris operating system as an open source project, under the name OpenSolaris. Sun indicated that the patents it is granting a royalty-free license to under the CDDL are all directly related to Solaris.

The volume of patents IBM and Sun have basically donated for at least parts of the Open Source community to us is impressive, but what does it all really mean for developers?

In general terms, a patent grants the patent owner the exclusive right, for a period of time, to restrict others from making or using the invention covered by the patent. It is often overlooked, or at least misunderstood, that merely owning a patent to a particular invention does not grant the patent owner the right to make the invention. That’s because the invention could also be covered by other patents not owned by the patent owner.

Thus, the effect for developers of these two sets of patent licenses or pledges is that as long as they adhere to the terms of these licenses, they are assured that only IBM and Sun will not go after them for patent infringement. That is not to say some other company might not take it upon themselves to assert patents they own against certain Open Source developers or distributors of Open Source software.

IBM’s pledge, which more broadly covers all open source software developed under OSI-approved licenses, including the popular GNU Public License and the Mozilla Public License, also includes one exception. Namely, the pledge to non-assert is withdrawn from any entity “who files a lawsuit asserting patents or other intellectual property rights against Open Source Software”. This exception may well prevent one Open Source developer from suing another one, but would have no impact on closed source proponents, of which Microsoft is perhaps the largest and most vocal (ignoring for the moment the inevitable cross license agreements in place between Microsoft and IBM in particular).

Sun’s license, which addresses only the CDDL, and thus apparently only the further open source development of OpenSolaris, is thus much narrower in scope and cover. Further limiting its utility are numerous exceptions to the patent license as documented in the CDDL.

The value of IBM’s and Sun’s patent license offer to the Open Source community should not be totally discounted, as both firms have extensive R&D and patenting efforts and do produce meaningful technology. But on the whole, without a host of other significant companies bellying up to the bar to offer their patents on royalty free terms to the Open Source community, IBM and Sun’s offers are but mere tokens – good for public relations, and possibly good for inciting other companies to follow their lead with patents of questionable utility and coverage, but not much else.

All it takes is one small patent holding company with a good patent or two to cause major upset in the Open Source community, and then all the patents IBM and Sun own would not make one bit of difference.



U.S. patents granted to Sun Microsystems by year – total granted since Sun’s founding = 4,288

Year U.S. Patents Granted
2004 651
2003 589
2002 511
2001 374
2000 483
1999 582
1998 453
1997 163
1996 117
1995 77

U.S. patents granted to IBM by year since 1981 – total granted since IBM’s founding = 41,372

Year U.S. Patents Granted
2004 3048
2003 3458
2002 3343
2001 3006
2000 2951
1999 2834
1998 2727
1997 1809
1996 1938
1995 1476
1994 1365
1993 1142
1992 875
1991 709
1990 638
1989 655
1988 575
1987 636
1986 632
1985 588
1984 618
1983 488
1982 453
1981 522

(Note – above information courtesy of Delphion. Includes patents of all subsidiaries and controlled entities.)

A New Age In Software Licensing

Tuesday, July 11th, 1995

(This column first appeared in the July 11, 1995 issue of PC Graphics Report)

Most, if not all, of you are in the hardware business. And, hopefully you realize that without software, your hardware is close to useless. Software is the fuel that makes hardware go. The bigger and unwieldier the software, the bigger and more expensive the hardware needed to make it run properly.

It’s a cycle, of course. Hardware improves in capabilities, which results in existing software being revised with more features and bulk to suck up the extra bandwidth the new hardware provides. That in turn helps drive the creation of even faster and more capable hardware, which creates even bulkier and feature laden software. And so on, and so on…

However intertwined hardware and software are, certain undeniable sales differences exist. A given user only needs one clump of hardware to run a multitude of different software packages, but as the SPA is quick to point out, a fair number of users don’t pay for those software packages. But, they have to pay for the hardware regardless, because it can’t be easily pirated.

Of course, both hardware and software companies make a significant part of their revenue by selling users newer versions of what the users already have. The percentages between new sales and upgrade sales differ between hardware and software, but not significantly on the whole.

The Cycle Wobbles
The perpetual upgrade cycle to have the latest and greatest hardware and software is starting to wobble a bit. Many users, especially in the corporate world, are increasingly resisting the purchase of upgrades to hardware and software solely for the sake of the upgrade. The bulk of today’s hardware and software sales appear to be coming from the consumer market, which has newly rediscovered the value of having a home PC.

Upgrade resistance is borne of a number of factors. On the software side, software companies have over the years made their offerings so feature rich and powerful that new upgrades appeal to an ever smaller, more specialized segment of the market, leaving the larger part of the market unenthused. Several examples include:

  • Word processing – most people using word processing software didn’t come close to using all the features their old word processing software had (some of us dinosaurs even still use, gasp!, the DOS versions of popular WP software);
  • Desktop publishing – the greatest volume of DTP usage is still black and white, since full color reproduction is still rather costly in comparison, but leading DTP software companies have run out of truly useful black and white features to add to their software, and are instead adding more and more full color manipulation and output support to new versions of their software.
  • CAD – All the big names are adding, almost exclusively, 3D capabilities galore to their upgraded CAD packages, even though the majority of CAD users are still 2D oriented.

Most users don’t care to spend the additional $500, $100, or even $50 an upgrade costs, because they can’t justify the expense based on features they won’t be using.

PC hardware is suffering similarly in markets where such hardware has existed for long periods of time (i.e. the corporate world), which is why sales of Pentium systems in such markets are a shadow of those into the new, booming consumer market. Users (and even some noted columnists) have been quoted in various national publications as finding the performance of their 486DX2/66 systems as adequate for their needs. It logically follows that since the overall system performance impact of the newest graphics technologies is minor when compared to the previous generation, that graphics hardware sales may suffer as well from the “upgrade apathy”. The only saving grace for hardware sales is that hardware will fail sooner or later and need to be replaced. Software can pretty much run forever under normal use, and therefore doesn’t benefit from failure.

In both cases, one obvious way to reverse the downward upgrade trend would be to significantly increase the benefit to the user of the upgrade, but this is the course that software companies have been trying to follow until now, without significant success. There is more hope for the display hardware industry however, with the increasing demand for real-time, high resolution digital video playback and the anticipated market for 3D – new features that aren’t truly satisfied by existing hardware capabilities.

Another way to increase upgrades is to obsolete the current implementation, usually in theform of OS-specific support. This minimizes the options a user has in adding necessary components to his system as his or her computer use expands. With no support for such expansion, the user is forced to have to upgrade his software, and frequently the operating environment. Alas, not too many companies can accomplish this entirely by themselves.

There’s only one company that has enough clout to do this by itself, and that’s Microsoft.

And the latest effort at getting those valuable software upgrade revenues is called Windows 95. Already we’re starting to see 32-bit applications being announced which will not run on previous PC operating systems, as well as hints that current PC OSes will be phased out of the channel eventually. Where other companies have helped is by hopping on  the bandwagon and developing Win32 applications which have no hope of running on Windows 3.1x.

The most ideal model for software companies would be to have “disposable” or “consumable” software, which would require users to repurchase software on a regular basis, as the previous software purchase gets used up. A variant of this is software that has a naturally limited life, like a tax package targeted at a specific year, a virus checker, or an annual compendium of current events. However, without something which expires tied to software, this model is impossible to achieve. Or is it…?

What’s Old is New
Back in the days of mainframes, software companies used to license software over a period of time. Software packages would be licensed annually, with a hefty advance payment. This practice was justified by virtue of the hundreds of users that might all be using the same software, as well as the on-going development that was frequently required to adapt a huge general purpose package to a company’s specific needs. There was also significantly less competition at the time.

With single user systems and an exploding software market, the mainframe licensing model went out of style rather rapidly, but now, some companies are bringing variants back.

First, one of the more innovative licensing models I’ve seen announced comes from Graphisoft, with their ArchiCAD product. ArchiCAD is one of many products in the higher end PC CAD market, selling for just under $5000. Since ArchiCAD has such a high price of entry, the company was apparently finding it difficult to attract as many new users as they would have liked. So, how do you sell a costly software package affordably without cheapening the product like CADKEY, another competitor did when it dropped its pricefrom over $3000 to less than $500? Graphisoft’s answer was to rent out the software in a program they call “Pay per use”. For a setup fee of $795, which covers administrative and materials costs and training, Graphisoft will send a customer a full ArchiCAD package, along with a special hardware lock. This hardware lock meters the use of the software, and is good for 50 hours of active use. The cost of using ArchiCAD works out to be $3.83/hour. When the hardware lock is used up, another needs to be purchased.

While Graphisoft’s solution is a very good one, it could be significantly improved if it did away with the added cost of the hardware locks, and instead offered monthly electronic billing, with the computer automatically dialing in to a central accounting office on a regular basis, much in the way that the pay-per-view feature works on the new breed of digital satellite systems (DSS) for TV. (I spoke with Graphisoft, and they indicated that they are working in this direction, as well as on a functional metering approach, which may charge different at different hourly rates for various program functions – rendering might incur a lower cost because it’s more machine dependent, while a vertical market specific function might cost more.)

Microsoft has entered the market with a new license model as well, namely the software subscription. Microsoft started this with its developers, via the Microsoft Developer’s Network, where, for an annual fee, you get 4 quarterly releases of CDs containing lots of valuable software resources, if you’re a developer. The license applies to a single named individual, and is not supposed to be shared among developers, but Microsoft currently has no way to verify or ensure that this requirement is being observed at all companies which have one or more subscribers to MSDN. However, the subscription does successfully create a significant annual revenue stream for Microsoft, something which under a traditional upgrade model would be less likely (and more of an administrative headache).

Microsoft has taken the subscription model to an application recently as well, namely its Visual C++ compiler. The subscription guarantees several updates to the compiler per year, which is important to developers, as Visual C++ is a moving target of sorts. The benefits to Microsoft are the steady revenue stream, as well as the ability to make changes and bug fixes as needed and update its active customer base, with that customer base paying for such fixes and enhancement.

Finally, Microsoft has promised to use a similar mechanism for Windows 95, although the general scuttlebutt is that this is to allow Microsoft to provide regular bug fixes under the auspice of “Update Packs”, all with customers paying for the privilege of getting a product that some say is not ready for retail sale. Whether the latter is true or not, I suspect we’ll start seeing more and more companies ship product earlier than it should be and use the “Update Pack” subscription strategy to correct any problems in the software at a later date.

By the way, for those of you that didn’t see the Cringely column in InfoWorld a couple of weeks ago, there’s an unconfirmed rumor floating around that Microsoft will put a hardware lock on each copy of Windows 95. I think it’s unlikely, but stranger things have happened.

The Future of Software Licensing
Both Graphisoft and Microsoft are pointing to the future of software licensing. The missing component to make it all work better, though, is global (or at least nation-wide) networking. I envision that within a couple of year’s time, you’ll be able to automatically download the latest version of your word processor in the background off the Net, under a combined subscription and metered license plan. For applications you use infrequently, more of a metering approach would make the most sense – only pay for when you use it, while for indispensable applications, you’d use a subscription service (or at least hope that your several-year old application works in some sort of compatibility mode).

A few more years down the road, and you’ll start using applications remotely in real time and be metered that way, and we’ll be back to the mainframe days, albeit with distributed processing in place.

You’ll definitely have some interest problems arise along the way though. First, once a large number of people are connected to the Net on a regular basis, software piracy will become more carefully monitored and hunted down, because the operating system (or at least the network connection) will in all probability allow the client software to peruse your system and see what you have installed, and report it back to some monitoring agency (or the actual software company whose software you’ve got installed). This will lead to the commercialization of local firewalls that prevent such utilities from detecting everything on your system, and we’ll be back to the whole vicious hack-crack-protect cycle again.

You may think that this is all just a tinge of paranoia, which, of course, brings us back to the present, and a thing called Microsoft Network. A recent issue of a major industry weekly described how the Microsoft Network connection software that is part of the Windows 95 Beta, and supposed to be part of the release version of Win95 if the Justice Department permits it, examines your system’s hard drive to locate all Microsoft applications installed on your system. This list of applications is quietly reported back to Microsoft electronically as part of the on-line registration process. Microsoft officials were quoted as saying that the MSN software is only doing what Microsoft would ask users to provide on a registration card anyway. This invasive procedure is only the start of what might come down the pike in terms of system software monitoring, stuff that wouldn’t be possible without a network connection (or modem). It will help software companies curb piracy though, which probably is a good thing (there’s a theory that some piracy actually increases legitimate software sales).

In any event, in exchange for the immense resources the proverbial Net provides, you will give up some system privacy.

What This Means To You
In the home, and in many offices, these software licensing changes and increased networking capabilities mean that the PC will become more like the phone or cable TV box or DSS dish in terms of billing. Users will just get a monthly bill for software services provided to them. This also more than likely means that they will spend much more on  software annually than they have in the past (but for more variety), which may mean less of a budget for hardware, unless the software company which receives the largest  monthly software revenues decides to loan/lease/rent PCs to consumers so that they can suck up more software (just like in the Cable world with set-top boxes). It’s all back to the Gillette approach – give away the razors to sell more razor blades.

Either way, it calls for more standardization of base hardware platforms so that a common set of features can be depended on for a minimum configuration. The next year or two will cause that base configuration to freeze, so get your dibs in now with the best 3D and video playback technology you can, and don’t forget to start looking at high speed communications technologies to integrate with your devices.

Intellectual Property, Part 2

Tuesday, March 8th, 1994

(This column first appeared in the March 8, 1994 issue of PC Graphics Report)

In my last column I discussed the lack of Intellectual Property (IP) disputes in the PC graphics industry, and some possible reasons for this blessing. In this week’s column, I’ll give you an overview of what IP really is so that you and your company can even better avoid any IP disputes in the future (in particular, the nasty scenario I presented at the end of the last column).

IP commonly covers three areas: Copyrights, Trademarks, and Patents.

Perhaps the most basic form of IP, copyrights imbue their owners with precisely what the name implies, namely the right to copy (as well as the right to prevent others from copying).

Every “work”, upon its creation, is automatically copyrighted, just by virtue of its “being”, in a tangible form. This applies to software, ad copy, articles (like this one), new user interfaces (much to my dismay), music, presentations, drawings, pictures, etc. Unless the creator of the work has explicitly transferred his or her rights in the work to another party, such as an employer, then the creator exclusively owns all rights in the work.

This last item is very important to any of you who use consultants, since without a consulting agreement that specifically assigns the rights in a work to you, the consultant owns all the rights to the work, even if it’s being created for you, and even though you are paying to have the work done. Consequently, without an agreement, the consultant would be free to sell the work to others.

There are two types of copyrights, unregistered and registered. Unregistered copyrights are what I just described above, while registering a copyright requires action on the part of the owner of the work. Registering a copyright is quite easy and inexpensive, just a two page form and a $20 registration fee. For written or visual works, a copy of the whole work needs to be sent in. For software, only the first 25 and last 25 pages of the source code need to be sent in for registration. It takes about 2-3 months, from my experience, to get a copyright registered. Copyright registrations are valid for the natural life of a person plus 50 years, or for 75 years after publication for works authored by a corporation (or employees of a corporation as “works made for hire”).

Now, you’re probably asking yourself why you should bother registering a copyright. Quite simply, if someone is infringing on your copyright (i.e. distributing your “work” or a derivative of your work, without your explicit permission), you can’t go after them legally unless your copyright is registered. And, cease and desist letters don’t carry a whole lot of weight if you can’t back them up with proper legal action.

On the flip side, if you don’t go after someone who you know is infringing on your copyright at some point, you may forfeit your right to go after him or her. If you don’t like being proactive, you can register your copyright once you have a dispute, as long as you’re aware that this adds at least another 2-3 month delay to the resolution of the copyright dispute.

To play it safe with your copyrights, it’s important to take some preventative action:

First, make sure you have a visible copyright statement on your work. This usually comes in the form:

Copyright (c) 1994 by Jake Richter – All Rights Reserved

The word “Copyright” must appear, or at least a real “circle-C” (the ASCII’ized version, “(c)” is not really valid without a “Copyright” in front of it). The year indicates the year in which you published the work. If you make changes, as you would in software, over several years, then it’s suggested you should put a range of years in the copyright statement. You must put your name or company name (depending on who holds the copyright) in the copyright statement. The “All Rights Reserved” describes the level of rights you have (i.e. all of them). If you are a non-exclusive licensee of copyrighted material, you can’t use the “All Rights Reserved” because you don’t have all the rights to the material.

Second, for works that are important to you, take the time to register them, just to play it safe.

Finally, if you are providing materials on magnetic media, I would strongly suggest you operate with either a direct license agreement with the people you are providing the software to, or at least a shrink wrap license, which more fully describes what can and cannot be done with the software. If you take a look at the cover of this newsletter, and all the fine print at the bottom, you’ll see something to this effect.

Now, I’ve covered copyrights owned by you, but left a rather large gap, and that is proper care and handling of other peoples’ copyrights. This is now more important than ever because of the explosion of multimedia hardware, software, and demonstrations of the same. So, if you’re thinking of showing video clips of Terminator 2:Judgement Day in your next presentation to demonstrate your great video decompression board, make sure you have written permission from Carolco Pictures. The same goes for those sound bites you may want to use. You may be safe using the prepackaged video clips and sound bites a number of companies are selling (because they licensed the rights from the proper sources), but check the license agreements to see if they can be used for public presentations. I also should mention that you, technically, need permission from the copyright holder to use background music in your trade show booth, your office waiting area, or even for your music-on-hold.

This latter item raises an important issue, namely the fair use doctrine. While it’s an infringement to duplicate copyrighted works for personal gain, or to use copyrighted material for general, public consumption without the copyright owners permission, you may make a limited number of copies of things like magazine articles under something called the Fair Use Doctrine. This permits teachers to photocopy certain copyrighted materials for distribution to his/her students, or use a video clip or sound bite for a private presentation. It does not permit copying of music, nor of software if the software is provided with a license agreement that forbids this. FreeWare and ShareWare software (as opposed to Public Domain software, which has no copyright) may be freely copied, but under their license terms, may require certain credit or payment for prolonged use.

To close out this whole discussion of copyrights, I mentioned earlier that copyrights have a certain longevity. After a copyright expires, the work goes into the public domain, but someone can then take the work, implement it in a different fashion, and copyright that implementation. That’s why a given publisher’s version of Charles Dickens’ A Tale of Two Cities can be copyrighted – that publisher has a copyright on the layout of the book, including the typeface used and the exact positioning of the words on the page, but not on the actual sequence of words (unless the publisher produces a revised, edited edition), since Dickens’ original work is public domain.

The next IP item of interest is the concept of trademarks. A trademark is a unique phrase, name, logo, graphic, or saying (i.e. all considered the “mark”), associated with a particular product or company, in a given industry (i.e. a “trade”). Hence the name, “trademark”.

A trademark allows companies to develop value and recognition in a mark, while protecting it from mis-use or mis-appropriation by other companies in the same, or a similar, industry.

In some cases, this protection applies even if the companies are in totally different industries, but the potentially infringing company might have enough exposure to cause confusion if they used the mark, or at least one similar to it. A good example of this latter situation is when NBC developed a new logo composed of two trapezoids back in the early ’80s. A small company, not even remotely associated to the broadcast industry, in the mid-west, had already developed and been using the logo for some time, and therefore owned the trademark to the logo. NBC had to buy rights to the logo to the tune of several million dollars in order to use it once the infringement become known. That’s because NBC’s use of the mark would have decreased the value that the smaller company had built in the mark by using it to represent something other than the smaller company.

This last item is the key issue in trademark infringements: will the use by Company B of a mark, or something quite similar to the mark owned by Company A cause confusion among Company A’s users? If the answer is “yes”, or even “maybe”, then trademark infringement could be a reality. By the way, ownership of a mark is based on its first use. As with copyrights, trademarks exist upon their creation, and can be registered to strengthen their value and uniqueness. However, creation of trademarks has to do with their exposure to the public versus a copyright being created just by the fixation of a work. Public exposure for product names requires that a product be sold across state lines in order to garner nationwide trademark protection, for example. Having trademarks appear in print, especially in publications that are read in multiple states also helps protect a trademark.

For non-registered trademarks, there are two types to contend with: a trademark as applied to a product, and a service mark, applying to a phrase or slogan. These are designated with a “TM” or “SM”, respectively (usually in smaller print to the upper right of the mark). Any mark you want to protect should bear the “TM” or “SM”, but only needs to do so for the first such reference in a given use (i.e. a brochure, book, advertisement). You should further support your claim to the mark by adding fine print somewhere visible which credits you as the owner of the mark (as well as the owners of other marks you have used in your materials). Take a gander through any issue of PC Week or InfoWorld and look at the full page ads. Most of them do a reasonable job of this.

Registering a trademark is much more laborious than registering a copyright. First, the filing fee is $245, and while it can be done personally, it’s sometimes easier to use a lawyer (which will run you another $500-1000). For trademarks on product names, you can file for registration on the product name at any time, but if there is a conflict, the deciding factor will be the first use date of the mark (i.e. in the case of a product name, the date that product was first sold across state lines). The trademark registration process is also a lot more time consuming than a copyright registration. You should count on a minimum of 14-18 months for a trademark registration. Upon filing your application for registration, your application gets put in a queue. Within a month or so you generally will receive some type of notification, either 1) that you need to make some changes in your application in order to get your application to be accepted (usually this is a request to provide better supporting documentation, or a request to change the scope of your application and more narrowly define the market your mark covers); 2) your application has been rejected because it’s too similar to another trademark or another application which has been accepted for the formal review process; or 3) your application has been accepted for formal review. Hopefully, you’ll sooner or later get your trademark application accepted for formal review. Further research is performed to ensure there is no conflict with existing trademarks, and then your mark is published for public review for 3 months. During this latter review process, any company or individual has the right to oppose your claim to the mark, with the proper supporting documentation. If your mark survives the public review process, then you’ll end up with a registered trademark, about a month after the public review cycle has ended. You have to renew your trademark registration every 10 years, or your mark will be abandoned.

Once you have a registered mark, you may use the “circle-R” in conjunction with your mark. Again, you only need to use it with the first instance of the mark in a given work, and when you credit it in the fine print, you should also mention it is a registered trademark. For example, if I created a brochure describing my company’s products and history, the fine print might look something like the following (note the distinction between the registered and unregistered trademarks):

WinSpeed, DLD, and Panacea are registered trademarks, and PanaSaver, TurboDLD, and The Big Picture are trademarks, of Panacea Inc. VESA is a registered trademark, and VBE, VL-Bus, DPMS, and XBE are trademarks, of the Video Electronics Standards Association. AutoCAD, Autodesk, and ADI are registered trademarks of Autodesk, Inc.

Now, I’ve gone on at length about conflicts, first use, etc., regarding trademarks. In order to save yourself lots of money in the long run, I’d recommend you use a search service like Thompson and Thompson, to see if your anticipated mark will conflict with either registered arks, or ones the service has logged from magazine articles, press releases, or advertisements. This will run you about $500, and a year long trademark watch for about $250 will let you keep an eye out for conflicting marks going through the registration process.

If you find someone infringing on your trademark, you have several options: 1) send them a cease and desist letter (in which you could force immediate withdrawal of the infringing mark from the market, or compromise on allowing existing infringing materials to be used up with a sticker on them explaining your rights); or 2) offer to license the infringer the name for a certain fee as well as clarification of the mark ownership. If neither one of these approaches work, you could take harsher legal measures. In any event, it’s important that you actively defend and protect your trademarks in order to avoid having them cheapened. And, even if the conflict is currently not in the same market, keep in mind that it conflict in the future – the best example I can think of regarding this is Apple Records infringement claim against Apple Computer, which was not a real issue until Apple Computer started producing multimedia (more specifically sound) peripherals.

The final and perhaps most far reaching IP item is the patent. While copyrights apply only to specific works, patents cover the broader field of inventions and processes. Patents offer much stronger protection than copyrights and trademarks, and are therefore much more difficult and costly to obtain. Patents only come about as a result of filing a formal, detailed application, which can cost many thousands of dollars and take 2 years to be approved in some form. A patent’s life is 17 years from the date it issues, regardless of when it was first filed. In the U.S., a patent application must be filed within one year of the sale or offering of a product incorporating the technology to be patented. However, in some countries, like Japan, selling a product before a patent application is filed will forfeit all patent rights. You can file a patent application yourself, and there are several books on how to do this, but again, an IP attorney is probably a better bet for timely filing (at greater expense, of course).

A patent consists of a series of claims which help define an invention, and usually a sample implementation of the invention. For example, Microsoft has a patent on the mechanism used to power mice off the electrical current available via a serial port, and Compton’s New Media has a patent (hopefully being turned over soon) on data access on a CD-ROM. Any unique, innovative process can be patented these days, although heated discussion is going on about continuing to allow software patents. Software patents were not allowed until just about 5 years ago, and now there are 10,000+ software patent applications pending for review and approval. All this with only 20-30 patent  examiners for software.

Once an application has been filed, a patent examiner reviews it to see if it might conflict with some “prior art”. Prior art refers to other patents and published materials on the topic a patent is being sought for, as well as other products that may clearly have implemented the technology prior to the filing date of the application. There is usually at least   one request by the PTO to the filer to make a change in the application because of one or several claims being too broad.

However, because of the lack of up to date information on all aspects, and the general unfamiliarity of the nuances of a given application, the patent examiners at the U.S. Patent and Trademark Organization (PTO) are woefully unequipped to make sure that only truly unique patents get approved. The Compton’s New Media patent is a perfect example of this, and only by massive pressure by hundreds of individuals and companies has the PTO decided to reevaluate the patent.

If you are threatened by patent infringement, you don’t have a whole lot of choices (as the final scenario in my first IP column a couple of weeks ago indicated). Other than being able to convince the PTO to reexamine a patent (which requires lots of published prior art), the only other way to fight a patent, once it’s been granted, is by spending lots of money in the courts, in addition to producing prior art (published or internal) which may help invalidate the patent. Often it’s just cheaper to pay a licensing fee, or find  something in your patent portfolio that the other company is infringing upon and then offering to cross-license your patent (or patent portfolio) in exchange for both companies being able to continue doing business on a normal basis. The other option is what I would consider a pre-emptive strike, namely licensing a wide range of patents in your company’s field, such as what Microsoft did a while back, namely licensing all of IBM’s software patents for a something in the mid-8 figures.

The United States Constitution protects the rights of inventors, resulting in our patent system following the “first to invent” convention, while the rest of the world pretty much follows “first to file” as its convention. In theory, “first to invent” is designed to protect small inventors, many of whom unfortunately don’t file patent applications, and whose products may never even make it to market.

This lack of filing, or at least public disclosure, on the behalf of smaller and/or ignorant companies and individuals allows other entities to file applications for patents on inventions that may have been developed by others at an earlier time. Proving this type of patent as being invalid is very difficult and costly, because of the lack of real prior art.

Because of these issues, there are discussions sponsored by the PTO to review our current patent system to see what, if anything, needs to be changed to make U.S. patent processing fall in line with the procedures dictated by the World Intellectual Property Organization (WIPO) of the United Nations. As I mentioned above, the first step would be changing our filing system to be a first to file system, which some experts argue might require a constitutional amendment. The other major change would be the publication of all patent applications for public comment and review to ensure that prior art is brought to the patent office’s attention. Since a patent application describes the process in great detail, this has some potentially negative repercussion on the filer of the application, as it discloses technology before it is fully protected, but as a result, forces the filer to do a lot more diligence to ensure that no conflicting prior art exists.

If you want summaries of all the patents being issued, a glimpse at the on-going debate about world wide patent harmonization, and software patent issues, send ‘help’ to on the Internet. This is Greg Aharonian’s Internet Patent News Service, which I’ve found to be an absolutely fascinating and timely overview of the latest in patent news.

Intellectual Property is a broad field, and the scary thing is that my extensive diatribe has only scratched the surface. I would also caution you that I am not an intellectual property lawyer, so you shouldn’t rely on just my advice and comments above (although I did have an IP lawyer review them for accuracy). In any event, I strongly encourage everyone who reads this column to start performing semi-annual reviews of their existing and potential IP assets, just as they might do an equipment inventory. In most cases, taking this step will help you defend against claims being made against you or your company, and in some cases protect against real theft and infringement on your intellectual property. But, if you go on the offensive with your IP, keep in mind that usually the only entities that benefit from such efforts are the lawyers.

Intellectual Property, Part 1

Tuesday, February 15th, 1994

(This column first appeared in the February 15, 1994 issue of PC Graphics Report)

Lately it seems that I’ve been seeing and hearing more and more threats to the stability of the PC market, all as a result of something called Intellectual Property.

Intellectual property (IP) refers to a wide range of items, including copyrights, trademarks, patents, and trade secrets. Most of the news breaking legal actions of the PC industry in the last decade have been based on intellectual property issues, such as Intel’s actions against AMD and Cyrix for alleged misuse of technology, patent infringement, etc.; Lotus suing Borland and Apple suing Microsoft and Hewlett-Packard for look and feel; the more recent Compton announcement of a universal patent covering accessing data on a CD-ROM; Dell threatening action against companies using VL-Bus designs because of patent infringement; and the list goes on…

So far, the PC graphics industry has been very lucky. Only a few IP disputes have arisen in our industry in the last few years, and fortunately none of them very crippling. But, this should also be seen as a sign of trouble ahead. With the rest of American society being so ridiculously litigious, can the PC graphics industry be far behind?

I don’t think so. I think that our industry has avoided litigious behavior because it knows how unreasonable such behavior can be. Anyone remember the CAD Track patent on XOR cursors? This company got a patent on something everyone knew how to use and do, but that no one actually publicly documented to create "prior art" (more on this in the next column). The result is that virtually every graphics hardware company has had to pay a sizable fee to license this patent to avoid being taken to court for patent infringement.

Unfortunately, the lack of IP disputes in our industry has created something of a lackadaisical attitude towards the whole issue of intellectual property. Witness, for example, Diamond’s Viper graphics board family, and Tseng’s VIPER technology – both share the same name and both are PC graphics products, key components for a legitimate trademark dispute. And, what about the handful of competing graphics boards that all have "Ultra" as part of their name? And, doesn’t someone out there have a patent on high-speed, interleaved DRAM access for graphics boards?

While I’m thrilled that the PC graphics industry is currently not even remotely as litigious as the rest of our society, I can’t help feeling a little nervous that something will change our state of near-serenity sometime real soon. Things like the Compton patent, and Microsoft’s patent acquisitions are just some of the warning flags I see.

Ignorance is NOT Bliss
After discussing the whole topic of IP with a number of people in our industry, it appears that at least one contributing factor to our carefree nature regarding IP is the result of plain old ignorance. However, ignorance of IP matters won’t help anyone avoid IP disputes. It should be noted that lowering the level of IP ignorance in our industry is also fraught with peril. While IP education may raise awareness about how to protect one’s own interests, it also provides the tools for companies to go on the offensive once they realize the value of their own IP holdings. Realizing that I may be opening a major can of worms, and even creating a self-fulfilling prophecy, I nevertheless think it’s important that everyone understand the basics of intellectual property, benefits and pitfalls alike. Therefore, in my next column (to appear in a week or two) I’ll tell you more than you ever thought you’d want to know about copyrights, trademarks, and patents.

In closing, let me present a scenario to further drive home the importance of understanding IP issues, and whet your appetite for my next column.

Scenario 1:

Company A invents a new technology/process that it uses in its new FrabbleStam(tm) product, but keeps the technology under tight wrap, not telling anyone how it works and also not patenting it (after all, they say, who needs patents?).


Company B sees the FrabbleStam product, figures out what Company A’s new technology is, maybe tweaks it a little here and there, and applies for and is subsequently granted a patent on the technology. The patent grant is given because no prior art was ever published or patented by Company A.


Company B develops a product similar to the FrabbleStam totally independently of Company A or the actual FrabbleStam product, also inventing a technology similar to Company A’s technology, but after Company A has done so. Company B applies for and is subsequently granted a patent on the technology because of a lack of prior art.


Company B then approaches Company A accusing them of patent infringement, demanding either:


Company A immediately cease shipping its FrabbleStam product and any products using Company B’s patent, and pay Company B some sum of money to pay for damages Company B has incurred from Company A’s pre-existing sales of FrabbleStam.


Company A may continue to sell its FrabbleStam product providing it pays Company B a royalty on each sale, as well as some sum of money to pay for damages Company B has incurred from Company A’s pre-existing sales of FrabbleStam.



Company A may attempt to prove that Company B’s patent is invalid, which would cost at least $100,000 in legal fees, and take years.


Company A may try to get Company B’s patent reexamined by the U.S. Patent and Trademark Office (PTO). To use this approach Company A must procure sizeable amounts of relevant data to back up the request. This may cost less than invalidating the patent, but is also being less likely to produce results favorable to Company A since no published prior art exists.


Company A may just throw in the towel and agree to Company B’s demands, since it can’t afford to fight the patent.

So, which company would you rather own or be part of?

Until next week (or the week after that)…