Well, it's clear at least that this whole area is a complex issue :-) I've noticed there is a tendency among some people leaving comments to confuse the comments with what I wrote. I am not responsible for what is written in the comments here, except to the extent that I don't delete them (which so far I have managed to refrain from doing :-))

It also seems that the discussion in the comment area after awhile always returns to the same tired themes of rhetoric about what is better, who is evil, whatever. Please, if you want to debate that, go somewhere else. I’ll take comments on topic, even (especially?) ones I don’t agree with, but let’s try to avoid philosophical battles and attacks. Also, before you make a belligerent retort about something you think I wrote - please check that I actually wrote it. It will save you from being jumped on yourself.

It is also clear that many people are not really familiar with exactly how patents work. For example, you can't simply patent "word processor" or “BASIC” - the actual patent has to be far more detailed than that, and you have to make specific claims about a design. Since just about anything that is developed is built on the work of others before, there is essentially no way an entire device or product can be patented. Much more common is that the new, unique, creative part of the new product that sets it apart from others is patented. Also, not every great idea can be patented – there are a set of criteria that have to be met – a big one is that it has to be “not obvious to one skilled in the art”. Note that means “not obvious *before* someone explains it to you”. Many things are obvious in hindsight. The idea also has to be a device, method or process or algorithm – not just an idea for something you can describe. I believe (haven’t checked) that the paper clip was patented in 1901 as “a device for binding sheets of paper or other flat items using a coiled piece of stiff wire” or something like that. There's an interesting synopsis here about paper clips - there were patents on the clips themselves, but the main patent was for a machine to make them. What I love about this article is that it describes even the lowly paper clip in the same way we think about design for products - a series of compromises to reach a globally optimal solution - people then always want to modify it to fix one flaw or another, but in so doing they invariably decrease the overall value. Also note that we think of a paper clip as obvious now that we have them, but until they appeared at the start of the 20th century, people were making do with much less satisfactory methods - string, pins, etc for decades. They had wire before that but despite some experiments no one seemed to come up with that “obvious” idea in all those years.

There were several comments that alluded to (paraphrasing) “real” inventions being things like airplanes and so on that are entire units, and that patents from software should only cover such macroscopic things as full applications and operating systems. Yet in the industrial world, patents cover every tiny little improvement inside machines that have been made over the years. The exact shape of the aluminum opener on soft drink cans, the exact shapes and way that little mechanical widgets move inside my camera, the shape of the drive components inside the continuously variable transmission (CVT) in my wife’s car – not the CVT itself mind you – there are separate patents now by different manufacturers on individual improvements and variations on its implementation (after the original patents on the main idea ran out). In fact it is totally normal that tiny bits are patented that are internal to a larger “whole” as perceived by the end user, just as in software the internal sub-elements that are inventive can be patented. I don’t think that is unreasonable, in the sense that it is matching the way other industries work.

There was a comment that patents used to have to include implementations - I believe that was true originally. I read somewhere the Smithsonian actually has a zillion "models" in storage that were submitted with patent applications during the 19th century, but that the requirement was dropped - maybe in the 1920's or so if I remember correctly. Anyway, any modern patent does not require a working model to be submitted mainly for pragmatic reasons: there are simply too many patent applications for a system that assumed only a few hundred or thousands of inventions per year. And in any case, the patent was always granted for the *invention*, NOT the physical implementation, which was included mainly to make it easier for someone to examine the patent and make use of the inventive idea, since a written patent can be pretty obscure (to say the least).

Hmm. With all this talk about patents, I’m surprised no patent lawyers have deigned to make a comment here yet to save us from ourselves…

There was a comment that maybe open source (particularly the GPL) is not anathema to a software business. I think that can be true if the business is very small - i.e. not worth cloning. However, in a free market, if there is money to be made and your source is available and unencumbered by IP protection, it is hard for me to imagine you can build a significant (e.g. >$100,000) software business if a competitor (not an end-user - that is not a significant risk) can take your code and build the same product as you have. You have the advantage that you are familiar with the code so maintaining it is easier for you, but you also had the start-up cost you had to cover (whether it was your time or it was $$$ - they both had value). Which is greater - the cost to build original code, or the cost to become familiar with someone else's code well enough to sell/support it? It may depend on the case, but generally learning is easier than creating from scratch. If you argue that code is sometimes really hard to understand except by the original author, then I will argue you are not doing a good job as an open source contributor if your code is that hard to understand - why bother to call it open source if others can't understand it easily?

So maybe I should amend my point to say that as far as I can see, open source makes it difficult to build a *significant* software business. You can grow a business to the extent that there is "friction" in the marketplace that makes it not worthwhile to clone your product and business (say, to $100,000). But if your business grew to a significant size (say, $1 million) then someone else will come along, covet that money, and use your source to kick start a clone of your business. This is true in hardware as we all know (all the cheap knockoffs you see of original products), and the friction costs are higher in most cases for hardware. If for example you make an open source accounting app that starts to do well, I can take that source, study it, and start selling and supporting it for less than you offer it, and we can have a price war until we're both paupers, or one quits. Even better, as you make improvements, I get to incorporate them in my product as well, so you can't really stay ahead of me for any length of time. Am I misunderstanding how this would work once the numbers get high enough to be interesting? With open source, what seems to work is to make the money in services, support, hardware or other ancillary businesses around the software.  If you believe that you can make a significant software business built on open source, please tell me how it would work (That’s not a rhetorical question - I’m serious about that. Did I miss something?)

A comment implied that by mentioning the "USA and other countries" I was somehow claiming the USA does and has always behaved better than others regarding IP issues. I made no such claim.

Regarding individual inventors and patents – I made no claim that only large companies can innovate. I re-read my post and I am not sure how my comments could be read that way. There are many valid patents granted to individual inventors or small companies. Just about any viable startup these days is a result of a patent or two and some funding based on the strength of that patent. Note that without IP protection, these startups would likely not receive any funding and those smart people would go do something else, since their idea could be ripped off immediately (by a competitor large or small). A fraction of these patents are used for “submarining” large companies - by no means all. I was just mentioning submarining as one of the many abuses of the patent system.

Copyright was mentioned as most appropriate for software, and patents for algorithms. I believe (not sure) that this is more or less how it works today (flaws and all). Source code is covered by copyright (can be waived of course). Patents are not awarded for source code or "particular implementations". Patents are awarded to algorithms or methods, and in the patent, the language is something like "one possible implementation of which is...” followed by a description of the implementation - usually not source code). This is done explicitly to make sure the patent covers the method and not a particular implementation, which would narrow a patent unnecessarily.

There was also some confusion among the comments (especially near the end) about what I said relative to the GPL. It seems a lot of people react defensively to any such mentions and are able to slide words and paragraphs together in such a way as to construe something they can take offense to - perhaps because they are expecting me to write something I am not? To be clear, I did not equate open source or the GPL with communism or looting - if anyone thought that, please reread my post – I had moved on to intellectual property before mentioning communism. In case you’re too busy to check, here’s what I wrote about GPL (not so bad, is it?):

“To the extent that open source is about people working together, contributing their time and effort to build something, I think its great….<snipped lots of relevant stuff here>…

…Certainly software companies can't get fully behind open source (especially the GPL), since it is anathema to their business model. So it has to be the hardware and services companies. If the open source people fully realized that they are effectively working zealously and for free to help one type of corporate entity over another, would they still be so dedicated to donating their time and energy? Interesting to ponder.

There are many different kinds of open source licenses. I don't have a problem with any of them - but things like the GPL need to be treated with caution by anyone hoping to build software in that area and later make money from the software. That shouldn't be a surprise to anyone, since the goal of the GPL was to make it hard to "own" software rights and therefore to make any money from it directly (services, support and consulting aside). Other types of open source are more about sharing and less about enforcing that sharing on everyone, and that's fine too.”

There was a fairly philosophical comment that software is not a product like a TV, and what the customer is really purchasing is the IP and the rights to create and modify the software. If those are not received, you have not received all you are entitled to. I am not sure why software gets this distinction frankly. When I buy a TV, I am not given the circuit diagrams and all the details of how it is constructed, including the image adjustment ASIC blueprints. Most people would say that to demand that from the manufacturer and expect to get it is absurd (although you used to be able to get repair manuals for some older sets that gave you some of it). So in effect (with or without those plans) since any reasonably normal person has no chance of ever being able to modify their TV even if they wanted to, do I then not own it? In fact, I can’t modify by myself just about any product I own that includes IP – I have to pay a service technician to make the modifications. Do I not really own any of them as a result? Why is software so different? Is it because unlike most hardware, there are some people in the general population now who are essentially “TV repairmen”, interested and willing and able to modify it? And because they want to do this, they are interested in redefining what software should be and who should own or have access to its IP? (that's a real question, not a loaded one)

BTW, if software is covered by copyright rather than patent, then surely the analogy is more like a book, in which case the public expectation is that you cannot do anything with that work except use it and briefly quote from it, certainly not freely modify it and republish it unless you have permission from the author. I don't see it as similar to scientific research - which is not bought and sold, rather freely donated. I am unclear why software has its own category, unless the argument is that software should *not* be bought and sold - there you've lost me - that's just an opinion or a wish. The market should decide what people want and are willing to pay for. If some people want to give away their work, that's fine. If others want to charge what the market will bear, that's fine too.

That said, I don’t plan to be drawn into some never-ending downward spiral of a debate about open source philosophy or its merits or demerits.  I am really just not that interested to be honest. Even if the topic were not so loaded that it is nearly impossible to have a serious, mature discussion about it, there are many other discussions to be had than the same old “open source” ones. There are many forums to go do that if that's what you want (try Slashdot for one), but this is not one of them.

Thanks for sticking with me this far. My next post will be back on more interesting topics I hope.