Toolkit

Story: UK Patent Office considers problem of trivial patents

• 
• 

The previous comment hits the nail on the head exactly. Surely we need to consider what patents and copyright are actually for? If someone patents the obviously best way of doing something, and then charges a large sum to licence that 'best way', the majority of products will use the non-patented 'second best way' and everyone ends up using something which is not the best possible.
Clearly if the 'best way' was very difficult or costly to determine, it's cheaper for others to make use of the research carried out by the inventor, and in this case protection of their IP is sensible.
If on the other hand, the 'best way' is cheap or simple to invent, which is the case with most software concepts (note 'concepts' not products - a product does take a long time and hence is expensive to develop, but most of the work is not innovative), the effect of the patent is to prevent others from using a solution they could come up with themselves.
What is expensive with software is the whole development process, and this is what generates the revenue - look at some of the best revenue generating products in the industry (eg large office productivity suites). Most of the cost of producing these is not person-years of research on a new concept in user interfaces, but person-years of writing source code, help files, UI graphics, testing etc. In this case, what you need to protect is the end-product of that expensive but substantially non-innovative process, the end-result of which is a set of files of code, graphics, text etc - exactly what copyright covers.
Compare software and pharmaceuticals - the software development is equivalent to the research into a new drug. In both cases, standard techniques are used, but in different combinations to produce an end-result which nobody has produced before. With a drug you might use different raw materials, temperatures, production processes, in different orders - with software you might use different routines from a standard software library and different tools. In both cases, you produce an end result which is much cheaper to copy than to invent from scratch. IP protection should be designed to prevent others from taking that end-result and simply copying it, bypassing the development process. For drugs, this is achieved by patents (I suspect you cannot copyright a molecule - it has been discovered not invented, so another one the same is indistinguishable). For software this is achieved by copyright - a free system which ensures others cannot take the end product of a developer's work and copy it to make money.
You only actually need patents where copyright cannot help because similarity between 2 items does not imply copying. 2 drugs may be the same molecule, but this doesn't prove whether one producer copied the other's research or not. The patent is needed to provide a legal basis to protect the originator's research investment.
If 2 pieces of software are identical, byte for byte, then the second must almost certainly have been copied from the first. There is so much flexibility in software design that for 2 people independently to come up with identical files is almost inconceivable. In this case you don't need the patent, as copyright protects the originator. In fact it does a better job than patents - if the 2nd developer creates their software independently, it may do the same job but you can tell they developed it independently because the files are not the same. This allows competition - for example, Firefox/internet explorer/opera etc - they all do the same job, but independently developed, and you can tell this by comparing the files. Patents don't provide this fine level of distinction.
Where you might need patents is for a way of doing something in software which is very difficult to discover, but once dicovered, easy to reproduce in different, apparently independently created, forms. An example is a compression algorithm or asymmetric encryption - it takes months

• 
• 

• 
• 
• 

Full Talkback thread