Last month I blogged about why it is important to participate in the study by parliament’s Industry Committee on science and technology. I also provided a link to my own submission (OpenDocument, PDF, HTML). Michael Geist has reported that the Canadian Association of Broadcasters has also included copyright proposals in its submission.
The study is already underway, with meetings being held on April 10 and April 17. The transcript for the April 10 meeting is already online, and audio is also available (April 10, April 17).The April 10′th meeting included officials from Industry Canada, andthe April 17′th meeting included witnesses representing the Associationof Canadian Community Colleges, Canadian Institutes of Health Research,Social Sciences and Humanities Research Council of Canada.
I found it interesting that one of the questions I’ve always hadabout metrics (criteria for success) came up in questions from Ms.Peggy Nash. She asked exactly how we measure how we are doing.
In his response Mr. Richard Dicerni (Deputy Minister, Department of Industry, said:
“Where we’ve ended up is that if you want to get a sense of theeffectiveness or usefulness of your investment, you have to look at abasket of indicators—for example, citations per capita. If you’relooking at the pure research part, how many of our Canadian researchersget citations? At the other end, how many patents per capita getproduced?”
It is interesting how these different metrics suggest two verydifferent (opposing) public policy directions. It is well known thatthe best way to increase citations to scientific papers it to use OpenAccess (OA) publication methods. If increasing citations became apolicy goal, then mandating OA publication as a condition for publicfunding would be appropriate. This is being done in an increasingnumber of jurisdictions and government funding agencies.
If increasing the number of patents per capita is a goal, thenclosing off access to research though increased exclusive rightsbecomes desirable. It may be seen as desirable to lower the barrier topatent granting by allowing more poor quality patents. This couldinclude removing or minimizing any subject matter exclusions,underfunding patent offices to make it impossible to do adequateexamination, as well as providing other incentives to ensuring thatpatent offices err on the side of granting questionable patents (IE:patent filers can sue the patent office if a patent was not granted,but there is no mechanism to sue the patent office if a patent wasimproperly granted). While all of these things will increase the numberof patents per capita, studies also indicate that these types ofpolicies chill innovation.
I found it ironic what I had to do to access the audio of these meetings. The ParlVu service streams the audio using Microsoft’s depreciated legacy proprietary mms protocol, with files in Microsoft’s proprietary Advanced Systems Format (ASF).While specifications are available, there are patented methods thatmust be licensed, and Microsoft does not offer the RF (Royalty Free)licensing required in order to implement the methods in Free/Libre andOpen Source Software. In other words, the only way I could access theaudio from this parliamentary committee is to potentially infringesoftware patents.
I used mplayer to dump the streamed audio to a local file, and then used ffmpeg to transcode to MP3 format which I could then listen to on my portable devices (My cell phone and my OLPC XO).
I say “potentially infringe” because this is currently a legal greyarea. Software Patents are often justified by claiming that they arenot patenting software,itself, but patenting methods embedded withinhardware that happen to be implemented in software. Does this mean thatthese software patents only apply to hardware manufacturers, or also todistributors of software unbundled with any specific hardware that isintended to be used on a general purpose computer?
The other question is about amateur uses. It has historically beenunderstood that patent law applies to industrial users of the methods,and does not impact activities that would be carried out by theeventual end-owner of the good, such as any tinkering. I know of nocase of an amateur software developer or any individual software userbeing sued for patent infringement, and I would be extremely surprisedif the patent holder would win such a case.
I think it is an important part of the committee’s study on scienceand technology to think about these issues, and to look towardsclarifying these areas of law. My submission to the industry committeeincluded the following on patent law:
Patent law is economic policy, intended to provide an incentive forinnovation by providing a temporary government granted monopoly on anart, process, machine, manufacture or composition of matter which areworkable, new, and ingenious (useful, novel, unobvious). While this iseconomic policy, changes in patent law have come as a result of legalanalysis rather than economic analysis.
There are subject matter where independent studies have indicatedthat patents stifle rather than provide an incentive for innovation.Governments must commission independent studies and make decisionsabout patentable subject matter based on these studies, rather thanlegal arguments presented by those who benefit from legal uncertaintyor stifling of innovation.
Patent law should have exclusions for practical, moral, and ethicalreasons. Canada needs only to look at the European Patent Convention’sarticle 92 exclusion list for an example.
Patent quality is very low in some subject matter, such as software,which creates a practical problem. Patent quality must be assured ineach subject matter area., possibly with a rule saying that 50% must begood quality in order for patents to be granted and/or enforceable inthat subject matter? It is competition and first mover advantage, notexclusive rights, that drive innovation in software
We need a “Fair Use” doctrine for patent law, possibly carving outinterface patents, royalty-free FLOSS implementations, and theactivities of private citizens in their home or as amateurs andtinkerers. Patents for producing hardware or other tangibles are oftenopened up to competitors using RAND (Reasonable And Non-Discriminatory)licensing terms where there is a low per-copy royalty which is offeredto anyone. While RAND is appropriate for tangibles, it is inappropriatefor software which has a natural marginal cost of zero. RAND excludesfrom implementation any businesses using peer production methodologies.