Fall of the US Dollar

I hope this is preserved, in case anyone ever wonders why the United States ever fell from glory. The US experiment may not fail, but this offers insight and renders the plausibility scarily close to home.

Hollywood Tripe

Rottentomatoes shows aggregate statistics collected from various review sites all over the internet. I use it to decide if a film is worth paying to see in theatres. I've effectively stopped seeing movies since I started reading this site.

More revealing, perhaps, is that in a given theatre with 10 films, typically only a single film will have an aggregate rating of greater than 60% positive ratings. One out of ten. The other 9 films were each considered by 60% of the reviews to be poor.

However, this is in contrast to the independent movies, which in spite of their limited distribution are often rated fantastic. This says something about either the distributors or the clientelle, or both. Or perhaps it says something about the standard of the review, but I am skeptical of that, based on personal experiences with both independent and Hollywood movies.

The value of 80 hour work weeks

Well, the Blog crashed and lost my last post. I will bear this in mind for some time.

The link goes to a slashdot article on the potential for 80 hour work weeks. Interestingly enough, it is very much in line with what I was thinking.

I often wonder how slave-driving organizations deal with the degradation in quality of work resulting from pathologically unsustainable working hours.

Things to do

Some things to lobby for, to promote effective government:

Problem: Representation in the electoral system is not fair.

• How: Condorcet method of voting.

Problem: the policitcal incentive; those attracted to the electoral system are of a certain type, not generally philanthropists; political power attracts those who desire power.

• Limited number of terms in office.
• Limited salaries and pensions.
  
• Ultimately: Lottery elections.

Problem: Regulation of environment, securities, etc is expensive and ineffective.

• How: Reward whistleblowers, and competing companies for providing evidence of breach of regulations.
  

Problem: Lobbying is private.

• How: Publish all lobbying.
  

Problem: The expensive of litigating against large corporations prohibits Rule of Law.

• How: Equalization payments for extraordinarily one-sided litigation. Eg. when you sue a large corporation for having wronged you, if there is a plausible case, then remedy should not be denied because you cannot afford to litigate.

Progammer incentives

I have spent a great deal of time pondering the incentives and disincentives of creation. This is my take on programmers, having been one myself.

(As posted on slashdot)

Programmers are a lot like lawyers, value-wise. Like lawyers, the value of programmers is, or traditionally has been, their creativity and intellect. Better tools have reduced the value of that personal asset in programmers, but not eliminated it.

It is notably different from most engineering in that the products do not require large capital to distribute, once the creativity is complete.

In this manner, I have often wondered if programmers would work better in limited liability partnerships rather than corporations. A small group of programmers who produce on contract to corporations would be, if well organized, very valuable.

The corporate structure lends itself to growth in traditional economy, whereas a larger programming companies have, in my limited experience, not been efficient. There are exceptions, like Electronic Arts, I think.

But the hierarchical view of corporations, looking down upon employees, is flawed in the programming world because the direction of the company is often better felt by the programmers themselves, and management has often had a terrible disconnect from the technical reality, and a tendency to dictate where they should listen. Good management isn't necessarily this way, but many people cling to this management style.

In a partnership, the partners would be responsible for bringing in clients, the design, the programming, and the effective reuse of code. In a corporation, they are typically responsible only for the programming. I believe savvy programmers would be much better at selecting appropriate clients and choosing the direction of the code. I believe, when it comes to the effective reuse of code, a partnership would have better structures adopted to accommodate it.

This sort of delegation among partners has been very effective, in my opinion, in lawyer partnerships. I believe the effectiveness could translate into programmer partnerships. Mind you, moving programmers into management positions in companies may have the same effect, but I think the hierarchial structure inherently causes problems. The distinguishing feature being that in a partnership, management would also be programmers, and vise versa. There wouldn't just be a "delegation to programmers" by management, so to speak.

Eerieness in human-likeness

It seems that an uncanny but slightly off anthropomorphism results in an eerieness described in Doctor Mori’s thesis, as described on the link. A fascinating observation I thought I would share.

Economist on Patent Reform

Well, here's to the obvious. The incentives in the patent system are so far from innovation that they are distorting the principles behind patents.

Patent commons

The skew of patents is well known. What isn't is how to address it adequately. The commons mentioned here is certainly interesting, though we shall see how effective it is.

Microsoft Indemnification against IP Claims

Well, Microsoft is indemnifying their clients. More precisely, their legal department has determined that the threat of litigation against Microsoft customers is negligible in comparison to the media benefit of claiming that you can indemnify your clients.

Microsoft would indemnify their clients only if their clients will never need it.

What I'd like to see is some sort of warranty or guarantee, such as "we will recoup your losses for worms, viruses, and malicious hacks, provided you show your due diligence". That's about as likely as Microsoft ever actually indemnifying a client.

UK Debate

The United Kingdom apparently has a "partisan role" in the software patent debate, promoting its own interests by promulgating incorrect and unsubstantiated assertions about the consequences of software patents.

On a tangential note, it is ironic that Microsoft hasn't spoken out yet. They are between a rock and a hard place. If they support software patents, they open themselves up to more Eolas like legislation. If they don't support patents, they lose their one lynch-pin over open source software, where they are consistently losing ground based already on comparisons of price, support, functionality, security, and good-will. The salvation by intellectual property is moot. They'll have to go the IBM Global Services route someday, I guess.

The only pure beneficiaries of software patents are the companies that produce only patents and so do not incur liability, the lawyers, and the patent office. I guess, in the legal profession, I shouldn't knock it too much.

Perens on Patents

Bruce Perens, in the linked article, offers an insightful view into the problem of Patents and standards. It's one sided but certainly gets the point across. Patents are a dangerous game for standards, and their pragmatic cousins, open source.

Trademarks and open source

The issue of trademark usage in open source software is noteworthy, particularly for idealistic projects like Debian.

Trademarks are primarily to protect consumers against passing off, so they don't buy Aspirin branded sugar pills or a Toyota branded Dacia. The conflict of trademarks does not seem, on its face, to be as contrary to innovation or freedom as the forthcoming patent issues. Nevertheless, it will be an issue.

As a matter of freedom, two issues are at hand:

1. Authors prohibiting "open source" software distribution by enforcing trademarks, and
   
2. Modifications to that free software make it different from what the author has trademarked.

The second problem came up with Jörg Schilling, where his cdrecord project was modified and distributed by Suse, but still called cdrecord. When users of Suse had problems with cdrecord, related to Suse's changes, they nevertheless went to Mr. Schilling for advice. As a matter of trademark, it should have been within his power to make Suse change the name of the software, as consumers were obviously confusing Suse's brand with his own work.

With respect to the first issue, an open source license shows an intention to not restrict distribution or modification. Prohibiting distribution by enforcing trademarks would be contrary to the intention laid out in this license, notwithstanding modification per the cdrecord example. As a matter of action, a trademark claim could possibly be estopped based on reliance upon the free software license.

As a matter of general goodwill, I suspect it unlikely that trademarks will substantially detriment the community. By virtue of open source software's freedom, the software can simply be renamed and redistributed.

Patent Buzz

There is no lack of buzz around patents. This article garnered opinions from some big names in the patent world, like Free Software Foundation counsel Dan Ravicher, law firms Phillips Fox and Baker & McKenzie.

Dan Ravicher of the FSF made this point: "patents pose less of a threat to open-source software than they do to proprietary software". He also states: "There are no patents that choose only to be infringed by open source. Any patent that imposes a threat to open-source software is going to impose a threat to proprietary software."

Well, the first point doesn't flow logically from the second point. Open source and proprietary software are in two different categories, from an evidientiary and a monetary point of view.

Access to source code
As a matter of evidence, violating patents in open source software is evident on its face: you can read the source code. Legible code makes a patent infringement case trivial. Proprietary software may require expensive reverse engineering, to devise how it operates and whether that operation violates the patent. There may be some legal questions regarding your capacity to reverse engineer legally, particularly with oppressive legislation such as the DMCA.

Access to software
Also, in this vein, to determine a violation of proprietary software, you must also have access to it. Proprietary software, particularly custom or enterprise software, may not be readily available to examine. Even if the software is available, it may require an onerous license that prohibits reverse engineering. Open source software is, almost by definition, accessible to anyone for examinition. Having a clause to prevent reverse engineering would be contrary to its object.

Licensing capacity
The lucid nature of open source software means that to obtain a mandatory license for a patent would be prohibitively expensive for two reasons. First, open source software does not have deep pockets or the capacity for a cross licensing agreement. Second, even if either were available, the nature of open source software would wholly undermine the purpose of the patent: an open source implementation of the patent would be available for free, unrestricted use.

A proprietary software company, on the other hand, has the economic means and an economic incentive to obtain a license or cross license, and would presumably do so only for the benefit of the company, and would not threaten the other economic interests of the patent.

Compulsory Licensing
Patent legislation provides for compulsory licensing, I understand, if it is in the public interest. However, even though a proprietary company could enforce this licensing by challenging it at the patent office, currently the cost would be prohibitively expensive to many, if not most, open source software developers. As well, compulsory licensing that undermines the patent, by creating an open source unrestricted implementation, would create contentious arguments about the real public interest. Patents protect the patentor, and as a secondary consideration they may have licensing imposed against the will of the patentor, if it is in the public interest. Their rights would likely trump.

For at least these reasons open source software is in a different situation than proprietary software, and I am not entirely convinced of Mr. Ravichers's assertion, as they are quoted in the linked article.

Patent Pools

Even though I am interested in protecting the infrastructure inherent to the internet and future computer innovation, it is not entirely dissimilar from the incentives applied to biotechnology, as the US Patent Office describes this potential patent pool.

I believe a public patent pool to defend open source software against being barred from innovation would act as a strong counterbalance to the forthcoming threats of software patent infringement.

Large corporations acquire patent protection not for innovation, but rather from large capital backing. As open innovation is often poorly dated, there is little capacity to submit verified prior art.

Ironically, Apple Computer's innovation in interfaces are often duplicated, but they have very little remedy for this. Nonetheless, their innovation in interfaces is a product of adhering to principles. The interface itself can be copyrighted, the look perhaps trademarked, but the methods leading to clean, simple and effective interfaces are inherent principles, and cannot be institutionally protected. Nevertheless, Apple is very creative and effective, and seems to be doing quite well without intellectual property protection of their interface.

Software Patent Reform

What forces make software patents a problem? The Electronic Frontier Foundation, linked above, is a good example. There are some very key considerations when looking at how patents should work:

1. Many people come up with an obvious idea simultaneously. Only one can have a patent.
2. The patent office churns through patents, and many of them are obvious.
3. Many patents, particularly those falling into the first two categories, are relied upon until such time as it is a de facto standard. For example, the GIF and MP3 patents.
   

In an ideal world, where patent reform is a political option, a number of suggestions have been made. First: shortening the term of the patents, as software changes so rapidly. I will consider this at a later date.

Another possibility, I would suggest, is providing a right to estop software patent infringment claims. The law should presume that patents are enforceable, but I think it should be rebuttable on grounds that:

• it can be shown on a balance of probabilities that the infringment was developed in isolation of knowledge of the patent, or
• the patent holder waited until it was a de facto standard before claiming infringement, and the results of the patent are so fundamental to society that reliance upon it is inherent in future technology.
  

Much the same way that trademarks must be enforced, I believe software patents could be more eloquent if their enforcement were poised in such a way as to protect non-patent holders who innovate in isolation, and society is not detrimentally harmed by the enforcement of the patent. This would provide protection against the plethora of useless and overly broad innovation-harming patents, such as those listed on the EFF web page.

However, to be clear, we do not live in an ideal world and this suggestion for reform is merely academic. I will soon discuss more reasonable and likely forms of addressing the patent issue.

Silly Patents

There seems to be no lack of silly patents. For example, the Patented European Webshop shows an assortment of ungodly simple patents. I am fairly certain many of these patents, if rendered enforceable, would purely burden society with little benefit even to the patent holder. The Electronic Frontier Foundation also cites a number of ludicrous patents.

There are cases where software patents can be economically viable. However, consider the case where even novel, non-obvious, and useful technologies such as TCP, a core protocol of the internet. A license fee would have 1. stymied the internet with unnecessary economic burden, and 2. set a precedent for economic recovery in base technologies. However, in their absence the internet has become ubiquitous and cheap. Its accessibility is precisely what makes it a viable technology. The whole Request For Comment process has created a commons for innovation in fundamental internet processes, that would arguably be impossible for a single corporation. There are very few RFC's that would not qualify for a patent, and yet its commercialization is essentially non-existent.
Where, then, can software patents be useful? Innovation seems to be happening just fine without the economic rewards of patents. Indeed, innovation seems to be substantially superior in their absence. Perhaps the most valuable use of software patents is not in asserting claims but as a cross-licensing mechanism with other commercial entities. Is there a positive economic incentive, however? What about cases like the GIF or ZIP or MP3 patents? They were certainly complex procedures. However, interestingly enough, their successors, JPEG/PNG and deflate/Z and OGG were designed precisely to avoid the patents, and nevertheless seem to achieve uniformly superior quality without the burden of patents.

Certainly, there must be cases where patents should reward innovation. I find myself hard pressed to come up with circumstances where they are appropriate, offhand. Nevertheless, I will ponder this problem. Having heard very little about cases where patents are appropriate, and an enormous amount about cases where they are not, I must educate myself to alleviate this parallax.

Conflicts of interest

When can our opinions be used against us? When is free speech limited not by a government censorship, but by economic, or political consequences?

Certainly, criticising the Law School at the University of Chicago limits my capacity to show up there applying for an LL.M., or a faculty position. Though that quality is ephemeral; the consequences of criticism may play out, but the dean will not, nor will anyone who remembers.

But what about publications? For example, publishing articles on patent reform. In a recent conversation, the capacity to publicly criticize patents was noteworthy because, in potentially working for a patent firm, it may create a conflict of interest with patent clients. This week, I'll talk about some patent reform, with this conflict of interest in mind. I have not yet publicly criticized patent reform, but I think it's something important to talk about.