mult.ifario.us

Albert Einstein is credited with saying that "Compound interest is the most powerful force in the universe." True or false, compound interest and its exponential brethren are certainly among of the most inscrutable forces in the universe, as evidenced by Ponzi schemes, flower frenzies, and most recently the Internet bubble.

Free and open source software (FOSS) is fascinating stuff, but it's also a red herring for the same reason that the Internet was a red herring during the bubble:

No technology or technique permits the suspension of fundamental economics for an infinitely long time.

Someone, at some point, has to pay more for something than that something cost to produce. There are a set of natural questions for the FOSS entrepreneur to answer, and the basics are the same as for any other business:

What do you sell? What does it cost you to manufacture or procure it? To whom do you sell it? By what means? For how much? How many potential customers are there? How do you generate repeat business or other recurring revenue? Who are your competitors? What are the barriers to entry?

The questions about competitors and barriers to entry are the most challenging for a would-be FOSS business -- What would prevent someone from creating something more free or more open and otherwise replicating your business model? How do you prevent a "commercial" business from profiting from your efforts without supporting them? The psychology of the customer is one of the keys to FOSS, so the challenge is that any defensive measures must be implemented in a way that does not topple with any of the pillars of good FOSS citizenship. The naive solutions aren't going to fool any of the faithful. For example, using traditional intellectual property protections like patents (in traditional ways) would be unthinkable, offering a low-end freebie under a GPL license as a lead-in to a traditionally licensed product would be heresy, and asking people to register for downloads or access to documentation would be in exceedingly poor taste.

So, where will the FOSS Amazon.com or Yahoo! come from? There are plenty of interesting case studies out there: RedHat and MySQL are well-enough studied that there are business journal articles about them, and companies like Sleepycat and Zope are a little more off the beaten path but still accessible. Venture capitalists appear somewhat willing to spend money on FOSS business strategies, and people steeped in the community (e.g., Bob) are thinking about the practical challenges in FOSS.

Personally, I don't think there's any big money to be made in FOSS. (In an analogy with the Internet, it's like being a bandwidth provider or hosting company...) People will become well-known and may be financially comfortable running lifestyle businesses, and some people may even become rich; but no one is going to get wealthy directly. (I'm using the Chris Rock scale of money here — Shaquille O'Neal is rich, but the guy who signs his paychecks is wealthy...) The FOSS Amazon.com or Yahoo! will have a business model based secondarily on open source, whether it's support, services, management, or (hopefully) something more creative, only time will tell.

You know what the funniest thing about Europe is? It's the little differences...

My wife and I spent the Thanksgiving holidays in Paris, which was a welcome break from work and all things American after the election; the Musée national Picasso Paris and a big SNCF protest which kept us from visiting Musée Rodin de Paris were a couple of trip highlights.

We flew over on SAS (the powered seats were cool, but no wi-fi until next year), and on the way back, I asked one of the Norwegian flight attendants what a danish is called in Denmark. The response was "Wienerbrød", which loosely translates to "bread from Vienna". Go figure.

I should preface this entry by noting that I am neither a lawyer nor an accountant. Rather, I've spent plenty of money on the services of both, and I'm willing to share my recollection of some of the things I've learned along the way.

Failure, i.e., non-success, is one of the things to plan for in any entrepreneurial activity, and when raising an angel round, i.e., investment from accredited individual investors, it is possible to engineer the round so that the investors get some generous downside protection from the tax code. Specifically, Section 1244 allows investors who sell their stock at a loss to deduct the loss. Under certain assumptions about the financial circumstances of the individuals, that means that they will lose no more than 65% of their original investment, providing that investment is less than $50,000 if they're filing singly or $100,000 if they're filing jointly.

If you're planning an angel round, ensuring that the investors are eligible for Section 1244 treatment would be a good idea. The burden of proof for Section 1244 rests on the company, but those records (Board minutes, financials, tax forms, annual reports, state and federal filings) are things that you should be keeping anyway.

With FiveSight's users almost entirely switched to non-Microsoft operating systems (namely, Mac OS X and debian Linux), it's an opportunity to ditch Microsfot Word entirely. (As little as I like Word, I like even less the use of .doc as the standard interchange format for business.) Using L_AT_EX for documentation makes good sense, since the output via PDFL_AT_EX is beautiful and portable. Moreover, both L_AT_EX and tools to edit it (i.e., Emacs with AUCT_EX) are generally available. Product documentation, Board minutes, and corporate letterhead were the first to go, and now contracts and other legal documents are following.

One of the challenges was signature blocks for contracts, but L_AT_EX's tabular* environment makes it easy:

\begin{tabular*}{\textwidth}% =396pt in this case
{@{} \p{180pt} @{} \p{36pt} @{} \p{180pt} @{}}
& & \
\includegraphics[scale=0.10]{../images/fmugwump_signature.pdf}&
& \ \cline{1-1}\cline{3-3}
{\small By} & & {\small By}\
& & \
Farquar T. Mugwump, Esq. & & \ \cline{1-1}\cline{3-3}
{\small (printed)} & & {\small (printed)}\
& & \
Chief Foobar Officer & & \ \cline{1-1}\cline{3-3}
{\small Title} & & {\small Title}\
& & \
Newcular, LLC & &  \ \cline{1-1}\cline{3-3}
{\small Company} & & {\small Company} \
\end{tabular*}

At least in my book, that beats mousing around for 15 minutes...

While it was a trickle at first, comment spam ramped up to 100+ per day in the past couple of weeks, so I bit the bullet and learned how to install WordPress plugins. That may not sound like such a big deal, but I don't know PHP and have absolutely no desire to learn. In fact, I have so little desire to learn PHP that I thought briefly about switching to NewsBruiser (which is written in Python, a language I know, and offers a bayesian filter for comments) or even Movable Type (which is written in Perl, a language I used to know), but in the end, suffering a little PHP was more palatable than the thought of migrating content, proxying permalinks, and learning yet another weblog server.

All in all, it turns out to be trivial — there is a WordPress "CAPTCHA" plugin called "authimage" that does the job. Even though I had to make some simple changes, installing it took all of 20 minutes, including installing a couple of necessary debian packages -- and I still don't know PHP.

Some deep thoughts from Eric Newcomer:

The most significant innovations are over. Some would call this the "death of software." But only as we know it. Software will continue. But we are not likely to have any new languages, or see any significant new inventions. Twenty years ago no one knew what a database was, or middleware, or Java. But now I think innovation like that has stopped because IT doesn't need it any more.

Maybe Eric's right. There comes a point in any given field of inquiry where all of the questions with good ratios between effort and effect are answered, and the remaining questions are sharply partitioned into trivial or imponderable. Knowing when a field is empty is more art than science, and the only heuristic that I can think of is when the same unsolved problems (e.g., P vs. NP) keep coming up over and over.

Or, maybe Eric's wrong. Databases, message queues, and application servers are everywhere, but enterprise software and the practices that surround it as a discipline are at the level of stone knives and bearskins -- things are different but not necessarily any better than when the The Mythical Man Month was written.

For example, there is a huge, open green field waiting for innovation in constructive approaches to model checking for distributed systems. See BLAST or BOOP to see what's being done for straight C programs, and there are similar efforts underway for OO systems and orchestration languages like BPEL.

I do think Eric's right in the sense that incremental advances, like the database evolving out of a progression of data management libraries, are more or less at an end. From here on out, incremental advances will provide incremental benefit. To take a leap forward, there is a real need for software engineering, by which I mean a sort of applied computer science that fuses the discipline, depth, and rigor required to attack hard problems with the practical concerns of software at large and the savvy to spot the real problems hiding in the woodwork.

Microsoft has the vision and resources needed, as do, to a lesser extent, other majors like IBM and HP, but the first open problem to solve may be to find a way to marry entrepreneurship with the economic and intellectual climate to create the next kind of innovation.

As the saying goes, "Live by the MSWord, die by the MSWord." FiveSight has accululated, between corporate infrastructure (articles and bylaws, filings, Board minutes, options plans, investments, employment agreements) and business relationships (licenses, support agreements, services agreements, distribution agreements) an entire filing cabinet full of legal documents, and all of it was written in Microsoft Word.

I'm no fan of Microsoft Word to begin with, and I share the sentiments of those people who think that the use of .doc as an exchange format is ridiculous.

I was swimming through a 45-page contract a couple of weeks back, and hours of time spent unraveling track changes, fixing romanette references, and other low-grade psychological torture over the years came flashing back. Contracts are just the tip of the iceberg, since more complex filings and briefs require tables of contents, references, and other things that are an absolute mess in Word. There must be some way to convince lawyers to start using TeX, although it looks like some already have. We could call it... L_AWT_EX...