Brad Feld

Month: June 2008

Thanks to my buddy Alex Iskold at Adaptive Blue for my new t-shirt.

$DO || ! $DO ; try

try: command not found

That describes my marathon and life philosophy perfectly.

The Latest Pile of Books

Jun 30, 2008
Category Books

It’s summer time and I’ve once again been powering through a bunch of books.  Amy and I are heading to our place in Homer, Alaska tomorrow where I’ll likely continue my pace of at least a book a day.  Look for regularly updates and quick reviews here.  In the meantime, here’s the latest set from the last week.

The Turnaround Kid: What I Learned Rescuing America’s Most Troubled Companies: This is the autobiography of Steve Miller, a well known turnaround executive.  He started his career at GM and progressed to be part of Lee Iacocca’s turnaround team at Chrysler (he was the CFO).  After Chrysler, he has been involved in a number of turnarounds including Federal-Mogul, Morrisson-Knudson, Bethlehem Steel, Waste Management, and Delphi.  Miller is a guy that’s not afraid of a Chapter 11 filing and appears to have skin as thick as the steel that Bethlehem Steel produced.  Good business history, especially if you enjoy reading about difficult situations.

Marathon: The Ultimate Training Guide: This is an update of Hal Higdon’s classic.  I picked it up at the Grandma’s Marathon Expo and wolfed it down.  I got a few new ideas from it – if you are a marathoner – especially a beginning, or aspiring, one – it’s definitely worth reading.

In Defense of Food: An Eater’s Manifesto: I read this one on a trip last week to Raleigh Durham (started on the flight out; finished on the flight back).  As I ate my Balance Bar on the plane, I realized that everything that Michael Pollan was saying rang gigantic bells in my head.  The first half of the book describes the devolution of "food" from "food" to "nutrients" and has a scathing analysis of how the food industry and our government have completely screwed the American diet.  The second half of the book tells you what you can do about it.  Eat food.  Not too much.  Mostly plants.  That’s what I’m going to do for all of July – let’s see what happens when I combine that strategy with > 40 miles / week of running.

My Life on the Run: The Wit, Wisdom, and Insights of a Road Racing Icon: A few weeks ago Dan Gannon at Newmerix told me to start running Yasso 800’s to get my marathon times down.  At Grandma’s as I was wandering around the Expo (apparently buying books) I ran into Bart Yasso.  I mumbled in a semi-star struck way something about Yasso 800’s and bought his book.  I also bought a copy for Dan that he should have by now.  Yasso is a fucking running maniac / hero / star.  I have a new running idol.  The book is a great story for anyone that likes to run.

Masters Running: A Guide to Running and Staying Fit After 40: Yeah – I picked this up at the Grandma’s Expo also.  I’m 42 so the subtitle (something about running over the age of 40) appealed to me.  This book was a no-op – Higdon seems to have slapped this one together and didn’t really do anything substantive.  I got nothing from it.  Oh well – two out of three ain’t bad.

Fear & Greed: After all the running books, I needed some mental floss.  I can’t remember when / why I bought this book (or maybe someone sent it to me) but it had reached the top of the infinite pile of unread books.  I got to page 100 before I quit – it should have been called "Dumb & Stupid" with a subtitle of "Poorly Written Mental Floss".

Today’s Wall Street Journal has an article titled Tech Giants Join Together To Head Off Patent Suits.  It describes the efforts of a new organization named Allied Security Trust who’s goal is to "buy up key intellectual property before it falls into the hands of parties that could use it against them."  The named companies that have joined Allied Security Trust are Verizon, Google, Cisco, Telefon, Ericsson, and HP.

Allied Security Trust appears to be an example of the emerging construct of a "patent commons".  There are already a number of existing patent commons such as the Patent Commons Project aimed at protecting open source software.

There are two types of patent commons – offensive and defensive.  So far the folks that have been putting together patent commons for potentially offensive purposes have kept a very low profile and often have denied publicly that they will use their patent portfolio’s offensively.  However, I’ve heard directly from a number of people involved in some of these organizations that the long term goal is to aggressively license the patent commons once it is large enough.

I’m not a fan of the offensive patent commons.  However, I am a huge fan of the defensive patent commons. As I’ve written in the past, I strongly believe that the entire ecosystem around software patents is completely fubared.  The courts – especially in the US – are poorly equipped to deal with the software patent issues and the USPTO has demonstrated that it’s either not up to the task or unable structurally to change the way things work.  Our government – especially Congress – has demonstrated that it lacks the political will to address the situation.  And, while the Supreme Court has finally waded in with a few key decisions, it still has an extremely long way to go if it really wants to address the underlying issues.

Having studied this for the last few years, it’s my strong belief that the software / computer industry has to solve the problem.  Recently, I’ve been advocating the idea of defensive patent commons – ones that are organized by clusters of large companies – but open to all that are interested.  There are lots of challenges in organizing this, including determining who can join, what the price of admission is, and what the ongoing costs of supporting the organization are, but these are solvable issues if the broad construct is adopted.

I’ll reserve judgement on the Allied Security Trust until I learn more about it, but it seems like it’s a step in the right direction if the brief description in the WSJ is accurate.  A key indicator to me will be whether organizations like Allied Security Trust vow to only use their patents defensively.  The absence of this will always raise suspicion that it’s a veiled effort to create a mega-patent-troll or that unintended consequences might result from future activity. 

Ultimately, a defensive patent commons is analogous to the idea of patent insurance, which is also starting to emerge.  I think a defensive patent commons is ultimately going to be a more powerful mechanism if organized correctly, but the analogy is a useful one to understand better how a defensive patent commons might operate.

Month 1 of this year’s TechStars program is over.  Andrew Hyde captures the sights and the people of TechStars in a fun three minute video.

While you are at it, if you were a Lego fanatic take a ride down Lego Memory Lane.

Finally, in the continued emergence of the Kindle as a serious thing, Princeton University is now publishing Kindle textbooks.  No more giant backpacks full of books and lecture notes.

I feel like bitching about FAS 157 today.  I was at the annual meeting for one of our LPs yesterday and there was a long discussion about the impact of FAS 157 on both the buyout and the venture capital business.  Once again everyone was in violent agreement that this was yet another accounting rule – promulgated by the accounting industry – to generate more fees for the accounting industry while burdening companies, especially entrepreneurial ones, with additional regulations that have no real impact on reality.

If you aren’t familiar with FAS 157, it’s officially known as the "fair value measurement" rule and unofficially known by some as the "mark to market" provision.  Before you ask, "wait – isn’t mark to market the thing that got Enron in trouble and started this whole wave of SOX regulatory stuff", I’ll simply answer "yes" and let you ponder that.

Like our dear friend 409A, FAS 157 has come out of the latest efforts by accountants to create more transparency in financial reporting.  Like 409A, I’m sure these are well intentioned ideas although my cynical side envisions an accountant in a sub-basement of a building NY with green eyeshades and a little green desk lamp sitting around dreaming up ways to torture entrepreneurs while accomplishing his accounting bosses goal of generating more work (and fees) for themselves.  Oops – sorry – back to the main story.

Since the beginning of the VC business, valuation methodologies were generally consistent and straightforward.  They were usually some variation of:

  1. Value your investments at your cost.
  2. If a financing happens at an increased valuation and is led by a new investor, write your investment up to the new price per share.
  3. If a financing happens at a decreased valuation regardless of whether or not there is a new investor, write your investment down to the new price per share.
  4. If bad things are happening, you can take a discretionary write down based on your best judgement.
  5. If good things are happening, you should not take a discretionary write up.  Only write things up in case #2.
  6. If the company is public, use the publicly traded price but discount it due to illiquidity (usually 25%).

Pretty straightforward.  Very conservative.  This almost always understates the value of a VC portfolio, which presumably is a good thing since it’s illiquid and the only fund performance information that should ultimately matter to a VC (and their LPs) should be the one linked to cash flows (draw downs from their LPs and distributions to their LPs.)

FAS 157 blows this up completely.  Under FAS 157, VC’s now have to mark all of their portfolio company values to market (er – "fair value measurement") qualify for GAAP (which is a requirement for every VC firm – our investors require we have audited GAAP financial statements.)

It gets worse.  Our LPs (who typically invest in multiple VC funds – in some case many multiples) also have to adopt FAS 157.  So they also have to mark their portfolios to market.  It used to be the case that they could simply rely on the VC valuations.  To comply with FAS 157, they theoretically have to look at all of the underlying assets in the VC portfolios and make an independent judgement on the values of those underlying assets.

Some VCs (and LPs) are just starting to implement FAS 157.  Ironically, some accounting firms wanted 2007 as the start year; others seems to want 2008 as the start year.  Many VC firms are viewing this as an annual exercise even though they report to their LPs quarterly.  Some VC firms (like us) have already built it into our quarterly reporting cycle (our accountants told us we needed to comply in 2007).  Yeah – it’s all over the map. 

But that’s not the real problem.  I’ll get to the real problem(s) in my next post on our new friend, FAS 157.

I heard a "superb" cynical statement today.  I have no idea if it is factually correct, have no data (empirical or anecdotal) to support it, but it is such a great potential example of unintended consequences that I thought it was worth putting out there.

The statement was "While hybrid vehicles make us feel better, they actually do more harm than good because they result in more driving." 

The follow up thought is that for hybrid cars to really work (at today’s efficiency levels), people still need to modify their behavior and drive less (e.g. relying on public transportation or carpooling.)  However, once you’ve bought a hybrid, you suddenly feel like you are doing your part and subsequently drive more!  This additional driving adds up across the system and increases total system fuel (and other resource) consumption.

Ponder that the next time you get in your hybrid.

I’ve got three great posts for you this morning to interrupt the long essays that my marathon addled brain has been pumping out.

First Mover vs Fast Follower – Who wins? Don Dodge reprints a story he wrote three years ago about whether being first wins or being a fast follower wins.  He adds some useful nuances for any entrepreneur (or VC) that is obsessed with the "we have to be first to market to win", "#1 takes most of the market, #2 takes the rest, and #3 to #n don’t matter", and other such cliches.

The Downturn Is a Rounding Error: Since I’m not a macro guy, I didn’t notice we were having an economic downtown.  I guess I noticed that the price of gas was higher, but as Amy is fond of saying, "don’t bother me about it until a gallon of gas costs more than a gallon of milk."  Oops – getting close (I think Amy meant "organic milk.")  Plus, our government is telling us that there is no "core inflation" (where "core inflation" doesn’t include fuel or food.)  Tom Peters helps us understand what he thinks really matters.

You gotta start somewhere: David Cohen posts an email to his business partner David Brown dated 2/08/06 that was the origin of TechStars.  I checked my calendar and the first time I met David was at 4pm on 6/06/06.  I find history to be fascinating.

There is a great Bill Gates email from January 2003 titled Windows Usability Systematic degradation flame that is making the rounds on the web.  I love a good rant and even though this one is dated, Gates says in great detail what a large number of Windows users have summarized over the years as "shit – why won’t my damn computer do <blah>."

I’m a heavy computer user and have some variation of this thought on a daily basis.  One of my special talents is finding bugs and breaking things – just ask any of the companies that I’ve invested in who their most "useful" (where useful is a euphemism for "annoying") alpha tester is.  Think of me as helping improve software quality on planet earth.

Now – software quality is a complicated thing to measure.  Not all bugs are overt ones.  Let me give you an example of a particular pernicious Microsoft one that no one seems to ever prioritize to fix (no – I’m not going to pick on Windows Calculator again, although I could.)

I use a Windows Mobile-based Dash.  I expect I’ll try the iPhone again on July 11th now that it actually syncs with Exchange, but until then I’m tethered to my Dash.  I love the form factor and have trained my muscle memory to deal with having to press multiple keys to do things that I should be able to do with one keystroke – mostly due to design flaws in Windows Mobile.  I’ve used some variant of Windows Mobile for the past eighteen months (I think starting with Windows Mobile 5; I’m currently using Windows Mobile 6.1.)  If I were Mr. Windows Mobile UI Designer, I’d change a bunch of things, but it works well for what I need it for, which is primarily email, calendar, tasks, contacts, phone calls, IM, and twitter.  And sync.  My data needs to transparently sync with my Exchange server without me having to do anything.  Oh – and my BlueAnt bluetooth headset.  And I’m sure there are a few other things.

Here’s the problem – the sort algorithm on contact lookup is terrible.  I have a large contact list (5048 as of today).  Searching for "Stan Feld" should be immediate since that’s how it’s listed in the address book.  Progressively typing S then T then A then N should bring up "Stan Feld" immediately.  Typing "Stan Feld" into the To: field on the email program should be immediate.

Nope.  The delay is anywhere from 10 to 30 seconds.  At some point I decided to try to figure out the underlying algorithm.  My guess is that it’s doing a full table scan of first_name + last_name for each letter typed.  There doesn’t appear to be an index – either fixed or dynamic – and as a result the time for most searches is approximately linear based on the number of letters typed.

Now – if this problem was in Windows Mobile 5 but fixed in an update, I’d let it slide.  I’ve done at least three (I think four) major updates of the software since I’ve had my Dash.  There has been virtually no improvement in this feature.

Whenever someone asks me about my Dash / Windows Mobile, I tell them that I generally like it except for this one thing.  I then describe the thing. Occasionally I’ll show the thing.  And then I feel stupid that I’m still using this phone since I spend so much time looking up contacts or completing names in email fields.

Having written my share of sort algorithms, I expect this is less than 50 lines of code regardless of which language it is written in.  It is sophomore in college computer science type stuff, not PhD stuff.  Optimizing this to improve performance by 10x – 100x is maybe a day or two of a single programmer’s time.

This is not a Microsoft-specific problem.  I could have picked on anyone.  I’ve got a long list of Apple issues like this, plenty of Google issues including some remarkably silly ones, and – well – don’t get me started on the Yahoo ones.  All of the companies I invest in have problems like this.  It’s just an endemic part of software.  And one that users shouldn’t have to put up with.

It’s also not limited to software.  When filling up my car recently, the gas pump clicked off at $75.  I’d noticed this happening periodically, but now it was happening every time.  Gas is now over $4 / gallon.  Each of my cars has a 20+ gallon gas tank.  $75 doesn’t fill up the tank in any of them (and in at least one it doesn’t come close.)  There was a point in time when I’m sure someone decided that a way to mitigate credit card fraud at the gas pump was to limit the amount of each transaction to $75.  Now all that does is inconvenience a large number of customers with a mysterious cut off point.

If you develop products (especially software) for a living, never forget that people remember the little things.

One of this year’s TechStar’s companies – Devver – is building web-service tools for Ruby developers. They are taking the tools that Rudy developers already use and putting them into the cloud, adding benefits like faster execution, easy setup and configuration, and change management.

Devver is currently looking for feedback from Rubyists on the types of tools that would be most useful. You can help them out by filling out their survey.

They’re also interested in talking to Ruby teams in the Denver/Boulder area. If you’re willing to talk to them and interested in getting an early look at what they are working on, send them email to set up a meeting.  The Devver guys are also going to be in the bay area on July 16th so if you are a bay area Ruby developer, they are interested in meeting you.