Philip Greenspun’s Weblog

Folks:

After four years of daily use, my 2006 Dell XPS 400 has finally become obsolete. It has been running 32-bit Windows XP and I want to use Adobe Premiere Pro with my new Sony camcorder. The hardware engineers who built what turned out to be a flawless Dell were no match for the world’s software engineers, who have created video editing software that will run only on 64-bit operating systems.

Here are my requirements:

• drive old Dell 30″ display (2560×1600 pixels through single DVI cable)
• be nearly silent
• be no taller or deeper than 18″ (the biggest current Dells are 19.4″ high)
• run Adobe Photoshop and Premiere Pro
• be fast at converting RAW photos to JPEG (I must admit that my old computer seems a bit pokey when dealing with HD video or the latest 20+MB camera RAW files)
• be responsive for video editing and reasonably fast for format conversion (the Sony captures in AVCHD, which cannot be used anywhere)
• run Windows 7/latest Microsoft Internet Explorer so that I can use FAA Web sites to sign off students
• run cygwin and PuTTY so that I can connect to and maintain some Linux-based servers
• run OpenVPN
• run software included with various peripherals, including Fujitsu ScanSnap scanner and a couple of printers
• hold a fair amount of data (1.5 TB hard drive? Possibly with a mirror so that a drive failure does not inconvenience me, though actually I cannot remember the last time that I had a home computer drive failure)
• burn a Blu-Ray disk for distribution of HD video

Due to all of the software that has been freighted onto this machine over the years, some of which requires license keys that I have surely forgotten, moving to a new computer is going to be painful and time-consuming. So I’d rather not do it too often and would be willing to spend some extra money now to delay the next replacement.

My actual current needs could probably be handled reasonably well by a $750 desktop PC, but in light of the labor involved in configuring the new computer as my daily desktop I am thinking of treating myself to a monster fast machine. I think that I could spend $2000 without feeling embarrassed.

The latest and greatest Intel Core i7 processor is available only on Dell’s XPS 9000, a machine that is physically too large, and the option costs $1000 extra. A 6-core AMD seems to be available. And then there are the Dell “workstations” with Xeon CPUs. I can’t figure out any of the differences among these CPUs. Is it too much to ask Intel to label their desktop CPUs “slow, medium, fast, super” instead of “Core i7-975 Bloomfield” (the complete list is daunting and it is just for one variant of “Intel Core”; Dell also sells “Core i3” and “Core i5”)? Or how about just give them numbers that correspond to performance benchmarks? That way I could know how much faster a new machine should accomplish a task such as video encoding.

My inclination is to stick with Dell, but I would be open to other brands as long as the acoustic engineering is good.

Ideas? Suggestions? What CPU? How much RAM?

Full post, including comments

I attended a conference today that was intended to match startup companies with angel investors and to promote the idea of seed-stage investment by individuals (i.e., “angels”).

Background: The popular understanding of venture capital is that the professional venture capitalists fund brand-new companies who are developing new products and services. In reality, professional VC funds are too big and too risk-averse to do seed-stage investments. A “small” fund with $200 million in assets must invest $5 million or more in each company so that it doesn’t end up with too many investments to keep track of and too many board meetings to attend. Three bright engineers working out of a basement aren’t capable of applying $5 million in capital and the VCs certainly don’t want to take the risk that the widget won’t work or that customers won’t need it. So a company getting VC funding usually has a working product and at least one paying customer.

Who then funds the startups? Rich folks who’ve been successful in some earlier enterprise and who may have some knowledge of the market or technology. These “angels” take risks that would cause professional VCs to wet their pants and, ideally, provide useful advice and introductions to the startups.

The speakers at the conference were very effective in discouraging anyone from becoming an angel investor. To be successful, one needs to make at least 25 investments and be actively engaged in each one. I.e., one needs to do nearly all the work of a professional venture capitalist, but take a lot of personal risk and not get paid anything (the professional VC will get “2 and 20”, i.e., 2 percent of the fund every year for expenses and then 20 percent of any returns, so a VC firm with a $1 billion fund gets $20 million every year in fees even if no investments are made).

Asked if it wouldn’t make more sense to apply capital in rapidly developing countries such as Brazil and China, the speakers responded that being an angel was more about having fun than getting a good return on investment. (Not sure whose idea of “fun” included sitting in board meetings with frustrated entrepreneurs, but personally I would rather be flying a helicopter or going to the beach.) In fact, the speakers said that it was quite likely that one’s angel investing returns would be lower than a passive investment in an S&P 500 index fund.

Nobody had thought about the question of whether Boston in fact needs more angel investors or venture capital. Nobody could point to an example of a good startup that had been unable to obtain funding. However, there were examples of startups, notably Facebook, that had moved to California because of superior access to capital and other resources out there.

Conference attendees noted that angel investors tend to come out of a related industry and that it was hard to fund consumer-oriented Internet services here in Boston because hardly anyone here had been successful with such a company. By contrast, in Silicon Valley the streets were littered with the wealthy idle founders of PayPal, eBay, Yahoo!, Google, and similar companies. Silicon Valley also had a deep well of startup management talent from such ongoing successes as Hewlett-Packard and Intel.

Nobody at the conference could answer a macro question: With the US private GDP shrinking, why do we need capital at all? Capital is required to finance growth. The only part of the U.S. that is growing significantly is government and the government can print money if it needs capital. With private GDP shrinking and billions in venture capital chasing the remaining startups, returns on investment are bound to be low (and indeed VC returns have been dropping). It is not clear why the U.S. needs even the VC funds that we have, much less additional angels piling in.

Nor could anyone answer a micro question: what evidence is there that the Boston area has ever been a sustainable place for startups to fluorish? When the skills necessary to build a computer were extremely rare, Digital Equipment and other minicomputer makers were successful in the Boston suburbs. As soon as the skills for making and programming computers became more widespread, nearly all of the new companies started up in California, Texas, Seattle, etc. When building a functional Internet application required working at the state of the art, the Boston area was home to a lot of pioneering Internet companies, e.g., Lycos. As soon as it became possible for an average programmer to download SQL Server and Microsoft Active Server Pages and work effectively, Boston faded to insignificance.

Currently the successful startups in Boston are in biotech, which requires the highest average skill and education level of any sort of company, and in high-speed networking gear, e.g., the equipment that telecoms purchase for their main switching centers. Could it be that as the pace of knowledge and skill dispersal accelerates, Boston will become even less prominent in the world of business?

Full post, including comments

Folks: Today was supposed to be a big day for East Coast Aero Club. Having sold 2600 coupons for a helicopter intro lesson at $69 back in March (posting), we figured that we’d sell a lot more sightseeing rides at $99 today. After all, there are far more people who go on helicopter sightseeing rides worldwide than there are who embark upon flight training. So far, however, the sales rate on the sightseeing rides is much slower. How to account for the comparatively sluggish sales?

Here are some theories:

1. Bostonians who love helicopters already bought the $69 intro lesson
2. everyone is busy packing up for the holiday weekend rather than buying stuff
3. the two-day deadline encourages people to wait to buy and it would actually be better if the deadline were tighter (though I think last time it was supposed to be open for two days and we cut it off at 2600)
4. people already have their summer plans in place and March is a better time of year to sell an activity than end of May
5. the marketing copy for this deal is not as strong; there is no mention of foliage season and no mention of a graduation present (the original copy, which we had changed at about noon, had some confusing language about how there was a 1 coupon/customer limit but 3 people were required to go on the flight)
6. a helicopter crash in Boxboro, Massachusetts on Wednesday, which killed an FAA examiner giving a checkride to a helicopter instructor applicant, put everyone in the Boston area off the idea of getting into a helicopter (this was in the news and was only the fourth fatal helicopter accident in Massachusetts since 1995)
7. the price of $99 is too high for an impulse purchase (we could do a $69 ride but it would have to be shorter)
8. something else?

Theories would be appreciated.

[Regarding the crash on May 26, the applicant reported that a simulated engine failure turned into a real engine failure and they weren’t over an airport or other open area where it is easy to make a good touchdown autorotation. They tried to make it to a road but came up short and hit trees. I believe that the helicopter was a fuel-injected Schweizer (1950s-design trainer) and we won’t use fuel-injected engines in our trainer Robinsons because they have a reputation for quitting when the throttle is rolled down to idle. Mike Wheeler, the FAA inspector who was killed, had most of his experience in turbine helicopters and in our Robinsons (the FAA rents the machines from us periodically so that their employees can maintain currency). The helicopters that he had flown would not have had this weakness so he very likely was unprepared for the engine stoppage. I flew with Mike a week ago for an FAA Part 135 checkride (to fly charters). I saw him the day of the accident as he got into this Schweizer helicopter on the Signature ramp (the helicopter had come up from a flight school based in Norwood, MA). He was tough, fair, and did his best to be helpful. We will miss him.]

Full post, including comments

A parent’s guide from New Yorker magazine, well worth reading and ties into earlier post about jobless college graduates.

Full post, including comments

Yesterday’s New York Times carries “Padded Pensions Add to New York Fiscal Woes”, the usual story about retired public employees in their 40s collecting pensions of more than $100,000 per year. These pension are automatically adjusted for inflation, exempt from state and local income tax, and these folks should live quite a long time since they retired young and no longer suffer from the stress and sedentary lifestyle of a worker. If these folks live 50 years post-retirement and investment returns in the crippled-by-pension-debt U.S. economy continue to be roughly equal with inflation, that’s up to $5 million of today’s dollars for every person who ever worked for the government.

Given the voting power of public employee unions, there does not seem to be any practical way for states and local governments to shed these obligations. As thousands of the highest earners in the state (i.e., the public employee pension collectors) are exempt from income tax, it is plain that these pensions will be paid for with dramatically higher property taxes. How to warn folks considering buying a house in one of these towns, then, that the value of their house will be sucked dry by taxes necessary to pay retired 45-year-olds? How about a law that towns and/or states with a significant pension overhang be required to change their name to something Greek?

New Jersey would be renamed “Epirus”; California “Peloponnese”. Yonkers would become “Kalamata”, etc.

Related: “Pensions: How states and local governments indulge in deficit-spending.”

Full post, including comments

I chatted with a 75-year-old man here in Massachusetts and listened as he sang the praises of Big Government and deficit spending. In his opinion, the government’s massive borrowing saved our economy because surely we would have fallen into another Great Depression otherwise. He thought that the crisis was over and the economy was in good shape now. How then did he feel about the government’s plans to borrow another nearly $10 trillion over the next ten years (graph)? He loved that too.

Did he have grandchildren? No.

So basically he had benefited from the Reagan tax cuts and extra helpings of government services for the last 30 years but wouldn’t live long enough to get hit with the new taxes necessary to repay the debt. He was protected on the Massachusetts Turnpike by state troopers earning an average of $150,000 per year (plus another $150,000 or so in pension and other benefits). He is getting back far more than he paid into Social Security. He is getting unlimited government-paid health care in the world’s most expensive hospitals (Medicare). All to be paid for one day by people who are currently children and denied the right to vote.

Now that the federal government’s biggest expenses are “entitlements” for old people and most of the money is coming from future generations, does it make sense to allow old people to vote in federal elections? As government workers and contractors are also allowed to vote, what opposition to federal government expansion could possibly prevail at the ballot box?

Full post, including comments

In my economic recovery plan (November 2008), I proposed that the U.S. adopt an immigration system that favors people who are likely to earn a lot of money. My basic theory was that Americans will never be able to pay back all of the money that our government is currently borrowing, so we need to find some foreigners who will agree to come in here and work to pay our debts.

A few recent news events seem to indicate that we’re going in the opposite direction. Barack Obama’s Aunt Zeituni was granted asylum and will be eligible for citizenship, not on the grounds that she might one day earn enough money to stop living off the taxpayers of Massachusetts but rather because she is too ill to work or travel (nytimes).

Meanwhile, Faisal Shahzad, granted citizenship in April 2009, went down to Times Square and tried to kill a few hundred of the Americans who had welcomed him just a year earlier. In looking at http://en.wikipedia.org/wiki/Faisal_Shahzad, it is hard to understand what economic or cultural benefit we could have expected from adding Mr. Shahzad to the ranks of U.S. citizens. Here are some of his credentials:

• D student in high school
• “mediocre student” in a Pakistani business school
• C, D, and F student at a college in Washington, D.C. that was stripped of its accreditation

By the time Mr. Shahzad was granted citizenship, there were 15 million unemployed Americans, many of whom had superb educations and skills. How was Mr. Shahzad supposed to thrive in a U.S. economy flooded with jobless native-born Americans? In this case, of course, we know that Mr. Shahzad did not thrive and that whatever taxes he may have paid are now dwarfed by the billions in antiterrorism costs his actions have imposed on the rest of us.

I would be interested to get comments from readers who live in countries such as Australia and New Zealand that use point systems to evaluate potential immigrants. Could Faisal Shahzad have earned citizenship under these systems? (And we might as well ask about Aunt Zeituni as well.)

Full post, including comments

Greece has been having some financial troubles lately and economists have been blaming Greece’s use of the Euro. A small country should have its own currency, they say, so that it naturally falls in value when the government prints money or spends all of the income that today’s infants might reasonably be expected to earn over their lifetime. If only Greece had kept the Drachma its future would be as bright as China’s or Australia’s.

Conventional economic wisdom in the Americas, however, seems to have been that small countries do better when they don’t have their own currency. Ecuador, for example, prospered after “dollarization” (reference). Now that the local government could not print money, investors were more likely to trust the place (though of course lately these countries are having their currency devalued by the U.S. government’s spendthrift ways (chart)).

How come what was supposedly good for Ecuador is bad for Greece?

Separately, how do the Europeans go about printing money? In the U.S., we issue Treasury Bonds and then have the Federal Reserve Bank buy them (older posting). How can this work in Europe, though? They seem to have a central bank, but how would it decide which country’s bonds to buy? Must it buy equal weights of bonds from all of the members of the monetary union? Or do Europeans simply not print money?

Full post, including comments

I’ve done a bit of planting this season, mostly by hollering at helicopter students to dig. So far we have about 1500 new plants in the ground and therefore it is time to think about irrigation. Given the high cost of water, either paid directly to a municipal water company or to the electric company when pumped up from a well, I figured it would be easy to find a smart spigot-end controller/timer. My timer would measure temperature, light, humidity, and rainfall then irrigate accordingly. I.e., it would shut itself off on cold damp days and run longer on hot bright dry days. As the controller/timer is already out in the elements, all of the sensors can be mounted directly on/in the box. Amazon sells desktop digital temperature/humidity sensors, complete with LCD screen and battery holders (parts already required for the controller/timer) for less than $10. So you’d think that a $40 or $50 hose-end controller/timer could have at least some of the obvious sensors. But they don’t. A friend pointed out that Make published an article on a home-made system that does some of this (link). But I’m surprised it isn’t offered at Home Depot.

Full post, including comments

It has become common to listen to radio stations on computer systems, e.g., a desktop or laptop PC or the Sonos system. The processors inside these computers are capable of processing (low quality) video in real time. So one would think that the processor is powerful enough to compute whether the station is playing music or if an announcer is talking. A consumer listening to a public radio station might want to hear the music but not the advertisements (generally all talk, e.g., “support for WCRB comes from BP, the world’s leading blah blah blah”). Why isn’t there an option in a Windows radio player or in the Sonos system to mute or kick the volume down 20 dB during the talk segments?

Full post, including comments