Weblog

Wednesday May 28, 2008

Americans getting on 2 wheels in favour of 4

In a story titled "No Radical Change Yet", Jim mentioned that Americans are still getting to work the same as they have been. Alas, there is already a change in the works: people are buying scooters to replace the car as the means to get to work. While there has been more coverage of this recently (see below), the issue first popped into the papers back in 2004 - "U.S. Scooter Sales Thriving With Gas Prices High".

So change is afoot, as evidenced by just this story, Vespa sales are on the rise, where a local Vespa dealer (in Ohio) has sold as many scooters in one month (April 2008) as he did in all of 2007. And it is happening throughout the country - Milwaukee: Vespa Scooter Sales Soaring On High Gas Prices.

But there's one other catch here: the sales, at present, are all occurring during the American spring/summer. In some of these states, such as WI and OH, use of the scooter during the colder months comes with additional challenges: the roads are more slipery and you're more exposed to the elements.

All figures below are in US Dollars.

So what are the economics involved? If the car is kept and the scooter added to the stable, there's an outlay of around $2000-$4500, depending on the model of scooter and local taxes. You may also need to get a motor cycle licence (add in some lessons plus the cost of the test.) For the sake of simplicity, lets say that the new price of super-inflated fuel for your car is $4/gl and the old "ok" pice is $2.50/gl. Some people report spending $80/week (vs $50) with the new prices, for me, it is now around $50 (vs $31.) For those now spending $80, the additional outlay is $1500/year. Given that the price will not go down between now and 2016 (8 years), the simple math is to say $32,000 is the cost of fuel over the that period. If a Vespa can do a week of travelling for $10 where the car is $80, that's a saving of around $3500/year, assmuing that maintenance costs are the same (and they may not be) - see below. Insurance also needs to come into play here, but I have no idea what this will be nor do I have any worthwhile references. To cut to the chase, the $4500 scooter may take upto 2 years to break even in cost effectiveness if the price of fuel remains the same. If the price of gas continues to increase, and there is no reason to suspect it won't, then the time frame for cost recovery comes in. If you can sell the car and live with only a scooter, then the equation shifts dramatically in favour of the scooter. (Corrections made)

But it isn't necessarily a bed of roses either, as one owner in San Francisco has found out, with their blog "Vespa Lx150 total cost of ownership". And lets not forget something else very important - safety.

References:

• High gas prices are driving scooter sales up
• Some are enjoying high gas prices
• ( May 28 2008, 02:52:12 AM PDT ) Permalink Comments [1]

Saturday May 24, 2008

Is America ready for what's coming?

This morning over breakfast, I heard a cook in a restaurant comment that a large container cooking oil had risen in price from around $13 to over $60 and that a bag of flour and risen from $8 to over $30 (if I heard right and a blog post here seems to suggest it's not far wrong.) Last year the minimum wage went up about 12% and this year it is set to do the same ($5.85 (2006) -> $7.25 (2008)) In a recent article in a Nevada newspaper, the state was casting doubts over its wage rise for state workers due to a drop in its revenue - a cost of living increase of 4%. Question is, how would those people make do without it?

What's the root cause? Hard to say. The price of gas is not tipped to drop until at least 2016 as the futures market for oil has already sold supply to that date at current prices. But maybe that is a dream and the more realistic picture is the price of oil is not ever going to go down to where it was. The oil honeymoon is over.

So what has this got to do with the title? America, as a nation, runs on oil. There are a few hold out examples, such as New York City, Washington DC, Chicago, Boston and San Francisco (city area only) that have functional mass transit systems. Expand the view to, for example, the entire San Francisco Bay Area, to places such as Silicon Valley, San Jose, there is barely any worthwhile public transport. Now get out to some of the other populous cities, such as Los Angeles, Las Vegas, etc, there is a bus system that kind of works, but anything else...where's your car man?

It would seem that the basic cost of a lot of things is going to rise this year and next, perhaps pushing many into poverty. But something that cannot be missed is the amount of money that these rises are going to take away from the average person on the street, giving them less discretionary spending. This won't necessarily show up in the US GDP figures, as people will still be spending the same (or more) money but the fraction that goes to transport and food will increase, meaning something else will have to give.

A shift away from driving (too expensive) and maybe smaller portions of food (keeping cost dost.) Is America ready for that? Or will it go for the same size food portions, albeit more expensive, and find more fuel efficient vehicles and keep on polluting?

For further reading:

• The Market Begins to Believe in Peak Oil
• Why Crude Oil price is zooming
• FACTS ON POLICY: Consumer Spending

Thursday May 01, 2008

Recovering postgres data

For a while now I've been toying around with postgres at home and from time to time, I seem to have reliability problems. Those problems do not appear to be postgres's, rather it is likely they are due to the operating system (NetBSD) or hardware (a small & cheap HP box.) Whatever the cause, what I see is postgres refusing to access data because an index file is borked or some other internal meta data file is borked and it refuses to load any data. My data was there but it wouldn't give me access. Grrrr....

With files of a couple of gigabytes in size (yeah, small, I know), it seemed obvious that the data was still there - somewhere. Using some standard tools such as strings and hexdump showed that the data was all there, intact, I just need to work out how to get it out of the files. With some experimentation, I had managed to work out a good amount of how it worked but still ran into a few problems. Then I saw a pointer to another utility, pg_filedump, that worked on the data/index files in postgres. The key to making real progress was looking at the header files for postgres that defined its data file formats. I had, up until this point, been refusing to do this, intent on reverse engineering as much of the data format as I could (this was to make the task challenging - just looking at the source code seemed like cheating.) After finding the header files, I started over and life was a bit easier.

The result has been something of a success. Not being content with the easy sailing, I've left priting out tuple data to what I could make sense of in the binary files. The only important types I haven't been able to decode yet are time, timestamp and date.

There is a bold warning at the top of the man page in this file for a very good reason: if the database has crashed and will not start up, there is no way of knowing what the status of your data really is. In a risk free environment, you would recover data from backups and rebuild the database from that point. This program requires you to make an educated guess about the sanity and correctness of the data that it retrieves: if you want to reuse it. i.e. no dba worth anything would ever use this with production data - that's what backups and application log files are for.