Category: computing

Tuesday, 13 September 2011

08:02 – I’m really thinking about abandoning Firefox for Google Chrome. The only reason I haven’t already done so is that I deeply mistrust Google, whom I rank right up there with Apple and Microsoft among corporations I consider to have abhorrent business practices and lack of respect for people’s privacy. I really don’t want Google keeping track of every web page I visit and every link I click, and then storing that information forever. Who knows what they do with it, and, more importantly, what they’ll eventually do with it. I’m convinced that Google never discards any data, even data that any reasonable person would consider ephemeral (and private). I don’t trust Google not to spy on me, but there’s no other alternative to Firefox that I’d consider using. In fact, I’m not entirely sure I trust Firefox. It’s too close to Google.


Here’s another Kindle book you might want to grab. It’s normally $50, on sale for $0.00. Genes, Chromosomes, and Disease: From Simple Traits, to Complex Traits, to Personalized Medicine. I grabbed a copy last night and read 20 or 30 pages from several chapters. So far, it look interesting. It’s written at a non-specialist level. That is, you don’t have to be a biologist to understand most of what the author talks about, but it’s helpful to have at least a basic grounding in science. If you’re interested, grab it immediately, because free offers like this tend to go away quickly.


Things continue to get worse for the US Postal Service. Much has been made of the decline in first-class mail and correspondingly smaller revenues, but the real problem is personnel costs. It wasn’t always that way. My senior year in high school, 1970, marked the transition from the government Post Office to the semi-private US Postal Service. At that time, the starting wage for USPS workers was, IIRC, about $2.80 per hour. The highest wage, which required more than 20 years to achieve, was something like $4.20 per hour. There were no lavish benefits, either. At the time, the minimum wage was $1.60 per hour, so entry-level USPS workers made about 175% of minimum wage and those who’d been there 20+ years made just over 250% of minimum wage. Minimum wage is currently $7.25 per hour, which means entry-level USPS workers should now be making about $26,000 per year and those with 20+ years should be making about $38,000 per year. Instead, ordinary letter carriers are paid about $45,000 to start, and top out at about $58,000. That excludes overtime, of course, but more importantly it ignores the gigantic increase in benefits costs. In 1970, retirement and medical benefits were a tiny percentage of compensation costs. Now, they’re a huge part of it.

If the USPS is to survive, they have no option but to cut personnel costs dramatically, including chopping pensions and benefits for current retirees. As things stand, the USPS will default this month, unable to make a required $5.5 billion deposit to fund retirement and health care benefits. That’s the least of the problem, though. At the current rate, the USPS will be literally bankrupt by next summer, unable to pay its operating costs. At that point, post offices close and the mail will no longer be delivered.

Read the comments: 30 Comments

Friday, 9 September 2011

08:42 – Well, the copy finally finished at 10:00 p.m. last night, after 13 hours of writing 300,000+ files totaling about 1,300 GB. I’ll probably do the same thing again to a second new 2 TB Barracuda drive and then pull both of the old 1.5 TB drives. Of course, that leaves me with a system drive that’s older than either of those drives. Barracuda drives are extremely reliable, but these guys are well past retirement age in dog years.


Jerry Coyne just posted an interesting article about science versus religion, Adam and Eve: theologians squirm and sputter. The whole of Christianity is based on the Adam and Eve myth and original sin. Without that, Christianity has no basis whatsoever. And yet science tells us, indisputably, that Adam and Eve never existed. It is amusing to watch accommodationists try to reconcile religious myth with the cold, hard light of science.


We’ll be using a lot of dropper bottles for the biology kits, so I ordered a couple hundred dozen from one of my wholesalers in 15 mL and 30 mL capacities. These bottles are Chinese-made, and they’re fine except for one thing. They arrive with the dropper tips and caps installed, which means that Barbara and I have to disassemble them all before filling and labeling them, and then turn around and reinstall the dropper tips and caps. Doing that for a few bottles is no big deal. Doing it for thousands involves some work.

We’ll also need 30 mL wide-mouth bottles, which none of my current wholesalers offer. So I contacted a bottle supplier about the wide-mouth bottles and decided as long as I was at it to see what they had to offer in the way of dropper bottles. Their bottles are US-made, which is good, and they come in bulk with the dropper tips and caps in separate plastic bags. The problem is, bottles, even Chinese bottles, aren’t cheap, and the US-made ones cost 30% to 40% more than the Chinese-made ones. Still, I was willing to consider paying more for the US-made bottles. Until the samples arrived.

The problem is the dropper tips. The Chinese bottles have dropper tips whose bodies are long plugs that friction-fit the mouth of the bottle. Seating a dropper tip is a simple matter of pressing the tip into the mouth of the bottle until it seats. The dropper tips on the US bottles have much shorter bodies, and snap into place. The problem is that it requires close attention to make sure the tip has actually seated and snapped into place. This would slow down processing significantly, so I decided to stick with the Chinese bottles.

Speaking of processing bottles, I looked into automated methods and concluded that our current manual method is actually better unless and until we reach the point where we need to produce hundreds of kits per month. Working together, Barbara and I can fill, insert dropper tips, cap, and label the bottles at a rate of about 150 bottles per hour. (I fill and insert the dropper tips; Barbara caps and labels.) The only part of that that can be automated at anything approaching a reasonable price is the filling operation, but that still requires individual attention to each bottle as it’s filled, and actually saves almost no time.


10:36 – Barbara saw an article in the paper this morning about using copper sulfate to kill the mildew that’s appeared on some of her shrubs. I buy copper(II) sulfate by the kilogram, so she asked if I could make her up some right here in the sink. Of course, I agreed. The problem is, what concentration?

Apparently, the concentration needed varies. One site mentioned ranges from 2 to 6 pounds per 100 gallons, which translates to something like 909 to 2727 grams per 379 liters, or about 2.4 to 7.2 grams per liter. Or, as I think of it, about 0.01 to 0.03 molar. Since I keep liters of 1.0 M copper(II) sulfate solution in inventory–the stuff takes forever to dissolve–I’ll just compromise on 0.02 M and dilute one part of the stock solution to 49 parts water.

I did wonder whether the high solubility of copper(II) sulfate would be a problem. If Barbara sprays on 0.02 M copper(II) sulfate, it’ll stick around only until the next good rain dissolves it and rinses it off the plants. That’s apparently why people use Bordeaux or Burgundy mixtures, which are solutions of copper(II) sulfate mixed with either calcium hydroxide (lime) or sodium carbonate (washing soda or soda ash) to form insoluble precipitates of either copper hydroxide or copper carbonate. Apparently these are more persistent because they don’t dissolve in rainwater, but I do wonder whether they’d clog up Barbara’s sprayer. What the insoluble copper salts really are is time-release treatments, because even “insoluble” compounds are very slightly soluble in water. So, apparently copper(II) ions kill fungi even in nanomolar or even picomolar concentrations. I think we’ll start with just 0.02 M copper(II) sulfate solution and see if that makes the fungi gag, clutch their tiny little chests, and drop off the plants.


Read the comments: 22 Comments

Thursday, 8 September 2011

09:47 – My main office system started making a bad sound yesterday afternoon. I hoped it was a fan bearing failing, but I feared it was one of the hard drives. This system is a quad-core that we built a few years ago, and it still has the original main hard drive, a 750 GB Seagate Barracuda. It also has two other hard drives that we installed later, both 1.5 TB Barracudas. Those drives were installed before the 1.5 TB Barracuda was officially released, which should give you some idea of the age of this system.

At any rate, it turned out it was one of the hard drives, of course. It was one of the 1.5 TB drives, which I use for on-line backup. Here’s what happened when I unmounted it and ran jfs_fsck on it.

thompson@darwin:~$ sudo jfs_fsck -a /dev/sdb5
jfs_fsck version 1.1.12, 24-Aug-2007
processing started: 9/8/2011 8.58.17
The current device is: /dev/sdb5
ujfs_rw_diskblocks: read 0 of 4096 bytes at offset 32768
ujfs_rw_diskblocks: read 0 of 4096 bytes at offset 61440
Superblock is corrupt and cannot be repaired
since both primary and secondary copies are corrupt.

CANNOT CONTINUE.
thompson@darwin:~$

Fortunately, the entire contents of the failed hard drive are replicated on the other 1.5 TB drive. I didn’t want to tear down this system to replace the drive, so I stuck a new 2 TB Barracuda in an external USB drive frame and formatted it jfs. I’m now copying about a third of a million files totaling about 1,300 GB from the working 1.5 TB drive to the external 2 TB drive. At USB 2.0 speeds of about 25 MB/s, that’s going to take 13 or 14 hours to complete.

I really do need to take the time to get our computer situation straightened out. Right now, Barbara has a 6-core Core i7 system with 6 TB of disk space in her office, which is gross overkill. She has that system, which was to be my new desktop system, because it was the only one ready to hand when her old system started having problems. She uses her system only for email, web browsing, and so on, so I think what I’ll do is build her a new system around an Intel Atom motherboard much like the one I’m currently using in my den system. Or I may just swap my den system into her office and replace it with another Atom system. At that point, I can strip down and discard my current main office system, which is nearing the end of its design life. Heck, I’m still running Ubuntu 9.04 on it, which has been unsupported for a year now. But all of that takes time, which is in very short supply right now.


Read the comments: 20 Comments
// ------------------------------------------------------------------------------- // end of file archive.php // -------------------------------------------------------------------------------