aubreyturner.org

I’ve heard of people being addicted to video games before, but this guy may be the first to show that what might sound like a fairly harmless addiction can be deadly.

A South Korean man who played computer games for 50 hours almost non-stop died of heart failure minutes after finishing his mammoth session in an Internet cafe, authorities said on Tuesday.

The 28-year-old man, identified only by his family name Lee, had been playing online battle simulation games at the cybercafe in the southeastern city of Taegu, police said.

Or perhaps he just had a bad ticker…

In my earlier ruminations on the changes to Daylight Savings Time I mainly considered computers.  However, the change will affect many other devices besides computers.

All of the clocks that you see in stores that bill themselves as “atomic clocks” are actually regular clocks that set themselves using the NIST WWVB reference signal.  WWVB broadcasts at 60kHz, using a BCD time code format.  I have three clocks that use this method (including my alarm clock) to keep themselves accurate.  Generally, they resync once per day, late at night (when the wave propagation is best for LF transmissions).  These clocks also have adjustments for DST and for your local timezone, so that they correctly compute the correct local time based on the UTC standard that is being sent.

I was initially worried that these clocks will not properly adjust to the new beginning and ending dates of DST.  However, it appears that the time code format includes two bits for signalling DST.  Page 21 of NIST Time and Frequency Services (NIST Special Publication 432) (PDF) gives the information on how these bits signal if DST is in effect or when DST is starting or ending:

Daylight saving time (DST) and standard time (ST) information is transmitted at seconds 57 and 58. When ST is in effect, bits 57 and 58 are set to 0. When DST is in effect, bits 57 and 58 are set to 1. On the day of a change from ST to DST bit 57 changes from 0 to 1 at 0000 UTC, and bit 58 changes from 0 to 1 exactly 24 hours later. On the day of a change from DST back to ST bit 57 changes from 1 to 0 at 0000 UTC, and bit 58 changes from 1 to 0 exactly 24 hours later.

So it appears that the data stream includes enough information for the clocks to adjust to the change as long as NIST sends the DST/ST transition bits on the correct days.  Of course, this also requires that the clock designers followed the standard and didn’t code the DST start/end rules into the firmware.  I’m hoping that they followed the standard.  Otherwise these clocks may require manual fiddling (provided it’s possible) to continue to show the correct time in 2007.

Update:  A short article on problems for VCR’s and cell phones, via Slashdot.

I made two trips to Radio Shack during my weekend adventure in home theater wiring.  The first was on Saturday morning to get more speaker wire.  I went to the local, Keller Radio Shack because it’s close and fairly convenient.  The problem was that they try way too hard to push you to buy other stuff.  To the point where I hate going there.  But, once again, the only reason I go there is because it’s close by.  If I’m not in a hurry, I’ll go somewhere else. 

This time, all I wanted was some speaker wire.  I checked the website and it showed it as being in stock.  First, the guy tried to mention that they had Monster wire, but I didn’t care about that.  Audiophiles may notice the difference, but Monster has always seemed overpriced for what you get.  Then, at checkout he tries to sell me batteries.  I’m used to this, because they do it all the time.  The problem was that this guy wouldn’t take “No” for an answer.  As far as I’m concerned, once I say “No” to an offer, that’s the end of it.

For my second trip, I checked the stock on the remote extender at the Watagua store since I was going to be in that area.  They were a lot less of a pain in the ass, so perhaps I’ll start using them more.  Still, I don’t go into a Radio Shack unless I’m in a hurry and don’t want to go all the way over to Fry’s.

As usual for me with any kind of home project, they always start simple and then turn into a long, drawn-out process…

I wanted to get a new stand for the TV in the bedroom so it would be high enough to be seen over the foot of the bed.  I also wanted to move it into a corner so that it wouldn’t be too close to the foot of the bed as I walked by.  The final impetus was when I rubbed up against it and the clip of a knife in my pocket scratched the wood of the footboard.

As I was looking at racks and stands it occurred to me that I had a bunch of leftover equipment that I wasn’t using.  Over the years I’d upgraded all the components in my home theater, to the point that I had a receiver, a set of speakers, and a VCR and DVD player.  So with this realization, the project went from “get the TV out of the way” to “put all this equipment to use and move the TV.” 

After looking at various catalogs and websites, I finally arrived at a suitable stand (in cherry).  But it occurred to me that there were going to be a lot of wires running out of this thing to the various speakers (five satellite speakers plus a subwoofer).  I didn’t want them laying on the floor or running under the carpet.  And I’d learned in the past that wires hung from the wall or the ceiling look like crap.  I’d previously used some self-adhesive wire-tracks, but I found that over time they’d eventually come unstuck.  So I found some on-wall nonmetallic raceway at Home Depot that screws into the wall. 

While these things look pretty good when they’re done, they’re a pain to put up.  You have to drill six holes for each 5-ft section, and put in the associated inserts and screws.  It took me over 5 hours to put up all the tracks and run most of the wire (I started around 7:30pm on Friday and had to stop round 1:00 am when I ran out of speaker wire, with only one speaker remaining). 

When I was all done, the setup looked good and sounded good.  There was just one problem:  the receiver and other components were below the foot of the bed, so their remotes wouldn’t work.  One more trip to Radio Shack (and there’s a rant involved there), and the problem was solved.

In the end I’m not sure if it was worth all the trouble and expense, but at least I’ve got a kick-ass system in the bedroom that can scare the dog (if you’ve experienced the THX “thing” at the beginning of the Star Wars: Episode IV DVD you’ll understand.). 

My office has two computers and a laser printer that all put out copious amounts of heat, making it pretty toasty in here, especially in the summer when it hits 100°F outside.  I’ve responded by putting up blackout panels behind the curtain.  They block out a lot of heat and most of the light.  Unfortunately, it meant that I couldn’t see out and that it’s too dark to see in here.  I got a lamp for the desk to allow me to read the occasional printed document.

But lately it’d really been bugging me that I couldn’t see what was going on outside.  I had an internet camera laying around so I installed it today.

So now I can see out the window again.  But it just seems weird to use a remote camera to see out of a window that’s three feet to my left.  I suppose when fall hits I’ll be able to “see with my own two eyes,” but until then the nuclear blast furnace known as the “Texas summer” keeps me safely indoors behind cover. 

If I were to believe in a diety, I think that I might have to pray to it for deliverance from projects where the customer has already tried to engineer a solution and estimate the cost before they bring me in. 

I’ve spent the past several days dealing with fallout over the fact that once I get involved and take a look at things that I tend to find that their proposed “solution” is in fact a big mess.  This leads to me spending lots of time explaining why the original “estimate” was wrong and why it can’t be done the way they thought it could.

I know in some of these cases they did these things because when I was previously in the DO that it was hard to get me assigned to work on these things.  But the other problem is that one of my customers is a frustrated designer/programmer who is now working in the business area.  His requirements specification documents read more like design outlines than requirements.

To give an analogy, imagine that you want to have a new building built and you have some ideas about what you want it to contain.  Now think about how realistic it would be if you (without much knowledge of architecture, construction methods/costs, or building codes) tried to create a blueprint and to estimate the cost.  Do you think your estimate would be very accurate?  Would your blueprints actually be usable for construction?  Would you actually try to use this estimate to obtain financing for construction?  Further, would you then estimate a completion date based on all this?

Most people wouldn’t expect an architect or contractor to accept these as binding, yet somehow when it comes to software they think this is OK.  They will come to me with a pseudo-design which won’t work and with an already established budget based on their finger-in-the-wind guesstimate.  And then they will complain bitterly when the real design and cost estimate are different!

This isn’t to say that I won’t try to accomodate their needs, but it sometimes boggles my mind the unrealistic expectations that people bring to the software development business.  But why should I and my development team be punished for bad assumptions on the part of other people?  I can only hope this gets better now that I’m dedicated to this project and will be involved in these sub-projects earlier.

The new Texas law that forces stores to lock up anything with pseudoephedrine in it goes into effect today.  It requires stores to get ID from purchasers and keep a log of all sales.  It’s supposedly “for the children” and to prevent people from cooking up meth in small labs.

Frankly, I don’t think this is going to put much of a dent in the meth supply.  First, the majority of it is coming over the Mexican border via couriers (who cross with the other illegals).  If I didn’t know better, I’d almost be inclined to believe that this bill was written at the behest of the big illegal drug labs south of the border.  If it has any effect on the small labs, it would likely cut down on their competition with the big ones in Mexico.

Given the law of unintended consequences, I wouldn’t be surprised if some of the meth cookers start targeting delivery trucks to get their supplies.  Alternately, they might also start using networks of “buyers” who go around to different stores.  I just don’t see this having much of a dent in the overall availability of meth.

In the end I suspect this will only end up inconveniencing people like me who rely on Claritin-D.  I use it year-round, since I seem to be allergic to a select group of things that always seem to be in the air (although I get it worse when the mold count goes up).

Ever since I completed CERT training I’ve gotten periodic invitations to participate in various mock disaster drills (such as the one at Texas Motor Speedway, the one at DFW airport, etc).  Usually they want us to act as casualties, although an upcoming one in Ft. Worth would allow us to practice in more active roles.

The main thing I’ve noticed about all of these drills is that they occur on weekdays during (or close to) regular business hours.  I imagine that part of this is cost driven, since having first-responders out on nights or weekends would likely involve bringing in people from first-shift who would incur overtime.  But it hampers participation by volunteers to those who don’t work during the day. 

Aside from the volunteer pool issue, though, wouldn’t a real disaster likely occur at any time?  Shouldn’t second and third shift responders be in on these?  Wouldn’t a good disaster drill test the system during the later shifts to see how well the system responds at times of lower staffing (i.e. it would stress their emergency callout capability)?

Somehow I don’t think bad weather, plane crashes, or terrorists will confine themselves to regular business hours on weekdays…

On thinking further about DST and the computer situation, I suspect that the astronomy example is probably contrived, as all calculations would likely be done in UCT and then only translated to local if needed.

However, I would expect that there would be cases where it would be necessary to calculate the correct local time of day at some time in the past, which would require that such calculations properly take into account changes in the DST rules over time.

The example that popped into mind had to do with an audit or dispute of the response time for a support call that had been handled in the past.  Consider that some support contracts have limitations on when support is provided and how long the company has to respond with a callback to the caller.  As an example, let’s assume that a contract specifies that a caller has support during normal business hours in their own timezone (e.g. 8:00am-5:00pm in the Central timezone) and is entitled to a callback within 2 business hours.  This means that if the caller opens a problem at 4:00 PM Central, then the company has to respond by 9:00 AM the next business day.  So if a caller comes back at some later time in the future to dispute the call and claim that the contract conditions were not met because the callback was late, it would be necessary to correctly translate the timestamps (problem open time, callback time) from UTC to the local timezone to determine if the contact conditions were actually met or not.  It is often the case that callers misunderstand the contract terms and complain about expectations not being met, which could possibly result in the company giving back revenue if the actual times are not properly calculated.

A while back I wrote about spying printers (and copiers), which embed identifying information on each page.  The EFF is now asking for assistance to gather information from as many printers as possible in an effort to identify which models have this technology and to try to break the encryption/encoding scheme they use so that the data that is being encoded can be shown.

Link via Slashdot.