Category: Geekery

What happened there?

StackOverflow reputation tracker

Somehow overnight, I got 80 new points from votes, but somehow my net reputation went down 133 points without any of my other questions or answers getting any down votes. I have no idea how that happened, but I don’t like it.

Update: It turns out that Jeff Atwood decided to implement some “fraud detection” and removed what he thought were “suspicious” upvotes. I should consider myself lucky – Konrad, who had the second most points on the system, lost 6,000 points and Daok, who was pretty high up in spite of the fact that English is obviously not his first language, lost 5,000.

First snow drive in the Prius

I drove my Prius in a snow storm today for the first time. It did a pretty good job, considering. I’ve never had a car with traction control before, and Vicki has frequently complained that hers doesn’t allow her to get out of the driveway when the snow is bad. I didn’t have that problem, but there was one point where a pickup truck in front of me fishtailed badly on some snow that I just sailed through. Sure, there were times when the car didn’t react to the accelerator, but it kept going straight and didn’t get stuck.

In other news, I decided to reboot my colo box for the first time in 327 days (I checked the uptime before) because I got a new libc-xen. Of course, being that long between reboots, it had to fsck every file system before it would boot, so that took some time.

And in other other news, the right shift key on my laptop is getting balky. I’m going to have to see if i can pry it off and clean out whatever is lodged under it.

My Stack Floweth Over

I spend most of the time that I’m waiting for things to happen (compiles, application start ups, etc) reading and answering questions on StackOverflow, the best site for programming related questions bar none. One of the things that makes StackOverflow so good is their “reputation” system (although I keep calling it “XP” just to see if anybody will start copying me). You get XP when other users vote up your questions or answers, and when you “level up” to certain levels you are trusted to do various book-keeping tasks on the site like editing other people’s posts or closing them. There are also various badges you can get (or “Achievements” ) when you accomplish various things like getting a certain amount of up-votes on an answer. It’s a great system, and I don’t just say that because I am in the top 15 for XP. But one of the things they do is limit the amount of XP you can get in a single day. And overnight, some of my older answers got a whole bunch of up-votes, so I started off the day with nearly my limit, and I’ve now hit my limit for the day. Which means that for the rest of the day, I’ll be answering questions out of the goodness of my heart (and a design to show-off) rather than for XP.

This also means that I might have to actually be productive at work today. Sigh.

It never fails

It never fails that when on “FAA Data Reload Day” (which occurs every 56 days on the ICAO cycle), I manage to screw something up and end up staying up late. It doesn’t matter how early I start.

Today’s screw up was after loading the data, I realized that I’d done something wrong, and needed to restore the database to the state it was before I started the load. For reasons too complicated to go into here, I load the data on my home Linux box, and then scp it up to my colo box where the web site lives. The database that lives on my home box doesn’t have all the same tables as the one on my colo box, just the tables that are important to data loading.

So, I thought, the easiest way to get back to the data as it was before the data load is to upload the script I use to export the appropriate tables on the home box to the colo box, run it there, copy the file back to the home box and load it. Except after I loaded it, I noticed a distinct lack of data on my home box. As a matter of fact, it appears that the load went way too fast, like it had no data at all. A quick look at the export file confirmed that there wasn’t any data in it, just some table deletion and creation stuff. Oh oh.

That’s when I realized that one of the consequences of having different versions of PostgreSQL on the two boxes was that “pg_dump … -t waypoint -t comm_freqs -t runway…” works on my home box, but not on the colo. Not thinking too straight, I then used a ‘for table in …” command to run pg_dump on each table individually. When I copied them home, I discovered that this messed up the foreign keys rather badly. So I tried to manually stitch all the files together. That wasn’t working very well, because I had things in the wrong order and the foreign key stuff still wasn’t right.

That’s when it suddenly hit me. Duh. The whole reason I have an external drive on my machine is so that I can do hourly rsync backups. I have a copy of the postgis.dump file that I copied over 56 days ago. As a matter of fact, I have dozens of copies of it. The only reason I was avoiding it was because I had done a few small manual modifications to the database since then. But those were still in the history buffer of psql, and so they were easy to reproduce. I restored the backup, made the changes at around 10:45, ran the updates again, and now here it is at 11:30 and everything is finally done.

I just hope this doesn’t happen again in 56 days, although I’m sure it will.