Posts from — June 2008

At work, we’re expanding our server room by bringing the entrance wall about a meter forward. This will give enough room so that both my boss and I can stand comfortably in there at the same time; it will also allow us to get behind and around servers as necessary. However, while the room is under construction, I can’t use my desk for all the plastic and dust and equipment flying about.

I spent yesterday camped out in a conference room with a laptop. Students tend to go there to take breaks from research (no food in the lab, dontchaknow), so I actually got to talk to them about something other than computers or their data or their pending conference presentation or thesis defense. It was also nice having the refrigerator nearby, and somebody left a bunch of very tasty Rainier cherries on the table for public consumption.

Today the laptop is inadequate for the work I’m doing, so I’ve commandeered a storage room with an Ethernet jack and made it into my temporary office. I’m going barebones here, though, with just one monitor on my beloved Gentoo box. It’s going to take some getting used to.

When the room is not under construction, we’re taking advantage of the summer lull as a chance to upgrade infrastructure. First on the list is a new server to operate the tape backups on a new jukebox. It’ll be using the mobo, CPU, and memory that we removed when we upgraded the offsite file server, a hard drive that has been collecting dust in a drawer, and a new external SCSI adapter. Hopefully I can fold it into the budding distcc cluster without too much trouble.

Last weekend Fred and I met a wonderful dog named Kooper during an adoption event at a local pet store. We haven’t formally adopted him yet, but this weekend we’re fostering him as a sort of trial run.

Kooper is about three and a half years old, and he appears to be a mix of German shepherd, Dalmatian, and either husky or malamute. According to one of his foster mothers, he had been adopted before by a couple, who apparently did some training (he knows “sit” and “stay” and is very well-behaved). When the couple split up, though, they gave Kooper back to Stray Rescue.

Six a.m. walks notwithstanding, Kooper has done an outstanding job adjusting to us and our home. Our major concern, though, is getting the two cats to warm up to the dog. When Kooper first arrived, both cats started for the back of the house; Kooper’s “chase” instinct kicked in and he pursued, which made the cats run faster, which made Kooper run faster, and so on. Since then we’ve set up a gate dividing the front half of the house from the back half, with Kooper’s stuff in front and the cats’ stuff in back. There has been a very tentative détente as a result.

Every now and then we’ll gently restrain Kooper, open the gate, and let the cats approach him on their own terms. As the cats tire of their relative confinement, they’re becoming more and more willing to do it. Kooper is basically fine with the cats; if anything, he’s eager to meet them and play with them, and sometimes this eagerness makes him jump and whimper, thus spooking the cats. Oddly, though, the normally-social cat (Chip) is still extremely skittish of the dog, and the normally-shy cat (Dale), while still skittish, is making cautious progress towards full acceptance.

We’re still a little uncertain about how this will all work out. Kooper’s pretty large compared to the cats, and combined with his excitability that may keep the cats from fully trusting him. Unfortunately, if the cats and Kooper can’t get along, then Kooper will need to find another home, and we’ll have to find a smaller dog to adopt. Having met Kooper, though, we don’t see him having any trouble finding somebody to love him and care for him. (And if we don’t take him, I am willing to bet that my mother-in-law will be the first one in line to adopt him; we took him to their place today and they got along immediately.)

As an administrator on the Wikimedia Foundation’s OTRS system, I’ve been asked a couple times what the admins look for in potential volunteers. I can’t speak for the other admins, but here is what I personally like to see in OTRS volunteers. [Of course, volunteers get bonus points if they can work in an underserved language, with tickets for an underserved project, or at a time of day when other volunteers are asleep.]

Some of this phrasing may apply best to the English Wikipedia, but the general concepts apply to all wikis.

A history of diplomacy

After a while, we start seeing patterns in tickets. One archetype is the “what do you mean X is not notable” message, where somebody connected to a recently-deleted article complains about deletion policy. A good reply to this category of ticket refers to deletion policy without sounding elitist; it gently points the user towards an appropriate remedy but does not give false hope. That sort of answer is tough, but possible, to write, and when I see that a user has been able to achieve this sort of tact on-wiki I get a little excited.

Experience dealing with thorny biographies

This isn’t an absolute requirement, but boy, does it help. Biographies of controversial figures are quite possibly the toughest articles to edit; not only are they high-profile targets for vandals, but they also get a lot of subtle POV added by supporters and critics alike. Somebody who can clean up these articles is just about a shoo-in for OTRS as far as I’m concerned.

An understanding that OTRS access is a responsibility, not a trophy for your user page

If you tell us that you want to help with OTRS, then we really do expect you to help with OTRS. I know OTRS can be challenging and time-consuming and stressful, and I know that it can take a while to really adapt, but don’t waste our time if you’re never even going to log in.

A willingness to step away from templates on user talk pages

Yes, templates are quick; yes, they’re easy; yes, we have plenty of stock answers written in OTRS; yes, I sometimes use templates on-wiki or boilerplates in OTRS. Sometimes, though, a more specific response is necessary. Remember, OTRS is essentially customer service, and as a result quality counts more than quantity.

“Administrator” status on a WMF wiki

It doesn’t have to be enwiki, but the sysop flag shows me that somebody somewhere trusts you. I do check out RfAs as well; they can be handy in spotting potential problems with a particular volunteer.

Discretion, discretion, discretion!

This one should go without saying. People who e-mail the Foundation expect that we keep their personal information private.

If you think you fit these criteria and still want to help, then check out the OTRS volunteering page on meta.

I’m posting some stuff that has been languishing in my “drafts” file for a while. A more typical update should come sometime this weekend.

The quick version of this post:

energy efficiency saves you money, stupidhead.

The long version:

Look. We all know gas prices are ridiculous, and that market forces are going to be scrambling to catch up for a while, and that there are environmental reasons to be energy-efficient, blah blah blah. Much of our collective problem, is that it’s apparently pretty easy to forget about energy efficiency when energy prices are low. So let’s crunch some numbers.

To calculate the per-hour cost of running a particular electrical item — a light, a refrigerator, a TV, whatever —

• Determine its electrical usage in watts. Some items, such as light bulbs, list this outright; those are easy. Other items, such as computers or refrigerators, list voltage and amperage; these use differing amounts of electricity depending on what they’re doing, but you can figure the theoretical maximum wattage by multiplying volts by amperes.
• Divide usage in watts by one thousand to get usage in kilowatts. (Americans: “kilo-” means “one thousand”. It’s that icky metric stuff, I know, but stick with me.)
• Find your local utility’s rate per kilowatt-hour. It should be on their Web site, though you may have to wade through several types of rate. (Right now, I would look at my utility’s summer residential rate, which comes to 7.92 cents per kilowatt-hour [PDF].)
• Multiply the per-kilowatt-hour rate by the usage in kilowatts. Your result is the per-hour cost to operate this item.

As an example, let’s discuss lighting in my kitchen, where there are five bulbs in a ceiling fan. On an average day, we’re in there using those lights for about three hours total; those lights are typically off when they aren’t in use. Throw in a thirty-day average billing cycle and here’s how much it used to cost to light that room, before we swapped incandescent bulbs for CFLs:

• Five bulbs, 60W per bulb = 300W total to use
• Three hours of usage per day, 300W while in use = 0.9kWh per day
• Thirty days in a billing cycle, 0.9kWh per day = 27kWh per cycle
• 27 kWh per cycle at the rates above ~= $2.14 per month in summer

Using 14W CFL bulbs in the same pattern providing the same amount of light:

• Five bulbs, 14W per bulb = 70W total to use
• Three hours of usage per day, 70W while in use = 0.21kWh per day
• Thirty days in a cycle, 0.21kWh per day = 6.3kWh per cycle
• 6.3kWh per cycle at the rates above ~= $0.50 per month in summer

And that’s just one fixture in one room, at rates that are relatively low for U.S. urban centers. The savings are even more pronounced at higher rates: in Juneau, where rates recently hit 56 cents/kWh, that one light fixture would have cost $15.12 per month with incandescents (ouch!), but only $3.53 per month with CFLs (much less ouch).

While it will cost around $10 to put five CFLs in that fixture (vs. $2.50 for five new incandescents, using multi-pack prices from a Home Depot near me), the extra up-front cost will be recaptured after about 4 months of average use at my house. It’s also worth noting that CFLs last about five times longer than incandescents, so you can often come out ahead just on the cost of the bulbs themselves.

(Aside: What’s the heat source on the original Easy-Bake Oven? An incandescent light bulb! “Incandescent” means “glowing hot”, after all. Think about that while your air conditioner runs up your utility bills.)

[And yes, LEDs are considerably more energy-efficient and longer-lasting than CFLs. Unfortunately, their up-front cost is appalling at the moment ($60 for a single bulb?!), so until prices come way, way, way, way, way down I don't see LEDs replacing CFLs.]