Working on an improv synth solo for So Talk About It at 3am

I have a strong desire to share all our new songs with everyone here, but a few things remain to be worked out. First, I don’t have an “official” CD–David does, and he’s in Africa for a month. Second, I’m not naive enough to think that the mix I have will be final. We usually go through rounds of tweaking, and, while things sound excellent, there’re always one or two minor changes to make. Third, the album has no title. Fourth, the album has no artwork. I’d like all of these things to be present before I unveil everything officially. So just hold tight for now! I’ll probably break down and put one or two up here anyway, since the only reason I feel secure is that I am validated by my peers.

Speaking of being validated by my peers, let’s go on a tangent: those of you who are e4 weblog owners or frequenters might have noticed that I have added trackback capability to the e4 weblogs. Trackbacks are a cool example of the way the web is supposed to work–now, when you write entries, e4 will check out any links you make, and attempt to send a trackback ping to that site. If the site is trackback-enabled, it’ll accept the ping, and people who visit it will see that your entry references the page! Likewise all the weblogs here are now trackback-enabled, so anyone can send pings from Movable Type, WordPress, or wherever. Let’s just hope for no horrible spam!

(For you e4 bloggers, you can play with the trackback preferences at your weblog config page)

Well, I am now officially back in California for a while, so hopefully I’ll have time to write some more reflective posts soon. Ciao!

I’m not sure exactly how many songs I had then, but it was probably in the vicinity of 5,000. So even if I rated one song a second, it would have taken an hour and a half. What I decided to do, though, to be fair, was to listen to each song fully before rating it. If an average song is 3 minutes, we need to multiply that 1.5 hours by 180–270 hours! I was somewhat optimistic that this 270 hours would pass by quickly while I was working or what not.

I soon discovered that rating songs one-by-one in iTunes was a somewhat inefficient task, given that the only way to rate something was to go into iTunes and click on the little star for that particular song. Not a bad method if you’re already in iTunes, but if you are working on something else and would rather not have your workflow interrupted every 3 minutes, it’s less than optimal. So I wrote some software (iRateTunes) that let me rate songs with keystrokes, without going to iTunes.

(I also created a little app that showed me info about my iTunes library in the OS X status bar, including the number of songs left to be rated, and the total amount of time I’d spent listening to songs in iTunes. This was so I could monitor my progress)

Still, even with these tools, it was slow going. It turned out that the mental energy required to rate song after song was somewhat exhausting, so if I could get 100 done per day, that was good. Most days I got far fewer songs rated, and some not at all, depending on the intensity required by my work. An even more exhausting task was keeping a consistent standard with which to rate the songs. There are all sorts of options–should the ratings be relative to just the songs in my library? Or should it be relative to all the songs in pop culture? Should I try to keep a nice bell curve, so that there are about as many 1-star songs as 5-star songs, and far more 3-stars than either? In the end, I’m not sure I was super-consistent.

But anyway. Today, over a year after I began, my iTunesCount status bar reads: “6794/0 | 52.74” This means that I have 6794 songs in my library, and 0 are unrated! Finally! The 52.74 lets me know that the total amount of song air time with this iTunes library is 52.74 days (I created the library in July of 2004). That’s 1,265 hours! And I don’t let iTunes run while I’m not there listening.

Stats per rating category: 1-star: 443 songs (6.5%) 2-star: 801 songs (11.8%) 3-star: 2752 songs (40.5%) 4-star: 1836 songs (27%) 5-star: 962 songs (14.2%)

• Semifreddis cinnamon swirl bread
• Drinking tea again, since it’s cold enough to
• The Velvet Teen
• Jimmy Eat World’s Stay on My Side Tonight EP
• Takk
• Jed’s Other Poem
• Installing Kubuntu and trying to get wireless drivers working
• Ndiswrapper (see previous item)
• Jonathan Strange & Mr. Norrell
• Ruby on Rails
• Cron + PHP for the win in easy archive/backup
• “tail -f access_log | grep IP” for tracking use patterns on this site in real time, and “tail -f access_log” in general for a good procrastination method
• The World of Warcraft grind
• Flaming Dr. Peppers
• The Palo Alto Moonlight 10k (for complicated reasons having to do with not paying the sign-up fee, I ran as “Ryan Frederick”, in the 30-39 division, though I am still only 23)
• Intense ultimate frisbee with friends
• Dropping temperatures and overcast skies
• Pad Thai all the time
• Harry Potter VI out loud with Nyffy (the Chris one)
• The Firefly Complete Series
• Arrested Development - Season One
• favicon.ico
• Pizza + beer + WoW with my brother at my family’s place in San Francisco
• Uncomfortable honesty
• A dream in progress
• Strong community
• Hope that doesn’t die for some reason

(OK, I did get a little deep at the end; I couldn’t help it!)

Podornot is a fun mash-up of the new podcasting trend (internet buzz, anyone?) and the old Internet time-wasting stand-by, Hot or not.

The idea is to generate a random audio experience; you never know what you’re going to get! But as you go through, you can rate the podcasts you hear, and maybe someday we will use this information to create lists of the hottest podcasts on the web. So go to the site and check it out!

Graphic/visual design right now done by me and Justin, hopefully to be improved in the near future.

Anyway, it’s fun to have an idea, to implement it rapidly, and to see it pretty much near completion in a short amount of time. Go visit!

Recently I’ve taken to a strategy which has been used, to great effect, by almost every software giant in existence. This strategy doesn’t so much solve the workday synchronization problem (as getting up at 4:30 am would, for example), but instead says that, actually, desynchronization is the solution! Namely: around-the-clock development.

This is how it works:

At 6am Eastern, Joe and Pavi get to work at Teleios headquarters. They put in a good 8 hours of work, writing new code and testing code that I’ve written previously. In the course of this writing and testing, they come up with a number of bugs having to do with my web app or with the integration process.

Then, around 12pm Pacific, I stumble out of bed and wander bleary-eyed in my underwear to my computer downstairs, where I check my e-mail and immediately wish I hadn’t. I’m greeted by the results of Teleios’ hard labor the previous day (well, it’s the same calendar day, but I got up so late it’s effectively after their work day), namely a huge list of bugs for me to squash.

So I spend a good 4-5 hours fixing bugs and wondering when it was that I wrote that particular line of code and was I on drugs at the time? About then it’s dinner time, so I’ll do that and hang out with people, play sports or games or talk or whatever. Then around 10pm-12am, I’ll put in another few hours of work, and this is when I code new features, since all the previous part of the day has been spent on fixing old ones.

Sometimes I’ll code (like I did a few days ago) to 3am, which is just about when Joe and Pavi are getting into the office in Florida. Now it’s their turn to be greeted by a bunch of new code which may or may not play nice with their own respective pieces of software!

In this way the proverbial sun never sets on the Teleios workday.