bantic

Yeah. I think I get a little extra subconscious background-thread-processing boost when I do this.

I took a month-long sabbatical last Fall to learn a new skill (machine learning). I had a project goal to keep me focused on making progress and staying practical: making an ios app that will use ML to "recognize" the letters on a Boggle game grid and "solve" it, printing out all the words on that grid. I made large strides during the month but the project wasn't quite working at the end. After the sabbatical it was back to work at my dayjob and I had a lot of motivation but little time, so I kept telling myself, "No time to work on it today, but I'll try tomorrow."

After a few weeks thinking this and not doing any work, I despaired that my project would end up as vaporware, blowing away in the smoke of my good intentions. So I started a new habit: Committing to work on it, every day, for just 15 minutes.

I use a pomodoro timer (Vitamin-R) and note-taking app (EverNote) to do my daily 15 minutes. Each time I start, I write down my goal for that session, and keep a running list of things I think of to do but don't get to in that particular session. Every day I successfully put in 15 minutes I mark off that day on a 365-day hanging wall calendar that is on the back of my apartment door.

Results:

* I have worked on the app for 90% of the days this year.

* I absolutely got it to a "working" state (I am now facing a diminishing returns problem where I'm putting in too much time fiddling with UX and UI polish, but that's a problem I'm happy to have).

* On over 50% of the days I end up working more than 15 minutes.

* On days when I feel like going longer, and have the time, I'll still occasionally hard-stop at 15 minutes. This gives me something I'm eager to come back to tomorrow, and also (I think) helps to engrain the habit. I want my subconscious to always think: yes, the cost of starting my session today really is only 15 minutes, I will do this even if I don't feel like I have a lot of willpower today.

I worked at a company that hired a developer who was slightly consumed by his regret at being an early employee and then getting out of Groupon pre-IPO. We'd be at a bar and the conversation would die down and he'd look sort of glum and say, "I should be on a yacht right now...". I felt bad for him.

The thing that made this click for me was realizing the "1" in the regex was not special. The regex is just looping through integers (starting with 2) and checking if the target number is divisible by any of them. It can use any digit if the target string uses the same digit. Ruby example using 5 instead of 1:

def is_prime(n); str = "5"*n; str !~ /^5?$|^(55+?)\1+$/; end

The regex essentially says, for integer n, "does 2 ones fit evenly into n ones", then "does 3 ones fit into n ones", etc., which is the same as "does 2 fit into n", "does 3 fit into n", etc.

Hey HNers, I created a game / prediction net where people can tweet in final score predictions for the upcoming March Madness NCAA games. To play you have to post a tweet that includes "@scorecaster" or "#scorecaster", the hashtag for the game (they are listed on the site), and the score prediction (in the same order as the teams in the hashtag), example: "@scorecaster #UMvMU 65-55".

I wanted to make something that adds a playful, social element to watching sports. I tweet and chat with friends back home when watching basketball games, and I wanted to have a way to sort of trash talk/compete with one another. Something more casual and less formal than betting, and more social.

I'd love to get feedback from all of you. My biggest concern is that it might not actually be very fun to try to predict the final score of a game. Sometimes it feels quite arbitrary to me. I'm looking for ways to make that more interesting. If I get a critical mass the first thing I'd want to add is the ability to restrict the leaderboard and predictions list to people that are connected to you on twitter.