andnam

On Code etc.: The Haskell / Snap ecosystem is as productive (or more) than Ruby/Rails.

Sun, 29 Apr 2012 07:40:42 +0200

On Code etc.: The Haskell / Snap ecosystem is as productive (or more) than Ruby/Rails. :

This may be controversial, and all of the usual disclaimers apply - this is based on my own experience using both of the languages/frameworks to do real work on real projects. Your mileage may vary. Because this is something that has the potential to spiral into vague comparisons, I am going to…

"Many kids with (attention deficit disorder) have been misdiagnosed, they’re actually bored with..."

Wed, 18 Jan 2012 17:24:52 +0100

“Many kids with (attention deficit disorder) have been misdiagnosed, they’re actually bored with teachers, as most won’t even blink at concentrating on video games for hours.”

- Nolan Bushnell - Atari founder (via it8bit)

hamburglr: First, go grab some headphones. The best ones you’ve...

Thu, 05 Jan 2012 09:14:09 +0100

hamburglr:

First, go grab some headphones. The best ones you’ve got. If the best ones you’ve got are these suckers (or something similar), you should really go buy new ones, but use the best you’ve got for right now.

Take a break from whatever you’re doing for 2 minutes and listen, but just listen to the whole thing, even if you have to multi-task.

Headphones on? Ok. Good.

Now, press play.

“Upular (3D Audio Version)” - Pogo

it8bit: Mario Perler Bead Art - by Black Rose2011

Fri, 01 Jul 2011 15:51:26 +0200

it8bit:

Mario Perler Bead Art - by Black Rose2011

Best of White Ninja: White Ninja and the fat turtle

Fri, 01 Jul 2011 15:36:44 +0200

Best of White Ninja: White Ninja and the fat turtle

littlebigdetails: YouTube - Nyan Cat Progress Indicator on...

Wed, 22 Jun 2011 21:56:57 +0200

littlebigdetails:

YouTube - Nyan Cat Progress Indicator on YouTube

/via @smorgasbrod

See it in action here

emmyc: dad nooo (Ryan why you so good at comics so much all the...

Wed, 22 Jun 2011 21:55:21 +0200

emmyc:

dad nooo

(Ryan why you so good at comics so much all the time)

Photo

Mon, 13 Jun 2011 01:34:26 +0200

My work flow process

Mon, 23 May 2011 02:44:48 +0200

I get easily distracted when I work. I easily get into tack ons or spend way to much time doing major work in areas that have little effect on what I’m supposed to do. I need a way to organize what I’m doing. Writing this blog post is somewhat of a tack on i suppose. I already failed. Blah.

Lets start out with three example projects. A web app, a web framework (which the app uses), and an IRC client (not related to the other projects at all).

Some thinking has created broad todos for each, probably use cases (such as “Usercreates an account” in the web app and “User connects to a database” in the framework).

I organize all these todos into the different projects.

I choose one project that I will work on exclusively to avoid losing focus. What project I choose depends on deadlines mostly, or if it’s hobby programming then I pick the one I would like to work on.

Say I choose the web app. Sounds fun.

I take the first todo. I don’t pick one that sounds fun, here or the hardest one or anything like that. I just arbitrarily pick the first one.

I start thinking about the todo, is it a lot of work (say >30 minutes)? If it is a lot then I try to split it up into parts that i estimate will take about that much time, if it’s less that’s completely fine. If it’s more than 30 minutes then I can split it up further later. All these new todos are added to *the end* of the todo list. I then move on to the next todo until I find one that looks small enough for me to do.

I start work on it. When working on it I’m bound to think of other things I haven’t thought of before, (oh the user needs to be able to log out as well, I forgot about that). I append that to the todo list. If I figure out this todo has other dependencies there are several possible categories here:

It’s a minor dependency in this project (Add a submit button), I can then do it without creating a new todo. If I think I may forget it i might make a separate todo for it and add it to the top of the list (above the todo I’m working on) since I’m going to finish it in conjunction with the current todo.
It’s a major dependency in this project. I create a todo for this, then stash all the current work I’ve done, move this todo to after the major dependency todo (so next time I get to it the major dependency is solved). I then proceed with the next todo.
It’s a major dependency in another project (Web framework needs database layer). I then add it to the end of that project’s todo list. Can’t do it now, working on another project.
It’s a minor dependency in another project (Web frameworks API for connecting to the database is somewhat verbose). Add it to the end of the todo list for that project. I don’t want to do it now since I’m working on another project. I can use the verbose interface for now and then refactor later when I’ve finished the framework todo.
It’s a completely unrelated todo (IRC client has a bug in it, I need to vaccum my room). I just append that to that projects todo list.

This keeps me going in the right track. I’m always working where I should be. It’s also very satisfactory to know that I don’t just ignore things, I add them to my lists. Another good thing about it that I just realized, which made me want to write this blog post is the fact that when something big and unplanned comes up (Forgot to implement proper authentication for web app) then I don’t have to immediately do it. Future Adam will take care of it. So I just append, and keep doing the much smaller and easier task (since it’s < 30 minutes of work). It really motivates me at this point, since I don’t have to think about “all that other stuff” yet.

I also never do anything that isn’t a todo. Fixing minor styling issues in the app is not something I need to be bothered with when I’m doing a todo. I’ll instead just add a todo for it, either specific (submit buttons have an ugly color), or a general one (review graphic design).

I have been working kind of like this for a long time, but for the last few days I’ve been following it blindly. I have been more productive than I have for very long.

I’m not sure if everyone will be efficient like this, or if everyone will find this rigid structure as motivating as I do. But try it out :-)

I’m using Things for OS X to organize my todos. It’s not really necessary to have a proper application, multiple text files (emacs org-mode comes to mind) will probably do equally well. The top down approach also prevents me from needing to specify dependencies between todos, they are sorted topologically as I move through the list.

tillvardags: Viktigt budskap! Uppsatt i Skelleftehamn.

Sun, 22 May 2011 19:18:45 +0200

tillvardags:

Viktigt budskap! Uppsatt i Skelleftehamn.

Error message subtleties

Sun, 22 May 2011 19:18:14 +0200

I had an object with a 1:M aggregate, and there were constraints on the aggregates being added, they could not match the other aggregates in some manner.

Example: A User has many addresses, he may not have several addresses that are the same.

I wrote the constraint check, and the error message ’User already has the address “%s”’.

The subtle error here is that the message is somewhat accurate, but it’s not as informative as you’d like. The error is too specific. The problem isn’t that that particular address already exists. The error is that two addresses are identical. So it’s better to write ’User may not have two identical addresses (got “%s”)’. It tells the client where the problem arises from, and tells him what he specified.

jennipoos: Is this what will happen in Toronto? I hope...

Tue, 03 May 2011 22:21:26 +0200

jennipoos:

Is this what will happen in Toronto? I hope not.

Wooster Collective

Browser navigator objects in JSON format

Wed, 27 Apr 2011 07:11:00 +0200

Here’s a JSON that contains the navigator objects for the latest version of ff/sf/chrome/op, and for ie 6/7/8. Perhaps useful, perhaps not. Either way it’s quite the entertaining read.

https://gist.github.com/943748

Runkeeper says: Abstract as much as possible.

Fri, 22 Apr 2011 23:49:39 +0200

Runkeeper says: Abstract as much as possible.

Struggle of the day

Fri, 22 Apr 2011 15:10:07 +0200

Struggle of the day

My Hangeul Transliteration lib/app

Fri, 22 Apr 2011 14:27:37 +0200

Wrote a JS lib and an accompanying app for hangeul transliteration, good for learning to read and good for writing if you don’t want to bother with keyboard layouts. If only the translate API would give me some dirt.

Piano Basics

Fri, 22 Apr 2011 14:23:44 +0200

So I finally bought a softsynth to learn to play the piano. It’s so much fun. Had to re-learn how to read notes since I haven’t since ~6th grade. Hard.

Obviously I zoom in on entirely the wrong aspects:

Lilypond is awesome for writing music, can export to pdf as well as midi
lilypond-mode for emacs ships with lilypond
Writing music in a GUI sucks. And Garageband sucks even more at it
The production quality of Garageband’s piano lessons is astounding
eMedia Piano Method felt like it was written in the 90’s. Ugh
Need to write a lilypond generation web site, like we did for latex. For funsies
Could use Lilypond to make some kind of music quiz (randomly generate notes, chords, ask user to identify)

Current musical project is the FF8 Ending Theme. Starting with the right hand, exporting to Lilypond as I go, to learn and to export to midi to compare.

Need to log more stuff on tumblr. Not blog mind you. But perhaps that’s a technicality.

Fri, 22 Apr 2011 14:14:10 +0200

Need to log more stuff on tumblr. Not blog mind you. But perhaps that’s a technicality.

Oops. Forgot that I usually write in english. Wow. What happened?

Sun, 10 Apr 2011 11:57:05 +0200

Oops. Forgot that I usually write in english. Wow. What happened?

OpenID-phishing

Sun, 10 Apr 2011 11:53:54 +0200

Sitter och pillar med OpenID. Kommer ihåg hur spotify använder facebook connect. En popup med facebookloggan och ett formulär kommer upp, direkt i spotify. Inget adressfält, inget sätt att verifiera att datan verkligen skickas till facebook. Det tog mig ungefär en månad innan jag gick emot det som var smart och loggade in i alla fall. Har inte behövt ångra det *ännu*.