the_list.txt

New blog
    * Markdown
    * distraction*free writing
    * better way to manage data?
    * Auto-saving
    * substance.io?
    * code samples
    * upload and insert assets
Javascript
    * proxies
    * async programming
    * function helpers
    * Backbone, Ember
node
    * streams (nodeconf)
    * http client
    * procstreams
        * new release
        * integrate glob
        * integrate nexpect?
    * domains

For your blog, have you used Mou for writing? Also, I highly suggest Jekyll.

Mou looks cool. Not sure I want a desktop app. It would be cool if it could upload to a server on save :) I also decided I didn't want a generated site. Or at least one that has a few more management tools than jekyll does.

@polotek In that case you could always fork my editor and do that on your server (if you do, use the 0.1.0-build-refactor branch). Icky, but you could also use Mou to edit files over FTP so upon saving it'd upload.

As for Jekyll, I totally understand. I ended up going with it personally because i always over engineered my blog. It forced me to keep it simple and put restrictions on what i could and should do with it.

Can you expand more on "proxies"?

@jamescarr I need to explore the harmony proxies spec and get more familiar with it. And then I want to examine some ideas I have about how I might use them. https://twitter.com/#!/polotek/status/189431189864579072

Interesting. Coming from a java background I have made heavy use of dynamic proxies and often find cases where they could be helpful in my node based apps

Very cool that you're hashing this out in public. Keep us appraised of the progress. It's an interesting topic.

Sometime this year, I got in the habit of keeping time slips for personal projects using Harvest. Today, I'm working on a timezone aware time and date library in JavaScript. I've worked on this library for 111.58 hours. I budgeted 40 hours.

I've spent hours upon hours looking for savings of speed and size where speed and size are already trivial. It's a dark feeling when I'm not making progress. I'm happy to work on it, but part of me knows that it is a bike shed, and that part of me is not happy about this project.

I'm also building an evented I/O b-tree in Node.js. I've spent 123.92 hours on it. It is outwardly more interesting that a time and date library, but inwardly, I could just as easily work on one or the other.

The time slips teach me that everything takes forever. Nothing takes 40 hours.

This give me permission to laugh out loud at anyone who talks about finishing something over a weekend. You can spike something in a couple hours, but you can't finish it. I now find the concept of the Hack-a-thon repellant. Why would you want to compress your working time in a jumbled mess of anxious hours? Most of my design takes place in the in between time. Sleep is part of my workflow.

Nothing is finished. Finishing something will be a new experience. I look forward to it.

My take away from my accounting is that I'm going to deep dive on anything I touch, so I may as well choose a difficult problem. Whatever I I work on, I splurge on time. There's no point in trying limit scope. I won't adhere to those limits.

@bigeasy I have the opposite problem. I never dig into anything too deeply before I'm distracted by something else interesting.