Anthony Panozzo's Blog

Getting Real: The smarter, faster, easier way to build a successful web application Author: 37signals Published: 2006 Pages: 177

This book describes the philosophy of development at 37signals when they created five successful web applications while taking no venture capital.

The biggest theme that I can see with this book is that you should do a few things exceptionally well as opposed to doing many things, only some of which are done well. I see this theme repeated in many works that I read. You don't want to be good at several things, you want to be world-class at whatever you choose to do.

I really liked the firm's approach to taking venture capital. They are big fans of bootstrapping by keeping functionality at a bare minimum to what is required. Having less features means a firm can spend time on the very most important features and ensures that they are agile enough to change plans once a page is actually out there. Every single feature that goes into your software is a liability, and users will not want it taken out. So think carefully about every feature. Be ready to say "no" a hundred times. If you keep hearing about something, you will know that users really need it. This comes back to designing something based on the actual underlying needs of users and doing continuous usability testing and getting continual feedback to ensure that designs work as desired.

I recently started a new supplementary coding technique that I have not heard about anywhere else. I'm not sure what to call it, although internally I've been calling it streaming, an abbreviation for stream of consciousness programming. It works by writing out how you are solving a problem.

Here's a real-life example that likely took place over the course of a couple of hours:

control_parsing
 Go through documents and figure out what appropriate values would be for the metadata and the control data to get it out
  Given a standard document, the ability to correctly parse out the fields
   Given a patient identifier, find the matching patient
    What API call do I use for this?
     Where is the system id of 387 coming from?  I could swear I've seen that before
      I don't think it's actually used, I think that we just use the patient identifier and the user's institution id
       Pretend we are calling the API call with the known patient identifier and user's institution id to ensure that I have the right mental model

Process

I use vim to type out my thoughts and save them a text file. I could see marking things off by putting '#' at the beginning of the line, which for me would gray the line out, but normally I just delete things when I get done. This results in a nice feeling of crossing things off of the list. If I'm in the heat of solving a problem, I often don't write or delete lines. But when I come to a stopping point or am considering what next to do, I usually write out what I am thinking. This minimizes overhead and preserves flow.

Awhile back I did a thought exercise. The exercise was a way of seeing what you valued most in life. It goes like this:

Every day for a week, write down the following snippet and answer it ten times:

Something I really want out of life is…

It seemed interesting, and I like it when things are spread out over a period of time so that you can see yourself in different ways. So I did this off and on for a few weeks. I expected that depending on the time of the day and my mood that this would fluctuate quite a bit, but I was surprised to see some recurring themes and a strong direction.

One of the things that struck me was "Something I really want out of life is to have meaningful conversations". I put this down a few times, but I never really defined "meaningful conversation." Perhaps I knew what I meant, but perhaps I didn't and was just being vague. So I thought about it for a bit.

Meaningful conversations are one of the most desirable events that I can think of. If I knew that I would have a great conversation but I had to miss something else to have it, there are few things that I would rather do with that time. Of course, this rarely happens in practice because there is a certain degree of spontaneity, it just so happens that two people are walking along similar paths in life, considering the same problems or running into the same obstacles. It is, of course, possible to be too analytical, and thereby spoil things. As with anything that is desirable, it is advisable to not become too attached to previous experiences and expectations.

When I was growing up, my relatives always seemed to ask me the same questions. "How is school going?", "Do you have any girlfriends?" and, of course, the time-honored classic: "What do you want to be when you grow up?"

While the second question usually flustered me, it was the third question that seemed to evoke the most interest from my relatives. They would say ominous things like, "If you want to be happy in life and make a lot of money, you had better figure out what you want to do." I usually said something like be a doctor or lawyer, and they seemed mollified and the conversation changed to adult topics.

Seeking a way to extend completion in zsh for Capistrano tasks for a Rails project, I found a script that apparently works for OS X. However, it didn't seem to work out of the box for Ubuntu, so I made a few modifications. Put this in a file and source in your .zshrc file, and you should be set!

_cap_does_task_list_need_generating () {
  if [ ! -f .cap_tasks ]; then return 0;
  else
    accurate=$(stat -c "%y" .cap_tasks)
    changed=$(stat -c "%y" config/deploy.rb)
    return $(expr $accurate '>=' $changed)
  fi
}
 
_cap () {
  if [ -f config/deploy.rb ]; then
    if _cap_does_task_list_need_generating; then
      cap -T | cut -d " " -f 2 | sed -e '/^ *$/D' -e '1,2D' >! .cap_tasks
    fi
    compadd `cat .cap_tasks`
  fi
}
 
compdef _cap capompdef _cap cap

The first time that you attempt completion with cap, the available tasks will be analyzed and cached. Then, future completions will work correctly and quickly. If you have any problems, deleting the .cap_tasks file will signal that the generating process needs to happen again.