Eduardo Grajeda's Profile

However, while pairing slows you down in the quantity of the work you create, it speeds up the quality of the work you create.  This leads to less bugs and more time for development as the project progresses.

Use PHPUnit for unit testing

Don’t invent your own standards, see what phpDocumentor has to offer. Doxygen also supports phpDoc tags

For documenting the software project, try DocBook - XML-based format that allows you to quickly publish a PDF document, or a Website with documentation

Support 3 environments - development, staging, and production

Everything that needs to be done on the production servers needs to be automated

Check out Monit and Supervisord for deployment monitoring

Long projects eat at you and you’re not even looking to do good stuff you just want to finish things and they don’t turn out well.

Create a situation where projects are short and there’s excitement and it’s a short 2 week project and it leaves people in excited mode.

2 week rule.

37Signals doesnt do roadmaps, specs, projections.

Planning is Vastly Overrated

Don’t want to be boxed into decisions you made 18 years ago.

Get rid of abstractions

Focus on what’s real right now.

Don’t worry about the things you might do.

Decisions are temporary

They think that what they decide today has to be that way forever.

Optimize for now. Don’t worry too much about what might happen. Focus on today.

Red Flag Words

Interruption is the enemy of productivity

Proximity invites collaboration.

“Average work day has been traded in for work moments.”

Strategy: on Thursdays nobody talks to each other. Passive collaboration instead of active collaboration. If someone is busy they can put it aside and come to it when they’re ready. Interruption points screw your days up.

Focus on what doesn’t change

Focus on what doesn’t change and think about the things that matter today and will matter 10 years from now.

Reinvest in the things that don’t change.

Worrying about things that don’t matter yet

ason finds that it’s better to sketch with a sharpie rather than a pen. With a sharpie you can’t get lost in the detail. You don’t want to worry about the details too early.

The longer it takes to develop something the less likely you’ll launch it.

Underdoing

Solve the simple problems because the big guys who have all the money won’t even notice you - you’re in a market they’ve already moved in.

Find the right size

Oxford University is one of the world’s best universities. Why aren’t there tons of them? Because the Oxford that exists is the right size.

Follow the Chefs

Always be Questioning:

Give up on hard problems

Work less

One lesson that you might pick up here is that is OK to give your tests extremely long method names.

Our test code is becoming easier to read since we have introduced contexts.

change from what we are physically doing to what we are trying to accomplish.

We should all strive for great readability, but don’t let the quest for English like tests take you to far from what you are actually doing: writing tests.

taking a mentoring approach, instead of dictating your opinions.

“I told you so”