DEV Community

Discussion on: Building Your Own Tools Is Dumb

Collapse
 
kspeakman profile image
Kasey Speakman

The classic application of the phrase is this kind of scenario.

  1. Developer salaries are expensive. ✔️
  2. We need to build a product. ✔️
  3. The product needs to integrate with X. ✔️
  4. We build X. ❌

When I wrote this I was thinking X is "database". Could be filesystem, web server, etc. There is sometimes a dev tendency to see existing tools such as these and think we could do "better". However, the time the team spends building X instead of our product is throwing money on a burn pile. (Unless the product is a database or web server.)

But then some organizations face unique challenges (or have sufficient resources) where building your own is the best play. The example that I heard a while back from Joel Spolsky is that MS Excel team built their own C compiler. And that was a strategic advantage for them.

For an established company, the question is whether the investment will eventually turn into revenue. For a startup, it's also a question of whether the company can keep the doors open long enough to see it.