3 ms·
Interesting. I encourage people to publish early/often. However, depending on the problem, I would rather see some tests early on which would be defining the
by mgkimsal 5y ago
Interesting. I encourage people to publish early/often. However, depending on the problem, I would rather see some tests early on which would be defining the approach being developed. It's not necessarily a black/white issue for me, but there are cases where having some tests (or even just some written docs with the thinking outlined) is preferable.
- ergocoder 5y agoI understand that there might be some cases that prefer a test. However, when communicating a guideline like this, it is better to make it simple. Otherwise, then it would be equivalent to no guideline. A guideline like this isn't meant to cover 100% of the cases anyway. If there are exceptions, I simply tell them to add some tests.