#fail-fast

Fail-Fast Reliable Software Strategy: Debug Failures Effectively

I love cooking and use my  Thermomix a lot. If you hadn’t heard about that amazing innovation, it’s a kitchen robot… Well, it’s a magical super cooking machine. When designing the Thermomix, its designers took the approach of fail-safe instead of fail-fast. This is a smart choice in this case, but it has its drawbacks. E.g. my machine tried to recover from a failure which sent it into an infinite…

Java: Difference between HashMap and Hashtable

HashMap and Hashtable both implements Map interface and used to store data in key and value pairs. Both use hashing techniques to get unique keys.

Apart from some similarities, there are many differences between HashMap and Hashtable classes as follows:

java.util.HashMap java.util.HashTable HashMap class introduced in JDK 1.2. Hashtable is a legacy class. HashMapinherits AbstractMap class.

First thing to mention about List interface is , List is a child interface to Collection interface. List is an ordered collection which is introduced in 1.2v of Java . List interface is also known to be Sequence collection. The hierarchy of List interface is shown below 

public Interface List<>

All Super Interfaces ( extends)

 Collection , Iterable

All known implementing classes

AbstractList,…

There's a neat little command you can add when you want to run the tests to your code.  I use rspec, and you can type:

rspec --fail-fast

And what will happen is that the system will start running your tests and then freak out as soon as something fails.  This is in contrast to running it without "fail fast", which runs all your tests and then tells you a summary of what failed.

One of the benefits of fail fast is speedy iteration.  If I catch something wrong early and I fix it, that fix might apply to later problems as well, hence saving time.  Also, in a situation where you are striving for perfection - as we do with automated testing in programming - one fail is as bad as 50 fails, so you might as well know earlier than later that something is wrong.

During one of my idle moments of thinking and pondering, I extended this philosophy beyond testing code and into the realm of the real world.  What if I added "fail fast" to my career?

Let's say I was rejected by someone I reached out to professionally.  Should I automatically regroup and change how I approached the person?  Not really.  Unlike coding, life operates more on statistics.  You may have the best pitch in the world but you just happened to pitch the one person that will never like your idea.  Better to gather more data points before calling in a fail.  Sort of like:

career --fail-enough-times-to-know-it-is-a-fail

Or:

career --fail-smart

Though we may never reach perfection - even 100% passing tests doesn't mean you wrote all the right tests or tested correctly - fast iteration can be smart if done correctly.