Francesco Allertsen - Issue #42

Probably you have head multiple time about “Blockchain”, and this “magical” technology that a lot of companies claim will solve a lot of our problems in the future.

I found this great question on Quora, answered by Terry Lambert (in case you don’t know, he has been working on the FreeBSD kernel, around 7% of Mac OSX kernel and Chrome OS kernel, so he is not someone who has no experience in software).

I’m not going to spoil the answer, but he is one of the few that approaches the question purely on the technical part, without thinking about any business behind it. It answers a lot of questions that every blockchain startup founder should ask himself.

I have learned this while working in the industry: there are so many databases, and the database choice that you are going to make will influence your whole business. It will influence how do you structure your codebase, how do you store your customer data and the performance of your whole system.

You can ask yourself “can’t there be one database which is the best of everyone else?” and all the DBAs in the world would reply you with “no”. Store and retrieve data is one of the most complex and crucial parts of your codebase, and if you want to store files like Dropbox or emails like Gmail or payment transactions like PayPal, the needs are completely different, hence why there are multiple databases and each solves one of this problem.

What is the best choice? It’s to think upfront about your pattern and how you would store and retrieve data from the database, then try few alternatives and test each one with some production traffic/data (you don’t want to test it only with a small subset, because you may find a lot of performance problem in the future). Also, it’s important to understand at least a bit how the database stores the information, to make sure that the database guarantees the correctness of the data. You don’t want to have corrupted data after some time because you haven’t thought about it upfront.

This problem is very true also in big companies, and AWS provides multiple database solutions according to your use case. This article by Werner Vogels (CTO of Amazon) explains some of the alternatives and options you can find on AWS and evaluate which one is best for your use case.

Heroku can be used almost exclusively from their CLI (Command Line Interface), and not long ago they launched the autocomplete feature on it.

There are always problems creating an autocomplete tool for your CLI, and they are explaining what type of problem they encountered and how they solved them.

What I really appreciate from Heroku is that even if they have an online page where you can configure everything, they still focus a lot of time in their CLI and the user experience and user interface of the CLI. I think a lot of other services don’t spend enough time thinking about how their customers are going to use their CLI and improve their use case, while only improving their frontend interface. Especially developer-focused services should have a very nice and extremely well usable CLI interface since most of the developers live in their terminal.

I think every entrepreneur or anyone who works in a small startup should read this article (and forward it to his/her colleagues).

For the ones who don’t know, Amazon has a list of 14 Leadership Principles, that every employee at Amazon should follow to create value for the customers and how to work together in the company.

They are very well written and explained, but what I really like from this article is that is written by someone with a lot of experience, so for each principle, he can show you a personal example, and that is extremely valuable. I suggest everyone read it to have a look at how a big company like Amazon works and how they can innovate at such speed. Most of these principles can be also applied to different companies and circumstances, they are very flexible.