Punk Programming - Tumblr Posts

SQL nearly always.

SQL is based on relational algebra, meaning you have a mathematical garantee for data validity.

And it is well tested, and many automation tools to do it for you. No-sql sacrifices that. ( Note, any category which starts with "no-" is a badly defined category. Always"

No-SQL really only have a use-case when:

1: Your database is so big it needs to be on multiple servers and cannot be broken down to smaller databases

2: You need the database to sync very quickly within itself.

If both of those are true, then SQL might not be the right choice.

And... 99% of databases do not need this. Unless you are making Facebook, twitter or similar you have no need for this.

No-SQL is mostly a bad choice being pushed by Google and Amazon selling cloud solutions ( "cloud solutions" meaning "You have no right to know anytjing about how we store your users data" ) which is done so the big corporations do not have to put too much work into their services and to keep their users ( developers ) in the dark about what is actually happening.

It is much easier to steal your users data when they have no idea about access and what is stored where. Google SPECIFICALLY writes that they WILL access the data you store on their cloud solutions for "business purposes".

Learn SQL, it is not hard, get a automation program to set up your database on a server you actually control. Which can include renting server space.

That is not only the cheaper and more efficient solution in nearly all cases, it comes with the added benefit of not selling YOUR users out to big corp.

If a corporation say they take good care of your data, and use Google, Amazon or certain Microsoft services... they are lying.

Software engineers, how often do you find you’re using SQL vs NoSQL solutions for your systems?