luckycharms810

Having worked in software in a few different domains, I've bumped in to this question quite a few times with different companies doing this different ways.

The traditional n-tier / mvc / DAL mindset is that data validation needs to happen in a set of native objects, and once that logic is evaluated data can be updated in the persistent data store.

Why is this preferred over maintaining data integrity at the datastore level, and having all the API endpoints getting to behave as pure RPCs with data retrieval and business logic mixed together?

I spent the first few years of my programming career as a VB.net developer. I worked on the systems side with multi-threaded applications. Over the course of the job I dealt with classes/objects/interfaces/events/cross thread delegates/threads/socket communication. I read about design patterns to structure my code well. I understand that there is an amount of abstraction that lets you achieve what you want to slightly easier, but I assume that this is just the mark of a well designed higher level language.

Since then, I have moved to an open source shop. The database design is a mess, the code base has been sloppy, frameworks have been used without any understanding of their benefits. The day I joined a sys admin asked me what my background was. When I told him I worked at a Windows shop, he asked me "Why did they hire you?"

Today, I received an email about how a company was interested in me, but they would continue their search somewhere else because my prior experience was all VB.net.

As a collective, open source developers seem to look down on Microsoft shop developers. Is there any substance to this?