Hacker News new | comments | show | ask | jobs | submit login

Letting go of ORMs is so easy with good lightweight model binding libraries. e.g. Dapper (https://github.com/StackExchange/Dapper)

We use something like Dapper at where I work. I think for some projects it's fine, but for some a full ORM can really pay dividends. Especially if you're still working out your problem domain, all the hand crafted SQL can be a pain to update if you find you need to add just 1 more field, or if you figure out you need to split a table up because of a new unforseen feature, etc.

Dapper is great for reads but it doesn't handle complex updates nearly as well, although I'd argue most of those are architectural/design issues in nature.

How do model binding libraries compare to ORM's w.r.t. performance, e.g. query optimization?

ORMs are doing more work and are naturally going to have more overhead. With Dapper , at least with the way I used it, you're going to write your own SQL queries so there is nothing to optimize. I've even run into some nasty SQL server parameter sniffing issues. But despite that I still prefer the style over ORMs.

MyBatis is great for Java as well.

Applications are open for YC Winter 2018

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact