3 ms·

I’m always surprised when I see people say things like this, based on my own experiences. In the last month or so, I’ve used: - linear algebra (transformation

I’m always surprised when I see people say things like this, based on my own experiences. In the last month or so, I’ve used: - linear algebra (transformation matrices and inverses primarily) - calculus - approximation methods to get “good enough/fast enough” solutions to the Chinese Postman Problem and the Traveling Salesman Problem - a self-derived plane sweep algorithm for doing something special with a ton of planar polygons The only conclusion that I’ve come up with so far is that it either varies dramatically by which industries you’re developing software for, or other organizations have people with specialized skill sets that work out the broad strike solutions to these types of problems before getting the software teams to build the production versions of them.

mkl95 6y ago
Don't get me wrong, I do plenty of exciting stuff on my side projects (involving linear algebra, some AI, etc.) I completely agree with your last paragraph. I am a backend engineer, a job that at its core involves creating and fixing CRUD apps, and writing relatively simple business code. No matter how exciting people try to make it seem.
- amw-zero 6y ago
  I think the "CRUD app" classification is never accurate. For example, your app almost certainly performs queries that have a where clause. That means that the R step has logic, which means you can (and I know I do) get the logic wrong. Also, there's usually complicated rules during C, where a user is only permitted to actually create data in certain cases, and when data is created other data needs to be updated to remain accurate, etc. So yea, of course data is Created and Read, in that every application maintains data. But that doesn't mean that it's a CRUD app. If you have any where clauses or if expressions, it's not a CRUD app. Finally, the most popular way of storing data is the relational model. Which, is just math applied to the notion of data: https://www.seas.upenn.edu/~zives/03f/cis550/codd.pdf https://www.seas.upenn.edu/~zives/03f/cis550/codd.pdf
amw-zero 6y ago
I'm with you. I feel like all actual business problems have a lot of applicable CS solutions. Maybe if you're building UI-heavy stuff, there's less theory required there?

mkl95 6y ago

Don't get me wrong, I do plenty of exciting stuff on my side projects (involving linear algebra, some AI, etc.) I completely agree with your last paragraph. I am a backend engineer, a job that at its core involves creating and fixing CRUD apps, and writing relatively simple business code. No matter how exciting people try to make it seem.

amw-zero 6y ago
I think the "CRUD app" classification is never accurate. For example, your app almost certainly performs queries that have a where clause. That means that the R step has logic, which means you can (and I know I do) get the logic wrong. Also, there's usually complicated rules during C, where a user is only permitted to actually create data in certain cases, and when data is created other data needs to be updated to remain accurate, etc. So yea, of course data is Created and Read, in that every application maintains data. But that doesn't mean that it's a CRUD app. If you have any where clauses or if expressions, it's not a CRUD app. Finally, the most popular way of storing data is the relational model. Which, is just math applied to the notion of data: https://www.seas.upenn.edu/~zives/03f/cis550/codd.pdf https://www.seas.upenn.edu/~zives/03f/cis550/codd.pdf

amw-zero 6y ago

I'm with you. I feel like all actual business problems have a lot of applicable CS solutions. Maybe if you're building UI-heavy stuff, there's less theory required there?