What are you working on, /g/?Previous: >>107060072
>>107088689Fortune’s algorithm implementation, then planning to implement an eikonal solver on an irregular triangular mesh and see how shitty it turns out.
>>107088689>not C+=ngmi
>>107088689New month, new project!This one's just a quickie to scrape a stream of presence data off a bunch of healthcare devices and pitch it into a DB for the customer to analyze later.
>not --Cngmi
>>107088734no C& do
What the fuck.I tried making a quad tree replacement using a simple grid bucketing. It was intended for small objects but I came up with an idea how to deduplicate query results of large objects spanning many cells.So I accidentally querying that easily beats quad tree and is barely faster than rtree even in cases that look like they should be pessimistic, at the cost of way slower removal of single elements and not being sparse.Is this actually a thing or did I make something new?
>>107088689i'm working on a bootstrapping compiler for my programming language. I chose C# to implement the bootstrap compiler because I didn't want to deal with memory management for what is essentially throw-away code. The whole project has turned into overly verbose object oriented boilerplate (from when I was trying to follow the OOP dogma) and spaghetti code (from when I just said "fuck it" and made every class member public and started poking at it directly), but I'm in too deep to start over.I hate the object oriented cult.
>>107089947Quite often the best way of doing things is directly, and then adding indirections if it is deemed nessecary.With that said I believe the best way to represent things that cannot be represented with a borrow checker would be to write that part in C. Everyone has a C compiler.
