I am working with Perforce for 10 years. Our repository has around 500.000 submits and our submits are either C/C++ source code or (binary) dependencies.
The lack of editor support for Perforce was always a downer. So recently I finally started evaluating Git and I enjoy the concept. I already see it's not the shiny world, but good enough to stay tuned.
So my biggest concern are dependencies. Over the entire depot lifetime we have around 200 GB of binary dependencies (new revisions, versions, etc...). The first solution is Git LFS. I made a few tests and instead of placing them in .git/objects
they ended up in ./git/lfs
as simple copies. My question now is, why do people still complain about Git LFS? Is that solution not good? What is the problem with Git LFS that this doesn't solve the problem git is accused of?
git lfs
not being a real solution or being half baked. Anyway, I will update my question