3 ms·
Ew. If you want to say something, say it. "I don't like that my contributions aren't accepted, so I'm forking the codebase". That's going to generate some disc
by outworlder 4y ago
Ew.
If you want to say something, say it. "I don't like that my contributions aren't accepted, so I'm forking the codebase". That's going to generate some discussion, but it's fine. Public domain and all.
However,
"look what happened to qemu, same thing will happen to sqlite (and I'm contributing to the problem and forking it)". One can't say "no contributions led to fragmentation" while _at the same time_ contributing to fragmentation by making a hard fork!
Also,
> "However, edge computing also means that your code will be running in many geographical locations, as close as possible to the user for the better possible latency. Which means that the data that hits a single SQLite instance now needs to be replicated to all the others."
Then replicate away. Leave that stuff out of SQLite. If that's a really important use-case, go use couchdb or something similar.
- chasil 4y agoIt's wonderful that there is a fork of SQLite. It's good to see new ideas. Is Airbus going to use it in the A350? No. Why not? It's not compliant with DO-178B, because it has not been confirmed correct with "Test Harness 3" (TH3).
- dathinab 4y agopeople are not annoyed by anyone forking SQLite they are annoyed by people misrepresenting facts in a subtle manipulative way to make their fork and the reasons around it look like something it isn't what the article intentionally or unintentionally conveys in its tone and formulation is: "sqlite is badly outdate beyond saving and needs to be replaced" what actually is the case: "we want something like SQLite but with some core changes incompatible with SQLites core principles which should be API compatible enough to allow reusing a lot of DB tooling".