Gerd Zellweger
Gerd Zellweger
Hi @alexandruradovici I don't mind relicensing the x86 crate as Apache 2 and MIT, I don't have strong views about it. But I'm also not exactly an expert on copyright...
When this is merged all the way, lets try to configure our CI instances with this feature as a "real-world" test
Maybe it could be removed only the next time the pipeline is started (and the binary has changed)? But I leave it up to you, feel free to close as...
Thanks, looking into PGO seems like a good idea.
I think it's more user friendly to have aliases which indicate bits like int8, int32, int64 ... I personally can't remember what is INTEGER or BIGINT, but I do know...
what dialect do we follow?
I think this is out of scope of the issue.. the issue is about adding `int32` which I don't think has a conflict with `int8`
>especially when the processes upon failure get auto-restarted this needs to be documented. I think it's fine once we have restart functionality that the log stream does not end. But...
that would be nice if we can run it in CI, thanks
note that we already can mock the cpuid instruction in tests and the library by swapping it out with a custom fn. What do you think SDE can do here...