Páll Haraldsson

Quaternion networks still interest me, I mean since two days ago when I first heard of them. :) I'm not sure you read my thread. but if numbers of parameters...

Looking at the code: https://github.com/JeffreySarnoff/ArbNumerics.jl/blob/d34c8d180ff303668c45e932c81100cae15fb75a/src/libarb/ArbComplex.jl#L73 while you allow base-10 it seems to be only to calculate precision (number of fractional bits you store), i.e. 0.1 is still not exact, nor...

>small benchmarks See me addition above. Yes, I figured the actual number of allocations are not a big deal, still surprising that the amount of bytes doesn't matter much. [Most...

The D language has `@nogc` for functions (and those functions can only call such functions). I've thought such might be useful for Julia (should be able to do in a...

Yes, presumably, I see you had a fix: https://github.com/JuliaLang/Downloads.jl/pull/190 1000 seems excessive, not sure I hit that mark, was this backported to 1.7.3? I found, that issue may or may...

I'm not on Windows, I'm on Linux and your excellent (precompiling) timing is strangely at odds with mine. It seems you didn't exit before the second timing though (the allocations...

>Well, I think I can re-add the support to Julia 1.0 in PrettyTables I don't personally need older than Julia 1.6 for any package, but if it's easy(er) to support...

For some uses PythonCall.jl [EDIT: renamed from Python.jl] may be preferable? I'm not yet clear about the pros and cons, or why that new package needed, but at least it...

FYI, fix here: https://www.dropboxforum.com/t5/Dropbox-installs-integrations/How-do-I-fix-the-Qt-platform-plugin-quot-windows-quot-error/td-p/91545 It's Windows/[ana]Conda related it seems, works for me on LInux.

>The C API is not considered part of the stable guarantees, only the Julia API That's surprising, I don't see that documented on its page only: "Julia has a C...