[01:21:17] <wm-bb>	 <u99of9> My messages were unclear, as I was flip-flopping while writing. I should have struck the earlier comments, sorry.  Anyway, my first impression is that the JS has worked, but the python isn't quite right. I'll keep investigating. (re @vrandecic: Done! I misunderstood one direction was sufficient)
[05:38:06] <wm-bb>	 <vrandecic> Thank you! Shall I revert the python one or didn't it make the situation worse?
[07:54:29] <wm-bb>	 <u99of9> Yes, revert the python one please. I'll look into it properly later. (re @vrandecic: Thank you! Shall I revert the python one or didn't it make the situation worse?)
[11:38:51] <wm-bb>	 <u99of9> Interestingly RustPython fixed their subnormals and max floats within 24 hours of our fix @vrandecic ! https://github.com/RustPython/RustPython/pull/6216 (re @u99of9: https://github.com/RustPython/RustPython/issues/5471)
[11:46:04] <wm-bb>	 <Al> “Interestingly”? No coincidence, methinks (but I haven’t looked at anyone’s code) 🤔 (re @u99of9: Interestingly RustPython fixed their subnormals and max floats within 24 hours of our fix @vrandecic ! https://github.com/RustPy...)
[12:01:27] <wm-bb>	 <u99of9> Haha. Well if it helps, I can hardly read his code: https://github.com/RustPython/RustPython/pull/6216/commits/4c88621305d976ec6675f560b32a41dfe8a57bef (re @Al: “Interestingly”? No coincidence, methinks (but I haven’t looked at anyone’s code) 🤔)
[12:16:02] <wm-bb>	 <Al> 🤷‍♂️ “The approach of scaling subnormals by 2^54, decomposing, and adjusting the exponent implements exactly what was needed”, apparently. (re @u99of9: Haha. Well if it helps, I can hardly read his code: https://github.com/RustPython/RustPython/pull/6216/commits/4c88621305d976ec6...)