Conversation

Notices

1. Embed this notice
   John Regehr (regehr@mastodon.social)'s status on Monday, 05-Feb-2024 00:32:28 JST John Regehr

   @jrose I have feelings about this! on one hand of course you're right this should be how it is. but then again, exceptional values like NaN and nullptr are just an absolute disaster, nobody expects the spanish inquisition or the exceptional value

   • Embed this notice
     Paul Cantrell (inthehands@hachyderm.io)'s status on Monday, 05-Feb-2024 00:32:27 JST Paul Cantrell

     in reply to

     • Steve Canon

     @steve @regehr @jrose
     (Except that none == none but NaN ≠ NaN, but that’s a disaster no matter how you treat it)

   • Embed this notice
     Steve Canon (steve@discuss.systems)'s status on Monday, 05-Feb-2024 00:32:28 JST Steve Canon

     in reply to

     @regehr @jrose how much of this is due to languages failing to expose the HW semantics in useful fashion, though? If we bound NaN to none in an optional, all of a sudden no one would be confused by it at all.

   • Embed this notice
     Paul Cantrell (inthehands@hachyderm.io)'s status on Monday, 05-Feb-2024 01:32:16 JST Paul Cantrell

     in reply to

     • Steve Canon

     @steve @regehr @jrose
     They lied to me, everything has been a lie

   • Embed this notice
     Steve Canon (steve@discuss.systems)'s status on Monday, 05-Feb-2024 01:32:17 JST Steve Canon

     in reply to

     • Paul Cantrell

     @inthehands @regehr @jrose you can make NaN == NaN so long as you also have an IEEE compare by some other name.