From: Pancho.Jones@protonmail.com   
      
   On 1/31/26 03:00, c186282 wrote:   
   >   
   > 'write python "pre-threaded" server client program'   
   >   
   > It wrote something quick enough, except it was a   
   > plain simple threaded server, not PRE-threaded.   
   >   
   > Pre-threaded creates/manages a 'bank' of threads at   
   > the start so they don't have to be created one at   
   > a time on demand - quicker, it's the highest-   
   > performing kind of server.   
   >   
   > 'write python "pre-threaded" "non-blocking" server client program'   
   >   
      
   As a non AI, I wouldn't have known what pre-threaded meant. I would have   
   thought what you describe was called using a thread pool. The AI may   
   have decided the overall performance hit of creating new threads was low   
   and not worth the complexity of a thread pool, or it may be the   
   threading library/OS is effectively doing a thread pool behind the scenes.   
      
   My naive solution would be to avoid threading, use IO with callbacks, or   
   use a handful of long-lived threads, like the "worker thread"/"gui   
   thread" pattern.   
      
   > It created a non-blocking server, but NOT threaded or pre-threaded.   
   >   
   > Seems AI doesn't always "get it"  :-)   
   >   
      
   Perhaps you need to have more of a conversation with the AI.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)