From: lcargill99@comcast.com   
      
   I have an embedded ( Tiny ) Linux platform that will, under conditions   
   not fully understood get error 32 on these two calls. We need to know   
   exactly what this means.   
      
   It also gets code-104 errors, but I believe those are 100% "the other   
   side hung up".   
      
   We can remove one particular communications component and these go away.   
      
   We are not *nominally* out of bandwidth. Budget for total bandwidth is   
   around .25 to .5 MBit/sec ( actual measured is .25/5 is a guess for   
   worst case)  over TXBase100 and 802.11B.   
      
   I believe our 802.11B solution is failing us. It's "downshifting" to 1   
   MBit and we're outta luck.   
      
   My story now is "an error 32 is a complaint about resources; the stack   
   has filled up a buffer and is giving up." This finds a "hole" in TCP   
   that results in the error code.   
      
   The team believes that we should never see these. I "refute them   
   thusly."   
      
   Thoughts?   
      
   --   
   Les Cargill   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)