Python 1.6 alpha 1 released

[James Logajan]

Fredrik Lundh wrote:
> 
> John Aycock wrote:
> > Although I don't have time to volunteer, it wouldn't be too difficult
> > to write a program which would look through existing Python code and
> > warn of potential uses of these changed constructs.
> 
> after reading your post, tim peters used barry's new "post
> via the time machine" facility:
> 
> http://www.deja.com/=dnc/getdoc.xp?AN=591243549
> 
> changing it to look for "connect" should be trivial (he's probably
> done that by now).
> 
> fwiw, while it may miss many real appends (see comments in
> script), I'm pretty the number of missed connect calls are very
> close to zero.
> 
> </F>

And just when you think you've nailed all the connect()s in your client code
(which one must admit will in general probably occur in only one or two
locations for most network applications) and hope you've blessed all
appends, you find your server code is throwing exceptions in the bind()
calls....

P.S. Still waiting for a new release of the Unix standard so that the
spelling of the creat() system call can be corrected. <wink>