Proposal for discussion - remove all Windows-port cruft

Eric S. Raymond esr at thyrsus.com
Sat May 14 11:26:38 UTC 2016


Mark Atwood <fallenpegasus at gmail.com>:
> it sounds like there is no cruft getting in the way of complexity
> headaround or reduction.  leave it be.

Unfortunately, your premise is not correct; Hal's report was
incomplete.  There are substantial amounts of Windows cruft in some of
the trickiest places outside the port directories, notably the
worker-thread code for async DNS lookup and the network-plumbing
hairball.

The reason I am pushing on this now is that I'm still casting around
for ways to simplify the hairball to the point where I can really
grasp it. Complexity headaround is the exact issue.

Removing the Windows cruft won't completely solve the problem, but any
complexity reductions we can get are good.  Enough of them might get us
to where my head doesn't hurt when I look at that code.
-- 
		<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>


More information about the devel mailing list