Re: fnord bugs I found + new httpd (was: A fnord fork (a fnork?))

Well that's an interesting wrinkle.

I found #7 trying httperf, which depends on the 1.1 default keepalive. 
But every other 1.1 client I tried sends a keepalive header, so in
practice (ie. not using a benchmark program) it probably doesn't matter
much.

#3 and #7 are surely not major bugs, and I will even begrudgingly accept
an assertion that they are not bugs at all.  But I suggest that since
fnord is not returning a 505 response to HTTP/1.1 requests, it should
contain the one extra line of code to turn on keepalive for 1.1 requests
too.

However these are surely not as important as, say, #5.