rsync 2.5.4 (probably 2.5.5 too) server handles SIGPIPE very poorly

[Jonathan Kamens]

Earlier today, I wrote:

>  I don't agree that this is a better solution, especially given your
>  comment in the last sentence above.

Please allow me to recant my previous testimony :-).

In the worst-case scenario, your solution causes one extra level of
recursion into rwrite, whereas with my solution, if someone calls
rwrite to attempt to write to a hosed output socket when they don't
know it's hosed, they may sit around forever waiting for the other end
to respond to a message that never got sent.

I will test your solution, but it looks fine to me at first glance.
You can assume that I didn't find any problems if I don't report any
today.

Thanks,

  jik