>I'm trying to compile nmh-1.0.2 under Cygwin 1.0, and having problems.
>
>The first is that in zotnet/tws/dtimep.c, the definition of timezone
>(long) conflicts with the one in cygwin (time_t).
>Next, there is a sed script that mucks with the output of lex, and the
>sed script fails on the flex output! I don't know what the script is
>trying to do to the lex output (okay, I can see that what text it's
>adding, but there aren't any comments about why the changes are being
>made).
>
>Even though configure finds flex, the tws makefile tries to use lex and
>fails.
I have a problem under RedHat Linux similar to your second. On my
system flex appears to work properly, but the sed script eats all of
its output.
In the distant past, I modified dtimep.lex to work under flex; but I
appear to have deleted my work. If you can find out what the sed script
is for, please write me -- I'm thinking seriously of rehashing dtimep.lex
(again) so that it is flex compatible. See my posting of yesterday.
Jonathan Ryshpan <[EMAIL PROTECTED]>
Those who have put out the eyes of the people
reproach them for their blindness. -- Milton
