FreeBSD Port: qmail-1.03_3

[Michael Sierchio]

Joe Marcus Clarke wrote:

>>-     while ((k > i) && (cmds.s[k - 1] == ' ') || (cmds.s[k - 1] == '\t'))
>>+     while ((k > i) && ((cmds.s[k - 1] == ' ') || (cmds.s[k - 1] == '\t')))


> Actually, if k is less than or equal to i and cmds.s[k-1] is a tab, then
> the first bit of code evaluates to true while the second evaluates to
> false.  I haven't looked at the rest of the code, but those statements
> are semantically different, and I think the second is more correct.

You're simply mistaken.

( 1 && 0 || 1 ) will ALWAYS evaluate to 1 on any ANSI C compiler.

There was no bug corrected by this mess:

http://www.freebsd.org/cgi/cvsweb.cgi/ports/mail/qmail/files/patch-qmail-local.c?rev=1.1&content-type=text/x-cvsweb-markup
ports/mail/qmail/files/patch-qmail-local.c - view - 1.1

only a perceived bug because of a lack of understanding of how
operators are evaluated.  The author should have been contacted
if there were indeed a bug.  I assure you, there is not.

4) Add qmail's sendmail emulation fix: "David Phillips noticed that
    sendmail's '-f' option sets a default From: header, and so should
    qmail's emulation."
	- Add distinfo checksum for the patch

This is also bogus, IMO -- feature addition to software which has been
well-audited and which makes security claims is simply a bad idea.