SHA256 checksums

[Jeremy Lavergne]

> I expressed my views in this post:
> <http://lists.macosforge.org/pipermail/macports-dev/2011-March/014108.html>

I'm not certain where the "overall length" issue came up originally. I'd
agree that's not a problem. The length issue I saw was wrapping Portfiles
after a certain length, such as avoiding a really long description line.
To that end, having a shorter checksum hash would be handy.

Responding to "having two possible formats seems more inconvenient than a
long line": the only inconvenience is a check to see if the length of the
expected SHA256 checksum is different than the expected hex output, or
ends in an invalid character (=) as recommended by the RFC.