[racket-dev] Racket 5.3 release branch doesn't build

From: Eli Barzilay (eli at barzilay.org)
Date: Thu Jul 19 10:22:20 EDT 2012

Two hours ago, Robby Findler wrote:
> I looked at this a little bit and it appears to be a bug in
> version/utils, specifically the version->integer function is not
> working correctly for the release version number.

Sorry, I forgot to push a version fix on the release branch -- done
now.  But note that there are some other issues with the release
branch at the moment.


As for `version->integer' it's working as it should -- returning #f on
an invalid version instead of throwing an error.  The problem is that
the previous version on the release branch was broken: it was
"5.2.900.0", and the version specs for X.Y.Z.W versions is W != 0.

Maybe `version->integer' should have thrown an error instead, but it
seems risky to change that behavior now.  On the other hand, it seems
reasonable to assume in code that (version->integer (version)) never
returns #f, since the currently running racket is assumed to be fine.
So I don't think that anything should be done (at least on the short
term).

There is a safety check though in the build script that verifies that
the version is valid which should detect such problems early, but the
release branch build failed for a different reason, and failed earlier
than that check.  (Just in case, I verified now that it would have
killed the build with an error.)

-- 
          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
                    http://barzilay.org/                   Maze is Life!

Posted on the dev mailing list.