Ok, I just think this is funny. We’ve had a new version of our software out for a couple of months. It is version 5.2.3. The problem is that on the "about box" it says 5.2.1 (stupid bug not caught in internal testing) and I’ve been getting more and more calls of people confused why their “upgrade is failing”. So far I’ve been resisting issuing a patch where the only change is “changed version number”, but I don’t know . . .

If it's good enough for the Lord, it's good enough for you.

Then put as an explination: "Fixed problem with upgrade where installation would not succeed".

It will make you look better (you did more coding) and it will give your users an excuse to upgrade!

I only wish I could get away with distributing as many patches as he does!

Release early, release often. What really bites is that such an overwhelming percentage of Mark's patches;
a) Implement a new feature.
b) Work as intended.
c) (first time)

I wouldn't be shocked if Versioning bug-fix patches outnumbered other bug-fix patches.