Sure, decisions were made on the battlefield that at the time and considering the circumstances were optimal (for local maxima.) This kind of glib remark doesn't really prevent anyone else from making the same mistakes.

I don't know what the correct approach should have been, but it seems that changes to support other architectures should have been made independently from other fixes and tagged such that they could have been backed out rather than contaminating your source and pushing it to gold. (Remember this is before the days of Windows Update!)

> This kind of glib remark doesn't really prevent anyone else from making the same mistakes.

Where's the mistake? Wouldn't backing out the workarounds introduce new potential for error due to other changes made after the workarounds were put in? Why do the workarounds need to be backed out? They don't break anything. If it ain't broke...

My point is that it wasn't really a mistake; it was a change in circumstance.