The Trouble with Changesets

Strange Strands is now public, by virtue of having removed /strands/ from inamidst's robots.txt file. Sadly, I messed it up slightly: the Changes and Updates page for 2006-03-23 shows me having removed Strands before I then published it. I had to do it that way because the way the chances script works is to look for any i) modifications, ii) additions, or iii) deletions. Problem is, Strands already existed—it was merely robots.txt protected; so if I'd've just removed the robots.txt protection then the pages wouldn't've shown up in the changes.

Now, this would have worked fine had I not removed /strands/ from the robots.txt file at the same time as having moved the folder; but that's such a significantly stupid process to have to go through that sometimes I mess it up. Unfortunately, I'm not sure if there's a decent solution for it. I could've got rid of strands in the reference copy, I suppose, but I'd rather not fiddle with the backups. Or I could go back and fiddle with the changesets, but that would be changing history. If there were some way of flagging it as an extra thing to be noticed in the diff as an addition, that would be great, but it would be a significant hack. Perhaps it's the robots.txt protected development directory principle which is flawed?

Though I'm calling this post "The Trouble with Changesets", using a capitalisation style meaning that all words are capitalised except for internal articles, prepositions and conjunctions, I actually prefer to capitalise everything in the title. It's so rare nowadays that it looks almost wrong, but I prefer the regularity of it. Admittedly sometimes it makes some minor words look intrusive, but I find that this occurs not as often as one would expect, and probably not even the majority of the time.