Friday, March 30, 2007

Some Calamitous Error Messages - Redux


I've bemoaned the sorry state of "modern" software before (as have many others at various "this product sucks" sites).

My pet peeve is silly, meaningless, inaccurate, incomplete, misleading, thoughtless, and just plain stupid error messages.

I strive to make my own code display error messages that are as meaningful and helpful as possible, but it seems that not everybody does --and this is the year 2007! What hope is there for mankind?

Aunt Calamity has her own ideas over at the "ghost in the machine" web site: http://www.auntcalamity.com/ghost.htm Here are illustrations of a mere a few that I've encountered just in the last few days. They come from a variety of software products (the vendors all do it).








  • Top image (Techsmith Snag It v8.2.2) -- if there there are 0 seconds left, are we going into some sort of hyper-time?
  • IBM Lotus Notes v7.0.2 -- What on earth does this mean? (Maybe there's a connection between the digits in the version number and in the error message body "7" and "0" and "2" ?)
  • Qualcomm Eudora (actually this is a deficiency in Windows dialog box architecture, it's not really Qualcomm's fault ) -- Apart from the gobbledygook message text, why can't the "OK button be labeled "Cancel" and the "Cancel" button be labeled "Debug"?)
  • NTI Shadow 3 -- which of several concurrent jobs is the one that has just been completed? (I've asked them to add the Job Name to each message, but they haven't done it.)
Device drivers, and the code that calls them, are consistently bad offenders. Here's one that I came across at THE DAILY WTF recently. "Very Strange: Attempt to disowned by driver after WIN32 Checks out. Disown attempt ignored."

Very strange, indeed!

I could go on and on, but won't bother, and I'm sure you could add your own beauties to the collection!

No comments:

Post a Comment