Currently, the tag error has been used mainly for two purposes on CV:
- Statistical measure of deviation;
- Notification messages shown when statistics-related software fails.
We have a residuals tag. According to Wikipedia: error (observed - true function value) and residual (observed - estimated function value) are related but not same.
We have the tag deviation, but it is a synonym to standard-deviation. We also have bias.
Is it worth it to fix this? If yes, how? Please, present a proposal considering two actions:
- looking forward (i.e. to stop error tag being used for different purposes from now and on);
- to fix this looking backwards (i.e., retag in a proper manner existing error questions).
Edit 1.
Looking forward the action which will result from this question, it was mapped (05th, July, 2013) among the 125 error questions, 8 relating to software error messages.
They are (from newest to oldest):
Status-completed (Edit 2).
As far as I could go, I asked to CV's moderators to create the error-message tag as suggested by @A.M. (and with no objection by @gung). @chl gently, helped me in this step;
As suggested by @gung in #1 item, the tag wiki excerpts were edited (with peer review) for error and error-message to avoid confusion in future questions;
From the eight questions above, five of them had the tag error replaced by error-message as suggested in both answers (but not at once, as suggested on @gung's answer, item #3). Two of them (2 and 7) crashed in the meantime, and one I overlooked about the tag usage (5).