On Fri, Oct 30, 2015 at 05:54:35PM -0700, Russ Allbery wrote: > Axel Beckert <a...@debian.org> writes: > > > I'd restrict it to packaging (i.e. the debian/ directory) and explicitly > > exclude the file debian/copyright as any appearance of the word "legal" > > or "illegal" in there is very likely a valid (sic!) usage. > > Indeed, encouraging people to get into arguments with upstream about this > seems like a bad idea. The wording choice portions of the GNU coding > standards are highly controversial, to say the least. >
I see no need to argue with maintainers upstream. If somebody wants a clarification, it suffices to ask them to explain their use in a way others can understand. And facts should not be ignored by them in their answers. I can't see why the wording of using invalid/valid versus illegal/legal is (highly) controversial. The only explanation I can find is a) The people do not understand (comprehend) the issue. From "The GNU coding standards", last updated April 23, 2015: "Please do not use the term "illegal" to refer to erroneous input to a computer program. Please use "invalid" for this, and reserve the term "illegal" for activities prohibited by law." b) They use themselves the wording illegal/legal (and do not really know (understand) why), continue to use it (and do not really know (understand) why), and do not state this from the beginning and in all their contributions thereafter. What would a teacher of the English language, writing, or literature say about this? What do you (plural) expect, want him to do (say)? Is such use semantically correct? > I'm worried about false positives from a rule like this. "Illegal" is a > valid English word that frequently is the word intended. (For example, > consider intrusion detection systems designed to look for illegal activity > from one's local system.) The GNU usage recommendation is in a specific > context around actions in software, and Lintian is not bright enough to > read for context. > False positives (negatives) are a common (general) side effect of decisions, so it also includes "lintian" (it decides (tests), does neither advise nor not advise). What (specific) harm do you expect them to do? What harm have other false positives (negatives) from "lintian" done? Are they still doing it? Are there no countermeasures? ### When programmers began to use illegal/legal, it was not a decision, but was (is) distributed through SEEING IS BELIEVING which is an automatic behaviour, when people have not learned otherwise. (Are children biologically programmed not to believe what they see (notice with their sense organs?) NB. If it was a decision to use illegal/legal, how was it done? Were there alternatives, conditions to consider? What criteria were used? -- Bjarni I. Gislason
