Hello, Michael Hekeler wrote on Thu, Apr 10, 2025 at 11:00:35AM +0200:
> Some times ago if you required to have a unique UUCP node name for your > computer it was defined in some way as `myname`. > I think that was the reason for OpenBSD's "/etc/myname" filename I'm not absolutely sure that is true - but the fact that the first time that "myname" appeared in BSD code in the sense of a host name was as a variable name inside the C code of some UUCP-related program in 2BSD in 1979 at least makes your statement somewhat plausible - even though it then took 13 years before usage of "myname" spread from a C vaiable name to a shell variable name in 386BSD 0.1 (1992). Either way, even though it is sometimes possible to argue that a name is objectively better than another because it is more descriptive and better expresses the purpose, the art of naming is hard and non-trivial. Languages develop all the time throughout history and words gain new meanings over time (semantic shift) or theit meanings become narrower (semantic specialization, pars pro toto). It raely makes sense to claim that a name is absolutely better or absolutely worse. So, can we please refrain from renaming stuff that is part of a public API, and hence cause disruption for lots of people, unless there is a very compalling reason, for example a generally accepted standard that requires a renaming? > BTW This is a mailinglist. So please please send your mail to the list > only. Please do NOT copy emailadresses in Cc: and/or To: headers. > No reason to send multiple duplicates to multiple people around the > globe. It is true that this is considered good nettiquette on *some* mailing lists, but it does *not* apply to OpenBSD mailing lists. Here, it is not considered rude to leave developers who have contributed to a thread in the Cc: (even though it is not strictly required either). For people receiving very large amounts of mail (like myself), it can lessen the risk of overlooking messages mentioning me. Yours, Ingo
