>Ken wrote: > >> Weeeelll .... the whole reason this started was because people aren't >> reading the documentation :-) > >Documentation is not the whole reason. The fact that post(8) behaves >differently from nearly all other MH/nmh programs requires an exception >when thinking about it. (SUBTLE!) That is validated by the fact that >three of us looked at the user's profile excerpt and two missed the use >of a post: profile component. (And I was one of those two and am very >well aware of this exception.) And that the documentation itself had >missed this until late 2011. > >Is there any reason that we should NOT provide a warning to help users? >Especially first-time users who haven't read all 19K lines of man pages? >Aren't we trying to minimize configuration impediments?
I've thought about this a bit. Here's the best answer I can give. If you have a profile entry under "post", it really isn't for the program post(8); they're for a program who's argv[0] is "post". I know that seems like nitpicking, and maybe it is ... but I think it's an important point; it's a commonly-used idiom to create a different set of options to commands by symlinking your custom name to the original nmh utility and putting the options under your custom name. I'm fine with a warning for a misconfiguration, but ... given the way the profile entries work, I'm not sure it IS a misconfiguration. Now, you're going to raise the entirely-reasonable point that "hey, 9 times out of 10 it's not going to do anything when the user thinks it will, and that is a misconfiguration." I get where you're coming from. It's just ... we have a situation where a program that is NOT misconfigured is warning about ANOTHER program that may or may not be misconfigured. This is of course unavoidable if you want to generate a warning because all nmh programs are independent of each other, and the program you really want to generate a warning can't, because it doesn't read the profile. But ... if you're not convinced, how about a compromise? send(1) is the normal front-end to post(8); how about have it check and issue a warning, rather than all programs? --Ken -- nmh-workers https://lists.nongnu.org/mailman/listinfo/nmh-workers