On Wed, Dec 03, 2014 at 20:36, Jason McIntyre wrote: > > i think in the past some developers expressed a desire to not demote > -and and -or, as they are traditional bsd. i think we changed the > examples as a desire to provide portable examples, but still wanted to > keep the idea that -and and -or are perfectly acceptable. > > that's my understanding, and the reason why i made those changes in that > way. > > if i have it wrong, or there is a sudden swing to pooh pooh -and and > -or, i guess we could change it a bit. still, i dislike the idea of > documenting that we support stuff but try to obfuscate it. it just makes > it harder for the reader (and therefore still think my diff was better ;)
I think when the BSD way is superior in some way, it makes sense to elevate it. For -and vs -a, it doesn't seem like -and is obviously superior. For madvise, we document the posix options as an alternative, which I think makes sense. They're longer and uglier. I am at first glance unsure if the two -a -and operators are equal or different when they are both documented together. Could they be synonyms or two variations of similar operators? I like the idea that one spelling is documented and then the second introduced as an alternative. If you think my diff is too wordy, we could move the text around a bit, perhaps later in the section and say "-and and -or are alternatives to -a and -o" or such.
