Peter.Memishian at Sun.COM wrote: > > But for the dladm subcommands, <link> is always last (if present at > > all), so this seems to be consistent and easy to remember/deal with. > > My zfs get example seems to demonstrate that this may work well. > > I worry a bit about the multiple-property case -- e.g.: > > $ dladm show-prop power-mode,radio ath0 > > ... feels more comfortable to me *with* the leading "-p": > > $ dladm show-prop -p power-mode,radio ath0 > > However, I do agree with your overall point. Does anyone else have any > strong opinions on this topic? If so, please speak up :-)
To me, requiring the -p allows the syntax to more easily evolve in the future. There's no ambiguity about what properties (if any) are contained on the command-line. It allows for a future syntax where no properties are given on the command-line without ambiguity about whether a link-name is a property, or vice versa, where the lack of a link-name leaves no ambiguity about whether a given property could be a link name. Besides, CLIP is fairly clear about this (see section 5.1 of the CLIP companion PSARC/2006/062). My apologies to those external to Sun, I don't know of a publicly available version of the CLIP companion document. For those inside Sun, see: http://sac.sfbay.sun.com/PSARC/2006/062/clip-companion-latest/clip.html -Seb
