On Thu, Jan 02, 2020 at 11:19:30AM +0100, Ingo Schwarze wrote: > Hi Kurt, > > Kurt Mosiejczuk wrote on Wed, Jan 01, 2020 at 08:21:04PM -0500: > > > cp(1) uses -R for recursive copy. scp(1) uses -r. This diff adds -R > > as an alias for -r to scp(1) for those assuming consistency with cp(1). > > even if cp -R and scp -r did the same thing - which, if i understand > tedu@ correctly, is not the case - i would still dislike introducing > an alias. > > Aliases make documentation longer for no benefit and sometimes > also cause confusion. They also block namespace for no benefit > (or rather, option space in this case). > > While it is nice when different commands can use the same options > for similar purposes, that is not always possible. And even when > it would have been possible when designing the tools in the past, > if a different decision was made back then, remaining consistent > with previous versions of the same command is usually worth more > than symmetry with a different command. > > Sometimes, aliases are needed because different other operating > systems use different names and we want to be compatible with both. > But whenever they can be avoided, they should. > > When working on documentation, i'm actively trying to deprecate > aliases where that is possible. > > Yours, > Ingo
Contrary point: I'm with Kurt on that one. Very often, I do mix up -r and -R, and I do hate it. Especially since our cp *does* have a -r which is subtly different. Not having -R in scp means you can do all kinds of fuckups. Once in three times, I type scp -R and go "oh fuck" when it doesn't work. And if I use scp enough, I'm also likely to use cp -r by mistake. Are we likely to actually remove cp -r so the second mistake doesn't happen ?
