On Sun, 2007-08-12 at 23:48 -0500, Tony Abernethy wrote: > Darryl Dixon wrote: > > > > Hi All, > > > > I've browsed the history of the list, but can't seem to find > > an answer to something that I find quite surprising - why > > isn't --numeric-ids the default when rsync is told to > > preserve permissions? It seems to me that the current > > behaviour runs against the grain of more than a quarter > > century of unix tradition, and certainly caught me by > > surprise the first time I encountered it. Can anyone enlighten me? > > > > regards, > > Darryl Dixon > > Winterhouse Consulting Ltd > > http://www.winterhouseconsulting.com > > If you have (mostly) the same set of users (names) > on two different systems, > what are the odds that the numeric ids are identical? >
Hi Tony, I'm sorry, I don't understand the gist of what you're saying: the case that you imply seems like an unlikely situation to cater for as the default: what are the odds, if one doesn't control the UIDs, that one may also not control the exact construction of the username, and hence the files get mapped to some other individual with a similar name that happens to end up with a username on the second system the same as yours on the first? As I mentioned, I found it extraordinarily surprising that a tool dedicated to maintaining such precise control over the state and content of a collection of files should default to doing string comparisons between usernames and mapping them back to UIDs instead of simply taking the UIDs as they are presented on the filesystem and transferring them intact to another. I don't argue the desirability of being able to do username mapping between systems, but it strikes me as truly extraordinary that it is the default - I'm just trying to understand why this is the case? regards, D
signature.asc
Description: This is a digitally signed message part
-- To unsubscribe or change options: https://lists.samba.org/mailman/listinfo/rsync Before posting, read: http://www.catb.org/~esr/faqs/smart-questions.html
