On Mon, Dec 17, 2007 at 09:16:21AM +0100, Abdelrazak Younes wrote: > Stephan Witt wrote: > > Abdelrazak Younes schrieb: > >> Enrico Forestieri wrote: > >>> On Fri, Dec 14, 2007 at 03:32:40PM -0000, > >>> [EMAIL PROTECTED] wrote: > >>> > >>>> Author: younes > >>>> Date: Fri Dec 14 16:32:37 2007 > >>>> New Revision: 22148 > >>>> > >>>> URL: http://www.lyx.org/trac/changeset/22148 > >>>> Log: > >>>> The original support::copy() was overwriting the target file by > >>>> default. This commit put the code in line with pre-rev 22119 and 22129. > >>> > >>> I think that you should remove the second parameter in copyTo() and > >>> always overwrite the copied to file. Otherwise the changed semantic > >>> may lead to other obscure bugs. > >> > >> Well, the copyTo() method is not used a lot ( times) and the system > >> level copy does not overwrite by default nor the Qt one; so why should > >> we? I prefer to stay with a wider spread semantics. > > > > What system level do you mean? > > > > The Unix command cp (copy) overwrites the target file. It asks politely > > if you add the command line option -i. So, the logic is reversed to your > > system.
I think that Abdel was referring to the Qt behaviour. > Really? I thought you needed the '-f' option in order to do this. At > least that was the case on slackware. Maybe the command was just mapped > to 'cp -i', dunno. Too many years on Windows I guess :-) I think so, given that you don't even remember that commands are aliased rather than mapped ;-) > I stand then corrected (and I already modified the code accordingly > following Enrico recommendation). I saw that. Thanks. -- Enrico
