Re: [boost] Review Request: Filesystem Library

From: Mattias Flodin (flodin_at_[hidden])
Date: 2002-08-01 10:44:13

On Thu, Aug 01, 2002 at 04:25:37PM +0100, Keith Burton wrote:
> 2) in the documented spirit of the filesystem library , i.e. as a simple
> to use a scripting language , then remove()
> not requiring the path to exist is the correct ( i.e. least surprise )
> definition.  Thus avoiding the need to document an inevitable problem.
> However it looks as though we will have to agree to differ on this.

Perhaps, but I will try it from one more angle first: As an end user,
when I type "rm <filename>" or select a file and drag it to the trash
can / recycle bin, I want the program to tell me if something went
wrong, e.g. if the network connection was lost. I want to feel assured
that the file is really gone. Any other behavior could even be a
security risk.

Wouldn't you agree? And with the current API, I can't really make this
assertion to the end user.

/Mattias

-- 
Mattias Flodin <flodin_at_[hidden]>  -  http://www.cs.umu.se/~flodin/
Room NADV 102
Department of Computing Science
Umeå University
S-901 87 Umeå, Sweden
--
"If nothing ever sticks to teflon,
how do they make teflon stick to the pan?"