Bug#442003: mailstat: -k behaviour should be default

2007-09-13 Thread Jon Dowland
On Thu, Sep 13, 2007 at 11:19:44AM +0200, Santiago Vila wrote: > If I move mailstat out of the PATH, people will submit a > bug because upgrading to the new package would break > existing scripts. > > The same would happen if I take mailstat out of the > procmail package. That's true, yes. > > W

Bug#442003: mailstat: -k behaviour should be default

2007-09-13 Thread Santiago Vila
On Thu, 13 Sep 2007, Jon Dowland wrote: > On Thu, Sep 13, 2007 at 10:11:28AM +0200, Santiago Vila wrote: > > Sorry, this is a 17-year old program. This is not time to change the > > defaults. > > Fair enough. Indeed it is, and googling around shows this to be a common > problem, too. There are ot

Bug#442003: mailstat: -k behaviour should be default

2007-09-13 Thread Jon Dowland
On Thu, Sep 13, 2007 at 10:11:28AM +0200, Santiago Vila wrote: > Sorry, this is a 17-year old program. This is not time to change the > defaults. Fair enough. Indeed it is, and googling around shows this to be a common problem, too. There are other possible solutions to minimize risk: * m

Bug#442003: mailstat: -k behaviour should be default

2007-09-12 Thread Jon Dowland
Package: procmail Version: 3.22-16 Severity: important Mailstat will, by default, truncate the input logfile. I think this is un-intuitive behaviour that can make unexpected data loss quite likely. I think the default behaviour should be as the -k switch toggles, with perhaps a new command-line sw