On Nov 20, 2007 10:11 AM, Gerhard Jaeger <gerhard at gjaeger.de> wrote: > Am Dienstag, 20. November 2007 14:01:05 schrieb m. allan noah: > > > On 11/20/07, Johannes Meixner <jsmeix at suse.de> wrote: > > > > > > Hello, > > > > > > On Nov 19 15:38 m. allan noah wrote (shortened): > > > > why are you trying to use the -T? > > > > > > I thought "scanimage -d... -T" is the right way to do a test? > > > > > > I use it in the YaST scanner config tool to provide a [Test] option > > > for the user to test if his scanner/backend combination works. > > > > > > Should I better use "scanimage -d... >/dev/null" for a test? > > > > the -T test just tries using a variety of buffer sizes during the data > > transfer, followed by a cancel (instead of completing the scan) IIRC. > > This might put some scanners/backends into a funk, as they dont > > properly implement the cancel. > > > > i would not assume that every scanner/backend combination will act > > properly with -T. > > > > But in fact they should! A real working cancel is essential for > a good working frontend - IMHO! > Otherwise you always have to wait until a scan has finished even > if you see that it's crap - not really user-friendly. >
i agree in principle, but i would say it is the purpose of -T for the backend author to test his code, not for the scanner owner to test the machine. allan -- "The truth is an offense, but not a sin"
