On 04/30/2012 02:12 AM, Jeffrey Ratcliffe wrote: > On 29 April 2012 21:27, sean darcy<seandarcy2 at gmail.com> wrote: >> scanadf is generating images that aren't PNM. Maybe the it's generating raw? >> Shouldn't it be generating a PNM file? Is there an option to have it >> generate a PNM file? > > I think this is problem with the backend not sending enough data - > i.e. the backend reports via params that a scan of n x m should > appear, and therefore the frontend writes the appropriate PNM header, > plus all the data it gets until it gets EOF, but that the x bytes that > appear don't make up n x m. > > Any post processing tools than complain that the PNM is not valid. > > I've had to put a workaround in the development version of gscan2pdf > to pad the end of the file with whitespace in order to get a valid > PNM. > > Regards > > Jeff >
I don't think that's the problem here. I looked at the scan file with a hex editor. All zeros. No magic number, no x or y. sean
