Re: short comment about graphical frontends

From: Nick Lamb (njl98r@ecs.soton.ac.uk)
Date: Sun Aug 13 2000 - 20:46:38 PDT

  • Next message: Harald Finster: "Re: ScanMaker X6EL vertical line artifact"

    On Thu, Aug 10, 2000 at 06:31:53PM +0200, mh wrote:
    > I first noticed this when I tried to implement my own frontend.
    > Sometimes the image size seemed to be wrong.
    > Therfore I made a short test with xscanimage1.0.2 and xsane0.59...
    > same problem :-(
    >
    > At least the backend I'm using (plustek V0.36-18)doesn't allow you
    > to set a tlx/tly value that's bigger than the brx/bry value, even
    > if this is only an intermediate state.

    On Fri, Aug 11, 2000 at 03:24:17PM +0100, I wrote:
    > Well, I would think that's probably best fixed in xsane/ xscanimage
    > for now, a quick and easy solution is to set tlx, tly, then brx, bry
    > and then tlx, tly again to be sure.
    >
    > In fact, I think I've seen this problem, but I couldn't swear which
    > backend I've seen the behaviour you describe on.

    I have now confirmed this behaviour in the HP backend too, in fact
    for this backend (or perhaps for XSane?) the tlx...bry options are
    reset under all sorts of conditions.

    Since this is quite legal, and might even be necessary for some
    backends (I can't say if it is for HP) it seems XSane should do what
    it can to improve the user experience,
    ie: Whenever the backend resets tlx...bry, try to change them back

    Also, as I said before, it would be nice if it set the top-left
    twice, so that the annoying behaviour mh and I saw stops.
    In the _very unlikely_ occasion that I have free time next weekend
    I will try to write a patch, but Oliver knows the code far better
    so hopefully he will be able to sort this out.

    Nick. -- Using an HP again for the first time in months

    --
    Source code, list archive, and docs: http://www.mostang.com/sane/
    To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com
    



    This archive was generated by hypermail 2b29 : Sun Aug 13 2000 - 20:43:42 PDT