Re: DBG(0, ...)

From: Henning Meier-Geinitz (henning@meier-geinitz.de)
Date: Mon May 14 2001 - 11:14:04 PDT

  • Next message: Paul Nixon: "IT'S BROKEN YEA!!!!"

    Hi,

    On Mon, May 14, 2001 at 10:16:35AM +0200, Petter Reinholdtsen wrote:
    > I believe there is a misfeature in the DBG() functions. Peter Fales
    > pointed out to me that the following code in a backend
    >
    > DBG(0, "foo");
    >
    > will print out "foo" even if SANE_DEBUG_BACKEND is not set. I believe
    > this is wrong. I believe DBG() should not print anything unless
    > debugging is enabled for that specific backend. I suggest applying
    > this patch to fix it. Is there anything I haven't thought about?

    I don't know if this was intentional but I like the current behaviour.
    It should be clear that DBG(0,...) should be used with care and only
    when the backend has detected a scanner that is supported (at least in
    principle) by the backend. E.g. I'm using this in mustek.c to print a
    warning when a scanner is used that nobody has tested yet but which is
    a Mustek scanner.

    So I'm voting for the old behaviour.

    > The patch changes the default level to be (in practise, not in
    > reality) -1, instead of the previous default level 0.

    If I remeber correctly at least one backend (microtek?) uses the debug
    level directly. So probably it will be confused about this.
    Also the "[sanei_debug] Setting debug level ..." isn't printed any more.

    Bye,
      Henning

    --
    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 : Mon May 14 2001 - 11:04:27 PDT