[sane-devel] Genesys HP scanjet 2400

[Parag N]

Hello,
    I have some questions again on HP2400 current development. How you
people are checking code? i mean i have also same code but neither scanimage
-L is recognising nor Xsane detects it.
then how can i check whatever features are implemented are correct or wrong?
   only sane-find-scanner command is working fine but not xsane. How can i
calso check head movement?
regards,
Parag.

On 12/2/05, Gerald Murray <gmurray at cloudnet.com> wrote:
>
> Hello all,
>
> Scanner: hp scanjet 2400
> using sane-backends-cvs20051124
>
> Expected completion date: unknown, it is ready when it is ready
>
> Update from previous post:
> key: - previously reported; + update :
>
> - Head moves two times the expected distance.
> - Excessive time in genesys_warmup_lamp()
> + excessive time appears to be a result of an incorrect formula.
>   (the source was patched to force SANE_STATUS_GOOD, thereby ignoring
>   the frontend setup, and attempting to be sure that head movements were
>   OK.  I am to look into the frontend setup later)
> - There was no image in the STDOUT output (file size 1523712)
> - Status register for lamp reports OFF, but lamp remains ON.
> - Head remains where the scan stopped.
> + This is a FEATURE, not a bug.  The scan process depends on every detail
>   being expected and correct.  When a status other than SANE_STATUS_GOOD
>   is encountered, all processing stops whereever it was.  This is good,
>   because it assures that there can be no damage to the scanner as a
>   result of an unexpected reply from the device.  The cause of the
>   problem was a result SANE_IO_ERROR, reported at the end
>   of the scan (miscalculation of bytes expected/received)
> - In lineart mode, depth:1 was what I expected.  depth:8 was reported.
>
> NEW:
> + The device responds inconsistently to the same scan request.
>   When the same request is executed sequentially many times, on the
>   odd-numbered attempts, the scan completes, and goes to end of task.
>   On even-numbered attempts, the scan head never moves off of the home
>   position.  (reading bulk data: Error during device I/O)
>   The error comes from not properly setting up the frontend during
>   the lamp warmup.
>   NOTE: to duplicate this problem, install from cvs referenced above.
>   The source code is patched (see below) to force SANE_STATUS_GOOD
>   in function genesys_warmup_lamp().
>   Use scanimage from SOURCE_TOP_DIR/frontend/scanimage.
>   Execute the bash command as shown below, substituting your
>   {BUS}/{DEVNUM} as shown in /proc/bus/usb/devices file for your scanner.
>   REMEMBER to keep the image size small until the size of the image
>   is corrected.
>
> #!/bin/sh
> scanimage -d genesys:libusb:{BUS}:{DEVNUM} --resolution 150 --mode color
> -t 1 -l
> 80 -x100 -y 30 2>scan.log 1>image.pnm
> #end
> PATCH (DO NOT APPLY) This patch was how SANE_STATUS_GOOD was forced
>   from genesys_warmup_lamp().  The frontend is not setup right, so
>   the image is not correct, but it demonstrates motor handling of
>   the scan head and the inconsistent response from the device.
>
> --- sane-backend-20051124/backend/genesys.c     2005-12-02 00:25:
> 55.000000000
> -0600
> +++ sane-backend-hack/backend/genesys.c 2005-12-02 00:24:45.000000000-0600
> @@ -3419,6 +3419,12 @@
>        DBG (DBG_error,
>            "genesys_warmup_lamp: warmup timed out after %d seconds. Lamp
> defective?\n",
>            seconds);
> +      if (seconds > 40) {      /* FIXME GPM: remove 6 lines here */
> +        DBG(DBG_info,
> +           "genesys_warmup_lamp: ** ENOUGH (FIXME) **\n");
> +        status = SANE_STATUS_GOOD;
> +        goto FIXME_hack_trick;
> +      }
>        status = SANE_STATUS_IO_ERROR;
>      }
>    else
> @@ -3427,6 +3433,7 @@
>            "genesys_warmup_lamp: warmup succeeded after %d seconds\n",
>            seconds);
>      }
> +FIXME_hack_trick:      /* FIXME remove this GPM */
>    free (first_line);
>    free (second_line);
>    return status;
> ---- end-of-patch
>
> best regards,
> Gerald
>
>
> --
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20051206/54a39ad7/attachment.html