My scanner, a Plustek OpticSlim 2400, works with scanimage (the cmdline tool) but not with xsane or kooka. This is what I get when I use the command line 'scanimage' program: http://img158.imageshack.us/my.php?image=scanimagecmdline8jq.png As you can see, no problem there. However, both xsane and kooka fails miserably! Preview in Kooka: http://img158.imageshack.us/my.php?image=kookapreview7hn.png Preview in xsane: http://img158.imageshack.us/my.php?image=xsanepreview0nn.png Reproducible: Always Steps to Reproduce: Use xsane or kooka to make a preview scan with a Plustek OpticSlim 2400 Actual Results: The preview image looks like this: (example) Preview in Kooka: http://img158.imageshack.us/my.php?image=kookapreview7hn.png Expected Results: It should have showed a proper preview like scanimage does: (example) http://img158.imageshack.us/my.php?image=scanimagecmdline8jq.png I checked the sane-backends sources and noticed that the sane-backends sources are patched by the ebuild to include version 1.0-67, whereas they usually contain version 1.0-66. At least that's what I think, I'm not very experienced with portage. However, I saw this in the changelog for the gt68xx sources: V 1.0.68 (2005-05-07) * Plustek OpticSlim 2400: Fixed list of resolutions. That fixed the preview. Removed "untested" warning. Adjusted scan area. * Updated gt68xx.TODO. I compiled a new .so-file with sane-backends 1.0.15 patched with gt68xx sources 1.0-71, and the problem was fixed. So... I humbly ask the maintainer to patch the sane-backend sources with gt68xx sources version 1.0-71 instead of 1.0-67, and the Plustek OpticSlim 2400 will work. Thanks in advance.
I'm reassigning this bug to phosphan right away, since his name is all over the sane-backends ChangeLog file.
(In reply to comment #1) > I'm reassigning this bug to phosphan right away, since his name is all over the > sane-backends ChangeLog file. Even worse, my name is in metadata.xml :-)
Made a patch for the latest gt68xx version (build 71) from SANE CVS and checked in as .15-r3 - this way I won't ever get this stable before .16 is out :-) Please test and re-open this bug if necessary.
Great! It works, thanks! (Two things that may interest you... 1. I had the SANE_BACKEND variable set to "gt68xx" when I first emerge -r3, and that caused scanimage -L NOT to find the scanner anymore. sane-find-scanner did, though. Since scanimage -L didn't find anything, neither did kooka or xsane. I removed the SANE_BACKEND variable (thereby increasing the time it took to build the package substantially, of course) and that fixed the problem. I guess this is not how it is supposed to be. 2. When the scanner aquires a preview with -r3, it makes a loud CLICK sound when the head reaches the end of the scanner bed. I've never heard this click sound before, not with 1.0.15, nor with my homebuilt vanilla-sane-backend-sources-patched-with-gt68xx 1.0-71. I don't like the click, it really doesn't sound healthy. I'll close this bug, you think I should open another bug report one about the click sound?)
More about that "click" thing... Only kooka does it, not scanimage nor xscan.
> 1. I had the SANE_BACKEND variable set to "gt68xx" when I first emerge -r3, and > that caused scanimage -L NOT to find the scanner anymore. sane-find-scanner did, > though. Since scanimage -L didn't find anything, neither did kooka or xsane. > > I removed the SANE_BACKEND variable (thereby increasing the time it took to > build the package substantially, of course) and that fixed the problem. I guess > this is not how it is supposed to be. Definitely not. I already had some similar report (bug #90598), but was unable to find any difference between all(?) relevant files when building with or without SANE_BACKENDS. Could you perhaps build both versions for your configuration (using "noclean" FEATURES and copying the build directory somewhere) and check what's different? That would be really helpful. > 2. When the scanner aquires a preview with -r3, it makes a loud CLICK sound when > the head reaches the end of the scanner bed. > > I've never heard this click sound before, not with 1.0.15, nor with my homebuilt > vanilla-sane-backend-sources-patched-with-gt68xx 1.0-71. Since I don't have such a scanner (and no CLICK therefore) I would be interested why it is like that. When you look at the ebuild you see all modifications done to the original sources. I know it is time consuming, but please try to find out what's going on there. > I don't like the click, it really doesn't sound healthy. It is indeed possible to damage scanner hardware with faulty backends. > I'll close this bug, you think I should open another bug report one about the > click sound?) Would be nice - as well as commenting on the other bug about SANE_BACKENDS if you've got something new.
