Miura's work does include some good new features, and most of the changes from my S1100 patch. I think he left out the serial number support, maybe something else, I don't recall. I have intended to merge his work into sane-backends mainline, but have not had time. If you set up a github branch, I will pull from it. I might skip some of the patches if they don't make sense for mainline.
allan On Sun, Mar 2, 2014 at 7:08 AM, Peter Marschall <peter at adpm.de> wrote: > Hi all, > > in my quest to get my S1300i working with sane (see [1]), I found Hiroshi > Miura's (miurahr) work to improve Sane's epjitsu backend at github [2] and > rebased it to Sane's git HEAD. > > Advantages of the patch series: > + got rid of some magic numbers by replacing them with meanig ful macros > + page width support > + support of tl_y > + refactoring of coarsecal() > + support of S1100 > > I cannot tell about the S1100 support, but I can confirm that a patched > variant > of the epjitsu backend works together with my S1300i (although - again - with > skanlite only, see [1]). > I did not find any regressions in comparison to Sane 10.24 nor to git HEAD. > > A big advantage is the page width and tl_y support which allows to restrict > the scan output to e.g. DIN A4 instead of the full width of the scan area. > > Instead of having this patch series in my repo only, I'd rather have these > patches included in sane upstream. > > In which form do you prefer to get the patches? > * a branch on github? > * a series of mails here on the list? > * ... > > In addition to these epjitsu patches, I found other patches too: > - Return \0 even when length is zero in sane_read > - typo patches from Debian > - some small patches from me > which in my opinion should go upstream too. > > Thanks for your answer and your work on Sane. > Peter > > > [1] http://lists.alioth.debian.org/pipermail/sane-devel/2014-March/032139.html > [2] https://github.com/miurahr/sane-backends > -- > Peter Marschall > peter at adpm.de > -- "The truth is an offense, but not a sin"
