See to-be-attached ebuild and license file. This is an ebuild which installs enough of Samsung's UnifiedLinuxDriver to let CUPS print to a newer Samsung printer. I have tested that the driver produces valid output as installed by my ebuild. It should be possible to extend the ebuild so that sane support can be made use of when the printer is plugged directly into a Linux machine. However, I have only yet had the opportunity to test printing using samba against a Windows XP machine. Thus, I think/hope that this is enough to print directly to the printer if attached via USB (which is better than being unable to do anything). To allow the CUPS filter to work, I have decided to install libscmssc.so into the system library search path and modify the CUPS filter to not use an absolute path when trying to dlopen() that shared object. (the rastertosamsungsplc filter tries to access libscmssc.so at an absolute path where the filter thinks that CUPS filters would be installed to, which isn't where Gentoo installs filters.) Perhaps bin_replace_string is overkill, but IMO it's cleaner than installing a lone file into /usr/lib/cups/filters.
Created attachment 256564 [details] samsung-unified-linux-driver-1.01.ebuild
Created attachment 256565 [details] licenses/SAMSUNG-ELECTRONICS-software
Fantastic! Worked perfectly with my Samsung CLP-315 printer, much better than foo2zjs which screws up paper margins. My printer also seems to drop documents randomly with foo2zjs as driver, that problem has yet to appear with this official driver. It's contributions like this that makes Gentoo top notch, keep up the good work! Really hoping for these ebuilds to enter either sunrise or the main tree.
For the record: My Samsung CLP-315 still drops documents, it seems if you send a print job to it meanwhile it's printing something, the document gets dropped. The printer does not tell CUPS if a job is printing, when a job comes in it just tells CUPS that it is finished, instantly. I guess this is a firmware issue (or something that lacks implementation in the driver), because it was the same with foo2zjs.
*** This bug has been marked as a duplicate of bug 139715 ***