On Tue, Apr 18, 2017 at 12:24 AM, Wolfgang Goetz <wolfgang.zt...@web.de> wrote: > Hello Roman, > sorry for the delay, > just back from easter holidays having some time to track down: > > The real easter-egg appears here: (272-ko.png) > commit c8383ed5604b426a5da31dfe9865be20cbf7c8ec > Author: Roman Lebedev <lebedev...@gmail.com> > Date: Thu Jan 19 18:16:22 2017 +0300 > > Rawspeed submodule update: Huffman rewrite, some speed changes. Yeah, i have already checked, and it is indeed a regression. That is why it is important to have all the samples... And to not change stuff until there are at least some changes.
> workflow: > > git checkout <sha1> > git submodule update > sudo rm -rf /opt/darktable > rm -rf build > ./build.sh > sudo cmake --build "/home/goetz/repos/darktable/build" --target install -- -j4 > /opt/darktable/bin/darktable --library ":memory:" --cachedir /tmp/d > ~/20151027-10431071-NIKON_1_J5.nef > *looking* at the processed picture because dt doesn't give any error/warning > in earlier version. > > newer/latest git simply crash which is best for catching the fault: yep, all that sounds completely sane. > darktable --library ":memory:" --cachedir /tmp/d > ~/20151027-10431071-NIKON_1_J5.nef > .. > darktable: > /data_b/portage/portage/media-gfx/darktable-9999/work/darktable-9999/src/external/rawspeed/src/librawspeed/io/BitStream.h:156: > rawspeed::uint32 rawspeed::BitStream<Tag, > Cache>::peekBitsNoFill(rawspeed::uint32) [with Tag = rawspeed::MSBBitPumpTag; > Cache = rawspeed::BitStreamCacheRightInLeftOut; rawspeed::uint32 = unsigned > int]: Assertion `nbits <= Cache::MaxGetBits && nbits <= cache.fillLevel' > failed. > [1]+ Done gitk > Aborted > to answer your first reply: > i do not let any software change my raw files. using aufs to separate things. > (nef on ro layer, xmps and other stuff on top, paranoid but necessary when > playing around with software) > raws are copied from memory card to PC. > no software inbetween. > not even gphoto and NEVER EVER Nikon stuff. (they(N*) did nasty things in the > past, but have learned) Good :) Anyway, i'm looking into it, thank you for reporting it. > ... > > have a nice easter monday late night :) > > greets > Wolfgang Roman. ___________________________________________________________________________ darktable developer mailing list to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org
