Re: [darktable-user] dt 4.4 on Debian 11.7, apt update issue

[Dieter Faulbaum]

Hello Michael,

if you look here:
https://download.opensuse.org/repositories/graphics:/darktable/
you can see, there is no Debian_11 directory (anymore?) only a 
Debian_12

(and Debian_Testing and Debian_Unstable) directory.

Maybe download.opensuse.org doesn't support a oldstable version 
(of

Debian)?

darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

[darktable-user] Darktable Manual EPUB version

[Dieter Faulbaum]

I have problems to read this format by atril. Atril crashes with a 
segmentation fault:


/tmp/atril-65485/darktable_user_manual_en-4.4.epubCR7961/OEBPS/toc.ncx:10: 
parser error : Opening and ending tag mismatch: meta line 5 and 
head

 
^
/tmp/atril-65485/darktable_user_manual_en-4.4.epubCR7961/OEBPS/toc.ncx:2245: 
parser error : Opening and ending tag mismatch: head line 4 and 
ncx


 ^
/tmp/atril-65485/darktable_user_manual_en-4.4.epubCR7961/OEBPS/toc.ncx:2246: 
parser error : Premature end of data in tag ncx line 2


^
[1]65485 segmentation fault  atril 
darktable_user_manual_en-4.4.epub



So I checked this EPUB by 'epubcheck', which doesn't look very 
nice:
epubcheck darktable_user_manual_en-4.4.epub 2>& 1|grep -E 
"^(FATAL|ERROR)"
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/content.opf(23,44): Error 
while parsing file: character content of element "meta" invalid; 
must be a string with length at least 1 (actual length was 0)
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/toc.ncx(6,53): Error while 
parsing file: element "meta" not allowed here; expected the 
element end-tag
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/toc.ncx(7,46): Error while 
parsing file: element "meta" not allowed here; expected the 
element end-tag
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/toc.ncx(8,55): Error while 
parsing file: element "meta" not allowed here; expected the 
element end-tag
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/toc.ncx(9,54): Error while 
parsing file: element "meta" not allowed here; expected the 
element end-tag
FATAL(RSC-016): 
darktable_user_manual_en-4.4.epub/OEBPS/toc.ncx(10,5): Fatal Error 
while parsing file: The element type "meta" must be terminated by 
the matching end-tag "".
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/toc.ncx(-1,-1): Error 
while parsing file: The element type "meta" must be terminated by 
the matching end-tag "".
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/toc.xhtml(6,10): Error 
while parsing file: element "title" not allowed here; expected the 
element end-tag
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/toc.xhtml(11,59): Error 
while parsing file: element "link" not allowed here; expected the 
element end-tag
FATAL(RSC-016): 
darktable_user_manual_en-4.4.epub/OEBPS/toc.xhtml(13,4): Fatal 
Error while parsing file: The element type "meta" must be 
terminated by the matching end-tag "".
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/toc.xhtml(-1,-1): Error 
while parsing file: The element type "meta" must be terminated by 
the matching end-tag "".
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/index.html(10,59): Error 
while parsing file: element "link" not allowed here; expected the 
element end-tag
FATAL(RSC-016): 
darktable_user_manual_en-4.4.epub/OEBPS/index.html(12,3): Fatal 
Error while parsing file: The element type "meta" must be 
terminated by the matching end-tag "".
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/index.html(-1,-1): Error 
while parsing file: The element type "meta" must be terminated by 
the matching end-tag "".
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/front-cover.html(3,30): 
Error while parsing file: value of attribute "xml:lang" is 
invalid; must be an RFC 3066 language identifier or must be equal 
to ""
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/front-cover.html(4,11): 
Error while parsing file: element "body" not allowed yet; missing 
required element "head"
ERROR(CSS-008): 
darktable_user_manual_en-4.4.epub/OEBPS/style.css(146,3): An error 
occurred while parsing the CSS: Token "@top-left-corner" not 
allowed here, expecting a property name.
ERROR(CSS-008): 
darktable_user_manual_en-4.4.epub/OEBPS/style.css(165,22): An 
error occurred while parsing the CSS: Token "}" not allowed here.
ERROR(CSS-008): 
darktable_user_manual_en-4.4.epub/OEBPS/style.css(212,3): An error 
occurred while parsing the CSS: Token "@top-left-corner" not 
allowed here, expecting a property name.
ERROR(CSS-008): 
darktable_user_manual_en-4.4.epub/OEBPS/style.css(231,20): An 
error occurred while parsing the CSS: Token "}" not allowed here.
FATAL(RSC-016): 
darktable_user_manual_en-4.4.epub/OEBPS/module-reference/utility-modules/shared/metadata-editor.html(25,129): 
Fatal Error while parsing file: Attribute name "unchanged" 
associated with an element type "leave" must be followed by the ' 
= ' character.
ERROR(RSC-005): 
darktable_user_manual_en-4.4.epub/OEBPS/module-reference/utility-modules/shared/metadata-editor.html(-1,-1): 
Error while parsing file: Attribute name "unchanged" associated 
with an element type "leave" must be followed by the ' = ' 
character.
ERROR(RSC-008): 
darktable_user_manual_en-4.4.epub/OEBPS/style.css(137,5): 
Referenced resource "OEBPS/darktable-logo.svg" is not declared in 
the OPF manifest.
ERROR(RSC-007): 

Re: [darktable-user] Print View loosing image layout settings

[Dieter Faulbaum]

François Delègue  writes:

...
For me (4.2.0, Linux) a running Cups at darktable launch is 
enough for the printers to be
present in printing view, and printers themselves can be plugged 
only at printing time (if
not, Cups keeps the printing task(s) which are launched when the 
printer is online— dt
may have been stopped). So recognizing printers means for 
darktable a running Cups
service at launch, and there is no need to change dt code in my 
opinion (Cups is always

running, the more often).
For me (on a Debian (12) system with cups (always running after 
startup)

and dt 4.2.0) it doesn't work.
I have one "airprint" printer, which I don't see e.g. with lpstat, 
when

it is switch off. But when I switch it on, I see this printer with
lpstat but not under dt (print view) without restarting dt.

But you can say this is my problem (I shouldn't use "airprint".-)

darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

Re: [darktable-user] Print View loosing image layout settings

[Dieter Faulbaum]

"Dan Torop"  writes:

It's done with montage and mogrify (from ImageMagick) and 
exiftool.

The code is here:-)
https://multimedia4linux.de/index.php/bildbearbeitung/darktable/darktable-plugin-digitaler-kontaktabzug


It looks quite nice -- a good range of features. So nice that 
I'm a bit hesitant to make a quick version in the print view 
which does a subset. But I suppose what print view could bring 
to it is that there's already a good GUI-oriented way to set up 
boxes and fill them with images. And that print code already 
does the work of scaling images to a common DPI and converting 
to a common colorspace, So there isn't that much remaining work 
to do.


Current print view creates a hand-constructed PDF and sends it 
to CUPS or Turboprint. Output to file would either 
hand-construct a composite image or ask Cairo library to do 
this, then can output with the help of existing imageio code.


Future work could allow for a more "automatic" way of creating 
boxes, e.g. in a grid. I'm not sure if the other things (image 
labels, title for the grid) are totally out of scope, though.


I've used all sorts of awkward ways to do similar work -- 
ImageMagick command line, dragging images around in GIMP, either 
Scribus or Inkscape (can't remember), Libreoffice Draw. This 
does really seem like something which it would be nice for 
darktable to do.


The examples from BeeRef (https://github.com/rbreu/beeref/) look 
nice (and flexible.-) But it crashes under Debian (bookworm).-(

And gThumb has a good working "Contact Sheet…".

Query is that once print view has a really layout code, does it 
cease to be print view, and become something else (multi-image 
layout, with options to print or export to a file)?
The "name" isn't very important for me.-) There are others (the 
main developers?), who should decide this?


Another query is whether the monolithic "print settings" in the 
right panel gets split into a couple modules, with one remaining 
"print settings" (with a print button), one handling "image 
layout" (what's currently in print settings under that name, 
plus any more options), and as well an export module (as in 
lighttable: file type, resolution, an export button, etc.). I 
seem to recall that the code is so tightly interwoven that the 
split would, internally, be a challenging one.
I think, multiple modules will make it more flexible. But do you 
have enough time for such a challenge?-)


darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

Re: [darktable-user] Print View loosing image layout settings

[Dieter Faulbaum]

"Dan Torop"  writes:

...
(I know that Holger Klemm (from multimedia4linux.de) made a 
lua-script
for such a thing.-) 


That sounds great. I'm curious how it is implemented.
It's done with montage and mogrify (from ImageMagick) and 
exiftool.

The code is here:-)
https://multimedia4linux.de/index.php/bildbearbeitung/darktable/darktable-plugin-digitaler-kontaktabzug

...
It seems for me, that the presets are stored in data.db as a 
BLOB,

so I can't find out quickly, what went wrong.


You may have some control of this. Try adjusting "store XMP tags 
in compressed format" in the storage tab of preferences to 
"never".

This settings doesn't help.-( I had "only large entries" before.
I "lost" the first column of the image areas but can't create new 
areas because I got the message "maximum image per page reached"
(Is there a small typo? Should it be written as "maximum images 
per page reached"?)


...
It depends... I'm very interested in the print view (though 
short of time). 
Good luck with that (I can wait.-)! 


darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

Re: [darktable-user] Print View loosing image layout settings

[Dieter Faulbaum]

"Dan Torop"  writes:
...
But is there other information to keep, e.g. for multi-image 
layouts?
Making a preset is a great idea, if that does work. But it 
shouldn't
be that hard to store this in the configuration system as well, 
so
that the most recent layout is maintained. 

...

After the hint from François Delègue I tried presets for 
multi-image

layouts.
First I noticed, that I can't create a "print", if there is no 
printer
reachable.-( After I switched one printer on, I have to restart 
dt, to
notice this printer. It seems for me that dt only recognizes 
printers,
which are "reachable" at startup. It would be nicer, if the search 
for
printers is done every time, when the print view is selected. I 
think
(or hope) it's not a long task. 

Second I noticed, that the maximum of 20 images per page are not 
very
much. I hoped, that I can use the multi-image layout for something 
like
a contact sheet. 
(I know that Holger Klemm (from multimedia4linux.de) made a 
lua-script
for such a thing.-) 

I created a (test) layout on a A4 (landscape) page with 4*3 
(quadratic

70*70) image areas. That preset works well.
But another preset on a A4 (portrait) page with 4*5 (quadratic 
52*52)
image areas is much scrambled after storing it and loading it 
again.

It seems for me, that the presets are stored in data.db as a BLOB,
so I can't find out quickly, what went wrong.

I think you are right: the multi-image layouts should be kept too.
However for me it would be "enough", that I can switch between 
multiple
views in one session without losing the print settings. But indeed 
it
would be (much) nicer, if those settings aren't lost between 
multiple

sessions.-)

I'm not sure if a feature request is a good idea (but I can try 
it).
François reckoned "dt developers have surely more important things 
to do ?"


darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

Re: [darktable-user] Print View loosing image layout settings

[Dieter Faulbaum]

François Delègue  writes:


Le Mon, 16 Jan 2023 16:01:08 +0100,
Dieter Faulbaum  a écrit :

When I switched from to print view to any other view and back to 
print

view the settings of the image layout are lost.
I think it would be nicer, if these settings would be retained. 
With multiple image
areas it's faster to "clear layout", if I need another one, than 
to create the last one

again.

What do others think of this "idea"?


Yes it should be good, but dt developers have surely more 
important things to do ? I
don't know. For a printing layout you can make a preset which 
will keep the printing
zones (their sizes & positions): this will solve a good part of 
the question.


Oh yes thanks, a good idea (I overlooked the presets in this 
view). This helps me.

Btw, I meant losing (not loosing.-)

darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

[darktable-user] Print View loosing image layout settings

[Dieter Faulbaum]

When I switched from to print view to any other view and back to 
print

view the settings of the image layout are lost.
I think it would be nicer, if these settings would be retained. 
With multiple image areas it's faster to "clear layout", if I need

another one, than to create the last one again.

What do others think of this "idea"?

darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

[darktable-user] Ignore EXIF Rating (in 4.2.0)

[Dieter Faulbaum]

It seems for me, that this parameter setting doesn't work anymore 
under 4.2.0 (Debian testing).
I always get the Rating from the camera (independently, if I 
switch this parameter "on" or "off").


Anyone else having this problem?

darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

Re: [darktable-user] Tags in xmp files

[Dieter Faulbaum]

Bruce Williams  writes:

Is there a way to get dt to read the tags from the xmp files and 
add those tags to whatever database file they should live in? I 
would prefer

NOT to have to enter all the tags again if I can avoid it!


There is an option in the storage part 'look for updated XMP file 
on startup'.

Is it enabled on your Linux system?
Maybe this will help?

darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

[darktable-user] Map View problem, sorry forgot the pictures

[Dieter Faulbaum]

I have a small problem with the map view.
I use version darktable 4.2.0 on a Debian sid with version 1.2.0 
of libosmgpsmap.


I created the location of "Berlin" and in that the sub-location 
"Wannsee"

(with polygons from "find location").

But sometimes these (red) borderlines aren't shown correct
(at edges of the map view).
The 2 examples are from the upper right corner of the map view:
the difference is only generated by moving the map a little bit to 
the right.


Can anyone else confirm this?
Because I think this is a (small, unimportant?) bug.
I have seen this behaviour with other borderlines too.
And I have seen this in the darktable version 4.0.1 too.
Maybe a libosmgpsmap bug?



darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

[darktable-user] Map View problem

[Dieter Faulbaum]

I have a small problem with the map view.
I use version darktable 4.2.0 on a Debian sid with version 1.2.0 
of libosmgpsmap.


I created the location of "Berlin" and in that the sub-location 
"Wannsee"

(with polygons from "find location").

But sometimes these (red) borderlines aren't shown correct
(at edges of the map view).
The 2 examples are from the upper right corner of the map view:
the difference is only generated by moving the map a little bit to 
the right.


Can anyone else confirm this?
Because I think this is a (small, unimportant?) bug.
I have seen this behaviour with other borderlines too.
And I have seen this in the darktable version 4.0.1 too.
Maybe a libosmgpsmap bug?

[2. Map View I --- image/png; Screenshot from 2022-12-22 
15-18-39.png]...


[3. Map View II --- image/png; Screenshot from 2022-12-22 
15-19-00.png]...


darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

[darktable-user] edit `new preset'

[Dieter Faulbaum]

can anyone tell me, why in this "module",
if I set "auto apply the preset to matching images"
in the exposure part the times '1/1000' and '1/15' are shown two 
times?


Is this a (little, unimportant) bug or intention?

darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

Re: [darktable-user] Question regarding .svg watermarks comparing Windows and Ubuntu Linux

[Dieter Faulbaum]

Patrick Shanahan  writes:


* Willy Williams  [02-28-22 21:42]:

Question:

    I have some .svg watermark files that were generated in 
Inkscape on the
Windows machines.  They work superbly when applied using 
darktable on the
Windows machines.  When I copy those same .svg files to the 
Linux machine
and put them in the correct place, then open them using the 
Inkscape
instance on the Linux machine, there is no apparent difference, 
compared to
the same files opened in Inkscape on the Windows machines.  
That said, when
I apply the .svg file watermarks to any image using darktable 
on the Linux
machine, the graphic "signature" is represented perfectly, but 
the text
portion is significantly shrunk side-to-side, but only when 
applied using
darktable.  Anybody have ideas regarding the issue? Are there 
any fonts that
do NOT work right when migrated from Windows to Linux?  
Inquiring minds want

to know...


check which font is used one each machine.  match them and you 
may solve

your problem.


I had a similar problem after upgrading to librsvg-2.48.3
(from version 2.44.10, where it works fine).
Now I have version 2.52.5 and it seems to me, that the problem 
isn't
solved, although the issue 
(https://gitlab.gnome.org/GNOME/librsvg/-/issues/599)
is marked as closed (maybe I should reopen it?-). 
The actual version of librsvg in Debian bookworm is 2.52.5

(the newest version is 2.53.1).
My actual libpango1.0 version is 1.50.4 (which is the newest).

darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

[darktable-user] Crash with retouch module

[Dieter Faulbaum]

Trying to learn a little bit more about the darktable modules 
(using the

user manual), I deploy the retouch module a little bit.

Every time I increased the wavelet decompose scales over 9 or 10,
darktable crashes (immediately).
Looking (only superficially) at the debug lines, there are many 
lines
which show the AMD OpenCL libraries, but disabling the OpenCL 
support

doesn't help for the crash behaviour.

I'm not sure, if I made an error, increasing the scales so high 
(without

doing something else).

Can anyone tell me, if she/he gets this crash too (before writing 
a bug

report) or if it's my fault.

Thanks of helping

darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

Re: Fwd: Fwd: Fwd: [darktable-user] Where are Collections stored?

[Dieter Faulbaum]

Patrick Shanahan  writes:
...

darktable --configdir ~/.config/darktable-3.4
darktable --configdir ~/.config/darktable-master


and if you want that these are "totally" independent from each 
other you

should set the cachedir (with --cachedir) accordingly like these:

darktable --configdir ~/.config/darktable-3.4 --cachedir 
~/.cache/darktable-3.4
darktable --configdir ~/.config/darktable-master --cachedir 
~/.cache/darktable-master



darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

[darktable-user] Initial rating

[Dieter Faulbaum]

It seems to me, that the initial rating (when importing a picture 
or folder) doesn't work anymore in darktable version 3.4 in a 
Debian/GNU Linux system (which works fine in version 3.2.1).
My tests are: remove pictures, delete the xmp files of them and 
import them again.
I changed the (default) value (in the global preferences) to '2' 
but this doesn't help either.
After setting the value back to '1' (and exit darktable) there is 
a line in the (actual) 'darktablerc': 
ui_last/import_initial_rating=1


Can anyone tell me if I made something wrong.
I didn't find any hint in the news of 3.4 that the behaviour of 
the (initial) rating changed.


Thanks for any help

darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

[darktable-user] Darktable 3.2.1

[Dieter Faulbaum]

I want to compile darktable "as complete as possible" on a Debian 
testing system.


But I get hints (from build.sh):

-- The following OPTIONAL packages have not been found:

* LLVM (required version >= 3.9)
* libavif (required version >= 0.7.2)
* GMIC

I have llvm (version 9) installed and
libavifile-0.7-dev (version 0.7.48~20090503.ds-20.1+b1) and
libgmic-dev (version 2.4.5-1.1) too

Can anyone tell me, what is needed to get rid of these hints?
Is it possible on Debian testing?


Cheers

darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

[darktable-user] rdf:Bag and/or rdf:Seq

[Dieter Faulbaum]

Hello,

it seems for me that the version 3.0.1 of dt uses rdf:Bag instead of
rdf:Seq (in the xmp-files).
Am I right or is there any difference, when the one will be used and when
the orther?

Thanks for clarifying this for me.

Dieter


darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

[darktable-user] Watermark module enhancement

[Dieter Faulbaum]

I normally use a watermark with the presets of alignment right-bottom  
and x- and y-offset of -0.003.
For me it would be very useful, when I change the alignment to maybe  
left-top the offsets will be changed "automatically" to +0.003.

I think it would be nice to have an option for that.
What do you think?


darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org

[darktable-user] compilation problem (with libexiv2)

[Dieter Faulbaum]

I have a problem under Debian sid:
because the "native" libexiv2 in sid is version 0.25 an this version is
a little bit old it doesn't know my "new" lens(es). So I installed the
latest version (0.26) under /opt/exiv2, which works fine.
I set (export) these Environment-variables:
PKG_CONFIG_PATH=/opt/exiv2/lib/pkgconfig:/usr/local/lib/x86_64-linux-gnu/pkgconfig:/usr/local/lib/pkgconfig:/usr/local/share/pkgconfig:/usr/lib/x86_64-linux-gnu/pkgconfig:/usr/lib/pkgconfig:/usr/share/pkgconfig
CMAKE_INCLUDE_PATH=/opt/exiv2/include/exiv2
CMAKE_LIBRARY_PATH=/opt/exiv2/lib
LD_LIBRARY_PATH=/opt/exiv2/lib
and run ./buid.sh under /tmp/darktable-2.4.2.

First it looks fine:

...
-- Found JsonGlib
-- Found EXIV2: /opt/exiv2/lib/libexiv2.so
-- Found FLICKCURL: /usr/lib/x86_64-linux-gnu/libflickcurl.so
...

but later on I get this:

...
[ 51%] Linking C shared module libtethering.so
libdarktable.so: undefined reference to  
`Exiv2::XmpParser::decode(Exiv2::XmpData&,  
std::__cxx11::basic_string const&)'
libdarktable.so: undefined reference to  
`Exiv2::ExifKey::ExifKey(std::__cxx11::basic_string const&)'
libdarktable.so: undefined reference to  
`Exiv2::ExifData::operator[](std::__cxx11::basic_string const&)'
libdarktable.so: undefined reference to  
`Exiv2::XmpProperties::registerNs(std::__cxx11::basic_string const&,  
std::__cxx11::basic_string const&)'
libdarktable.so: undefined reference to  
`Exiv2::readFile(std::__cxx11::basic_string const&)'
libdarktable.so: undefined reference to  
`Exiv2::XmpParser::encode(std::__cxx11::basic_string&, Exiv2::XmpData  
const&, unsigned short, unsigned int)'
libdarktable.so: undefined reference to  
`Exiv2::XmpKey::XmpKey(std::__cxx11::basic_string const&)'
libdarktable.so: undefined reference to  
`Exiv2::Xmpdatum::operator=(std::__cxx11::basic_string const&)'
libdarktable.so: undefined reference to  
`Exiv2::XmpTextValue::XmpTextValue(std::__cxx11::basic_string const&)'
libdarktable.so: undefined reference to  
`Exiv2::XmpData::operator[](std::__cxx11::basic_string const&)'
libdarktable.so: undefined reference to  
`Exiv2::Metadatum::print[abi:cxx11](Exiv2::ExifData const*) const'
libdarktable.so: undefined reference to  
`Exiv2::ImageFactory::open(std::__cxx11::basic_string const&, bool)'
libdarktable.so: undefined reference to  
`Exiv2::XmpTextValue::read(std::__cxx11::basic_string const&)'
libdarktable.so: undefined reference to  
`Exiv2::IptcKey::IptcKey(std::__cxx11::basic_string const&)'
libdarktable.so: undefined reference to  
`Exiv2::Exifdatum::operator=(std::__cxx11::basic_string const&)'
libdarktable.so: undefined reference to  
`Exiv2::Value::toString[abi:cxx11]() const'
libdarktable.so: undefined reference to  
`Exiv2::PreviewImage::mimeType[abi:cxx11]() const'

collect2: error: ld returned 1 exit status
src/CMakeFiles/darktable.dir/build.make:181: recipe for target  
'src/darktable' failed

make[2]: *** [src/darktable] Error 1
CMakeFiles/Makefile2:1145: recipe for target  
'src/CMakeFiles/darktable.dir/all' failed

make[1]: *** [src/CMakeFiles/darktable.dir/all] Error 2
make[1]: *** Waiting for unfinished jobs
[ 51%] Built target print
[ 51%] Built target slideshow
[ 51%] Linking C shared module libmap.so
[ 51%] Linking C shared module libknight.so
[ 51%] Built target tethering
[ 51%] Built target knight
[ 51%] Built target map
[ 51%] Linking C shared module libdarkroom.so
[ 51%] Built target darkroom
[ 51%] Linking C shared module liblighttable.so
[ 51%] Built target lighttable
Makefile:151: recipe for target 'all' failed

The cmake_link_script (CMakeFiles/darktable.dir/link.txt) is this:
/usr/bin/cc -fPIC  -Wall -fno-strict-aliasing -Wformat  
-Wformat-security -Wshadow -Wtype-limits -Wvla -Wold-style-declaration  
-Wno-error=varargs -Wno-format-truncation -Wframe-larger-than=32768  
-Wstack-usage=32768 -Wlarger-than=524288 -std=c99 -fopenmp  
-march=native -msse2 -g -mfpmath=sse -O2 -g -DNDEBUG -O2   
-Wl,--as-needed -Wl,--as-needed -Wl,--as-needed -shared  -o  
libtethering.so CMakeFiles/tethering.dir/tethering.c.o  
-Wl,-rpath,"\$ORIGIN/../lib/darktable" ../libdarktable.so  
../external/LuaAutoC/liblautoc.a -lpthread -lm -lglib-2.0 -lgtk-3  
-lgdk-3 -lpangocairo-1.0 -lpango-1.0 -latk-1.0 -lcairo-gobject -lcairo  
-lgdk_pixbuf-2.0 -lgio-2.0 -lgobject-2.0 -lglib-2.0 -lxml2 -lsoup-2.4  
-lgio-2.0 -lgobject-2.0 -lgphoto2