I have what looks like a very similar setup: `pkg-config --libs libpng` returns the same thing, I have symlinks for those two files, but matplotlib finds libpng just fine. Did you do 'make distclean' after reinstalling libpng and pkg-config? What does "brew list" say?
On Thursday, November 21, 2019 at 1:43:05 PM UTC-8, Paul Masson wrote: > > Both are present as symlinks > > On Thursday, November 21, 2019 at 1:35:24 PM UTC-8, Dima Pasechnik wrote: >> >> >> >> On Thu, 21 Nov 2019, 21:09 Paul Masson, <[email protected]> wrote: >> >>> After installing both libpng and pkg-config, and making sure both are >>> completely updated, building Sage still fails on matplotlib with the same >>> error. Here is an output requested yesterday: >>> >>> pkg-config --libs libpng >>> >>> -L/usr/local/Cellar/libpng/1.6.37/lib -lpng16 -lz >>> >>> >>> I added /usr/local/Cellar to my $PATH. Build still fails for matplotlib. >>> >>> >>> Before I ignore the system libpng when building, is there anything else >>> I should try? >>> >> >> do you have /usr/local/include/png.h and /usr/local/lib/libpng.dylib >> present? >> >> >>> On Wednesday, November 20, 2019 at 4:36:02 PM UTC-8, Dima Pasechnik >>> wrote: >>>> >>>> On Wed, Nov 20, 2019 at 11:56 PM Paul Masson <[email protected]> >>>> wrote: >>>> > >>>> > I'd like to understand why this is now necessary. I've been building >>>> Sage on this Mac for several years without having this happen. I think the >>>> last "make distclean" I did was in April. What might have changed in the >>>> meantime? Perhaps this is relevant: >>>> https://trac.sagemath.org/ticket/27186 >>>> >>>> Sure, now a system version of libpng is used by default. >>>> We're trying to unvendor as much as possible from Sage the >>>> distribution, see >>>> https://trac.sagemath.org/ticket/27330 >>>> - something which is a necessary step towards make sagelib a "normal" >>>> Python library, >>>> which you could install by running pip. >>>> >>>> So, yes, if you have a stale library somewhere it might break the >>>> build, as it's not really >>>> possible to test on all the possible broken configurations. >>>> >>>> You may force building of libpng by doing >>>> >>>> ./configure --without-system-png >>>> >>>> before running make >>>> >>>> Hope this explains. >>>> > >>>> > On Wednesday, November 20, 2019 at 3:38:12 PM UTC-8, Dima Pasechnik >>>> wrote: >>>> >> >>>> >> you can do >>>> >> >>>> >> brew install libpng >>>> >> >>>> >> which would install libpng 1.6.37 into /usr/local/ >>>> >> (I think, by looking at >>>> >> >>>> https://github.com/Homebrew/homebrew-core/blob/master/Formula/libpng.rb) >>>> >>>> >> >>>> >> you can also install a lot of other homebrew packages, such as gmp, >>>> etc >>>> >> to account for more packages on >>>> https://trac.sagemath.org/ticket/27330 >>>> >> >>>> >> >>>> >> I've tweaked/wrote some Homebrew formulas to get flint and arb >>>> >> >>>> >> brew install dimpase/science/flint >>>> >> brew install dimpase/science/arb >>>> >> >>>> >> from https://github.com/dimpase/homebrew-science >>>> >> >>>> >> (probably many more can be done this way too) >>>> >> >>>> >> >>>> >> >>>> >> On Wed, Nov 20, 2019 at 11:22 PM Paul Masson <[email protected]> >>>> wrote: >>>> >> > >>>> >> > I have Homebrew installed but have rarely used it so assumed it >>>> wasn't part of the problem. It needed updating, but then the Sage build >>>> stopped with the same matplotlib error. >>>> >> > >>>> >> > libpng-config --cflags >>>> >> > >>>> >> > -I/usr/local/Cellar/libpng/1.6.34/include/libpng16 >>>> >> > >>>> >> > >>>> >> > libpng-config --ldflags >>>> >> > >>>> >> > -L/usr/local/Cellar/libpng/1.6.34/lib -lpng16 >>>> >> > >>>> >> > >>>> >> > Guess I'll do a manual install with Homebrew unless there is a >>>> better idea. >>>> >> > >>>> >> > >>>> >> > On Wednesday, November 20, 2019 at 2:38:21 PM UTC-8, Dima >>>> Pasechnik wrote: >>>> >> >> >>>> >> >> So you do have old (the current libpng there is 1.6.37, not >>>> 1.6.34) >>>> >> >> Homebrew (?) installed (or perhaps not 100% de-installed), with >>>> libpng >>>> >> >> installed in Cellar only (?) for some reason. It causes trouble, >>>> as it >>>> >> >> is (as well as matplolib's handling >>>> >> >> of dependence discovery) is a bit broken. >>>> >> >> >>>> >> >> more our of curiosity, what are outputs of >>>> >> >> >>>> >> >> libpng-config --cflags >>>> >> >> libpng-config --ldflags >>>> >> >> >>>> >> >> If you don't need it libpng, you can just remove it. >>>> >> >> Else, you can remove it only for the Sage building. >>>> >> >> Or you can update your Homebrew, properly install libpng etc... >>>> >> >> >>>> >> >> >>>> >> >> On Wed, Nov 20, 2019 at 10:17 PM Paul Masson < >>>> [email protected]> wrote: >>>> >> >> > >>>> >> >> > For the first: >>>> >> >> > >>>> >> >> > otool -L local/lib/libgd.dylib >>>> >> >> > >>>> >> >> > local/lib/libgd.dylib: >>>> >> >> > >>>> >> >> > /Users/Masson/Downloads/GitHub/sage/local/lib/libgd.3.dylib >>>> (compatibility version 4.0.0, current version 4.0.0) >>>> >> >> > >>>> >> >> > /usr/lib/libiconv.2.dylib (compatibility version 7.0.0, current >>>> version 7.0.0) >>>> >> >> > >>>> >> >> > /usr/lib/libz.1.dylib (compatibility version 1.0.0, current >>>> version 1.2.11) >>>> >> >> > >>>> >> >> > >>>> /Users/Masson/Downloads/GitHub/sage/local/lib/libfreetype.6.dylib >>>> (compatibility version 23.0.0, current version 23.1.0) >>>> >> >> > >>>> >> >> > /usr/lib/libSystem.B.dylib (compatibility version 1.0.0, >>>> current version 1281.0.0) >>>> >> >> > >>>> >> >> > >>>> >> >> > And the second: >>>> >> >> > >>>> >> >> > >>>> >> >> > libpng-config --prefix >>>> >> >> > >>>> >> >> > /usr/local/Cellar/libpng/1.6.34 >>>> >> >> > >>>> >> >> > >>>> >> >> > On Wednesday, November 20, 2019 at 2:12:17 PM UTC-8, Dima >>>> Pasechnik wrote: >>>> >> >> >> >>>> >> >> >> also please post the output of >>>> >> >> >> >>>> >> >> >> libpng-config --prefix >>>> >> >> >> >>>> >> >> >> >>>> >> >> >> On Wed, Nov 20, 2019 at 10:05 PM Dima Pasechnik < >>>> [email protected]> wrote: >>>> >> >> >> > >>>> >> >> >> > hmm, strange. So it is found somewhere, not clear where. One >>>> way to >>>> >> >> >> > see is to check linking of libgd >>>> >> >> >> > >>>> >> >> >> > otool -L local/lib/libgd.dylib >>>> >> >> >> > >>>> >> >> >> > or perhaps in >>>> >> >> >> > >>>> >> >> >> > logs/pkgs/libgd* >>>> >> >> >> > >>>> >> >> >> > On Wed, Nov 20, 2019 at 9:49 PM Paul Masson < >>>> [email protected]> wrote: >>>> >> >> >> > > >>>> >> >> >> > > Here it is >>>> >> >> >> > > >>>> >> >> >> > > On Wednesday, November 20, 2019 at 1:39:19 PM UTC-8, Dima >>>> Pasechnik wrote: >>>> >> >> >> > >> >>>> >> >> >> > >> please post config.log >>>> >> >> >> > >> you have a mysterious copy of libpng somewhere... >>>> >> >> >> > >> >>>> >> >> >> > >> On Wed, Nov 20, 2019 at 9:36 PM Paul Masson < >>>> [email protected]> wrote: >>>> >> >> >> > >> > >>>> >> >> >> > >> > I do not have anaconda installed. Building from the git >>>> repository with no package managers. >>>> >> >> >> > >> > >>>> >> >> >> > >> > On Wednesday, November 20, 2019 at 1:31:10 PM UTC-8, >>>> Dima Pasechnik wrote: >>>> >> >> >> > >> >> >>>> >> >> >> > >> >> Probably anaconda conflict, do you have pkg-config >>>> installed, so that >>>> >> >> >> > >> >> >>>> >> >> >> > >> >> pkg-config --libs libpng >>>> >> >> >> > >> >> >>>> >> >> >> > >> >> outputs a path to anaconda? >>>> >> >> >> > >> >> >>>> >> >> >> > >> >> I guess matplotlib does not use pkg-config, so it >>>> can't find libpng in >>>> >> >> >> > >> >> an unusual location. >>>> >> >> >> > >> >> >>>> >> >> >> > >> >> On Wed, Nov 20, 2019 at 7:57 PM Paul Masson < >>>> [email protected]> wrote: >>>> >> >> >> > >> >> > >>>> >> >> >> > >> >> > On macOS 10.15.1, after "make distclean" I get >>>> >> >> >> > >> >> > >>>> >> >> >> > >> >> > The following package(s) may have failed to build >>>> (not necessarily >>>> >> >> >> > >> >> > >>>> >> >> >> > >> >> > during this run of 'make all-start'): >>>> >> >> >> > >> >> > >>>> >> >> >> > >> >> > >>>> >> >> >> > >> >> > * package: matplotlib-2.2.4.p0 >>>> >> >> >> > >> >> > >>>> >> >> >> > >> >> > log file: >>>> /Users/Masson/Downloads/GitHub/sage/logs/pkgs/matplotlib-2.2.4.p0.log >>>> >> >> >> > >> >> > >>>> >> >> >> > >> >> > build directory: >>>> /Users/Masson/Downloads/GitHub/sage/local/var/tmp/sage/build/matplotlib-2.2.4.p0 >>>> >>>> >>>> >> >> >> > >> >> > >>>> >> >> >> > >> >> > >>>> >> >> >> > >> >> > In the attached log, it says that libpng is not >>>> installed. I've never seen this error before, but I didn't think I'd have >>>> to install anything manually that was mandatory for the build. Thoughts? >>>> >> >> >> > >> >> > >>>> >> >> >> > >> >> > >>>> >> >> >> > >> >> > On Sunday, November 17, 2019 at 4:42:15 PM UTC-8, >>>> Volker Braun wrote: >>>> >> >> >> > >> >> >> >>>> >> >> >> > >> >> >> As always, you can get the latest beta version from >>>> the "develop" git branch. Alternatively, the self-contained source tarball >>>> is at http://www.sagemath.org/download-latest.html >>>> >> >> >> > >> >> >> >>>> >> >> >> > >> >> >> This is the first Python 3 beta, so incremental >>>> builds probably aren't going to work. Use "make distclean" or the >>>> appropriate configure flag to keep using Python 2 (noooo) >>>> >> >> >> > >> >> >> >>>> >> >> >> > >> >> >> >>>> >> >> >> > >> >> >> 232ae1bddc (tag: 9.0.beta6) Updated SageMath >>>> version to 9.0.beta6 >>>> >> >> >> > >> >> >> 026a216fa4 Trac #28104: Openblas build failure >>>> >> >> >> > >> >> >> 1e7795c1f2 Trac #28726: py3: cysignals fails its >>>> test suite >>>> >> >> >> > >> >> >> c53e82c055 Trac #28734: a few miscellanous details >>>> >> >> >> > >> >> >> a55e518077 Trac #28721: Document sage-dist-helpers >>>> >> >> >> > >> >> >> c87212c10b Trac #28715: remove some deprecated >>>> stuff in combinat after 23319 and 15597 >>>> >> >> >> > >> >> >> 2adaf8bc28 Trac #28686: Demote some optional >>>> packages to experimental >>>> >> >> >> > >> >> >> 381ddacd15 Trac #15077: Inconsistency in polynomial >>>> .reverse(n) >>>> >> >> >> > >> >> >> af4d15a8e1 Trac #28725: fix lawrence_extension and >>>> lawrence_polytope for backend field >>>> >> >> >> > >> >> >> a8e457e711 Trac #28719: remove deprecated function >>>> aliases >>>> >> >> >> > >> >> >> e46f2a5383 Trac #28717: some cleanup in __init__ >>>> files >>>> >> >> >> > >> >> >> 9101b22829 Trac #28709: Python executable in >>>> doctest/control.py >>>> >> >> >> > >> >> >> 6544bf8e40 Trac #28707: More control on the >>>> numerical ODE solver for integrated curves and geodesics >>>> >> >> >> > >> >> >> 50034a0ae9 Trac #28652: Rework the constructor of >>>> PermutationGroupElement >>>> >> >> >> > >> >> >> edc1360355 Trac #28606: CombinatorialPolyhedron: >>>> unbounded -> is_bounded >>>> >> >> >> > >> >> >> 3a35510660 Trac #28605: CombinatorialPolyhedron: >>>> replace attributes by methods >>>> >> >> >> > >> >> >> 133435b38d Trac #27302: Cubic Braid Groups >>>> >> >> >> > >> >> >> 8d41eaeed4 Trac #27089: >>>> CPLEXBackend.solver_parameter: Handle CPX_PARAMTYPE_LONG >>>> >> >> >> > >> >> >> eb5d0df260 Trac #25183: Bug in the associahedron >>>> object >>>> >> >> >> > >> >> >> 94b6adde10 Trac #24108: >>>> quadratic_form_from_invariants() >>>> >> >> >> > >> >> >> e9e14ad864 Trac #28730: cvxopt fails its test suite >>>> >> >> >> > >> >> >> a9146cd62b Trac #28714: clarify docstring of >>>> plot.colors.float_to_integer >>>> >> >> >> > >> >> >> f94189b015 Trac #28708: 5 cplex doctests failed in >>>> cplex_backend.pyx >>>> >> >> >> > >> >> >> 92a0045e31 Trac #28700: Fixing missing optional >>>> flag in base_QQ.py >>>> >> >> >> > >> >> >> 6c5f93dfb4 Trac #28689: py3: fix doctests with cbc >>>> >> >> >> > >> >> >> 3dd1d44175 Trac #28660: Make Python 3 the default >>>> >> >> >> > >> >> >> 0a86718461 Trac #28382: Correct types in cplex >>>> backend >>>> >> >> >> > >> >> >> 82b36b2ead Trac #27761: DES Block Cipher >>>> >> >> >> > >> >> >> 93db4cc438 (tag: 9.0.beta5) Updated SageMath >>>> version to 9.0.beta5 >>>> >> >> >> > >> >> >> >>>> >> >> >> > >> >> > -- >>>> >> >> >> > >> >> > You received this message because you are subscribed >>>> to the Google Groups "sage-release" group. >>>> >> >> >> > >> >> > To unsubscribe from this group and stop receiving >>>> emails from it, send an email to [email protected]. >>>> >> >> >> > >> >> > To view this discussion on the web visit >>>> https://groups.google.com/d/msgid/sage-release/b5a898e9-75d3-4bfe-a276-c273df8844d8%40googlegroups.com. >>>> >>>> >>>> >> >> >> > >> > >>>> >> >> >> > >> > -- >>>> >> >> >> > >> > You received this message because you are subscribed to >>>> the Google Groups "sage-release" group. >>>> >> >> >> > >> > To unsubscribe from this group and stop receiving >>>> emails from it, send an email to [email protected]. >>>> >> >> >> > >> > To view this discussion on the web visit >>>> https://groups.google.com/d/msgid/sage-release/8aa8a9a8-1cb5-4d54-bb08-188f02dfd78f%40googlegroups.com. >>>> >>>> >>>> >> >> >> > > >>>> >> >> >> > > -- >>>> >> >> >> > > You received this message because you are subscribed to >>>> the Google Groups "sage-release" group. >>>> >> >> >> > > To unsubscribe from this group and stop receiving emails >>>> from it, send an email to [email protected]. >>>> >> >> >> > > To view this discussion on the web visit >>>> https://groups.google.com/d/msgid/sage-release/4739aecd-5670-425c-b30b-2158db34bebf%40googlegroups.com. >>>> >>>> >>>> >> >> > >>>> >> >> > -- >>>> >> >> > You received this message because you are subscribed to the >>>> Google Groups "sage-release" group. >>>> >> >> > To unsubscribe from this group and stop receiving emails from >>>> it, send an email to [email protected]. >>>> >> >> > To view this discussion on the web visit >>>> https://groups.google.com/d/msgid/sage-release/dfc961b1-4cf5-4922-b77d-16b158a99838%40googlegroups.com. >>>> >>>> >>>> >> > >>>> >> > -- >>>> >> > You received this message because you are subscribed to the Google >>>> Groups "sage-release" group. >>>> >> > To unsubscribe from this group and stop receiving emails from it, >>>> send an email to [email protected]. >>>> >> > To view this discussion on the web visit >>>> https://groups.google.com/d/msgid/sage-release/0d8abc59-9cc4-48ec-a539-082e2dd0fed7%40googlegroups.com. >>>> >>>> >>>> > >>>> > -- >>>> > You received this message because you are subscribed to the Google >>>> Groups "sage-release" group. >>>> > To unsubscribe from this group and stop receiving emails from it, >>>> send an email to [email protected]. >>>> > To view this discussion on the web visit >>>> https://groups.google.com/d/msgid/sage-release/fe421ed4-2a7c-4fb0-ae42-e2b2015b7bf1%40googlegroups.com. >>>> >>>> >>>> >>> -- >>> You received this message because you are subscribed to the Google >>> Groups "sage-release" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to [email protected]. >>> To view this discussion on the web visit >>> https://groups.google.com/d/msgid/sage-release/8f2476ef-43cd-496b-be31-52fc133f0863%40googlegroups.com >>> >>> <https://groups.google.com/d/msgid/sage-release/8f2476ef-43cd-496b-be31-52fc133f0863%40googlegroups.com?utm_medium=email&utm_source=footer> >>> . >>> >> -- You received this message because you are subscribed to the Google Groups "sage-release" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/0de7806e-e274-4976-9824-bb3729cfaa9a%40googlegroups.com.
