in view of problems with building gcc/gfortran on osx 10.15 and latest
Xcode, one must wonder what Homebrew does to make it work.
Do they patch system headers, or something like that?


On Mon, 4 Nov 2019 at 07:37, Andrew <andrew.mat...@gmail.com> wrote:

> Great, thanks John. I have essentially the same version of brew:
> Homebrew 2.1.15
> Homebrew/homebrew-core (git revision 89bf3; last commit 2019-11-04)
> Homebrew/homebrew-cask (git revision e317e; last commit 2019-11-04)
>
> but I have many more brew packages installed. I reinstalled all of the
> brew packages last week but I have just uninstalled all of the brew
> packages. I will now add everything back in one at a time, starting with
> things on your list, and see if I can compile sage. I'll report back what I
> find.
>
> Andrew
>
>
> On Monday, 4 November 2019 16:35:55 UTC+11, John H Palmieri wrote:
>>
>> I am using the same versions of OS and Xcode as you. I am now doing this
>> with a third computer. This one had an old Xcode and some homebrew stuff
>> already installed, and it gave me some problems. I ended up uninstalling
>> and then reinstalling various parts of homebrew, and Sage now builds for
>> me. (I had also first installed some things by hand, like gfortran many
>> months ago. So I had to delete things like that when doing the homebrew
>> installation.)
>>
>> $ brew list
>> autoconf        libdvdcss       p7zip
>> automake        libevent        pari
>> boost           libffi          pcre
>> cabextract      libidn2         pkg-config
>> cimg            libmpc          python
>> cmake           libomp          readline
>> emacs-mac       libpng          shared-mime-info
>> gcc             libtasn1        sqlite
>> gdbm            libtool         texinfo
>> gettext         libunistring    unbound
>> glib            mpfi            unrar
>> gmp             mpfr            wine
>> gnutls          nettle          winetricks
>> gpatch          ntl             x265
>> icu4c           openexr         xz
>> ilmbase         openssl         yasm
>> isl             openssl@1.1
>> libde265        p11-kit
>>
>>
>> $ brew --version
>> Homebrew 2.1.15
>> Homebrew/homebrew-core (git revision d191f; last commit 2019-11-03)
>> Homebrew/homebrew-cask (git revision cbd79; last commit 2019-11-03)
>>
>>
>>
>>
>> On Sunday, November 3, 2019 at 8:53:57 PM UTC-8, Andrew wrote:
>>>
>>> I uninstalled xcode and brew's gcc and then reinstall them both but I
>>> still the the same failure with givaro.
>>> I don't know if this is related, but even though I installed ntl using
>>> brew it is not used when I compile sage as the log file reports:
>>>
>>> checking NTL/ZZ.h usability... no
>>> checking NTL/ZZ.h presence... yes
>>> configure: WARNING: NTL/ZZ.h: present but cannot be compiled
>>> configure: WARNING: NTL/ZZ.h:     check for missing prerequisite headers?
>>> configure: WARNING: NTL/ZZ.h: see the Autoconf documentation
>>> configure: WARNING: NTL/ZZ.h:     section "Present But Cannot Be
>>> Compiled"
>>> configure: WARNING: NTL/ZZ.h: proceeding with the compiler's result
>>> configure: WARNING:     ## ------------------------------------------ ##
>>> configure: WARNING:     ## Report this to sage-...@googlegroups.com ##
>>> configure: WARNING:     ## ------------------------------------------ ##
>>> checking for NTL/ZZ.h... no
>>> checking whether we can link a program using NTL... no
>>> checking NTL version >= 10.3... 11.4.1
>>> using Sage's ntl SPKG
>>> configure: === checking whether to install the flint SPKG ===
>>> checking installing mpfr or ntl? ... yes; install flint as well
>>> using Sage's flint SPKG
>>> configure: === checking whether to install the arb SPKG ===
>>> checking installing flint? ... yes; install arb as well
>>> configure: === checking whether to install the bzip2 SPKG ===
>>>
>>> That sage wants compile flint and arb is expected because I try to
>>> install them with brew I get errors like:
>>>
>>> ==> make
>>> Last 15 lines from /Users/andrew/Library/Logs/Homebrew/flint/02.make:
>>>       ~~^
>>> /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/../include/c++/v1/cmath:324:9:
>>> error: no member named 'islessgreater' in the global namespace
>>> using ::islessgreater;
>>>       ~~^
>>> /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/../include/c++/v1/cmath:325:9:
>>> error: no member named 'isunordered' in the global namespace
>>> using ::isunordered;
>>>       ~~^
>>> /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/../include/c++/v1/cmath:326:9:
>>> error: no member named 'isunordered' in the global namespace
>>> using ::isunordered;
>>>       ~~^
>>> 13 errors generated.
>>> make[2]: *** [build/interfaces/NTL-interface.lo] Error 1
>>>
>>> These are similar to the errors that I get with givaro when compiling
>>> sage.
>>>
>>> John: what version of xcode and the command line tools are using and if
>>> you first do
>>>
>>> make distclean
>>>
>>> are you still able to compile sage? As I reinstalled xcode from the
>>> appsrtore I am using the latest versions av I am using:
>>>
>>> ProductName:    Mac OS X
>>> ProductVersion:    10.15.1
>>> BuildVersion:    19B88
>>> Xcode 11.2
>>> Build version 11B52
>>>
>>> Andrew
>>>
>>>
>>>
>>>
>>> On Saturday, 2 November 2019 05:45:08 UTC+11, Andrew wrote:
>>>>
>>>> Thanks John. I'll give this a try on Monday.
>>>> Andrew
>>>>
>>>> On Saturday, 2 November 2019 04:20:08 UTC+11, John H Palmieri wrote:
>>>>>
>>>>> I just upgraded a different machine to Catalina. This one didn't have
>>>>> Xcode or homebrew installed beforehand, so I installed Xcode, its
>>>>> command-line tools, and homebrew's gcc. Then I built Sage and it worked. I
>>>>> have now installed a bunch of other homebrew packages relevant to Sage, 
>>>>> and
>>>>> the build is going fine. (I had to log out of the machine, so I had to
>>>>> interrupt the build, but it had made it past givaro and ntl, which were
>>>>> sticking points for another user in this thread.)
>>>>>
>>>>> Thinking back on the other machine I upgraded, that one was also
>>>>> relatively "clean": no old Xcode or old homebrew stuff. So for those who
>>>>> are having problems, I wonder if uninstalling Xcode first would help. 
>>>>> Maybe
>>>>> there are some other old libraries which should be removed before
>>>>> installing the new Xcode.
>>>>>
>>>>>   John
>>>>>
>>>>>
>>>>> On Monday, October 14, 2019 at 10:43:09 AM UTC-7, Volker Braun wrote:
>>>>>>
>>>>>> Means: I successfully compiled Sage 8.9 on OSX Catalina with Xcode
>>>>>> 11.0
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Sunday, October 13, 2019 at 9:24:44 PM UTC+2, Samuel Lelievre
>>>>>> wrote:
>>>>>>>
>>>>>>> Sun 2019-10-13 08:26:52 UTC, Volker Braun:
>>>>>>>>
>>>>>>>> I just upgraded the OSX buildbot and Sage works
>>>>>>>>
>>>>>>> in case anybody is wondering. This is the first time in years
>>>>>>>>
>>>>>>> that the latest OSX release doesn't introduce major toolchain
>>>>>>>>
>>>>>>> issues and/or scrambles the filesystem, hope this starts a trend ;-)
>>>>>>>>
>>>>>>>
>>>>>>> That's good to hear, and there were people wondering.
>>>>>>> See for instance this question on Ask Sage about SageMath
>>>>>>> and macOS 10.15 "Catalina":
>>>>>>>
>>>>>>> - How to unlock SageMath 8.8 on macOS 10.15 Catalina?
>>>>>>>   https://ask.sagemath.org/question/48293
>>>>>>>
>>>>>>> By "Sage works" do you mean any of the following?
>>>>>>>
>>>>>>> - there are / there will be binaries for macOS 10.15 "Catalina"
>>>>>>> - the SageMath 8.9 binaries for macOS 10.14 "Mojave"
>>>>>>>   also work for macOS 10.15 "Catalina"
>>>>>>> - on macOS 10.15 "Catalina" building SageMath 8.9 from
>>>>>>>   source gives a version that works and is not "locked"
>>>>>>>
>>>>>>>
>>>>>>> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/sage-devel/2c4485e7-1310-429c-9098-b0c3e91fb16e%40googlegroups.com
> <https://groups.google.com/d/msgid/sage-devel/2c4485e7-1310-429c-9098-b0c3e91fb16e%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAAWYfq1orHOLbL2%3DD9K4ER0sa5nAAcx2jLhXnV6EcoCtJc-COQ%40mail.gmail.com.

Reply via email to