[sage-devel] Re: use Sage!
["Followup-To:" header set to gmane.comp.mathematics.sage.devel.] On 2013-05-06, rjf wrote: > --=_Part_1579_8214636.1367802043271 > Content-Type: text/plain; charset=ISO-8859-1 > > > > On Thursday, May 2, 2013 1:27:30 PM UTC-7, William wrote: >> >> Hi Sage-Developers, >> >> There is a big series of small books about R that Springer publishes: >> >> http://www.springer.com/series/6991?detailsPage=titles >> >> The editorial director of that series at Springer just talked with me >> on the phone for a while, and he says these are among "Springers best >> selling books"; moreover, he believes they have a major impact on >> making R a really viable platform for computational statistics. >> >> >> Am I the only one who finds this implausible? > Here's an article on R > http://www.nytimes.com/2009/01/07/technology/business-computing/07program.html?pagewanted=all > and it doesn't mention Springer. > > How many copies do you need to be among the best-sellers for Springer? > Maybe 3,000? > How many copies of R are there, and how many people actually buy a book to > use free software? R is often used to teach statistics at universities, so there may be hundreds of thousands, if not millions, of installations around. Here are the Amazon data on "Introductory Time Series with R (Use R!)" published by Springer (272 pages, 45$) Amazon Best Sellers Rank: #72,412 in Books (See Top 100 in Books) Statistical or "R by Example (Use R!)" (350+ pages, 44$) Amazon Best Sellers Rank: #72,627 in Books which looks good to me. But notice that "The Art of R Programming: A Tour of Statistical Software Design" published by No Starch Press fares much better: (400pages, 25$) Amazon Best Sellers Rank: #9,454 in Books (See Top 100 in Books) Mathematics Note that "R Cookbook (O'Reilly Cookbooks)" (400+ pages, 25$) has even better scores: Amazon Best Sellers Rank: #8,835 in Books (See Top 100 in Books) IMHO this data shows that Springer wants to charge much more for its books (and its sales are perhaps 10 times less, at least on Amazon, see http://www.fonerbooks.com/surfing.htm), for reasons that do not look convincing to me. Looks like Springer behaves in this market as Apple in personal computers, charging more for reasons not always clear... Dima -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Sage 5.9 Fails to build in conway_polynomials
Hi List, I tried to upgrade sage via sage -upgrade but it fails in conway_polynomials. For more Details see attachment. On: Linux neo 3.2.0-41-generic #65-Ubuntu SMP Wed Apr 10 18:23:37 UTC 2013 i686 i686 i386 GNU/Linux Ubuntu 12.04 bg Johannes -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out. Testing mirrors... [15] sage.mirror.garr.it 140 [ms] [8 ] mirror.switch.ch 162 [ms] [4 ] ftp.sh.cvut.cz 174 [ms] [19] www.mirrorservice.org187 [ms] [17] sunsite.rediris.es 205 [ms] [18] www-ftp.lip6.fr 209 [ms] [10] mirror.yandex.ru 242 [ms] [16] servingzone.com 252 [ms] [12] mirrors.fe.up.pt 260 [ms] [7 ] mirror.clibre.uqam.ca367 [ms] [6 ] jambu.spms.ntu.edu.sg539 [ms] [3 ] ftp.riken.jp 678 [ms] [11] mirror.yongbok.net 803 [ms] [2 ] ftp.kaist.ac.kr 818 [ms] [9 ] mirror.unej.ac.id878 [ms] [5 ] ftp.tsukuba.wide.ad.jp 1162 [ms] [13] mirrors.tuna.tsinghua.edu.cn1411 [ms] [14] sage.Igor.onlineDirect.bg 5175 [ms] [1 ] boxen.math.washington.edu 5612 [ms] Automatically selected server mirror.switch.ch (http://mirror.switch.ch/mirror/sagemath/). Downloading packages from 'http://mirror.switch.ch/mirror/sagemath//spkg'. Reading package lists... Done! The following packages will be upgraded: extcode-5.9 r-2.15.2.p2 sagetex-2.3.4 ** WARNING: This is a source-based upgrade, which could take hours, ** fail, and render your Sage install useless!! Do you want to continue [y/N]? http://mirror.switch.ch/mirror/sagemath//spkg/standard/extcode-5.9.spkg --> extcode-5.9.spkg 'extcode-5.9.spkg' is already present. http://mirror.switch.ch/mirror/sagemath//spkg/standard/r-2.15.2.p2.spkg --> r-2.15.2.p2.spkg 'r-2.15.2.p2.spkg' is already present. http://mirror.switch.ch/mirror/sagemath//spkg/standard/sagetex-2.3.4.spkg --> sagetex-2.3.4.spkg 'sagetex-2.3.4.spkg' is already present. http://mirror.switch.ch/mirror/sagemath//spkg/standard/VERSION.txt --> VERSION.txt [.] http://mirror.switch.ch/mirror/sagemath//spkg/install --> install [...] http://mirror.switch.ch/mirror/sagemath//spkg/standard/deps --> deps [...nothing changed nothing changed ] make base env SAGE_BUILD_TOOLCHAIN=yes make toolchain make toolchain-deps make[1]: Betrete Verzeichnis '/usr/lib/sagemath/spkg' make /usr/lib/sagemath/spkg/installed/zlib-1.2.6.p0 make /usr/lib/sagemath/spkg/installed/mpir-2.6.0.p2 make /usr/lib/sagemath/spkg/installed/mpfr-3.1.0.p2 make /usr/lib/sagemath/spkg/installed/mpc-1.0.p0 make[1]: Verlasse Verzeichnis '/usr/lib/sagemath/spkg' make all-sage make[1]: Betrete Verzeichnis '/usr/lib/sagemath/spkg' /usr/lib/sagemath/spkg/pipestatus "sage-spkg ${SAGE_SPKG_OPTS} conway_polynomials-0.4 2>&1" "tee -a /usr/lib/sagemath/logs/pkgs/conway_polynomials-0.4.log" make[1]: Verlasse Verzeichnis '/usr/lib/sagemath/spkg' *** ALL ENVIRONMENT VARIABLES BEFORE BUILD: *** AR=ar AS=as CCACHE_BASEDIR=/usr/lib/sagemath CC=gcc CHMOD=chmod CPATH=/usr/lib/sagemath/local/include:/usr/include/i386-linux-gnu CP=cp CPP=cpp CXXFLAGS= CXX=g++ DISPLAY=:0.0 DOT_SAGE=/home/j_schn14/.sage/ ECLDIR=/usr/lib/sagemath/local/lib/ecl/ EDITOR=emacs F77=gfortran F90=gfortran F95=gfortran GP_DATA_DIR=/usr/lib/sagemath/local/share/pari GPDOCDIR=/usr/lib/sagemath/local/share/pari/doc GPHELP=/usr/lib/sagemath/local/bin/gphelp HGENCODING=utf8 HGPLAIN=yes HOME=/home/j_schn14 IPYTHONDIR=/home/j_schn14/.sage//ipython-0.12 LANG=de_DE@euro LANGUAGE=de_DE:de:en_GB:en LC_ALL=de_DE.UTF-8 LDFLAGS= LD=ld LD_LIBRARY_PATH=/usr/lib/sagemath/local/lib:/usr/lib/sagemath/local/lib/R/lib LESSOPEN=| /usr/share/source-highlight/src-hilite-lesspipe.sh %s LESS= -R LIBRARY_PATH=/usr/lib/sagemath/local/lib:/usr/lib/i386-linux-gnu LN=ln LOGNAME=root MAIL=/var/mail/root MAKE=make MAXIMA_PREFIX=/usr/lib/sagemath/local MKDIR=mkdir MPLCONFIGDIR=/home/j_schn14/.sage//matplotlib-1.1.0 MV=mv PATH=/usr/lib/sagemath/spkg/bin:/usr/lib/sagemath/local/bin:/usr/lib/sagemath/spkg/bin:/usr/lib/sagemath/local/bin:/home/j_schn14/studium/ss11/infoPruefung/tanagra/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/opt/android-sdk/tools:/opt/android-sdk/platform-tools/ PKG_CONFIG_PATH=/usr/lib/sagemath/local/lib/pkgconfig PREREQ_OPTIONS=--disable-compiler-checks PWD=/usr/lib/sagemath/spkg PYTHONHOME=/usr/lib/sagemath/local PYTHONPATH=/usr/lib/sagemath/local RANLIB=ranlib __sage__= SAGE64=no SAGE_BROWSER=fi
[sage-devel] Re: Singular fails to build in Sage 5.9
Oh... I see. I am ignorant on what various flags mean. Let me summarize my problems and see if I get them right: 1) my gcc compiler confuses c and c++ programs. 2) I want to solve 1) by forcing sage to use the gcc-4.7.2. but the problem is that the copy of gcc that come with sage won't build from my system gcc... (see the my 3rd message in this discussion) Am I right? Any ideas on what else can I try? On Sunday, May 5, 2013 9:20:00 PM UTC-7, leif wrote: > > Pong wrote: > > Report: buliding Singular by itself seems to be completely fine. > > > > ./configure > > make -j4 > > results in > > > > Well, you'd have to (at least) configure with '--with-apint=gmp' as well > (cf. Sage's spkg-install script). > > > -leif > > > /usr/bin/install -c -s solve_IP > > /home/pong/Downloads/Singular-3-1-5/x86_64-Linux > > /usr/bin/install -c -s change_cost > > /home/pong/Downloads/Singular-3-1-5/x86_64-Linux > > /usr/bin/install -c -s toric_ideal > > /home/pong/Downloads/Singular-3-1-5/x86_64-Linux > > /usr/bin/install -c -s gen_test > > /home/pong/Downloads/Singular-3-1-5/x86_64-Linux > > /usr/bin/install -c -s LLL > /home/pong/Downloads/Singular-3-1-5/x86_64-Linux > > make[2]: Leaving directory > > `/home/pong/Downloads/Singular-3-1-5/IntegerProgramming' > > echo "* do not forget to install Singular-3-1-5-share.tar.gz" > > * do not forget to install Singular-3-1-5-share.tar.gz > > make[1]: Leaving directory `/home/pong/Downloads/Singular-3-1-5' > > > > and it gives me back the prompt. I'm not sure if this "test" is helpful > > in bugging what's going on. But perhaps it is not a Singular bug. > > > > On Sunday, May 5, 2013 4:31:10 PM UTC-7, Pong wrote: > > > > Well... it said "prerelease" > > gcc version 4.8.0 20130502 (prerelease) (GCC) > > > > I see now, obviously 4.8.0 is still a prerelease... > > > > Just out of curiousity, I have just downloaded a copy of > > Singular-3.1.5 from the offical site and see if it compiles on my > > machine. > > Will keep everyone posted. > > -- > () The ASCII Ribbon Campaign > /\ Help Cure HTML E-Mail > > -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: Singular fails to build in Sage 5.9
Pong wrote: Report: buliding Singular by itself seems to be completely fine. ./configure make -j4 results in Well, you'd have to (at least) configure with '--with-apint=gmp' as well (cf. Sage's spkg-install script). -leif /usr/bin/install -c -s solve_IP /home/pong/Downloads/Singular-3-1-5/x86_64-Linux /usr/bin/install -c -s change_cost /home/pong/Downloads/Singular-3-1-5/x86_64-Linux /usr/bin/install -c -s toric_ideal /home/pong/Downloads/Singular-3-1-5/x86_64-Linux /usr/bin/install -c -s gen_test /home/pong/Downloads/Singular-3-1-5/x86_64-Linux /usr/bin/install -c -s LLL /home/pong/Downloads/Singular-3-1-5/x86_64-Linux make[2]: Leaving directory `/home/pong/Downloads/Singular-3-1-5/IntegerProgramming' echo "* do not forget to install Singular-3-1-5-share.tar.gz" * do not forget to install Singular-3-1-5-share.tar.gz make[1]: Leaving directory `/home/pong/Downloads/Singular-3-1-5' and it gives me back the prompt. I'm not sure if this "test" is helpful in bugging what's going on. But perhaps it is not a Singular bug. On Sunday, May 5, 2013 4:31:10 PM UTC-7, Pong wrote: Well... it said "prerelease" gcc version 4.8.0 20130502 (prerelease) (GCC) I see now, obviously 4.8.0 is still a prerelease... Just out of curiousity, I have just downloaded a copy of Singular-3.1.5 from the offical site and see if it compiles on my machine. Will keep everyone posted. -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: use Sage!
On Thursday, May 2, 2013 1:27:30 PM UTC-7, William wrote: > > Hi Sage-Developers, > > There is a big series of small books about R that Springer publishes: > > http://www.springer.com/series/6991?detailsPage=titles > > The editorial director of that series at Springer just talked with me > on the phone for a while, and he says these are among "Springers best > selling books"; moreover, he believes they have a major impact on > making R a really viable platform for computational statistics. > > > Am I the only one who finds this implausible? Here's an article on R http://www.nytimes.com/2009/01/07/technology/business-computing/07program.html?pagewanted=all and it doesn't mention Springer. How many copies do you need to be among the best-sellers for Springer? Maybe 3,000? How many copies of R are there, and how many people actually buy a book to use free software? How many would buy a bunch at US$65 each? (Certainly some, but perhaps because the book teaches something other than documenting R. Like statistics, data analysis, etc. ) Regarding just "R" how good is the free documentation available here: http://www.r-project.org/ ? While I would not like to discourage anyone from writing documentation on anything, all the traditional motivations for publication seem to be lacking for a Springer series. Here are some of the traditional motivations. 1. The author hopes to make money from royalties. 2. The author hopes to become famous (or gain tenure) from the book. 3. There is a gap in the literature and the author and publisher wish to remedy this for non-financial reasons, perhaps political, moral, religious; making friends, influencing enemies. 4. Some kind of creative impulse. You may have some other ideas for motivation, but these come to mind. Insisting on Creative Commons means, it seems to me, that you would be working to make sure that the only people who might make money from sales would be Springer. Of course Springer would be biased in favor of ANY way of adding to their list, as long it has some prospect of making some money for them, and they are used to sales of very short runs. It falls short of their activities with journals -- the author surrenders all control in return for no royalty, and Springer charges big bucks for library subscriptions -- but it is something. An intermediate stance, where an author writes a good/popular book and gets paid royalties on sales, is a common although rarely lucrative (for the author) choice. I personally doubt that combinatorics with Sage would be a big seller, but if you have already written it and Springer would like to publish it, and it doesn't interfere with your ownership, it seems like you have nothing to lose. But make sure you don't give away something unintentionally, whether it is right or your valuable time and energy. RJF -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
Re: [sage-devel] Re: Singular fails to build in Sage 5.9
Report: buliding Singular by itself seems to be completely fine. ./configure make -j4 results in /usr/bin/install -c -s solve_IP /home/pong/Downloads/Singular-3-1-5/x86_64-Linux /usr/bin/install -c -s change_cost /home/pong/Downloads/Singular-3-1-5/x86_64-Linux /usr/bin/install -c -s toric_ideal /home/pong/Downloads/Singular-3-1-5/x86_64-Linux /usr/bin/install -c -s gen_test /home/pong/Downloads/Singular-3-1-5/x86_64-Linux /usr/bin/install -c -s LLL /home/pong/Downloads/Singular-3-1-5/x86_64-Linux make[2]: Leaving directory `/home/pong/Downloads/Singular-3-1-5/IntegerProgramming' echo "* do not forget to install Singular-3-1-5-share.tar.gz" * do not forget to install Singular-3-1-5-share.tar.gz make[1]: Leaving directory `/home/pong/Downloads/Singular-3-1-5' and it gives me back the prompt. I'm not sure if this "test" is helpful in bugging what's going on. But perhaps it is not a Singular bug. On Sunday, May 5, 2013 4:31:10 PM UTC-7, Pong wrote: > > Well... it said "prerelease" > gcc version 4.8.0 20130502 (prerelease) (GCC) > > I see now, obviously 4.8.0 is still a prerelease... > > Just out of curiousity, I have just downloaded a copy of Singular-3.1.5 > from the offical site and see if it compiles on my machine. > Will keep everyone posted. > > > On Sunday, May 5, 2013 4:14:05 PM UTC-7, François wrote: >> >> On Sun, 05 May 2013 16:07:11 Pong wrote: >> > Nope. I got the exact seem error. See the lastest log attached. >> > >> > One more thing that puzzled me is that the arch users just got an >> update >> > (including my machine) on gcc. So it's not the pre-release any more. >> > ~ [i]> pacman -Qi gcc >> > Name : gcc >> > Version: 4.8.0-4 >> > Description: The GNU Compiler Collection - C and C++ frontends >> > Architecture : x86_64 >> > >> And if you do "gcc --version" what does it says? >> >> Francois >> > -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
Re: [sage-devel] Re: Singular fails to build in Sage 5.9
Well... it said "prerelease" gcc version 4.8.0 20130502 (prerelease) (GCC) I see now, obviously 4.8.0 is still a prerelease... Just out of curiousity, I have just downloaded a copy of Singular-3.1.5 from the offical site and see if it compiles on my machine. Will keep everyone posted. On Sunday, May 5, 2013 4:14:05 PM UTC-7, François wrote: > > On Sun, 05 May 2013 16:07:11 Pong wrote: > > Nope. I got the exact seem error. See the lastest log attached. > > > > One more thing that puzzled me is that the arch users just got an update > > (including my machine) on gcc. So it's not the pre-release any more. > > ~ [i]> pacman -Qi gcc > > Name : gcc > > Version: 4.8.0-4 > > Description: The GNU Compiler Collection - C and C++ frontends > > Architecture : x86_64 > > > And if you do "gcc --version" what does it says? > > Francois > -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
Re: [sage-devel] Re: Singular fails to build in Sage 5.9
On Sun, 05 May 2013 16:07:11 Pong wrote: > Nope. I got the exact seem error. See the lastest log attached. > > One more thing that puzzled me is that the arch users just got an update > (including my machine) on gcc. So it's not the pre-release any more. > ~ [i]> pacman -Qi gcc > Name : gcc > Version: 4.8.0-4 > Description: The GNU Compiler Collection - C and C++ frontends > Architecture : x86_64 > And if you do "gcc --version" what does it says? Francois -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: Singular fails to build in Sage 5.9
Pong wrote: Hi leif, thank you for looking into this issue as well. Attached is my cf_gmp.h after a failed build attempt with sage -f singular-3-1-5.p7 Well, that version doesn't include bits/c++config.h. (It doesn't have as many lines as the previous one where the error occurred, and at first glance looks pretty sane.) Do you get a different error now? -leif On Sunday, May 5, 2013 12:02:26 PM UTC-7, leif wrote: The problem is not that a C compiler is used to compile weight0.c, but that cf_gmp.h includes in the first place. (Although bits/c++config.h is valid C code as well, it's not in the [C++] include path if you compile a file as C code which includes it.) But it seems ArchLinux's broken gcc/g++ behaviour triggers the problem. Pong, could you upload/attach your cf_gmp.h? (Should be located in $SAGE_ROOT/spkg/build/singular-3-1-5.p7/src/factory/, but only after a failed build or if you use 'sage -f -s ...', or alternatively set SAGE_KEEP_BUILT_SPKGS=yes.) I'll take a look at the full Singular build log you attached earlier. -leif -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: Singular fails to build in Sage 5.9
leif wrote: The problem is not that a C compiler is used to compile weight0.c, but that cf_gmp.h includes in the first place. (Although bits/c++config.h is valid C code as well, it's not in the [C++] include path if you compile a file as C code which includes it.) But it seems ArchLinux's broken gcc/g++ behaviour triggers the problem. Though it might still be a genuine Singular bug. Pong, could you upload/attach your cf_gmp.h? (Should be located in $SAGE_ROOT/spkg/build/singular-3-1-5.p7/src/factory/, but only after a failed build or if you use 'sage -f -s ...', or alternatively set SAGE_KEEP_BUILT_SPKGS=yes.) I'll take a look at the full Singular build log you attached earlier. The relevant lines seem to be these: echo "#!/bin/sh" >gen_cf_gmp.sh echo "GMP_H_T=\"`g++ -Wall -fno-implicit-templates -I. -I.. -I. -I/home/pong/sage-5.9/local -I/home/pong/sage-5.9/local/include -DHAVE_CONFIG_H -I/home/pong/sage-5.9/local/include -I/home/pong/sage-5.9/local/include -O2 -g -fPIC -M gen_cf_gmp.cc|grep gmp.h` \"" >>gen_cf_gmp.sh cat gen_cf_gmp.template >>gen_cf_gmp.sh /bin/sh gen_cf_gmp.sh generating cf_gmp.h from /home/pong/sage-5.9/local/include/gmp.h /usr/include/c++/4.8.0/iosfwd (Note the iosfwd.) -leif -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: Singular fails to build in Sage 5.9
Francois Bissey wrote: On 05/05/13 21:12, Wai Yan Pong wrote: Hum I see what you are saying. Well my case is a bit tricky but I think I'm not alone. I am using Archlinux, so there is not g++ separately. And gcc decides whether the program as a C or C++ program in the following way. |gcc file.C # uppercase .C (or .cpp) will process as C++ gcc file.c # lowercase .c will process as C| see http://unix.stackexchange.com/questions/34654/c-compile-in-arch-linux-with-no-g-package-in-repositories for details. So the question is what can be changed so that the gcc in my machine will interpret compile weight0.c as a C++ program... I wonder why I didn't encounter such a program earlier... perhaps it's the pre-release reversion that causes that problem. Also can I use force it to use the gcc (I suppose gcc-4.7.2.p1.spkg is the one) that comes with sage to compile? Yes it looks tricky. It could very well be a regression as it compile with the C compiler and not the C++ compiler on my box with gcc 4.6.3 (gentoo). That being said gcc is very permissive and one day someone may decide to cut that one off. Would you believe that 2 days ago I accidentally configured octave with CXX=gcc-4.7.1 and I almost completely compiled it before being shot down by such a problem [octave is mostly c++]. The problem is not that a C compiler is used to compile weight0.c, but that cf_gmp.h includes in the first place. (Although bits/c++config.h is valid C code as well, it's not in the [C++] include path if you compile a file as C code which includes it.) But it seems ArchLinux's broken gcc/g++ behaviour triggers the problem. Pong, could you upload/attach your cf_gmp.h? (Should be located in $SAGE_ROOT/spkg/build/singular-3-1-5.p7/src/factory/, but only after a failed build or if you use 'sage -f -s ...', or alternatively set SAGE_KEEP_BUILT_SPKGS=yes.) I'll take a look at the full Singular build log you attached earlier. -leif -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
Re: [sage-devel] Singular fails to build in Sage 5.9
On 05/05/13 21:12, Wai Yan Pong wrote: > Hum I see what you are saying. Well my case is a bit tricky but I think > I'm not alone. I am using Archlinux, so there is not g++ separately. And > gcc decides whether the program as a C or C++ program in the following way. > > |gcc file.C # uppercase .C (or .cpp) will process as C++ > gcc file.c # lowercase .c will process as C| > > see > http://unix.stackexchange.com/questions/34654/c-compile-in-arch-linux-with-no-g-package-in-repositories > for details. > > So the question is what can be changed so that the gcc in my machine > will interpret compile weight0.c as a C++ program... > I wonder why I didn't encounter such a program earlier... perhaps it's > the pre-release reversion that causes that problem. > Also can I use force it to use the gcc (I suppose gcc-4.7.2.p1.spkg is > the one) that comes with sage to compile? > Yes it looks tricky. It could very well be a regression as it compile with the C compiler and not the C++ compiler on my box with gcc 4.6.3 (gentoo). That being said gcc is very permissive and one day someone may decide to cut that one off. Would you believe that 2 days ago I accidentally configured octave with CXX=gcc-4.7.1 and I almost completely compiled it before being shot down by such a problem [octave is mostly c++]. Anyway you can certainly force sage's gcc by setting SAGE_INSTALL_GCC=yes before building sage. Francois -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
Re: [sage-devel] Singular fails to build in Sage 5.9
Hum I see what you are saying. Well my case is a bit tricky but I think I'm not alone. I am using Archlinux, so there is not g++ separately. And gcc decides whether the program as a C or C++ program in the following way. gcc file.C # uppercase .C (or .cpp) will process as C++ gcc file.c # lowercase .c will process as C see http://unix.stackexchange.com/questions/34654/c-compile-in-arch-linux-with-no-g-package-in-repositoriesfor details. So the question is what can be changed so that the gcc in my machine will interpret compile weight0.c as a C++ program... I wonder why I didn't encounter such a program earlier... perhaps it's the pre-release reversion that causes that problem. Also can I use force it to use the gcc (I suppose gcc-4.7.2.p1.spkg is the one) that comes with sage to compile? On Sun, May 5, 2013 at 1:47 AM, Francois Bissey < francois.bis...@canterbury.ac.nz> wrote: > On 05/05/13 19:32, Pong wrote: > > Looking for help in compiling SAGE 5.9 > > > > I encountered an error in buliding Singular-3-1-5.p7 > > > > I suspect the following may be the relevant part of the log file > > > > In file included from ../kernel/si_gmp.h:4:0, > > from ../kernel/structs.h:15, > > from weight0.c:13: > > ../factory/cf_gmp.h:2501:28: fatal error: bits/c++config.h: No such file > > or directory > > #include > > ^ > > compilation terminated. > > make[4]: *** [weight0.o] Error 1 > > > > However, I also attached the full file for the full details. > > Thank you for the help in advance. > > > > A fuller section of the log: > gcc -O2 -g -fPIC -pipe -I. -I.. -I/home/pong/sage-5.9/local > -I/home/pong/sage-5.9/local/include -I/home/pong/sage-5.9/local/include > -I/home/pong/sage-5.9/local/include -I/usr/local/include -DNDEBUG > -DOM_NDEBUG -Dx86_64_Linux -DHAVE_CONFIG_H -c weight0.c > In file included from ../kernel/si_gmp.h:4:0, > from ../kernel/structs.h:15, > from weight0.c:13: > ../factory/cf_gmp.h:2501:28: fatal error: bits/c++config.h: No such file > or directory > #include > > so bits/c++config.h is a c++ header provided by the compiler in your > case a pre-release of gcc-4.8.0. But the killer thing I think is that > weight0.c is a justly compiled with the c compiler but if you want > to use this header, g++ should be used. > One man's opinion at any rate. > > Francois > > -- > You received this message because you are subscribed to a topic in the > Google Groups "sage-devel" group. > To unsubscribe from this topic, visit > https://groups.google.com/d/topic/sage-devel/hqH8t69ixxA/unsubscribe?hl=en > . > To unsubscribe from this group and all its topics, send an email to > sage-devel+unsubscr...@googlegroups.com. > To post to this group, send email to sage-devel@googlegroups.com. > Visit this group at http://groups.google.com/group/sage-devel?hl=en. > For more options, visit https://groups.google.com/groups/opt_out. > > > -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
Re: [sage-devel] Singular fails to build in Sage 5.9
On 05/05/13 19:32, Pong wrote: > Looking for help in compiling SAGE 5.9 > > I encountered an error in buliding Singular-3-1-5.p7 > > I suspect the following may be the relevant part of the log file > > In file included from ../kernel/si_gmp.h:4:0, > from ../kernel/structs.h:15, > from weight0.c:13: > ../factory/cf_gmp.h:2501:28: fatal error: bits/c++config.h: No such file > or directory > #include > ^ > compilation terminated. > make[4]: *** [weight0.o] Error 1 > > However, I also attached the full file for the full details. > Thank you for the help in advance. > A fuller section of the log: gcc -O2 -g -fPIC -pipe -I. -I.. -I/home/pong/sage-5.9/local -I/home/pong/sage-5.9/local/include -I/home/pong/sage-5.9/local/include -I/home/pong/sage-5.9/local/include -I/usr/local/include -DNDEBUG -DOM_NDEBUG -Dx86_64_Linux -DHAVE_CONFIG_H -c weight0.c In file included from ../kernel/si_gmp.h:4:0, from ../kernel/structs.h:15, from weight0.c:13: ../factory/cf_gmp.h:2501:28: fatal error: bits/c++config.h: No such file or directory #include so bits/c++config.h is a c++ header provided by the compiler in your case a pre-release of gcc-4.8.0. But the killer thing I think is that weight0.c is a justly compiled with the c compiler but if you want to use this header, g++ should be used. One man's opinion at any rate. Francois -- 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 post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.