Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-05-06 Thread Volker Braun
If "make distclean" fails then $SAGE_LOCAL is not being deleted (it is the last command in the "distclean" target) On Tuesday, May 6, 2014 9:36:59 AM UTC+2, Andrew wrote: > > Well, I'm completely confused as to why one of your recommendations above > was: > I think you half-built Sage gcc conf

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-05-06 Thread Andrew
Well, I'm completely confused as to why one of your recommendations above was: I think you half-built Sage gcc conflicts with Apple gcc. Just "rm -rf /usr/local/src/sage/local" and try again. A. On Tuesday, 6 May 2014 17:32:49 UTC+10, Volker Braun wrote: > > From the Makefile: > > distclean: cle

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-05-06 Thread Volker Braun
>From the Makefile: distclean: clean doc-clean lib-clean bdist-clean @echo "Deleting all remaining output from build system ..." rm -rf local On Tuesday, May 6, 2014 9:30:03 AM UTC+2, Andrew wrote: > > > Can you clarify what you mean here? "make distclean" should indeed return

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-05-06 Thread Andrew
> Can you clarify what you mean here? "make distclean" should indeed return > the code to its pristine download state, except that it won't clean up > files in the "upstream" directory. Are there other things it doesn't clean > up which it should? > It doesn't remove the $SAGE_ROOT/local sub

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-05-05 Thread John H Palmieri
On Sunday, May 4, 2014 4:57:48 PM UTC-7, Andrew wrote: > > I now have a working version sage! I should qualify this: I have a working > version of 6.2.rc0 but I am still unable to compile 6.1.1. I think that > this is because of #15937 . This > patch is i

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-05-04 Thread Andrew
I now have a working version sage! I should qualify this: I have a working version of 6.2.rc0 but I am still unable to compile 6.1.1. I think that this is because of #15937 . This patch is included in 6.2.rc0 and it fixes issues between singular and the ve

[sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-05-01 Thread leif
kcrisman wrote: Though Merged in is still a useful field ;-) Couldn't a script update all newly merged tickets after each release (including devel releases)? I have no idea, but recall that Minh used to run a script to update the Milestone fields (everytime a new stable version had been rel

[sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-30 Thread kcrisman
On Tuesday, April 29, 2014 3:56:10 PM UTC-4, Volker Braun wrote: > > Ok, I changed it to: > > $ git trac find b42c581 > Commit has been merged in 6.2.beta8. <- new > addition > Yes, thanks! That will help a lot down the way. Though Merged in is still a useful field

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-30 Thread Volker Braun
I think you half-built Sage gcc conflicts with Apple gcc. Just "rm -rf /usr/local/src/sage/local" and try again. On Wednesday, April 30, 2014 10:56:11 PM UTC+1, Andrew wrote: > > On Thursday, 1 May 2014 00:47:20 UTC+10, Volker Braun wrote: >> >> Or just do a "git pull", since you are on the mas

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-30 Thread Andrew
On Thursday, 1 May 2014 00:47:20 UTC+10, Volker Braun wrote: > > Or just do a "git pull", since you are on the master branch this will give > you 6.1.1 which afaik also builds on OSX 10.9. > As detailed above, I have the latest master branch but for some reason it has the wrong version of singul

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-30 Thread Volker Braun
Or just do a "git pull", since you are on the master branch this will give you 6.1.1 which afaik also builds on OSX 10.9. On Wednesday, April 30, 2014 3:38:29 PM UTC+1, Jean-Pierre Flori wrote: > > To get the latest betas you should be tracking the "develop" branch (from > trac or github, they

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-30 Thread Jean-Pierre Flori
To get the latest betas you should be tracking the "develop" branch (from trac or github, they are the same). You can issue something like: """ git checkout -b develop trac/develop """ to make a local copy of it and directly swith to it. On Wednesday, April 30, 2014 4:30:44 PM UTC+2, Andrew wrote

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-30 Thread Andrew
OK, so the singular version seems to be in $SAGE_ROOT/build/pkgs/singular/package-version.txt and it is indeed set to 3.1.5.p9. Updating this by hand to 3.1.6.p1 and then re-making gave a checksum error on upstream/singular-3.1.6.p1 Deleting this and then re-making still gives a checksum erro

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-30 Thread Andrew
On Wednesday, 30 April 2014 20:48:01 UTC+10, Volker Braun wrote: > > On Wednesday, April 30, 2014 9:50:27 AM UTC+1, Andrew wrote: >> >> singular-3.1.5.p9 >>> >> > Sage-6.2.rc0 uses Singular 3.1.6.p1, you are building an old branch > (possibly older than when we fixed the OSX 10.9 build). Are you

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-30 Thread Volker Braun
On Wednesday, April 30, 2014 9:50:27 AM UTC+1, Andrew wrote: > > singular-3.1.5.p9 >> > Sage-6.2.rc0 uses Singular 3.1.6.p1, you are building an old branch (possibly older than when we fixed the OSX 10.9 build). Are you sure you didn't accidentally check out 6.2.beta0? Check "git log" -- You

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-30 Thread Volker Braun
Homebrew/Macports/Fink/etc are not supported and likely to conflict with Sage. What kind of error do you get from "make distclean"? You can try to exclude non-system stuff from the PATH to ignore 3rd party software: export PATH=/usr/bin:/bin:/usr/sbin:/sbin make distclean make On Wednesday, Ap

[sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-29 Thread Volker Braun
Ok, I changed it to: $ git trac find b42c581 Commit has been merged in 6.2.beta8. <- new addition commit e2ff979af93ea5f118e46621768efd923deaf2e5 Merge: bf56254 b42c581 Author: Release Manager Date: Mon Apr 14 00:00:20 2014 +0200 Trac #16099: Update libpng to 1.2

[sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-29 Thread kcrisman
On Tuesday, April 29, 2014 1:44:18 PM UTC-4, leif wrote: > > Volker Braun wrote: > > The issue is that libpng was updated, not libm4ri > > > > There is still something wonky because the two libpng versions ought to > > be compatible. There is something wrong in how m4ri links to libpng on >

[sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-29 Thread leif
Volker Braun wrote: The issue is that libpng was updated, not libm4ri There is still something wonky because the two libpng versions ought to be compatible. There is something wrong in how m4ri links to libpng on OSX, causing it to depend on the version patchlevel. On Tuesday, April 29, 2014 4:

[sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-29 Thread Volker Braun
The issue is that libpng was updated, not libm4ri There is still something wonky because the two libpng versions ought to be compatible. There is something wrong in how m4ri links to libpng on OSX, causing it to depend on the version patchlevel. On Tuesday, April 29, 2014 4:47:37 PM UTC+1, kcri

[sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-29 Thread kcrisman
On Tuesday, April 29, 2014 9:44:54 AM UTC-4, Volker Braun wrote: > > libpng was upgraded recently, and if you switch between different versions > then you have to rebuild (in the worst case: make distclean && make will > always work). > > >> I figured, but that does take some time. And more to

[sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-29 Thread Volker Braun
libpng was upgraded recently, and if you switch between different versions then you have to rebuild (in the worst case: make distclean && make will always work). On Tuesday, April 29, 2014 2:33:54 PM UTC+1, kcrisman wrote: > > >> File >> "/usr/local/src/sage/local/lib/python2.7/site-pack

[sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-29 Thread kcrisman
> > > File > "/usr/local/src/sage/local/lib/python2.7/site-packages/sage/categories/category.py" > , line 2293, in _repr_ > if len(categories) == 2 and Sets().Facades() in categories: > > --- > > ***

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-28 Thread Volker Braun
Does your compile always fail in different places with different errors? Maybe its more relevant to look at the output of "diskutil info /" and/or install smartmontools... On Monday, April 28, 2014 4:02:39 PM UTC+1, Volker Braun wrote: > > ld: symbol dyld_stub_binding_helper not found, normall

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-28 Thread Volker Braun
ld: symbol dyld_stub_binding_helper not found, normally in crt1.o/dylib1.o/bundle1.o for architecture x86_64 fatal error: libtool: internal link edit command failed sounds like your gcc install is hosed. Post the full log, not just the singular log. -- You received this message because you are

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-28 Thread Volker Braun
So it works. Your real error is elsewhere. Run "make distclean && make -j1" On Monday, April 28, 2014 2:55:53 PM UTC+1, Andrew wrote: > > On Monday, 28 April 2014 23:49:47 UTC+10, Volker Braun wrote: >> >> whats the output of "xcrun -v strip" >> > > xcrun: note: PATH = > '/Users/andrew/bin:/Libr

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-28 Thread Andrew
On Monday, 28 April 2014 23:49:47 UTC+10, Volker Braun wrote: > > whats the output of "xcrun -v strip" > xcrun: note: PATH = '/Users/andrew/bin:/Library/Developer/CommandLineTools/usr/bin:/bin:/sbin:/usr/local/bin:/usr/local/sbin:/usr/bin:/usr/sbin:/usr/texbin:/usr/local/bin' xcrun: note: SDKROOT

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-28 Thread Volker Braun
whats the output of "xcrun -v strip" On Monday, April 28, 2014 2:40:54 PM UTC+1, Andrew wrote: > > > > On Monday, 28 April 2014 22:20:28 UTC+10, Volker Braun wrote: >> >> On Monday, April 28, 2014 10:53:53 AM UTC+1, Andrew wrote: >>> >>> Here's another bizarre feature: >>> >>> 538-6.1.1-master-s

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-28 Thread Andrew
On Monday, 28 April 2014 22:20:28 UTC+10, Volker Braun wrote: > > On Monday, April 28, 2014 10:53:53 AM UTC+1, Andrew wrote: >> >> Here's another bizarre feature: >> >> 538-6.1.1-master-sage: /Applications/Xcode.app/Contents/Developer/usr/bin >> /xcodebuild -sdk / -find strip 2> /dev/null >> /Ap

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-28 Thread Volker Braun
On Monday, April 28, 2014 10:53:53 AM UTC+1, Andrew wrote: > > Here's another bizarre feature: > > 538-6.1.1-master-sage: > /Applications/Xcode.app/Contents/Developer/usr/bin/xcodebuild > -sdk / -find strip 2> /dev/null > /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault. > xcto

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-28 Thread Andrew
On Monday, 28 April 2014 20:09:53 UTC+10, Volker Braun wrote: > > Whats xcodebuild -version > > It's: Xcode 5.1.1 Build version 5B1008 Andrew -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emai

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-28 Thread Volker Braun
Whats xcodebuild -version On Monday, April 28, 2014 10:53:53 AM UTC+1, Andrew wrote: > > Here's another bizarre feature: > > 538-6.1.1-master-sage: > /Applications/Xcode.app/Contents/Developer/usr/bin/xcodebuild > -sdk / -find strip 2> /dev/null > /Applications/Xcode.app/Contents/Developer/Tool

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-28 Thread Andrew
Here's another bizarre feature: 538-6.1.1-master-sage: /Applications/Xcode.app/Contents/Developer/usr/bin/xcodebuild -sdk / -find strip 2> /dev/null /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault. xctoolchain/usr/bin/strip\539-6.1.1-master-sage: ls -l /Applications/Xcode. ap

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-28 Thread Andrew
Yes I have strip and the command line tools: 533-6.1.1-master-sage: which strip strip is /Library/Developer/CommandLineTools/usr/bin/strip strip is /usr/bin/strip On Monday, 28 April 2014 16:28:16 UTC+10, yomcat wrote: > > On 28/04/2014, at 1809, Andrew > wrote: > > > > sh: line 1: 45940 Tra

Re: [sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-27 Thread Michael Welsh
On 28/04/2014, at 1809, Andrew wrote: > > sh: line 1: 45940 Trace/BPT trap: 5 > /Applications/Xcode.app/Contents/Developer/usr/bin/xcodebuild -sdk / -find > strip 2> /dev/null > xcrun: error: unable to find utility "strip", not a developer tool or in PATH > install: child process failed:

[sage-devel] Re: Problem compiling sage 6.2.rc0-develop with macosx/mavericks

2014-04-27 Thread Andrew
Update: It's not a 6.2.rc0 issue after all. My compile of 6.1.1 failed with: ... sh: line 1: 45940 Trace/BPT trap: 5 /Applications/Xcode.app/Contents/ Developer/usr/bin/xcodebuild -sdk / -find strip 2> /dev/null xcrun: error: unable to find utility "strip", not a developer tool or inPATH ins