Re: [sage-devel] Re: OSX Yosemite
William Stein wrote: > This pattern has happened with literally every single OS X release: So true! Searching sage lists for '_scproxy' reveals that the failure to build the _scproxy module happened before on new releases of OS X. See the search results for sage-release [0] and sage-devel [1], with especially thread [2], where Volker [3] suggested a workaround: The following workaround should help to get around Apple's broken headers: CFLAGS="-DOS_OBJECT_USE_OBJC=0" export CFLAGS make Maybe this workaround could also help in the present situation? In the meanwhile, I sent a problem report to Apple via the 'Feedback assistant'. [0] https://groups.google.com/forum/#!searchin/sage-release/_scproxy [1] https://groups.google.com/forum/#!searchin/sage-devel/_scproxy [2] https://groups.google.com/d/topic/sage-devel/9B1GyPe6MAM/discussion [3] https://groups.google.com/d/msg/sage-devel/9B1GyPe6MAM/lQTo4ZrbjlAJ -- 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. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Graphs and order of vertices
On Mon, 20 Oct 2014, Erik Massop wrote: Shortly, do (di)graphs have some kind of order of vertices? If the vertices happen to have a total ordering everything should be fine. OK. Then I can use it on poset, because "labels" for vertices in Hasse diagram are just natural numbers. But... especially useful if there is no total ordering on the vertices, or when comparisons throw exceptions. Shall I open a ticket for this? Here is an example of a comparison throwing an exception: sage: G = Graph({complex(3):[complex(4)]}) sage: G.vertices() TypeError: no ordering relation is defined for complex numbers At least for me this seems as an error; but on the other hand, I don't know how it should behave. What makes this interesting is G = Graph({complex(3):[complex(4)]}) for x in G.vertex_iterator(): print x works as i guessed, but however G=Graph() for i in range(0,10): G.add_vertex(10-i) for x in G.vertex_iterator(): print x G.vertices() Does print numbers from 1 to 10 in order. On the other hand G=Graph() for i in range(0,10): G.add_vertex(randint(1,1000)) for x in G.vertex_iterator(): print x G.vertices() gives them in about random order. I started thinking this when doing http://trac.sagemath.org/ticket/17173 . Nathann suggested using directly (sub)graphs instead of (sub)poset. It is OK with .vertices() but not with .vertex_iterator(); try G=Graph() for i in range(2,50): G.add_vertex(i) for i in range(2,50): if not is_prime(i): G.delete_vertex(i) for x in G.vertex_iterator(): print x This seems to be quite open door for nasty bugs. .vertex_iterator might give vertices in order when graph is small or has not been modified many times. -- Jori Mäntysalo -- 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. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: OSX Yosemite
> > > > This pattern has happened with literally every single OS X release: > > Agreed, this is really annoying, and +A to the person who is the hero each time! -- 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. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Graphs and order of vertices
On Mon, 20 Oct 2014 20:11:31 +0300 (EEST) Jori Mantysalo wrote: > When I hit this, I was making a code for posets, but this is actually more > general question. Shortly, do (di)graphs have some kind of order of > vertices? Sort of. The code seems to assume this in various places, and uses python's sorted liberally. It makes sense that the code should assume this, because list and matrices need an order. If the vertices happen to have a total ordering everything should be fine. If the vertices happen to have a total preorder, most things should be fine because python uses stable sort. Otherwise things get scary. Many functions accept a parameter called key that gets used for custom ordering of the output. IMHO it would be a very good idea to allow a default key-function to be passed to the constructor. That way the user can force their own (total) ordering if this is wanted. I think this would make the assumption much more acceptable. Such a parameter is especially useful if there is no total ordering on the vertices, or when comparisons throw exceptions. Shall I open a ticket for this? Here is an example of a comparison throwing an exception: sage: G = Graph({complex(3):[complex(4)]}) sage: G.vertices() --- TypeError Traceback (most recent call last) in () > 1 G.vertices() /home/erik/sage/local/lib/python2.7/site-packages/sage/graphs/generic_graph.pyc in vertices(self, key, boundary_first) 8986 """ 8987 if not boundary_first: -> 8988 return sorted(list(self.vertex_iterator()), key=key) 8989 8990 from sage.misc.superseded import deprecation TypeError: no ordering relation is defined for complex numbers > To start, will > > Graph({'b':['a']}).vertices() > > always print ['a', 'b'], not ['b', 'a']? Yes, but only because strings compare lexicographically by ascii value in python: https://docs.python.org/2/tutorial/datastructures.html#comparing-sequences-and-other-types > How about directed graphs? I think .vertices is the same method, so should be the same. > What > about adding or deleting vertex (or edge), can it change order of > remaining vertices? How about list of incoming or outgoing list? If there is a total (pre)order, everything should be fine. If there is not, my bet is as good as yours. > Is there even some kind of general rule about this kind of thing at Sage? If there is, I'd like to know. Regards, Erik Massop -- 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. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: OSX Yosemite
On Mon, Oct 20, 2014 at 10:05 AM, Samuel Lelievre wrote: > Hi all, > > I am using OS X 10.10 Yosemite. > > I was beta-testing it since July, and failing to build Sage. > > (Strangely I never thought about using the OSX-10.9 version. > I just tried and it works! Thanks Stefan for asking the > question, and Volker for answering it works!) This pattern has happened with literally every single OS X release: - Sage builds on version A - Version A+1 is coming soon and Sage doesn't build on it, but binaries built for A *mostly* work. Watch out, since usually the mostly work, but there will be some misc failures. Do run doctests to see how bad the situation is. - Version A+1 gets released. Somebody -- e.g., Craig Citro a few years ago, and most recently Andrew Ohana -- makes a *truly heroic effort* to get Sage to fully build on version A+1. It's never easy. I remember being completely stumped after a day of trying before Craig Citro finally got one of the ports to work. With Andrew Ohana's porting, I got him a MBP retina laptop in exchange for porting Sage to run on 10.9 :-). Maybe you're the hero this time :-) > > About building Sage on Yosemite, here is my experience. > > - While beta-testing Yosemite, I tried to build Sage 6.3. > My report on sage-release [1] has links to a few log files. > - Incidentally I asked on sage-devel [2] about the Sage project > getting developer previews of future versions of OS X. > - Having understood that the bug was with gcc, I then installed > homebrew, and used it to install gcc 4.9.1, hoping this would > help building Sage, but that failed at a further step, which > I reported on sage-release [3] where Volker helped to point > the failure to a specific header: > """ > That header is Objective-C, which should never have been > included when compiling C sources. Most likely an OSX bug. > You can send them a bug report if you like talking at walls... > """ > - I don't mind talking at walls and I wanted to report to Apple > with some details, but I thought I should first make sure this > was not a homebrew thing, so I emailed the homebrew list [4] > and someone answered [5] suggesting an experiment (not done). > - I have now installed the finalized version of Yosemite, and > I still have Apple's "Feedback Assistant" from beta-testing. > (Plus, Apple tells me I will get betas of the next releases.) > So I would love to send feedback to Apple about this issue, but > I'm unsure how to formulate the issue, could someone help here? > One option would be to point to this thread, but maybe a > concise description of the issue would have more chances > of being dealt with? > > [1] Discussion on sage-release after Sage 6.3 was released (Aug 2014). > https://groups.google.com/d/msg/sage-release/Ilt8TZdy3H8/YxXy6mDyFEsJ > [2] Post on sage-devel asking if the Sage project has or could have > an Apple developer account, so as to test developer previews of > future releases of OS X. > https://groups.google.com/d/topic/sage-devel/-sm2ejzcjY8/discussion > [3] Discussion on sage-release after Sage 6.4.beta3 was released (Sep 2014). > https://groups.google.com/d/msg/sage-release/d7wrbZBZvrI/QtUT295ycVMJ > [4] My question on the homebrew mailing list: > > http://librelist.com/browser//homebrew/2014/10/11/homebrew-bug-or-osx-bug/#4f32863d45a19e4f340e1f1efd08e897 > [5] An answer to my question on the homebrew mailing list > > http://librelist.com/browser//homebrew/2014/10/11/homebrew-bug-or-osx-bug/#d2586d0f8ca2f240ee3eb51498a9fbcf > > More related links for reference. > > [6] Sage trac #17169: Upgrade to GCC 4.9.1 > http://trac.sagemath.org/ticket/17169 > [7] Sage trac #17174: Python _scproxy build fails on Yosemite > http://trac.sagemath.org/ticket/17174 > [8] Sage trac: non-closed tickets with keyword Yosemite > http://trac.sagemath.org/query?status=!closed&keywords=~yosemite > > -- > 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. > For more options, visit https://groups.google.com/d/optout. -- William Stein Professor of Mathematics University of Washington http://wstein.org -- 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. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: OSX Yosemite
The email looks good as far as bug reports go. I reported it to Apple but the more reports they get the more likely they are to act on it. /usr/include/dispatch/object.h has an Obj-C #if block but the dispatch_block_t declaration is outside of the #if/#endif. I can compile Python with gcc on Yosemite, but it does not include the _scproxy module. This is needed in Cython. So Python superficially seems to compile, but some modules fail. There is more info at #17169 On Monday, October 20, 2014 6:05:18 PM UTC+1, Samuel Lelievre wrote: > > Hi all, > > I am using OS X 10.10 Yosemite. > > I was beta-testing it since July, and failing to build Sage. > > (Strangely I never thought about using the OSX-10.9 version. > I just tried and it works! Thanks Stefan for asking the > question, and Volker for answering it works!) > > About building Sage on Yosemite, here is my experience. > > - While beta-testing Yosemite, I tried to build Sage 6.3. > My report on sage-release [1] has links to a few log files. > - Incidentally I asked on sage-devel [2] about the Sage project > getting developer previews of future versions of OS X. > - Having understood that the bug was with gcc, I then installed > homebrew, and used it to install gcc 4.9.1, hoping this would > help building Sage, but that failed at a further step, which > I reported on sage-release [3] where Volker helped to point > the failure to a specific header: > """ > That header is Objective-C, which should never have been > included when compiling C sources. Most likely an OSX bug. > You can send them a bug report if you like talking at walls... > """ > - I don't mind talking at walls and I wanted to report to Apple > with some details, but I thought I should first make sure this > was not a homebrew thing, so I emailed the homebrew list [4] > and someone answered [5] suggesting an experiment (not done). > - I have now installed the finalized version of Yosemite, and > I still have Apple's "Feedback Assistant" from beta-testing. > (Plus, Apple tells me I will get betas of the next releases.) > So I would love to send feedback to Apple about this issue, but > I'm unsure how to formulate the issue, could someone help here? > One option would be to point to this thread, but maybe a > concise description of the issue would have more chances > of being dealt with? > > [1] Discussion on sage-release after Sage 6.3 was released (Aug 2014). > https://groups.google.com/d/msg/sage-release/Ilt8TZdy3H8/YxXy6mDyFEsJ > [2] Post on sage-devel asking if the Sage project has or could have > an Apple developer account, so as to test developer previews of > future releases of OS X. > https://groups.google.com/d/topic/sage-devel/-sm2ejzcjY8/discussion > [3] Discussion on sage-release after Sage 6.4.beta3 was released (Sep > 2014). > https://groups.google.com/d/msg/sage-release/d7wrbZBZvrI/QtUT295ycVMJ > [4] My question on the homebrew mailing list: > > http://librelist.com/browser//homebrew/2014/10/11/homebrew-bug-or-osx-bug/#4f32863d45a19e4f340e1f1efd08e897 > [5] An answer to my question on the homebrew mailing list > > http://librelist.com/browser//homebrew/2014/10/11/homebrew-bug-or-osx-bug/#d2586d0f8ca2f240ee3eb51498a9fbcf > > More related links for reference. > > [6] Sage trac #17169: Upgrade to GCC 4.9.1 > http://trac.sagemath.org/ticket/17169 > [7] Sage trac #17174: Python _scproxy build fails on Yosemite > http://trac.sagemath.org/ticket/17174 > [8] Sage trac: non-closed tickets with keyword Yosemite > http://trac.sagemath.org/query?status=!closed&keywords=~yosemite > > -- 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. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Graphs and order of vertices
Hi, For these kind of questions, please use sage-support (an other google groups) or better http://ask.sagemath.org (as it may be useful for other users). To create a digraph, you need to use sage: G = DiGraph({'b': ['a']}) sage: G.edges() And you have methods G.outgoing_edges() and G.incoming_edges(). Vincent 2014-10-20 17:11 UTC, Jori Mantysalo : > When I hit this, I was making a code for posets, but this is actually more > general question. Shortly, do (di)graphs have some kind of order of > vertices? To start, will > > Graph({'b':['a']}).vertices() > > always print ['a', 'b'], not ['b', 'a']? How about directed graphs? What > about adding or deleting vertex (or edge), can it change order of > remaining vertices? How about list of incoming or outgoing list? > > Is there even some kind of general rule about this kind of thing at Sage? > > -- > Jori Mäntysalo > > -- > 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. > For more options, visit https://groups.google.com/d/optout. > -- 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. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Graphs and order of vertices
When I hit this, I was making a code for posets, but this is actually more general question. Shortly, do (di)graphs have some kind of order of vertices? To start, will Graph({'b':['a']}).vertices() always print ['a', 'b'], not ['b', 'a']? How about directed graphs? What about adding or deleting vertex (or edge), can it change order of remaining vertices? How about list of incoming or outgoing list? Is there even some kind of general rule about this kind of thing at Sage? -- Jori Mäntysalo -- 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. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: OSX Yosemite
Hi all, I am using OS X 10.10 Yosemite. I was beta-testing it since July, and failing to build Sage. (Strangely I never thought about using the OSX-10.9 version. I just tried and it works! Thanks Stefan for asking the question, and Volker for answering it works!) About building Sage on Yosemite, here is my experience. - While beta-testing Yosemite, I tried to build Sage 6.3. My report on sage-release [1] has links to a few log files. - Incidentally I asked on sage-devel [2] about the Sage project getting developer previews of future versions of OS X. - Having understood that the bug was with gcc, I then installed homebrew, and used it to install gcc 4.9.1, hoping this would help building Sage, but that failed at a further step, which I reported on sage-release [3] where Volker helped to point the failure to a specific header: """ That header is Objective-C, which should never have been included when compiling C sources. Most likely an OSX bug. You can send them a bug report if you like talking at walls... """ - I don't mind talking at walls and I wanted to report to Apple with some details, but I thought I should first make sure this was not a homebrew thing, so I emailed the homebrew list [4] and someone answered [5] suggesting an experiment (not done). - I have now installed the finalized version of Yosemite, and I still have Apple's "Feedback Assistant" from beta-testing. (Plus, Apple tells me I will get betas of the next releases.) So I would love to send feedback to Apple about this issue, but I'm unsure how to formulate the issue, could someone help here? One option would be to point to this thread, but maybe a concise description of the issue would have more chances of being dealt with? [1] Discussion on sage-release after Sage 6.3 was released (Aug 2014). https://groups.google.com/d/msg/sage-release/Ilt8TZdy3H8/YxXy6mDyFEsJ [2] Post on sage-devel asking if the Sage project has or could have an Apple developer account, so as to test developer previews of future releases of OS X. https://groups.google.com/d/topic/sage-devel/-sm2ejzcjY8/discussion [3] Discussion on sage-release after Sage 6.4.beta3 was released (Sep 2014). https://groups.google.com/d/msg/sage-release/d7wrbZBZvrI/QtUT295ycVMJ [4] My question on the homebrew mailing list: http://librelist.com/browser//homebrew/2014/10/11/homebrew-bug-or-osx-bug/#4f32863d45a19e4f340e1f1efd08e897 [5] An answer to my question on the homebrew mailing list http://librelist.com/browser//homebrew/2014/10/11/homebrew-bug-or-osx-bug/#d2586d0f8ca2f240ee3eb51498a9fbcf More related links for reference. [6] Sage trac #17169: Upgrade to GCC 4.9.1 http://trac.sagemath.org/ticket/17169 [7] Sage trac #17174: Python _scproxy build fails on Yosemite http://trac.sagemath.org/ticket/17174 [8] Sage trac: non-closed tickets with keyword Yosemite http://trac.sagemath.org/query?status=!closed&keywords=~yosemite -- 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. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: OSX Yosemite
Yes, if you have compiled it with 10.9 then you can still run it with 10.10. Its just Apple's SDK having issues, I think. On Monday, October 20, 2014 4:26:45 PM UTC+1, Stefan wrote: > > Does a compiled Sage still run? > -- 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. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: OSX Yosemite
Does a compiled Sage still run? --Stefan On Friday, October 17, 2014 2:53:29 PM UTC-5, Volker Braun wrote: > > I tried gcc 4.8.3 with the OSX 10.10 patch but that dies with "Bootstrap > comparison failure!" > > More info: http://trac.sagemath.org/ticket/17169 > > > > On Friday, October 17, 2014 6:14:07 PM UTC+1, Jean-Pierre Flori wrote: >> >> Relevant? >> https://groups.google.com/d/msg/sage-devel/-sm2ejzcjY8/x27G0nMRdY4J >> >> On Friday, October 17, 2014 6:58:16 PM UTC+2, Volker Braun wrote: >>> >>> I upgraded my OSX buildbot, now it doesn't compile Sage any more ;-) >>> >>> Checking multilib configuration for libgcc... >>> Configuring stage 1 in x86_64-apple-darwin14.0.0/libgcc >>> configure: creating cache ./config.cache >>> checking build system type... x86_64-apple-darwin14.0.0 >>> checking host system type... x86_64-apple-darwin14.0.0 >>> checking for --enable-version-specific-runtime-libs... no >>> checking for a BSD-compatible install... /usr/bin/install -c >>> checking for gawk... awk >>> checking for x86_64-apple-darwin14.0.0-ar... ar >>> checking for x86_64-apple-darwin14.0.0-lipo... lipo >>> checking for x86_64-apple-darwin14.0.0-nm... >>> /Users/vbraun/Sage/local/var/tmp/sage/build/gcc-4.7.3.p1/gcc-build/./gcc/nm >>> checking for x86_64-apple-darwin14.0.0-ranlib... ranlib >>> checking for x86_64-apple-darwin14.0.0-strip... strip >>> checking whether ln -s works... yes >>> checking for x86_64-apple-darwin14.0.0-gcc... >>> /Users/vbraun/Sage/local/var/tmp/sage/build/gcc-4.7.3.p1/gcc-build/./gcc/xgcc >>> >>> -B/Users/vbraun/Sage/local/var/tmp/sage/build/gcc-4.7.3.p1/gcc-build/./gcc/ >>> -B/Users/vbraun/Sage/local/x86_64-apple-darwin14.0.0/bin/ >>> -B/Users/vbraun/Sage/local/x86_64-apple-darwin14.0.0/lib/ -isystem >>> /Users/vbraun/Sage/local/x86_64-apple-darwin14.0.0/include -isystem >>> /Users/vbraun/Sage/local/x86_64-apple-darwin14.0.0/sys-include >>> checking for suffix of object files... configure: error: in >>> `/Users/vbraun/Sage/local/var/tmp/sage/build/gcc-4.7.3.p1/gcc-build/x86_64-apple-darwin14.0.0/libgcc': >>> configure: error: cannot compute suffix of object files: cannot compile >>> See `config.log' for more details. >>> make[5]: *** [configure-stage1-target-libgcc] Error 1 >>> make[4]: *** [stage1-bubble] Error 2 >>> make[3]: *** [all] Error 2 >>> >>> real1m12.951s >>> user4m28.718s >>> sys 1m0.100s >>> >>> Error installing package gcc-4.7.3.p1 >>> >>> >>> -- 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. For more options, visit https://groups.google.com/d/optout.