Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Jeroen Demeyer]

On 2016-02-04 13:02, Jan Groenewald wrote:

Does it matter for the PPA? Isn't that installed in a fixed location
anyway?


Yes, changes to this it is what broke the PPA since 6.9.
Sorry, I didn't understand to what you answered "yes". Is the PPA 
installed in a fixed location? If so, *you* should run the relocation 
script when creating the PPA. I think that would solve the problems with 
the PPA.


--
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Failure to Compile Sage 7.0 Stable on openSUSE Linux

[Dima Pasechnik]

What is gcc installed on your system?
And how do you start the compilation?
(these things are not clear from your posts)


On Thursday, 4 February 2016 04:33:43 UTC, Martin Vahi wrote:
>
>
> The source was downloaded on 2016_02_04 from 
>
> http://www.sagemath.org/download-source.html
>
> and the sha256 sum of the downloaded
>
> sage-7.0.tar.gz
>
> is
>
> 8eb8d48b78a29163ba4d89ab3f781a5476da4d2e119ba3662e84445fe3752bae
>
> An excerpt of the console output resides at the attached text files.
>
>
>

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Jan Groenewald]

Hi

On 4 February 2016 at 14:30, Jeroen Demeyer  wrote:

> On 2016-02-04 13:02, Jan Groenewald wrote:
>
>> Does it matter for the PPA? Isn't that installed in a fixed location
>> anyway?
>>
>>
>> Yes, changes to this it is what broke the PPA since 6.9.
>>
> Sorry, I didn't understand to what you answered "yes". Is the PPA
> installed in a fixed location? If so, *you* should run the relocation
> script when creating the PPA. I think that would solve the problems with
> the PPA.
>

Yes, changes to sage's patching of files broke the part of the PPA.

Yes, we now plan to run relocation before rather than after creating
uploading to the PPA since I yesterday submitted a patch to give it a
destination folder as a command line argument. It was not possible before.
We will not need to move it again after that.

Regards,
Jan







-- 
  .~.
  /V\ Jan Groenewald
 /( )\www.aims.ac.za
 ^^-^^

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: [sage-release] Re: recent woes with relocation of SageMath installation directory

[Jan Groenewald]

Hi

On 4 February 2016 at 12:20, HG  wrote:

> Following the ppa sage link, I tried to install sage-7, it works fine.
> Time for patching, but no problem.
> I use two version of sagemath the ppa and developer one (which I compile)
> and after I compile sagemanifolds (which I can't do at the moment with the
> ppa because it's installed system-wide).
> Thanks for this good work
> Henri
>

If you run sagemath-upstream-binary-full it is technically possible to run
sage manifolds. We had this deployed on campus last week. But you have to
reinstall it after every upgrade. It is bad because:

1) you should not compile as root. There is a past bug where compiling some
package would delete a system file [1]
. Malicious code could hose or
rootkit your system. We plan in future to have the sage tree owned by a
sagemath user, which will alleviate that (only) if people remember to
change to that user before customizing / adding optional packages and
rebuilding.

2) bad practice since dpkg (or your distro package manager) is supposed to
manage the system files, but it is technically possible.

It will not work with sagemath-upstream-binary (without -full) since the
SAGE_ROOT/src folder is missing and many files you need for rebuilding sage.

I will try to release a 7.0.1 (rather than waiting for 7.1) which also
fixes the smaller sagemath-upstream-binary in the PPAs. Note 7.0 is now in
the stable PPA for 14.04.

Regards,
Jan




-- 
  .~.
  /V\ Jan Groenewald
 /( )\www.aims.ac.za
 ^^-^^

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Jan Groenewald]

Hi

On 4 February 2016 at 13:42, Jeroen Demeyer  wrote:

> On 2016-02-04 12:32, Jan Groenewald wrote:
>
>> Is there any plan to support this in future?
>>
> I don't think so. This used to be semi-supported in the past (before Sage
> 7.0), but it is too difficult to do correctly.
>

I'm under the impression almost all software works relocatable, and only
start off hardcoded due to time pressure, but then get fixed.
Is it difficult because of legacy code?

Does it matter for the PPA? Isn't that installed in a fixed location anyway?


Yes, changes to this it is what broke the PPA since 6.9. I didn't catch it
between dev and stable because I run the -full version.

My head explodes that sage patches binaries after compilation to fix
hardcoded paths.

If I google "hardcoded paths" then there are mainly results about how to
fix having hardcoded paths.
I don't know which parts of Sage or packages in Sage do not support
prefixes. It seems bizarre.
Are there just not enough people working on the build side of Sage? Does
most of the sub-packages not support
prefixes?

Regards,
Jan






>
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-support" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-support+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-supp...@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-support.
> For more options, visit https://groups.google.com/d/optout.
>



-- 
  .~.
  /V\ Jan Groenewald
 /( )\www.aims.ac.za
 ^^-^^

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] [Bug] Sage 7.0 from the PPA fails to install

[Jan Groenewald]

Hi

This should be fixed by 7.0.1 or 7.1.

In the meantime use the larger version

sudo apt-get install sagemath-upstream-binary-full

Regards,
Jan

On 4 February 2016 at 11:28, Marcin M.  wrote:

> The log follows:
>
> $ sudo apt-get install sagemath-upstream-binary
>> Reading package lists... Done
>> Building dependency tree
>> Reading state information... Done
>> Recommended packages:
>>   gfortran dvipng texlive
>> The following NEW packages will be installed:
>>   sagemath-upstream-binary
>> 0 upgraded, 1 newly installed, 0 to remove and 0 not upgraded.
>> Need to get 0 B/510 MB of archives.
>> After this operation, 3477 MB of additional disk space will be used.
>> Selecting previously unselected package sagemath-upstream-binary.
>> (Reading database ... 303545 files and directories currently installed.)
>> Preparing to unpack
>> .../sagemath-upstream-binary_7.0~aimsppa1~qa201602031255151_amd64.deb ...
>> Unpacking sagemath-upstream-binary (7.0~aimsppa1~qa201602031255151) ...
>> Processing triggers for gnome-menus (3.10.1-0ubuntu2) ...
>> Processing triggers for desktop-file-utils (0.22-1ubuntu1) ...
>> Processing triggers for mime-support (3.54ubuntu1.1) ...
>> Setting up sagemath-upstream-binary (7.0~aimsppa1~qa201602031255151) ...
>> Running Sage once as root to set paths
>>
>> Rewriting paths for your new installation directory
>> ===
>>
>> This might take a few minutes but only has to be done once.
>>
>> patching /usr/lib/sagemath/src/build/cythonized/sage/rings/real_arb.c
>> Traceback (most recent call last):
>>   File "/usr/lib/sagemath/relocate-once.py", line 85, in 
>> p('src/build/cythonized/sage/rings/real_arb.c').substitute().save()
>>   File "/usr/lib/sagemath/relocate-once.py", line 70, in __call__
>> return FilePatch(self, filename)
>>   File "/usr/lib/sagemath/relocate-once.py", line 21, in __init__
>> self.st = os.stat(filename)
>> OSError: [Errno 2] No such file or directory:
>> '/usr/lib/sagemath/src/build/cythonized/sage/rings/real_arb.c'
>> ┌┐
>> │ SageMath Version 7.0, Release Date: 2016-01-19 │
>> │ Type "notebook()" for the browser-based notebook interface.│
>> │ Type "help()" for help.│
>> └┘
>>
>> ERROR:  The Sage installation tree has moved
>>
>> from
>> /mnt/highperf/buildbot/slave/binary_pkg/build/source/SageMath/jc4b6yulaujayb9sr94ia88eourzeqip0oidma
>>   to /usr/lib/sagemath
>>
>> This is not supported, and Sage will not work. To install Sage from a
>> binary package:
>>
>> 1. Open the .tar.bz2 archive (or .dmg on OSX)
>>
>> 2. Move the SageMath folder/app to where you want it to be. You can
>>also rename the directory now.
>>
>> 3. Start sage for the first time. This will then automatically patch
>>paths in binaries.
>>
>> After starting Sage for the first time you cannot change the
>> installation any more. To install Sage elsewhere, start over from the
>> binary package. Or recompile Sage from scratch in the new location
>> ("make distclean && make")
>>
>> dpkg: error processing package sagemath-upstream-binary (--configure):
>>  subprocess installed post-installation script returned error exit status
>> 1
>> Errors were encountered while processing:
>>  sagemath-upstream-binary
>> E: Sub-process /usr/bin/dpkg returned an error code (1)
>>
>>
> --
> 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 https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.
>



-- 
  .~.
  /V\ Jan Groenewald
 /( )\www.aims.ac.za
 ^^-^^

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Dima Pasechnik]

On Thursday, 4 February 2016 12:30:36 UTC, Jeroen Demeyer wrote:
>
> On 2016-02-04 13:02, Jan Groenewald wrote: 
> > Does it matter for the PPA? Isn't that installed in a fixed location 
> > anyway? 
> > 
> > 
> > Yes, changes to this it is what broke the PPA since 6.9. 
> Sorry, I didn't understand to what you answered "yes". Is the PPA 
> installed in a fixed location? If so, *you* should run the relocation 
> script when creating the PPA. I think that would solve the problems with 
> the PPA. 
>

my understanding is that PPA is providing a "normal" .deb, and they are 
installed in fixed locations
(typically /usr/***), unless I miss something here...
Then indeed PPA should contain hardcoded paths already.

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: recent woes with relocation of SageMath installation directory

[HG]

Following the ppa sage link, I tried to install sage-7, it works fine. Time 
for patching, but no problem.
I use two version of sagemath the ppa and developer one (which I compile) 
and after I compile sagemanifolds (which I can't do at the moment with the 
ppa because it's installed system-wide).
Thanks for this good work
Henri 

Le mardi 2 février 2016 18:16:12 UTC+1, slelievre a écrit :
>
> Dear sage-support (cc: sage-release, sage-devel) 
> (please reply on sage-support only) 
>
> I received the following question after announcing 
> Sage 7.0 on sage-announce. 
>
> This is one of many such questions I am reading on 
> various lists and sites recently. 
>
> Something needs to be more clearly documented, 
> or refactored. 
>
> Can someone answer this question by replying to 
> this post on sage-support? 
>
> Thanks! 
> Samuel 
>
> -- Forwarded message -- 
> Date: Mon, 1 Feb 2016 23:32:31 -0600 
> To: Samuel Lelievre  
> Subject: Installing upgraded SageMath 
>
> I tried installing sagemath-upstream-binary on my Ubuntu Trusty Tahr 
> system and saw this: 
>
> Rewriting paths for your new installation directory 
> === 
>
> This might take a few minutes but only has to be done once. 
>
> patching 
> /usr/lib/sagemath/src/build/cythonized/sage/structure/list_clone.c 
> Traceback (most recent call last): 
>   File "/usr/lib/sagemath/relocate-once.py", line 85, in  
> 
> p('src/build/cythonized/sage/structure/list_clone.c').substitute().save() 
>   File "/usr/lib/sagemath/relocate-once.py", line 70, in __call__ 
> return FilePatch(self, filename) 
>   File "/usr/lib/sagemath/relocate-once.py", line 21, in __init__ 
> self.st = os.stat(filename) 
> OSError: [Errno 2] No such file or directory: 
> '/usr/lib/sagemath/src/build/cythonized/sage/structure/list_clone.c' 
>
> Any idea what is up? 
>
> Dale Amon 
>

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Jeroen Demeyer]

On 2016-02-04 13:02, Jan Groenewald wrote:

Is it difficult because of legacy code?

No, it's difficult because Sage is like a distribution.

Relocating Sage is like relocating /usr to /otherpath/usr in your 
favorite Linux distro. I bet things will break down completely.


Jeroen.

--
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Volker Braun]

On Thursday, February 4, 2016 at 1:03:11 PM UTC+1, Jan Groenewald wrote:
>
> I'm under the impression almost all software works relocatable
>

Your impression is incorrect. Try to install any non-trivial (including 
shared libraries) deb/rpm into a prefix, good luck. The Fedora packaging 
guidelines even specifically tell you to not to bother with relocatable 
rpms. You can only build relocatable software if 
1. you have everything under a single build system and
2. you don't need to execute the built binaries while you are building it.

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] [Bug] Sage 7.0 from the PPA fails to install

[Marcin M.]

The log follows:

$ sudo apt-get install sagemath-upstream-binary
> Reading package lists... Done
> Building dependency tree   
> Reading state information... Done
> Recommended packages:
>   gfortran dvipng texlive
> The following NEW packages will be installed:
>   sagemath-upstream-binary
> 0 upgraded, 1 newly installed, 0 to remove and 0 not upgraded.
> Need to get 0 B/510 MB of archives.
> After this operation, 3477 MB of additional disk space will be used.
> Selecting previously unselected package sagemath-upstream-binary.
> (Reading database ... 303545 files and directories currently installed.)
> Preparing to unpack 
> .../sagemath-upstream-binary_7.0~aimsppa1~qa201602031255151_amd64.deb ...
> Unpacking sagemath-upstream-binary (7.0~aimsppa1~qa201602031255151) ...
> Processing triggers for gnome-menus (3.10.1-0ubuntu2) ...
> Processing triggers for desktop-file-utils (0.22-1ubuntu1) ...
> Processing triggers for mime-support (3.54ubuntu1.1) ...
> Setting up sagemath-upstream-binary (7.0~aimsppa1~qa201602031255151) ...
> Running Sage once as root to set paths
>
> Rewriting paths for your new installation directory
> ===
>
> This might take a few minutes but only has to be done once.
>
> patching /usr/lib/sagemath/src/build/cythonized/sage/rings/real_arb.c
> Traceback (most recent call last):
>   File "/usr/lib/sagemath/relocate-once.py", line 85, in 
> p('src/build/cythonized/sage/rings/real_arb.c').substitute().save()
>   File "/usr/lib/sagemath/relocate-once.py", line 70, in __call__
> return FilePatch(self, filename)
>   File "/usr/lib/sagemath/relocate-once.py", line 21, in __init__
> self.st = os.stat(filename)
> OSError: [Errno 2] No such file or directory: 
> '/usr/lib/sagemath/src/build/cythonized/sage/rings/real_arb.c'
> ┌┐
> │ SageMath Version 7.0, Release Date: 2016-01-19 │
> │ Type "notebook()" for the browser-based notebook interface.│
> │ Type "help()" for help.│
> └┘
>
> ERROR:  The Sage installation tree has moved
>
> from 
> /mnt/highperf/buildbot/slave/binary_pkg/build/source/SageMath/jc4b6yulaujayb9sr94ia88eourzeqip0oidma
>   to /usr/lib/sagemath
>
> This is not supported, and Sage will not work. To install Sage from a
> binary package:
>
> 1. Open the .tar.bz2 archive (or .dmg on OSX)
>
> 2. Move the SageMath folder/app to where you want it to be. You can
>also rename the directory now.
>
> 3. Start sage for the first time. This will then automatically patch
>paths in binaries.
>
> After starting Sage for the first time you cannot change the
> installation any more. To install Sage elsewhere, start over from the
> binary package. Or recompile Sage from scratch in the new location
> ("make distclean && make")
>
> dpkg: error processing package sagemath-upstream-binary (--configure):
>  subprocess installed post-installation script returned error exit status 1
> Errors were encountered while processing:
>  sagemath-upstream-binary
> E: Sub-process /usr/bin/dpkg returned an error code (1)
>
>

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Jan Groenewald]

Hi

On 4 February 2016 at 23:04, Jan Groenewald  wrote:

> Hi
>
> On 4 February 2016 at 22:58, Volker Braun  wrote:
>
>> On Thursday, February 4, 2016 at 9:45:49 PM UTC+1, Jan Groenewald wrote:
>>>
>>> Most software go ./configure --prefix=path; make; make install
>>> and installs to elsewhere with known paths for shared libraries, or
>>> relative paths for in-package libraries and you can move the tree.
>>>
>>
>> The above command will install *absolute paths* for libraries not in
>> standard locations.
>>
>> The actual use is also more complicated, distribution generally try to
>> use "make DESTDIR=/stagingpath install" to package files without installing
>> them in the prefix.
>>
>> If the package buildsystem does not support this, distributions often
>> fall back to post-link binary modifications as well (e.g. using patchelf)
>>
>> Conda (which is probably the closest to what Sage is trying to do, namely
>> distribute a pre-compiled python stack that can be installed anywhere in
>> the home directory) is doing exactly what Sage is now doing.
>>
>
So it seems (with difficulty) Conda patches binaries as well:

http://conda.pydata.org/docs/building/meta-yaml.html#relocatable

To make them use relative paths...?

Regards,
Jan

-- 
  .~.
  /V\ Jan Groenewald
 /( )\www.aims.ac.za
 ^^-^^

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: error rebuilding sage

[Dima Pasechnik]

On Thursday, 4 February 2016 21:00:03 UTC, jhonrubia6 wrote:
>
> Ok, that explains it (I keep learning, even the obvious). So if I 
> understand correctly I should have merged first, before rebuilding the 
> source. Now, it seems there is no  way back I guess, other than getting a 
> new copy of the source code.
> How do I know the version of a given ticket which needs review?
>

typically for reviewing you don't even want to know, just merge over the 
current develop branch.
(or use git trac)

anyhow, if you don't want to check out the branch, 
on the ticket click on "Commits" next to "branch", and then on "Summary"
e.g. for this ticket you will come to
http://git.sagemath.org/sage.git/?h=a9a45bf4314373cda000c34cf8296b457acb5fde
Then in "Age" section you will see: 
Updated Sage version to 6.8.beta7

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Volker Braun]

On Thursday, February 4, 2016 at 9:45:49 PM UTC+1, Jan Groenewald wrote:
>
> Most software go ./configure --prefix=path; make; make install 
> and installs to elsewhere with known paths for shared libraries, or 
> relative paths for in-package libraries and you can move the tree.
>

The above command will install *absolute paths* for libraries not in 
standard locations. 

The actual use is also more complicated, distribution generally try to use 
"make DESTDIR=/stagingpath install" to package files without installing 
them in the prefix. 

If the package buildsystem does not support this, distributions often fall 
back to post-link binary modifications as well (e.g. using patchelf)

Conda (which is probably the closest to what Sage is trying to do, namely 
distribute a pre-compiled python stack that can be installed anywhere in 
the home directory) is doing exactly what Sage is now doing.

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: error rebuilding sage

[jhonrubia6]

Ok, that explains it (I keep learning, even the obvious). So if I 
understand correctly I should have merged first, before rebuilding the 
source. Now, it seems there is no  way back I guess, other than getting a 
new copy of the source code.
How do I know the version of a given ticket which needs review?


El jueves, 4 de febrero de 2016, 21:49:36 (UTC+1), Volker Braun escribió:
>
> Just to point out the obvious: If you check out #18408 then you get 
> Sage-6.8.beta7
>
> You should merge in the latest Sage if you want to review that ticket; Its 
> likely that there are merge conflicts anyways which need to be resolved 
> first.
>
> On Thursday, February 4, 2016 at 7:52:20 PM UTC+1, Jeroen Demeyer wrote:
>>
>> On 2016-02-04 19:32, jhonrubia6 wrote: 
>> > I passed that point and now fails building ecm,... 
>> > I do not understand what I am making wrong. I had a working 
>> installation 
>> > on 7.0, then after checking out a ticket to review (#18408)I tried to 
>> > build again and I'm having a lot of problems to get the installation 
>> > working again 
>>
>> This should be fixed by 
>> http://trac.sagemath.org/ticket/19233 
>>
>> Which version of Sage is that? 
>>
>

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Jan Groenewald]

Hi

On 4 February 2016 at 22:58, Volker Braun  wrote:

> On Thursday, February 4, 2016 at 9:45:49 PM UTC+1, Jan Groenewald wrote:
>>
>> Most software go ./configure --prefix=path; make; make install
>> and installs to elsewhere with known paths for shared libraries, or
>> relative paths for in-package libraries and you can move the tree.
>>
>
> The above command will install *absolute paths* for libraries not in
> standard locations.
>
> The actual use is also more complicated, distribution generally try to use
> "make DESTDIR=/stagingpath install" to package files without installing
> them in the prefix.
>
> If the package buildsystem does not support this, distributions often fall
> back to post-link binary modifications as well (e.g. using patchelf)
>
> Conda (which is probably the closest to what Sage is trying to do, namely
> distribute a pre-compiled python stack that can be installed anywhere in
> the home directory) is doing exactly what Sage is now doing.
>

OK, so Sage is trying to make something you can install anywhere in the
homedir (like on a laptop for a single user) not a system-wide install via
a package manager (like in a computer lab with low personal disk space
quotas; or laptop users with inexperience with self-compile skills)? And
places like sagemathcloud or sage servers for multiple users all compile in
place?

Regards,
Jan



-- 
  .~.
  /V\ Jan Groenewald
 /( )\www.aims.ac.za
 ^^-^^

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Relocating sage

[John H Palmieri]

Should the model when building from scratch be

./configure --prefix=/target/location
make
make install

? If not, should we aim for that? Does "make install" do the right thing 
these days? It's marked as "experimental" in Makefile.

And for people downloading pre-built binaries, should there be a similar 
command that they run to install Sage? I mean something other than "mv 
sage-7.0 ", because using "mv" suggests that you can later move the 
directory again. We could have a top-level shell script "install-sage" that 
takes a directory as an argument (and does not copy the script 
"install-sage" when it installs everything else).

-- 
John

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Relocating sage

[Jeroen Demeyer]

On 2016-02-05 00:24, John H Palmieri wrote:

Should the model when building from scratch be

./configure --prefix=/target/location
make
make install


One thing which we could try is to make it such that

./configure --prefix=/target/location
make

installs in /target/location. Like Volker said, we cannot have separate 
build and install steps.


To implement this, we should set SAGE_LOCAL to /target/location. I'm 
sure that changing SAGE_LOCAL will break stuff, but maybe not so much 
these days...


Jeroen.

--
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Relocating sage

[Jeroen Demeyer]

On 2016-02-05 01:44, John H Palmieri wrote:

run the relocation script.


There is no such thing anymore. Sage can no longer be relocated.

--
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Relocating sage

[Volker Braun]

On Friday, February 5, 2016 at 12:24:27 AM UTC+1, John H Palmieri wrote:
>
> Should the model when building from scratch be
> ./configure --prefix=/target/location
> make
> make install
>

This basically doesn't work if you compile your own dependencies; You have 
to "make install" you dependencies before you can use them in later stages 
of "make".

It can work for sage-the-library only, and we should aim for that 
eventually. 

For sage-the-distribution not so much. There is a long list of desirable 
things that our current build system can't do: No reliable incremental 
builds, no modular binary packages, ... One possibility would be to use 
hashdist for our source builds and extend it to produce conda-compatible 
binary packages (as Aron proposed before) for private $HOME installs. And 
publish our own rpm/debs for system-wide installation, possibly using 
distro packages for dependencies when possible.

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Relocating sage

[John H Palmieri]

On Thursday, February 4, 2016 at 3:43:47 PM UTC-8, Volker Braun wrote:
>
> On Friday, February 5, 2016 at 12:24:27 AM UTC+1, John H Palmieri wrote:
>>
>> Should the model when building from scratch be
>> ./configure --prefix=/target/location
>> make
>> make install
>>
>
> This basically doesn't work if you compile your own dependencies; You have 
> to "make install" you dependencies before you can use them in later stages 
> of "make".
>

I guess that I'm asking whether "make install" should just copy the 
appropriate parts and then run the relocation script. That is, we have a 
way of installing Sage, so why not make it a make target?
 

>
> It can work for sage-the-library only, and we should aim for that 
> eventually. 
>
> For sage-the-distribution not so much. There is a long list of desirable 
> things that our current build system can't do: No reliable incremental 
> builds, no modular binary packages, ... One possibility would be to use 
> hashdist for our source builds and extend it to produce conda-compatible 
> binary packages (as Aron proposed before) for private $HOME installs. And 
> publish our own rpm/debs for system-wide installation, possibly using 
> distro packages for dependencies when possible.
>

Same question here.

-- 
John
 

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Jeroen Demeyer]

On 2016-02-04 15:09, Jan Groenewald wrote:

But can I say most binaries should not have hard-coded paths patched
just after compilation?
You are right, but only because "most binaries" are compiled with the 
correct path in the first place. You have that option with Sage: if you 
compile everything in the destination directory to begin with, 
everything will work fine.


The binary patching is a second option if you don't want hardcoded paths 
in the binary you download.


Jeroen.

--
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[kcrisman]

>
>
> I also agree with Volker that most software is *not* meant to be 
> relocatable. I wonder why you think the opposite. 
>
>
To be fair (though this has nothing to do with pipedream's situation), if 
you are a GUI-only Mac user (say), you might think software is infinitely 
relocatable because it lets you drag and drop things all over the place 
with (usually) no apparent side effects.  Even though most people just 
leave things in the default Applications directory, and even though 
presumably some software would barf if you did try to relocate it.  I'm not 
sure about the Windows situation but I wouldn't be surprised if at least 
some applications "look" like you can do it there too.

But as long as we are very clear about where it needs to be installed on a 
given system, I think most users could live with that. 

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: error rebuilding sage

[jhonrubia6]

... and the new error  (on make distclean && make) is 

/Users/J_Honrubia/Sage/sage-6.10/sage/build/pipestatus "sage-spkg -f 
setuptools-12.4 2>&1" "tee -a 
/Users/J_Honrubia/Sage/sage-6.10/sage/logs/pkgs/setuptools-12.4.log"

/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-env: line 434: 
/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/R/share/Makevars.site: No 
such file or directory

Found local metadata for setuptools-12.4

Attempting to download package setuptools-12.4.tar.gz from mirrors

Downloading the Sage mirror list

Traceback (most recent call last):

  File "/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-download-file", 
line 457, in 

tarball.download()

  File "/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-download-file", 
line 392, in download

for mirror in MirrorList():

  File "/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-download-file", 
line 152, in __init__

with contextlib.closing(urllib.urlopen(self.URL)) as f:

  File "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/urllib.py", 
line 87, in urlopen

return opener.open(url)

  File "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/urllib.py", 
line 213, in open

return getattr(self, name)(url)

  File "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/urllib.py", 
line 350, in open_http

h.endheaders(data)

  File "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", 
line 997, in endheaders

self._send_output(message_body)

  File "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", 
line 850, in _send_output

self.send(msg)

  File "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", 
line 812, in send

self.connect()

  File "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", 
line 793, in connect

self.timeout, self.source_address)

  File "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/socket.py", 
line 555, in create_connection

for res in getaddrinfo(host, port, 0, SOCK_STREAM):

IOError: [Errno socket error] [Errno 8] nodename nor servname provided, or 
not known

make[2]: *** 
[/Users/J_Honrubia/Sage/sage-6.10/sage/local/var/lib/sage/installed/setuptools-12.4]
 
Error 1

make[1]: *** [all] Error 2


real 729m8.876s

user 123m22.857s

sys 17m48.141s

***

Error building Sage.


The following package(s) may have failed to build:


The build directory may contain configuration files and other potentially

helpful information. WARNING: if you now run 'make' again, the build

directory will, by default, be deleted. Set the environment variable

SAGE_KEEP_BUILT_SPKGS to 'yes' to prevent this.


make: *** [all] Error 1

O

El miércoles, 3 de febrero de 2016, 18:41:43 (UTC+1), Volker Braun escribió:
>
>
>
> On Wednesday, February 3, 2016 at 6:25:46 PM UTC+1, jhonrubia6 wrote:
>>
>> after make dist-clean && make I've got an error 
>>
>
> its "make distclean" 
>

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Jeroen Demeyer]

On 2016-02-04 13:58, Jan Groenewald wrote:

I thought it would be more like moving / since it was completely
self-contained.

It doesn't matter: I'm sure that moving / will break everything too.

I also agree with Volker that most software is *not* meant to be 
relocatable. I wonder why you think the opposite.


Jeroen.

--
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: error rebuilding sage

[jhonrubia6]

As the compilation dos through the night, maybe there was some internet 
downtime that I am not aware of. I will try again and keep track of the 
router log, thank you for your help

El jueves, 4 de febrero de 2016, 17:36:46 (UTC+1), Dima Pasechnik escribió:
>
> did the internet connection from the machine die during the build, or 
> perhaps you are behind some big bad firewall?
>
> On Thursday, 4 February 2016 15:21:59 UTC, jhonrubia6 wrote:
>>
>> ... and the new error  (on make distclean && make) is 
>>
>> /Users/J_Honrubia/Sage/sage-6.10/sage/build/pipestatus "sage-spkg -f 
>> setuptools-12.4 2>&1" "tee -a 
>> /Users/J_Honrubia/Sage/sage-6.10/sage/logs/pkgs/setuptools-12.4.log"
>>
>> /Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-env: line 434: 
>> /Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/R/share/Makevars.site: No 
>> such file or directory
>>
>> Found local metadata for setuptools-12.4
>>
>> Attempting to download package setuptools-12.4.tar.gz from mirrors
>>
>> Downloading the Sage mirror list
>>
>> Traceback (most recent call last):
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-download-file", line 
>> 457, in 
>>
>> tarball.download()
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-download-file", line 
>> 392, in download
>>
>> for mirror in MirrorList():
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-download-file", line 
>> 152, in __init__
>>
>> with contextlib.closing(urllib.urlopen(self.URL)) as f:
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/urllib.py", line 
>> 87, in urlopen
>>
>> return opener.open(url)
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/urllib.py", line 
>> 213, in open
>>
>> return getattr(self, name)(url)
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/urllib.py", line 
>> 350, in open_http
>>
>> h.endheaders(data)
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", line 
>> 997, in endheaders
>>
>> self._send_output(message_body)
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", line 
>> 850, in _send_output
>>
>> self.send(msg)
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", line 
>> 812, in send
>>
>> self.connect()
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", line 
>> 793, in connect
>>
>> self.timeout, self.source_address)
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/socket.py", line 
>> 555, in create_connection
>>
>> for res in getaddrinfo(host, port, 0, SOCK_STREAM):
>>
>> IOError: [Errno socket error] [Errno 8] nodename nor servname provided, 
>> or not known
>>
>> make[2]: *** 
>> [/Users/J_Honrubia/Sage/sage-6.10/sage/local/var/lib/sage/installed/setuptools-12.4]
>>  
>> Error 1
>>
>> make[1]: *** [all] Error 2
>>
>>
>> real 729m8.876s
>>
>> user 123m22.857s
>>
>> sys 17m48.141s
>>
>> ***
>>
>> Error building Sage.
>>
>>
>> The following package(s) may have failed to build:
>>
>>
>> The build directory may contain configuration files and other potentially
>>
>> helpful information. WARNING: if you now run 'make' again, the build
>>
>> directory will, by default, be deleted. Set the environment variable
>>
>> SAGE_KEEP_BUILT_SPKGS to 'yes' to prevent this.
>>
>>
>> make: *** [all] Error 1
>>
>> O
>>
>> El miércoles, 3 de febrero de 2016, 18:41:43 (UTC+1), Volker Braun 
>> escribió:
>>>
>>>
>>>
>>> On Wednesday, February 3, 2016 at 6:25:46 PM UTC+1, jhonrubia6 wrote:

 after make dist-clean && make I've got an error 

>>>
>>> its "make distclean" 
>>>
>>

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Jan Groenewald]

Hi

On 4 February 2016 at 14:23, Jeroen Demeyer  wrote:

> Relocating Sage is like relocating /usr to /otherpath/usr in your favorite
> Linux distro. I bet things will break down completely.
>

I thought it would be more like moving / since it was completely
self-contained.

Regards,
Jan




-- 
  .~.
  /V\ Jan Groenewald
 /( )\www.aims.ac.za
 ^^-^^

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Jan Groenewald]

Hi

On 4 February 2016 at 15:04, Jeroen Demeyer  wrote:

> On 2016-02-04 13:58, Jan Groenewald wrote:
>
>> I thought it would be more like moving / since it was completely
>> self-contained.
>>
> It doesn't matter: I'm sure that moving / will break everything too.
>
> I also agree with Volker that most software is *not* meant to be
> relocatable. I wonder why you think the opposite.


OK, I'll retract that, perhaps I was thinking of software "trees",
including all shared libs and so on.

But can I say most binaries should not have hard-coded paths patched just
after compilation?

Regards,
Jan



-- 
  .~.
  /V\ Jan Groenewald
 /( )\www.aims.ac.za
 ^^-^^

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: error rebuilding sage

[Dima Pasechnik]

did the internet connection from the machine die during the build, or 
perhaps you are behind some big bad firewall?

On Thursday, 4 February 2016 15:21:59 UTC, jhonrubia6 wrote:
>
> ... and the new error  (on make distclean && make) is 
>
> /Users/J_Honrubia/Sage/sage-6.10/sage/build/pipestatus "sage-spkg -f 
> setuptools-12.4 2>&1" "tee -a 
> /Users/J_Honrubia/Sage/sage-6.10/sage/logs/pkgs/setuptools-12.4.log"
>
> /Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-env: line 434: 
> /Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/R/share/Makevars.site: No 
> such file or directory
>
> Found local metadata for setuptools-12.4
>
> Attempting to download package setuptools-12.4.tar.gz from mirrors
>
> Downloading the Sage mirror list
>
> Traceback (most recent call last):
>
>   File "/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-download-file", 
> line 457, in 
>
> tarball.download()
>
>   File "/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-download-file", 
> line 392, in download
>
> for mirror in MirrorList():
>
>   File "/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-download-file", 
> line 152, in __init__
>
> with contextlib.closing(urllib.urlopen(self.URL)) as f:
>
>   File "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/urllib.py", 
> line 87, in urlopen
>
> return opener.open(url)
>
>   File "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/urllib.py", 
> line 213, in open
>
> return getattr(self, name)(url)
>
>   File "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/urllib.py", 
> line 350, in open_http
>
> h.endheaders(data)
>
>   File 
> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", line 
> 997, in endheaders
>
> self._send_output(message_body)
>
>   File 
> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", line 
> 850, in _send_output
>
> self.send(msg)
>
>   File 
> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", line 
> 812, in send
>
> self.connect()
>
>   File 
> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", line 
> 793, in connect
>
> self.timeout, self.source_address)
>
>   File "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/socket.py", 
> line 555, in create_connection
>
> for res in getaddrinfo(host, port, 0, SOCK_STREAM):
>
> IOError: [Errno socket error] [Errno 8] nodename nor servname provided, or 
> not known
>
> make[2]: *** 
> [/Users/J_Honrubia/Sage/sage-6.10/sage/local/var/lib/sage/installed/setuptools-12.4]
>  
> Error 1
>
> make[1]: *** [all] Error 2
>
>
> real 729m8.876s
>
> user 123m22.857s
>
> sys 17m48.141s
>
> ***
>
> Error building Sage.
>
>
> The following package(s) may have failed to build:
>
>
> The build directory may contain configuration files and other potentially
>
> helpful information. WARNING: if you now run 'make' again, the build
>
> directory will, by default, be deleted. Set the environment variable
>
> SAGE_KEEP_BUILT_SPKGS to 'yes' to prevent this.
>
>
> make: *** [all] Error 1
>
> O
>
> El miércoles, 3 de febrero de 2016, 18:41:43 (UTC+1), Volker Braun 
> escribió:
>>
>>
>>
>> On Wednesday, February 3, 2016 at 6:25:46 PM UTC+1, jhonrubia6 wrote:
>>>
>>> after make dist-clean && make I've got an error 
>>>
>>
>> its "make distclean" 
>>
>

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: error rebuilding sage

[Jeroen Demeyer]

On 2016-02-04 19:32, jhonrubia6 wrote:

I passed that point and now fails building ecm,...
I do not understand what I am making wrong. I had a working installation
on 7.0, then after checking out a ticket to review (#18408)I tried to
build again and I'm having a lot of problems to get the installation
working again


This should be fixed by
http://trac.sagemath.org/ticket/19233

Which version of Sage is that?

--
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: error rebuilding sage

[jhonrubia6]

I passed that point and now fails building ecm,... 
I do not understand what I am making wrong. I had a working installation on 
7.0, then after checking out a ticket to review (#18408)I tried to build 
again and I'm having a lot of problems to get the installation working again

Now building GMP-ECM...

make  all-recursive

Making all in x86_64

m4 -I../ -DOPERATION_mulredc1 `test -f mulredc1.asm || echo 
'./'`mulredc1.asm >mulredc1.s

/bin/sh ../libtool   --mode=compile gcc  -march=native -g -O3  -fPIC -c -o 
mulredc1.lo mulredc1.s

libtool: compile:  gcc -march=native -g -O3 -fPIC -c mulredc1.s -o 
mulredc1.o

m4 -I../ -DOPERATION_mulredc2 `test -f mulredc2.asm || echo 
'./'`mulredc2.asm >mulredc2.s

/bin/sh ../libtool   --mode=compile gcc  -march=native -g -O3  -fPIC -c -o 
mulredc2.lo mulredc2.s

libtool: compile:  gcc -march=native -g -O3 -fPIC -c mulredc2.s -o 
mulredc2.o

mulredc2.s:40:8: error: invalid alignment value

.align 64 # Opteron L1 code cache line is 64 bytes long

   ^

mulredc2.s:150:8: error: invalid alignment value

.align 32,,16

   ^

make[5]: *** [mulredc2.lo] Error 1

rm mulredc2.s mulredc1.s

make[4]: *** [all-recursive] Error 1

make[3]: *** [all] Error 2

Error building GMP-ECM.


real 0m20.742s

user 0m8.886s

sys 0m8.923s



Error installing package ecm-6.4.4



Please email sage-devel (http://groups.google.com/group/sage-devel)

explaining the problem and including the relevant part of the log file

  /Users/J_Honrubia/Sage/sage-6.10/sage/logs/pkgs/ecm-6.4.4.log

Describe your computer, operating system, etc.

If you want to try to fix the problem yourself, *don't* just cd to

/Users/J_Honrubia/Sage/sage-6.10/sage/local/var/tmp/sage/build/ecm-6.4.4 
and type 'make' or whatever is appropriate.

Instead, the following commands setup all environment variables

correctly and load a subshell for you to debug the error:

  (cd 
'/Users/J_Honrubia/Sage/sage-6.10/sage/local/var/tmp/sage/build/ecm-6.4.4' 
&& '/Users/J_Honrubia/Sage/sage-6.10/sage/sage' --sh)

When you are done debugging, you can type "exit" to leave the subshell.



make[2]: *** 
[/Users/J_Honrubia/Sage/sage-6.10/sage/local/var/lib/sage/installed/ecm-6.4.4] 
Error 1

make[1]: *** [all] Error 2


real 0m22.540s

user 0m9.866s

sys 0m9.371s

***

Error building Sage.


The following package(s) may have failed to build:


package: ecm-6.4.4

log file: /Users/J_Honrubia/Sage/sage-6.10/sage/logs/pkgs/ecm-6.4.4.log

build directory: 
/Users/J_Honrubia/Sage/sage-6.10/sage/local/var/tmp/sage/build/ecm-6.4.4


The build directory may contain configuration files and other potentially

helpful information. WARNING: if you now run 'make' again, the build

directory will, by default, be deleted. Set the environment variable

SAGE_KEEP_BUILT_SPKGS to 'yes' to prevent this.


make: *** [all] Error 1



El jueves, 4 de febrero de 2016, 17:36:46 (UTC+1), Dima Pasechnik escribió:
>
> did the internet connection from the machine die during the build, or 
> perhaps you are behind some big bad firewall?
>
> On Thursday, 4 February 2016 15:21:59 UTC, jhonrubia6 wrote:
>>
>> ... and the new error  (on make distclean && make) is 
>>
>> /Users/J_Honrubia/Sage/sage-6.10/sage/build/pipestatus "sage-spkg -f 
>> setuptools-12.4 2>&1" "tee -a 
>> /Users/J_Honrubia/Sage/sage-6.10/sage/logs/pkgs/setuptools-12.4.log"
>>
>> /Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-env: line 434: 
>> /Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/R/share/Makevars.site: No 
>> such file or directory
>>
>> Found local metadata for setuptools-12.4
>>
>> Attempting to download package setuptools-12.4.tar.gz from mirrors
>>
>> Downloading the Sage mirror list
>>
>> Traceback (most recent call last):
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-download-file", line 
>> 457, in 
>>
>> tarball.download()
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-download-file", line 
>> 392, in download
>>
>> for mirror in MirrorList():
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/src/bin/sage-download-file", line 
>> 152, in __init__
>>
>> with contextlib.closing(urllib.urlopen(self.URL)) as f:
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/urllib.py", line 
>> 87, in urlopen
>>
>> return opener.open(url)
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/urllib.py", line 
>> 213, in open
>>
>> return getattr(self, name)(url)
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/urllib.py", line 
>> 350, in open_http
>>
>> h.endheaders(data)
>>
>>   File 
>> "/Users/J_Honrubia/Sage/sage-6.10/sage/local/lib/python/httplib.py", line 
>> 997, in 

[sage-devel] Re: Failure to Compile Sage 7.0 Stable on openSUSE Linux

[Dima Pasechnik]

Nobody ever built Sage with llvm, so you are in totally unchartered waters 
here.
It would be interesting to know what goes wrong.
It's also a bit unclear how this would play together with gfortran (which 
is part of gcc suite of compilers), 
which is needed for a part of Sage.

HTH,
Dima

On Thursday, 4 February 2016 17:50:52 UTC, Martin Vahi wrote:
>
>
> The compilation of the Sage 7.0 was with LLVM, not GCC. I usually add the 
>
> uname -a
>
> and 
>
> compilername --version
>
> to the end of the console output examples. The LLVM Clang has a version of:
>
> clang version 3.8.0 (trunk 249969)
> Target: x86_64-unknown-linux-gnu
> Thread model: posix
> InstalledDir: 
> /home/ts2/m_local/bin_p/LLVM/vd_2015_10_11/vd_2015_10_11_installed/bin
>
> The GCC has a version of:
>
> gcc (SUSE Linux) 4.8.3 20140627 [gcc-4_8-branch revision 212064]
> Copyright (C) 2013 Free Software Foundation, Inc.
> This is free software; see the source for copying conditions.  There is NO
> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
>
>
> I also tried to build the Sage 6.10. This time with the GCC.
> That failed to compile and the problematic 
> package at the 6.10 compilation seems to be 
>
> ecl-15.3.7p0
>
> The details reside at yet another set of attachments, but the 
> key phrase seems to be 
>
>
> checking whether we can use the existing Boehm-Weiser library ... no 
> configure: Configuring included Boehm GC library:
> /opt/2dot7TiB_k8vaketas/ts2/mittevarundatav/_home/m_local/bin_p/Sage/kompileerimine/sage-6.10/local/var/tmp/sage/build/ecl-15.3.7p0/src/src/configure:
>  
> line 6034: cd: 
> /opt/2dot7TiB_k8vaketas/ts2/mittevarundatav/_home/m_local/bin_p/Sage/kompileerimine/sage-6.10/local/var/tmp/sage/build/ecl-15.3.7p0/src/src/gc-unstable:
>  
> No such file or directory
> /opt/2dot7TiB_k8vaketas/ts2/mittevarundatav/_home/m_local/bin_p/Sage/kompileerimine/sage-6.10/local/var/tmp/sage/build/ecl-15.3.7p0/src/src/configure:
>  
> line 6042: 
> /opt/2dot7TiB_k8vaketas/ts2/mittevarundatav/_home/m_local/bin_p/Sage/kompileerimine/sage-6.10/local/var/tmp/sage/build/ecl-15.3.7p0/src/src/gc-unstable/configure:
>  
> No such file or directory
> configure: error: Unable to configure Boehm-Weiser GC
> Error - Failed to configure ECL ... exiting
>
>
>
>

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Failure to Compile Sage 7.0 Stable on openSUSE Linux

[Dima Pasechnik]

On Thursday, 4 February 2016 17:50:52 UTC, Martin Vahi wrote:
>
>
> I also tried to build the Sage 6.10. This time with the GCC.
> That failed to compile and the problematic 
> package at the 6.10 compilation seems to be 
>
> ecl-15.3.7p0
>
> The details reside at yet another set of attachments, but the 
> key phrase seems to be 
>
>
> checking whether we can use the existing Boehm-Weiser library ... no 
> configure: Configuring included Boehm GC library:
> /opt/2dot7TiB_k8vaketas/ts2/mittevarundatav/_home/m_local/bin_p/Sage/kompileerimine/sage-6.10/local/var/tmp/sage/build/ecl-15.3.7p0/src/src/configure:
>  
> line 6034: cd: 
> /opt/2dot7TiB_k8vaketas/ts2/mittevarundatav/_home/m_local/bin_p/Sage/kompileerimine/sage-6.10/local/var/tmp/sage/build/ecl-15.3.7p0/src/src/gc-unstable:
>  
> No such file or directory
> /opt/2dot7TiB_k8vaketas/ts2/mittevarundatav/_home/m_local/bin_p/Sage/kompileerimine/sage-6.10/local/var/tmp/sage/build/ecl-15.3.7p0/src/src/configure:
>  
> line 6042: 
> /opt/2dot7TiB_k8vaketas/ts2/mittevarundatav/_home/m_local/bin_p/Sage/kompileerimine/sage-6.10/local/var/tmp/sage/build/ecl-15.3.7p0/src/src/gc-unstable/configure:
>  
> No such file or directory
> configure: error: Unable to configure Boehm-Weiser GC
> Error - Failed to configure ECL ... exiting
>
>
> the error is earlier, in building Boehm-Weiser library (Sage package gc). 
In a working build one will see
"checking whether we can use the existing Boehm-Weiser library ... yes"
rather than 'no'

HTH,
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: hardcoded paths [Was: [sage-support] after a long compiling time i got this]

[Jan Groenewald]

Hi

Thanks for taking the time to help me learn.


On 4 February 2016 at 16:15, Jeroen Demeyer  wrote:

> On 2016-02-04 15:09, Jan Groenewald wrote:
>
>> But can I say most binaries should not have hard-coded paths patched
>> just after compilation?
>>
> You are right, but only because "most binaries" are compiled with the
> correct path in the first place. You have that option with Sage: if you
> compile everything in the destination directory to begin with, everything
> will work fine.
>

Most developer and science software are compiled in place.

Nothing on my Debian, Ubuntu, Fedora, Redhat, etc. systems are compiled in
place.  The files are put there by a package manager.

I can't move a binary but I can move a package.

0 jan@kontiki:~$cp /usr/lib/firefox/firefox .; ./firefox # <-- one binary
without libs
Could not find the Mozilla runtime.
255 jan@kontiki:~$rm firefox
0 jan@kontiki:~$cp -R /usr/lib/firefox .; ./firefox/firefox # folder with
interdependent libraries and relative paths
0 jan@kontiki:~$# worked!
0 jan@kontiki:~$

Most software go ./configure --prefix=path; make; make install

and installs to elsewhere with known paths for shared libraries, or
relative paths for in-package libraries and you can move the tree.

Does sage not ever hope to get to that (despite me not knowing the build
process and non-trivial work involved in this?

The binary patching is a second option if you don't want hardcoded paths in
> the binary you download.


hardcoded absolutenot relative (to SAGE_ROOT) paths??

Regards,
Jan



-- 
  .~.
  /V\ Jan Groenewald
 /( )\www.aims.ac.za
 ^^-^^

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: error rebuilding sage

[Volker Braun]

Just to point out the obvious: If you check out #18408 then you get 
Sage-6.8.beta7

You should merge in the latest Sage if you want to review that ticket; Its 
likely that there are merge conflicts anyways which need to be resolved 
first.

On Thursday, February 4, 2016 at 7:52:20 PM UTC+1, Jeroen Demeyer wrote:
>
> On 2016-02-04 19:32, jhonrubia6 wrote: 
> > I passed that point and now fails building ecm,... 
> > I do not understand what I am making wrong. I had a working installation 
> > on 7.0, then after checking out a ticket to review (#18408)I tried to 
> > build again and I'm having a lot of problems to get the installation 
> > working again 
>
> This should be fixed by 
> http://trac.sagemath.org/ticket/19233 
>
> Which version of Sage is that? 
>

-- 
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 https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.