Re: [sage-devel] Re: SPKG Maintainers??

[Jori Mäntysalo]

On Wed, 23 Sep 2015, Simon King wrote:

In other words, you suggest that we will in future have "maintainers" 
for functions or modules in the SageMath library, although we just got 
rid of the good old "SPKG maintainers"?


Good point. Forget this idea.

--
Jori Mäntysalo

[sage-devel] Re: SPKG Maintainers??

[Simon King]

Hi Jori,

On 2015-09-23, Jori =?ISO-8859-1?Q?M=E4ntysalo?=  wrote:
> On Tue, 22 Sep 2015, Andrey Novoseltsev wrote:
>
>> I had some random people contact me with questions about the modules 
>> I've written and I suspect it was due to the AUTHORS block, so it is 
>> useful to users.
>
> On Tue, 22 Sep 2015, Nathann Cohen wrote:
>
>> code is not static like a scientific paper. Contrary to it, code is
>> modified over time and after a while an authors block does not make
>> any sense.
>
> I think that both of the above are correct. So I wild(?) idea...
>
> Could we have some kind of "contact"-chain? Are developers ready to answer 
> questions about their code and documentation? Then the point would shift 
> from "I made this." to "I can debug or clarify this, if needed."?

In other words, you suggest that we will in future have "maintainers"
for functions or modules in the SageMath library, although we just got rid of
the good old "SPKG maintainers"?

Cheers,
Simon

-- 
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] SPKG Maintainers??

[Jori Mäntysalo]

On Tue, 22 Sep 2015, Andrey Novoseltsev wrote:

I had some random people contact me with questions about the modules 
I've written and I suspect it was due to the AUTHORS block, so it is 
useful to users.


On Tue, 22 Sep 2015, Nathann Cohen wrote:


code is not static like a scientific paper. Contrary to it, code is
modified over time and after a while an authors block does not make
any sense.


I think that both of the above are correct. So I wild(?) idea...

Could we have some kind of "contact"-chain? Are developers ready to answer 
questions about their code and documentation? Then the point would shift 
from "I made this." to "I can debug or clarify this, if needed."?


I have no specific idea about how to do it. Something like "contact" for 
whole SageMath would be sage-support, then there could have 
sage-partX-support or person X taking care of large part of Sage, other 
would have "responsibility" for sage-partX_1, and lastly there could be 
function level?


An example of chain:

is_supersolvable < lattices < posets < combinat < SageMath

A person writing .is_supersolvable? would get, say, name of somebody 
"responsible" for posets, if there is nobody handling specially just 
is_supersolvable or lattices.


Is something like this done in any software before?

--
Jori Mäntysalo

[sage-devel] Re: git trac frustration...

[Volker Braun]

Its because the ticket is closed but not yet released; you can't fetch 
commits by sha1. It works if you happen to have the commit in your local 
copy of the trac repo; e.g. run

git fetch trac public/ticket/18594

Then afterwards both 

git fetch trac f257d8224aa79c85a84ced0e69f378fafedbe546

and 

git trac checkout 18594

will work.




On Tuesday, September 22, 2015 at 11:56:20 PM UTC+2, Aram Dermenjian wrote:
>
> Was this every able to get fixed? I'm getting the same problem here on a 
> different ticket
>
> aram:/media/ubuntudata/Programming/sage (develop)$ git trac checkout 18594
> Loading ticket #18594...
> Checking out Trac #18594 remote branch 
> f257d8224aa79c85a84ced0e69f378fafedbe546 -> local branch 
> t/18594/f257d8224aa79c85a84ced0e69f378fafedbe546...
> Traceback (most recent call last):
>   File "/home/aram/bin/git-trac-command/git-trac", line 18, in 
> cmdline.launch()
>   File "/home/aram/bin/git-trac-command/git_trac/cmdline.py", line 209, in 
> launch
> app.checkout(args.ticket_or_branch, args.branch_name)
>   File "/home/aram/bin/git-trac-command/git_trac/app.py", line 115, in 
> checkout
> self._checkout_ticket(int(ticket_or_branch), branch_name)
>   File "/home/aram/bin/git-trac-command/git_trac/app.py", line 139, in 
> _checkout_ticket
> self.repo.checkout_new_branch(ticket.branch, branch)
>   File "/home/aram/bin/git-trac-command/git_trac/git_repository.py", line 
> 123, in checkout_new_branch
> self.git.fetch('trac', remote)
>   File "/home/aram/bin/git-trac-command/git_trac/git_interface.py", line 
> 341, in meth
> return self.execute(git_cmd, *args, **kwds)
>   File "/home/aram/bin/git-trac-command/git_trac/git_interface.py", line 
> 328, in execute
> popen_stderr=subprocess.PIPE)
>   File "/home/aram/bin/git-trac-command/git_trac/git_interface.py", line 
> 263, in _run
> raise GitError(result)
> git_trac.git_error.GitError: git returned with non-zero exit code (1) when 
> executing "git fetch trac f257d8224aa79c85a84ced0e69f378fafedbe546"
> aram:/media/ubuntudata/Programming/sage (develop)$ git remote -v
> originhttps://github.com/sagemath/sage.git (fetch)
> originhttps://github.com/sagemath/sage.git (push)
> tracgit://trac.sagemath.org/sage.git (fetch)
> tracg...@trac.sagemath.org:sage.git (push)
> aram:/media/ubuntudata/Programming/sage (develop)$ git fetch trac 
> f257d8224aa79c85a84ced0e69f378fafedbe546
> aram:/media/ubuntudata/Programming/sage (develop)$ echo $?
> 1
> aram:/media/ubuntudata/Programming/sage (develop)$ git branch
> * develop
>   master
>
> Seems as though it's not wanting to fetch that particular ticket as when I 
> tried to pull another ticket it seemed to work properly. Is it maybe some 
> head conflict that is happening where there is a potentially cached version 
> of the head somewhere and so the conflict is causing us not to be able to 
> download the branch even though it's been deleted? (BTW my git version is 
> 1.9.1)
>
>
> On Monday, April 6, 2015 at 10:49:37 AM UTC-4, Volker Braun wrote:
>>
>> $ git fetch trac e070dd191bb189b83bae5a2a5c12bf4d70981040
>> From trac.sagemath.org:sage
>>  * branche070dd191bb189b83bae5a2a5c12bf4d70981040 -> 
>> FETCH_HEAD
>> $ echo $?
>> 0
>>
>> Interesting, I get return code 0. Did it actually fetch anything (ie. 
>> FETCH_HEAD is set)? Why didn't it print anything to stdout?
>>
>> $ git --version
>> git version 2.1.0
>>
>>
>>
>>
>> On Monday, April 6, 2015 at 4:36:40 PM UTC+2, William wrote:
>>>
>>> I took a not-so clean Sage-6.5 install, installed git-trac (with no 
>>> keys or login/password) and tried to check out the code from ticket 
>>> 17826 (upgrade Ipython).  I just get some mysterious errors.How am 
>>> I supposed to debug this? 
>>>
>>> salvus@devel1dc5:/usr/local/sage/current$ git trac checkout 17826 
>>> Loading ticket #17826... 
>>> Checking out Trac #17826 remote branch 
>>> e070dd191bb189b83bae5a2a5c12bf4d70981040 -> local branch 
>>> t/17826/e070dd191bb189b83bae5a2a5c12bf4d70981040 
>>> ... 
>>> Traceback (most recent call last): 
>>>   File "/home/salvus/git-trac-command/bin/git-trac", line 18, in 
>>>  
>>> cmdline.launch() 
>>>   File "/home/salvus/git-trac-command/git_trac/cmdline.py", line 205, in 
>>> launch 
>>> app.checkout(args.ticket_or_branch, args.branch_name) 
>>>   File "/home/salvus/git-trac-command/git_trac/app.py", line 98, in 
>>> checkout 
>>> self._checkout_ticket(int(ticket_or_branch), branch_name) 
>>>   File "/home/salvus/git-trac-command/git_trac/app.py", line 122, in 
>>> _checkout_ticket 
>>> self.repo.checkout_new_branch(ticket.branch, branch) 
>>>   File "/home/salvus/git-trac-command/git_trac/git_repository.py", 
>>> line 123, in checkout_new_branch 
>>> self.git.fetch('trac', remote) 
>>>   File "/home/salvus/git-trac-command/git_trac/git_interface.py", line 
>>> 341, in meth 
>>> return self.execute(git_cmd, *args, **kwds) 
>>>   File "/home/salvus/git-trac-command/git_trac/git_inter

[sage-devel] Re: git trac frustration...

[Aram Dermenjian]

Was this every able to get fixed? I'm getting the same problem here on a 
different ticket

aram:/media/ubuntudata/Programming/sage (develop)$ git trac checkout 18594
Loading ticket #18594...
Checking out Trac #18594 remote branch 
f257d8224aa79c85a84ced0e69f378fafedbe546 -> local branch 
t/18594/f257d8224aa79c85a84ced0e69f378fafedbe546...
Traceback (most recent call last):
  File "/home/aram/bin/git-trac-command/git-trac", line 18, in 
cmdline.launch()
  File "/home/aram/bin/git-trac-command/git_trac/cmdline.py", line 209, in 
launch
app.checkout(args.ticket_or_branch, args.branch_name)
  File "/home/aram/bin/git-trac-command/git_trac/app.py", line 115, in 
checkout
self._checkout_ticket(int(ticket_or_branch), branch_name)
  File "/home/aram/bin/git-trac-command/git_trac/app.py", line 139, in 
_checkout_ticket
self.repo.checkout_new_branch(ticket.branch, branch)
  File "/home/aram/bin/git-trac-command/git_trac/git_repository.py", line 
123, in checkout_new_branch
self.git.fetch('trac', remote)
  File "/home/aram/bin/git-trac-command/git_trac/git_interface.py", line 
341, in meth
return self.execute(git_cmd, *args, **kwds)
  File "/home/aram/bin/git-trac-command/git_trac/git_interface.py", line 
328, in execute
popen_stderr=subprocess.PIPE)
  File "/home/aram/bin/git-trac-command/git_trac/git_interface.py", line 
263, in _run
raise GitError(result)
git_trac.git_error.GitError: git returned with non-zero exit code (1) when 
executing "git fetch trac f257d8224aa79c85a84ced0e69f378fafedbe546"
aram:/media/ubuntudata/Programming/sage (develop)$ git remote -v
originhttps://github.com/sagemath/sage.git (fetch)
originhttps://github.com/sagemath/sage.git (push)
tracgit://trac.sagemath.org/sage.git (fetch)
tracg...@trac.sagemath.org:sage.git (push)
aram:/media/ubuntudata/Programming/sage (develop)$ git fetch trac 
f257d8224aa79c85a84ced0e69f378fafedbe546
aram:/media/ubuntudata/Programming/sage (develop)$ echo $?
1
aram:/media/ubuntudata/Programming/sage (develop)$ git branch
* develop
  master

Seems as though it's not wanting to fetch that particular ticket as when I 
tried to pull another ticket it seemed to work properly. Is it maybe some 
head conflict that is happening where there is a potentially cached version 
of the head somewhere and so the conflict is causing us not to be able to 
download the branch even though it's been deleted? (BTW my git version is 
1.9.1)


On Monday, April 6, 2015 at 10:49:37 AM UTC-4, Volker Braun wrote:
>
> $ git fetch trac e070dd191bb189b83bae5a2a5c12bf4d70981040
> From trac.sagemath.org:sage
>  * branche070dd191bb189b83bae5a2a5c12bf4d70981040 -> FETCH_HEAD
> $ echo $?
> 0
>
> Interesting, I get return code 0. Did it actually fetch anything (ie. 
> FETCH_HEAD is set)? Why didn't it print anything to stdout?
>
> $ git --version
> git version 2.1.0
>
>
>
>
> On Monday, April 6, 2015 at 4:36:40 PM UTC+2, William wrote:
>>
>> I took a not-so clean Sage-6.5 install, installed git-trac (with no 
>> keys or login/password) and tried to check out the code from ticket 
>> 17826 (upgrade Ipython).  I just get some mysterious errors.How am 
>> I supposed to debug this? 
>>
>> salvus@devel1dc5:/usr/local/sage/current$ git trac checkout 17826 
>> Loading ticket #17826... 
>> Checking out Trac #17826 remote branch 
>> e070dd191bb189b83bae5a2a5c12bf4d70981040 -> local branch 
>> t/17826/e070dd191bb189b83bae5a2a5c12bf4d70981040 
>> ... 
>> Traceback (most recent call last): 
>>   File "/home/salvus/git-trac-command/bin/git-trac", line 18, in  
>> cmdline.launch() 
>>   File "/home/salvus/git-trac-command/git_trac/cmdline.py", line 205, in 
>> launch 
>> app.checkout(args.ticket_or_branch, args.branch_name) 
>>   File "/home/salvus/git-trac-command/git_trac/app.py", line 98, in 
>> checkout 
>> self._checkout_ticket(int(ticket_or_branch), branch_name) 
>>   File "/home/salvus/git-trac-command/git_trac/app.py", line 122, in 
>> _checkout_ticket 
>> self.repo.checkout_new_branch(ticket.branch, branch) 
>>   File "/home/salvus/git-trac-command/git_trac/git_repository.py", 
>> line 123, in checkout_new_branch 
>> self.git.fetch('trac', remote) 
>>   File "/home/salvus/git-trac-command/git_trac/git_interface.py", line 
>> 341, in meth 
>> return self.execute(git_cmd, *args, **kwds) 
>>   File "/home/salvus/git-trac-command/git_trac/git_interface.py", line 
>> 328, in execute 
>> popen_stderr=subprocess.PIPE) 
>>   File "/home/salvus/git-trac-command/git_trac/git_interface.py", line 
>> 263, in _run 
>> raise GitError(result) 
>> git_trac.git_error.GitError: git returned with non-zero exit code (1) 
>> when executing "git fetch trac 
>> e070dd191bb189b83bae5a2a5c12bf4d70981040" 
>> salvus@devel1dc5:/usr/local/sage/current$ git fetch trac 
>> e070dd191bb189b83bae5a2a5c12bf4d70981040 
>> salvus@devel1dc5:/usr/local/sage/current$ echo $? 
>> 1 
>> salvus@devel1dc5:/usr/local/sage/current$ git branch 
>> * deve

[sage-devel] Where are the results of "make test"?

[Simon King]

Hi!

I have mentioned the following problem before, but didn't get an answer
(and the problem has not been fixed).

If I do "make test", then $SAGE_ROOT/logs/test.log tells which tests have
failed. However, it does *not* tell *how* they have failed. I.e., it
does not show the expected output versus the received output. That
information is visible in the terminal, but is *not* visible in the log
file.

I opened #19271 for it. Just to draw the attention of test framework
experts to the problem...

Best regards,
Simon

-- 
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: SPKG Maintainers??

[William Stein]

On Tue, Sep 22, 2015 at 2:14 PM, Nathann Cohen  wrote:
> Simon,
>
> By reading your message I am under the impression that you actually
> believe that the information contained in 'authors' block is somehow
> consistent or even remotely correct.
>
> It is not.
>
> *because* it is mainained manually.

This argument is unconvincing to me; it's too black and white.  You
could also make the same argument about the majority of content in
every research paper ever written.  Just because "people" do things
doesn't automatically imply that what they do is not even remotely
correct.   And our extensive testing framework in Sage -- which at
least helps -- doesn't at all imply that the code in Sage is correct.

William

> Which means that we forget to
> update it, that those who contribute a lot are those who will not
> leave their name. Because it is done manually, it is totally
> unreliable.
>
> The trac server, at least, keeps its promises: each ticket has an author 
> field.
> Git blame, at least, keeps its promises: it tells you who last modified a 
> line.
>
> You will not solve the problem by having us maintain manually a
> changelog in every single function.
>
> I prefer to do it wikipedia-style: "what matters here is the code, not
> who wrote it" (except for debugging purposes -> git blame). And if I
> need to explain to someone that I am a Sage contributor, they will
> find what they need on my home page.
>
> Nathann
> http://www.steinertriples.fr/ncohen/
>
> --
> 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 (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.

Re: [sage-devel] updating sage-env LD_LIBRARY_PATH from spkg-install

[François Bissey]

One design that comes to mind is how distro populate their default
environment. Instead of shoving everything in a single file there is a
top file that source the content of a directory. Depending on your
distro it could be in /etc/env.d or /etc/profile.d.
So individual package can put particular settings for their package
in this folder and the next time you run you have an updated environment.

It probably would serve sage well but it is a redesign and it may hurt
a bit at first. Especially while the work is in progress.

Francois

On 09/23/15 05:43, Bill Janssen wrote:
> Yes, but in the absence of an achieved utopia, is there something we can
> do here?  I see that sage-env already has a lot of LD_LIBRARY_PATH
> hacking in it.  Might be good to design a system for regularizing this,
> and moving all this hacking to that system.
> 
> My solution has been to add Python code to detect this, and put out a
> warning to the user, telling them to add the required LD_LIBRARY_PATH
> setting to their sagerc file.  Somewhat unsatisfactory.
> 
> Bill
> 
> On Monday, September 21, 2015 at 10:49:05 AM UTC-7, Jeroen Demeyer wrote:
> 
> On 2015-09-21 19:23, Bill Janssen wrote:
> > On OS X, The CasADi subsystem now installs its dynamic
> > libraries in a subdirectory of JModelica, a subdirectory which
> then has
> > to be put on the LD_LIBRARY_PATH the Python interpreter is run with
> That is clearly an issue with the build system. Instead of requiring
> Sage to change sage-env, it would be better to report this problem
> upstream. Libraries should be installed in the standard places, such
> that no special LD_LIBRARY_PATH hacks are needed.
> 
> -- 
> 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.

Re: [sage-devel] Re: SPKG Maintainers??

[Nathann Cohen]

Simon,

By reading your message I am under the impression that you actually
believe that the information contained in 'authors' block is somehow
consistent or even remotely correct.

It is not.

*because* it is mainained manually. Which means that we forget to
update it, that those who contribute a lot are those who will not
leave their name. Because it is done manually, it is totally
unreliable.

The trac server, at least, keeps its promises: each ticket has an author field.
Git blame, at least, keeps its promises: it tells you who last modified a line.

You will not solve the problem by having us maintain manually a
changelog in every single function.

I prefer to do it wikipedia-style: "what matters here is the code, not
who wrote it" (except for debugging purposes -> git blame). And if I
need to explain to someone that I am a Sage contributor, they will
find what they need on my home page.

Nathann
http://www.steinertriples.fr/ncohen/

-- 
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: SPKG Maintainers??

[Simon King]

Hi Nathann,

On 2015-09-22, Nathann Cohen  wrote:
>> I usually add my name as author when I feel that I'm contributing something
>> new, not just fixing bugs.
>
> It can make sense to add your name when you add something new, but
> code is not static like a scientific paper. Contrary to it, code is
> modified over time and after a while an authors block does not make
> any sense.

I totally disagree with that.

If the history of changes to the code gets messy, then "git blame" will
be hardly able to answer the question who wrote the original code. In
order to keep a record of the authorship of changes, it is essential to
have an AUTHORS block, IMHO.

> The best we can have when it comes to acknowledgement is our trac
> server: it stores what each of us did, and this history does not
> change. Then we have 'git blame' when we want to know who wrote one
> specific line.

You assume that you know a specific line to look at. How do you find a
line that is significant for the authorship of a whole function or
even module?

> With the trac history and the git history, I believe that we have
> trustworthy information in both directions.

I disagree with your notion of "to have".

Assume that you go to a shop and the shop keeper asks you to pay 100
Euros for something. Will the shop keeper accept it as payment if
you tell him that he "has" the money, namely there are 100 Euros for
him in one drawer in one piece of furniture in one room of your
apartment (he "only" has to find your apartment and search through
it...)?

No, he will not.

And I don't accept if someone tells me that I "have" the information
in the trac history and the git history, since I first need to analyse
code, use git, open a trac ticket and so on, before I REALLY have the
information.

Cheers,
Simon


-- 
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: SPKG Maintainers??

[Simon King]

Hi William,

On 2015-09-22, William Stein  wrote:
>  People probably don't remember the early days of Sage, but there was
> often a belief that the project would be *impossible* due to the math
> culture in which authorship had to be clearly maintained.  Many people
> sincerely thought that the GAP or PARI developers would be very angry
> that their system was used in Sage, and that people wouldn't
> contribute since their contributions wouldn't be acknowledged.

I do remember that some Singular developers did get seriously angry if
people cited Sage without mentioning that the real work was in fact done
by Singular called from Sage.

Best regards,
Simon

-- 
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] SPKG Maintainers??

[Andrey Novoseltsev]

On Tuesday, 22 September 2015 14:22:14 UTC-6, Jeroen Demeyer wrote:
>
> On 2015-09-22 21:24, Dima Pasechnik wrote: 
> > I guess he cared more about acknowledgements 
>
> There is nothing wrong with caring about acknowledgements. 
>
> I usually add my name as author when I feel that I'm contributing 
> something new, not just fixing bugs. 
>
> Note that the "git blame" doesn't tell much about the effort that people 
> put into the code: usually bugfixes or incremental additions are a lot 
> easier than writing the code in the first place. Also, sometimes code is 
> just moved around or refactored which completely changes the "git blame" 
> output. 
>
> I think it's theoretically possible for a person X to be a significant 
> author to some piece of code, even though "git blame" doesn't list a 
> single line authored by X. 
>
> Exactly. I think this is the only time I've listed myself as an author of 
a function:
https://github.com/sagemath/sage/blame/master/src/sage/geometry/hasse_diagram.py
It required me quite a bit of time to read a paper to understand the 
algorithm, throw away optimizations that I judged irrelevant for Python, 
and write the code which was orders of magnitude faster than its previous 
implementation in Sage (the current bottleneck is poset constructor, 
getting input data for it is fast). I am not visible in git blame at all 
among 5 people that are. Plus github mysteriously shows only "1 
contributor" (Nathann visible after clicking the link) under the name of Marc 
Mezzarobba (who was the last contributor and whose contribution was to move 
a reference to another file). From this git does not seem to be designed 
for showing who wrote the code.

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

[sage-devel] Re: SPKG Maintainers??

[Simon King]

Hi Dima,

On 2015-09-22, Dima Pasechnik  wrote:
> One way or another, there are Copyright notices with names in many files, 
> and Author entries, sometimes they
> have to do something with each other, sometimes not, sometimes there are 
> huge additions done
> by people without adding their names to either of these fields...
> It's a mess.

IMHO, for a user, it is still easier to find his/her way through that
information mess than to use version control to find the same
information in an even bigger mess.

Cheers,
Simon

-- 
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] SPKG Maintainers??

[Nathann Cohen]

> I usually add my name as author when I feel that I'm contributing something
> new, not just fixing bugs.

It can make sense to add your name when you add something new, but
code is not static like a scientific paper. Contrary to it, code is
modified over time and after a while an authors block does not make
any sense.

The best we can have when it comes to acknowledgement is our trac
server: it stores what each of us did, and this history does not
change. Then we have 'git blame' when we want to know who wrote one
specific line.

With the trac history and the git history, I believe that we have
trustworthy information in both directions.

Authors block, however, are maintained manually and represent more
accurately how we feel with respec to acknowledgement than the
work/sweat we put into it.

You, Jeroen, believe that it is not worth adding your name when you
add a bugfix. If you read the 'authors' block of the graph.py file
(the module, not some specific function) you will see that some people
added their name there after fixing a docstring!

http://doc.sagemath.org/html/en/reference/graphs/sage/graphs/graph.html#graph-format

On the other side of the spectra, Vincent's name and mine appear twice
in the whole combinat/design/ folder, and we basically wrote (or
rewrote) everything that is in there. I spent months for what it is
that folder.

If we start adding our names to authors block at every patch, I will
have to decide whether I want everybody else's name but mine to
appear, or whether I want every single function of the graph/ folder
to have an 'AUTHORS' block with my name on it.

And I will hate myself.

Nathann

-- 
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: SPKG Maintainers??

[Simon King]

Hi Andrey, hi all,

On 2015-09-22, Andrey Novoseltsev  wrote:
> I had some random people contact me with questions about the modules I've 
> written and I suspect it was due to the AUTHORS block, so it is useful to 
> users.

I did the same as that user. And still do now.

Using "git blame" to find out the commit that changed an interesting part of
code, then looking up that commit and then finally finding the author's name
is a very awkward procedure. Looking up the name in the source file is a
lot simpler.

> Regarding git, it took me a while to get reasonably comfortable with it

+1.

> it is unreasonable to expect someone to learn using it 
> just for the sake of figuring out who wrote what.

+2.

> And for git blame specifically - it is often ugly when there were multiple 
> layers of semi-automated changes in the file and each line has its own 
> commit (rather than some logical blocks).

+1.

I started to wonder if the above procedure---browsing the code in "git
blame" (I configured it so that the code is opened in "less"), finding out by
reading and understanding code (!) which lines are interesting, coping the
commit number from the start of an interesting line, leaving "less", browsing
"git log" (again opening "less", searching for the commit number, and finally
reading the author name---really is the simplest way to solve the problem in 
git.

If it is---and Andrey seems to confirm---then git isn't suitable for that task,
and we should keep using AUTHORS blocks. We can expect users to be able to read
(the documentation). But we can not expect users (in contrast to developers) to
be familiar with git.

Best regards,
Simon


-- 
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] SPKG Maintainers??

[Jeroen Demeyer]

On 2015-09-22 21:24, Dima Pasechnik wrote:

I guess he cared more about acknowledgements


There is nothing wrong with caring about acknowledgements.

I usually add my name as author when I feel that I'm contributing 
something new, not just fixing bugs.


Note that the "git blame" doesn't tell much about the effort that people 
put into the code: usually bugfixes or incremental additions are a lot 
easier than writing the code in the first place. Also, sometimes code is 
just moved around or refactored which completely changes the "git blame" 
output.


I think it's theoretically possible for a person X to be a significant 
author to some piece of code, even though "git blame" doesn't list a 
single line authored by X.


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

Re: [sage-devel] SPKG Maintainers??

[Dima Pasechnik]

On Tuesday, 22 September 2015 12:00:36 UTC-7, Volker Braun wrote:
>
> On Tuesday, September 22, 2015 at 8:12:26 PM UTC+2, Dima Pasechnik wrote:
>>
>> One way or another, there are Copyright notices with names in many files,
>>
>
> Copyright notices are an entirely different topic from this thread. No / 
> unclear copyright means: No right for Sage to distribute. So they are 
> essentially a legal requirement. License information is also not tracked by 
> git, only authorship is
>

many files have no copyright... Should we fix this in an automated way?

-- 
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] SPKG Maintainers??

[Dima Pasechnik]

On Tuesday, 22 September 2015 11:49:37 UTC-7, William wrote:
>
> On Tue, Sep 22, 2015 at 11:12 AM, Dima Pasechnik  > wrote: 
>
> >> Who? 
> >> Is he removing his own name or other people's name? 
> > 
> > see done by the request of the reviewer 
>
> Nathann Cohen seems to have written: 
>
>"what is this?" about this in your patch: "+- cf. ``git blame`` for 
> the others involved" 
>
> Then there was some bizarre banter starting with "It committed suicide 
> last month, but through an exorcist I was..." 
>
> > 
> http://git.sagemath.org/sage.git/commit/?id=edb886e9f9d0f505f272fe769e69e8a51ab265df
>  
> > on http://trac.sagemath.org/ticket/19098 
> > 
> > One way or another, there are Copyright notices with names in many 
> files, 
> > and Author entries, 
>
> True.  As there should be.  Contributors keep their copyright. 
> Authors should get as much and as explicit of credit as we can 
> possibly give them. 
>
>  People probably don't remember the early days of Sage, but there was 
> often a belief that the project would be *impossible* due to the math 
> culture in which authorship had to be clearly maintained.  Many people 
> sincerely thought that the GAP or PARI developers would be very angry 
> that their system was used in Sage, and that people wouldn't 
> contribute since their contributions wouldn't be acknowledged. 
>
> > sometimes they 
> > have to do something with each other, sometimes not, sometimes there are 
> > huge additions done 
> > by people without adding their names to either of these fields... 
> > It's a mess. 
>
> I don't think we should impose a ton of bureaucracy on the AUTHORS blocks. 
>
> However, we should definitely not have reviewers asking people to 
> remove their names from it, and not removing other people's names 
> either. 
>
> this is not enough, as non-mentioning your name has a viral effect - 
people won't feel like putting their names 
in the AUTHORS block if there is none, or if it is obsolete.
One example is provided by src/sage/graphs/generators/families.py
(featured in the commit above)
 more examples:
sage/combinat/matrices/hadamard.py
 (Only 50% is by wdj, in AUTHORS)

more extreme:
sage/numerical/mip.pyx

(it has Nathann's name in copyright - I guess he cared more about
acknowledgements, as he was still a postdoc, IIRC...)
The only AUTHORS listed is my former student Risan, now at Facebook...

I don't know how widespread this is, absence/incompleteness of AUTHORS.

Probably some automated check should be run,  comparing output of 'git 
blame' with AUTHORS.

Dima

 -- William 
>
> > 
> > 
> >>> 
> >>> IMHO we need an official policy on this. 
> >>> 
> >>> 
>  
>   -- William 
>  
>  > documentation but these information, like TODO and TESTS, is not 
>  > usually 
>  > what the user expects to see in the doc.  These developer-oriented 
>  > blocks 
>  > are better to be hidden or placed in less prominent place for the 
> end 
>  > users. 
>  > 
>  > -- 
>  > 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+...@googlegroups.com. 
>  > To post to this group, send email to sage-...@googlegroups.com. 
>  > Visit this group at http://groups.google.com/group/sage-devel. 
>  > For more options, visit https://groups.google.com/d/optout. 
>  
>  
>  
>  -- 
>  William (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+...@googlegroups.com . 
> >>> To post to this group, send email to sage-...@googlegroups.com 
> . 
> >>> Visit this group at http://groups.google.com/group/sage-devel. 
> >>> For more options, visit https://groups.google.com/d/optout. 
> >> 
> >> 
> >> 
> >> -- 
> >> Sent from my massive iPhone 6 plus. 
> > 
> > -- 
> > 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+...@googlegroups.com . 
> > To post to this group, send email to sage-...@googlegroups.com 
> . 
> > Visit this group at http://groups.google.com/group/sage-devel. 
> > For more options, visit https://groups.google.com/d/optout. 
>
>
>
> -- 
> William (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://group

Re: [sage-devel] SPKG Maintainers??

[Volker Braun]

On Tuesday, September 22, 2015 at 8:12:26 PM UTC+2, Dima Pasechnik wrote:
>
> One way or another, there are Copyright notices with names in many files,
>

Copyright notices are an entirely different topic from this thread. No / 
unclear copyright means: No right for Sage to distribute. So they are 
essentially a legal requirement. License information is also not tracked by 
git, only authorship is.

-- 
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] SPKG Maintainers??

[William Stein]

On Tue, Sep 22, 2015 at 11:12 AM, Dima Pasechnik  wrote:

>> Who?
>> Is he removing his own name or other people's name?
>
> see done by the request of the reviewer

Nathann Cohen seems to have written:

   "what is this?" about this in your patch: "+- cf. ``git blame`` for
the others involved"

Then there was some bizarre banter starting with "It committed suicide
last month, but through an exorcist I was..."

> http://git.sagemath.org/sage.git/commit/?id=edb886e9f9d0f505f272fe769e69e8a51ab265df
> on http://trac.sagemath.org/ticket/19098
>
> One way or another, there are Copyright notices with names in many files,
> and Author entries,

True.  As there should be.  Contributors keep their copyright.
Authors should get as much and as explicit of credit as we can
possibly give them.

 People probably don't remember the early days of Sage, but there was
often a belief that the project would be *impossible* due to the math
culture in which authorship had to be clearly maintained.  Many people
sincerely thought that the GAP or PARI developers would be very angry
that their system was used in Sage, and that people wouldn't
contribute since their contributions wouldn't be acknowledged.

> sometimes they
> have to do something with each other, sometimes not, sometimes there are
> huge additions done
> by people without adding their names to either of these fields...
> It's a mess.

I don't think we should impose a ton of bureaucracy on the AUTHORS blocks.

However, we should definitely not have reviewers asking people to
remove their names from it, and not removing other people's names
either.

 -- William

>
>
>>>
>>> IMHO we need an official policy on this.
>>>
>>>

  -- William

 > documentation but these information, like TODO and TESTS, is not
 > usually
 > what the user expects to see in the doc.  These developer-oriented
 > blocks
 > are better to be hidden or placed in less prominent place for the end
 > users.
 >
 > --
 > 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+...@googlegroups.com.
 > To post to this group, send email to sage-...@googlegroups.com.
 > Visit this group at http://groups.google.com/group/sage-devel.
 > For more options, visit https://groups.google.com/d/optout.



 --
 William (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.
>>
>>
>>
>> --
>> Sent from my massive iPhone 6 plus.
>
> --
> 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 (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.

Re: [sage-devel] SPKG Maintainers??

[Dima Pasechnik]

On Tuesday, 22 September 2015 01:38:03 UTC-7, William wrote:
>
>
>
> On Monday, September 21, 2015, Dima Pasechnik  > wrote:
>
>>
>>
>> On Monday, 21 September 2015 21:42:31 UTC-7, William wrote:
>>>
>>> On Mon, Sep 21, 2015 at 7:58 PM, Kwankyu Lee  wrote: 
>>> > I have written a couple of AUTHORS-blocks, but I think I did it not to 
>>> have 
>>> > a credit but to be responsible for the code. 
>>> > 
>>> > AUTHORS blocks are rather for developers, not for end users. 
>>> These blocks 
>>> > are rather hindrance for end users as  they usually appear at the head 
>>> of a 
>>>
>>> When I "invented" this idea of AUTHORS blocks, they were definitelynot 
>>> only for developers but also for end users.They were partly meant 
>>> as an antidote to what happened with Magma when "they" deleted all the 
>>> names of many of the contributors to Magma from the beginnings of the 
>>> relevant sections of the reference manual, which pissed a lot of 
>>> contributors off. 
>>>
>>> When you read a mathematics paper you don't say "this statement of the 
>>> authors of this paper at the top is not for readers and is a 
>>> hindrance". 
>>>
>>> If anything, I think we should systematically do vastly *more* to 
>>> clearly acknowledge and appreciate the code contributors to Sage. 
>>> They are by far the most important people to the existence of Sage. g
>>>
>>
>> I know at least one prolific Sage contributor who does not add himself to 
>> AUTHORS block,
>> and actually is removing these blocks.
>>
>
> Who?
> Is he removing his own name or other people's name?
>
see done by the request of the reviewer

http://git.sagemath.org/sage.git/commit/?id=edb886e9f9d0f505f272fe769e69e8a51ab265df
on http://trac.sagemath.org/ticket/19098 

One way or another, there are Copyright notices with names in many files, 
and Author entries, sometimes they
have to do something with each other, sometimes not, sometimes there are 
huge additions done
by people without adding their names to either of these fields...
It's a mess.



>> IMHO we need an official policy on this. 
>>
>>
>>
>>>  -- William 
>>>
>>> > documentation but these information, like TODO and TESTS, is not 
>>> usually 
>>> > what the user expects to see in the doc.  These developer-oriented 
>>> blocks 
>>> > are better to be hidden or placed in less prominent place for the end 
>>> users. 
>>> > 
>>> > -- 
>>> > 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+...@googlegroups.com. 
>>> > To post to this group, send email to sage-...@googlegroups.com. 
>>> > Visit this group at http://groups.google.com/group/sage-devel. 
>>> > For more options, visit https://groups.google.com/d/optout. 
>>>
>>>
>>>
>>> -- 
>>> William (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.
>>
>
>
> -- 
> Sent from my massive iPhone 6 plus.
>

-- 
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] SPKG Maintainers??

[Andrey Novoseltsev]

I had some random people contact me with questions about the modules I've 
written and I suspect it was due to the AUTHORS block, so it is useful to 
users.

Regarding git, it took me a while to get reasonably comfortable with it and 
while I like it now, it is unreasonable to expect someone to learn using it 
just for the sake of figuring out who wrote what.

And for git blame specifically - it is often ugly when there were multiple 
layers of semi-automated changes in the file and each line has its own 
commit (rather than some logical blocks). Personally I find "History" on 
github for a particular module much more useful than blame display.

-- 
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] updating sage-env LD_LIBRARY_PATH from spkg-install

[Bill Janssen]

Yes, but in the absence of an achieved utopia, is there something we can do 
here?  I see that sage-env already has a lot of LD_LIBRARY_PATH hacking in 
it.  Might be good to design a system for regularizing this, and moving all 
this hacking to that system.

My solution has been to add Python code to detect this, and put out a 
warning to the user, telling them to add the required LD_LIBRARY_PATH 
setting to their sagerc file.  Somewhat unsatisfactory.

Bill

On Monday, September 21, 2015 at 10:49:05 AM UTC-7, Jeroen Demeyer wrote:
>
> On 2015-09-21 19:23, Bill Janssen wrote: 
> > On OS X, The CasADi subsystem now installs its dynamic 
> > libraries in a subdirectory of JModelica, a subdirectory which then has 
> > to be put on the LD_LIBRARY_PATH the Python interpreter is run with 
> That is clearly an issue with the build system. Instead of requiring 
> Sage to change sage-env, it would be better to report this problem 
> upstream. Libraries should be installed in the standard places, such 
> that no special LD_LIBRARY_PATH hacks are needed. 
>

-- 
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] SPKG Maintainers??

[Kwankyu Lee]

I am not that developer :-) 

How about this system?

We trim the AUTHOR blocks by just listing the names of contributors to the 
given module (to give credits to the contributors); 
And we recommend a new contributor to add his/her name to the end of the 
list (to give more credits to the initial creators); 
And this AUTHORS block only appears at the end of the documentation of the 
module (not to clutter the doc for the end users, but still give them 
motivations to contribute).

The incomplete information about "who did what" in the current AUTHORS 
blocks are removed but can be recovered in git.  


Kwankyu

-- 
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] Recompile .pyx when its .cc dependency changed

[Jeroen Demeyer]

On 2015-09-22 13:59, Vincent Delecroix wrote:

On 22/09/15 08:57, Jeroen Demeyer wrote:

On 2015-09-22 13:23, Jean-Pierre Flori wrote:

Or does %runfile just run cython on the file without further thoughts?


No, it's worse: it runs Cython with some additional *bad thoughts*.


Do we just need to get rid of this use_cache=True?


Maybe, but we might also need to change the way how cython is called 
from Sage. I haven't really looked at the code.


--
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] Recompile .pyx when its .cc dependency changed

[Vincent Delecroix]

On 22/09/15 08:57, Jeroen Demeyer wrote:

On 2015-09-22 13:23, Jean-Pierre Flori wrote:

Or does %runfile just run cython on the file without further thoughts?


No, it's worse: it runs Cython with some additional *bad thoughts*.


Do we just need to get rid of this use_cache=True?

Vincent

--
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] Recompile .pyx when its .cc dependency changed

[Jeroen Demeyer]

On 2015-09-22 13:23, Jean-Pierre Flori wrote:

Or does %runfile just run cython on the file without further thoughts?


No, it's worse: it runs Cython with some additional *bad thoughts*.

Cython does dependency tracking just fine, it's the Sage wrapper which 
breaks 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 http://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Recompile .pyx when its .cc dependency changed

[Jean-Pierre Flori]

On Tuesday, September 22, 2015 at 1:16:46 PM UTC+2, Nathann Cohen wrote:
>
> > This works 
> > 
> > sage: !touch my_file.pyx 
> > sage: %runfile my_file.pyx 
>
> Indeed, but I wouldn't mind if it were automatic somehow :-/ 
>
> Nathann 
>
Isn't it possible to  add some distutils directive to track dependencies?
Or does %runfile just run cython on the file without further thoughts?

-- 
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] Recompile .pyx when its .cc dependency changed

[Vincent Delecroix]

On 22/09/15 08:16, Nathann Cohen wrote:

This works

sage: !touch my_file.pyx
sage: %runfile my_file.pyx


Indeed, but I wouldn't mind if it were automatic somehow :-/


It would be reasonable to me that %runfile does execute (or compile) the 
file whatever the filestamp says. For automatic recompilation when it 
does change we have the %attach.


It seems to be because of Cython (there might be a force=True option to 
do that)... you can have a look at sage/repl/ipython_extension.py and 
sage/repl/load.py


Vincent

--
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] Recompile .pyx when its .cc dependency changed

[Nathann Cohen]

> This works
>
> sage: !touch my_file.pyx
> sage: %runfile my_file.pyx

Indeed, but I wouldn't mind if it were automatic somehow :-/

Nathann

-- 
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] Recompile .pyx when its .cc dependency changed

[Vincent Delecroix]

This works

sage: !touch my_file.pyx
sage: %runfile my_file.pyx

Vincent

On 22/09/15 06:27, Nathann Cohen wrote:

Hello everybody,

I work on a c++ file which I call in Sage, through a .pyx file. More
practically, I have a .pyx file that contains these lines:

 cdef extern from "/home/ncohen/lexbfs.cc" :
 void hello();
 hello()

Now, the code of the .pyx file never changes (it only calls the c++
function) and so typing "%runfile my_cython_file.pyx" never recompiles
anything even if the .cc file changed.

Is there a way to change this behaviour? I have to go back to the .pyx file
and save a trivial change each time I want to recompile it.

Thanks,

Nathann



--
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] access denied on git

[Vincent Delecroix]

Then do

 $ git remote add trac g...@trac.sagemath.org:sage.git

And then push to trac when you push (and not origin). With your current 
version you are *not* using ssh.


Note that you can have a look at the relatively good documentation

  http://doc.sagemath.org/html/en/developer/manual_git.html

Vincent

On 21/09/15 22:10, Mike Zabrocki wrote:



$ git remote -v

origin git://trac.sagemath.org/sage.git (fetch)

origin git://trac.sagemath.org/sage.git (push)



On Monday, 21 September 2015 21:02:16 UTC-4, vdelecroix wrote:


How did you configure git? What gives

$ git remote -v

Vincent

On 21/09/15 21:54, Mike Zabrocki wrote:

I mistakenly installed XCode 7 and so to save myself I started working

on a

fresh copy of sage where I don't do too much compiling.

In doing so I seem to have messed up my permissions to access git (or
something).  I can checkout any branch I like, but I can't seem to push.

$ git push

fatal: remote error: access denied or repository not exported: /sage.git

I thought it might be my ssh key, but I can authenticate:

$ ssh g...@trac.sagemath.org  info

hello zabrocki, this is git@trac running gitolite3 (unknown) on git

1.7.9.5



   R W sage

Does anyone have an idea about what I am doing wrong?

Thanks.
-Mike







--
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] What can we assume about our C compiler

[Samuel Lelievre]

2015-09-22 08:43:13 UTC+2, Ralf Stephan:
>
> On Tuesday, September 22, 2015 at 7:15:56 AM UTC+2, François wrote:
>>
>> OK but I don’t think you should hold back personally. 
>>
> It's just a matter of minimizing problems. There is no pressing need
> for override in Pynac.
>
> As to the link I gave, the new cryptominisat not only needs C++11
> but also, according to http://trac.sagemath.org/ticket/15674
>
>- cmake and not autotools, but Sage does not include cmake
>- more of Boost than Sage includes.
>
> For more Boost, see ticket 17966

http://trac.sagemath.org/ticket/17966
 

> So if Nathann was motivated by that he will have more on the plate.
> I'm second-guessing because you refused to be specific 8P
>
>

-- 
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] Recompile .pyx when its .cc dependency changed

[Nathann Cohen]

Hello everybody,

I work on a c++ file which I call in Sage, through a .pyx file. More
practically, I have a .pyx file that contains these lines:

cdef extern from "/home/ncohen/lexbfs.cc" :
void hello();
hello()

Now, the code of the .pyx file never changes (it only calls the c++
function) and so typing "%runfile my_cython_file.pyx" never recompiles
anything even if the .cc file changed.

Is there a way to change this behaviour? I have to go back to the .pyx file
and save a trivial change each time I want to recompile it.

Thanks,

Nathann

-- 
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] SPKG Maintainers??

[William Stein]

On Monday, September 21, 2015, Dima Pasechnik  wrote:

>
>
> On Monday, 21 September 2015 21:42:31 UTC-7, William wrote:
>>
>> On Mon, Sep 21, 2015 at 7:58 PM, Kwankyu Lee  wrote:
>> > I have written a couple of AUTHORS-blocks, but I think I did it not to
>> have
>> > a credit but to be responsible for the code.
>> >
>> > AUTHORS blocks are rather for developers, not for end users.
>> These blocks
>> > are rather hindrance for end users as  they usually appear at the head
>> of a
>>
>> When I "invented" this idea of AUTHORS blocks, they were definitelynot
>> only for developers but also for end users.They were partly meant
>> as an antidote to what happened with Magma when "they" deleted all the
>> names of many of the contributors to Magma from the beginnings of the
>> relevant sections of the reference manual, which pissed a lot of
>> contributors off.
>>
>> When you read a mathematics paper you don't say "this statement of the
>> authors of this paper at the top is not for readers and is a
>> hindrance".
>>
>> If anything, I think we should systematically do vastly *more* to
>> clearly acknowledge and appreciate the code contributors to Sage.
>> They are by far the most important people to the existence of Sage. g
>>
>
> I know at least one prolific Sage contributor who does not add himself to
> AUTHORS block,
> and actually is removing these blocks.
>

Who?
Is he removing his own name or other people's name?


>
> IMHO we need an official policy on this.
>
>
>
>>  -- William
>>
>> > documentation but these information, like TODO and TESTS, is not
>> usually
>> > what the user expects to see in the doc.  These developer-oriented
>> blocks
>> > are better to be hidden or placed in less prominent place for the end
>> users.
>> >
>> > --
>> > 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+...@googlegroups.com.
>> > To post to this group, send email to sage-...@googlegroups.com.
>> > Visit this group at http://groups.google.com/group/sage-devel.
>> > For more options, visit https://groups.google.com/d/optout.
>>
>>
>>
>> --
>> William (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.
>


-- 
Sent from my massive iPhone 6 plus.

-- 
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: SPKG Maintainers??

[kcrisman]

>
>
> As a code contributor I feel that core contributors should deserve more 
> thanks, not me... 
>
>
I feel similarly.
 

> But what if I start to write my name to comments? Then they can later be 
> semi-automatically extracted, if wanted? Like 
>
> def foo(): 
>  """ 
>  Do foo. 
>  """ 
>  # Author: Jori Mäntysalo 
>
> for a new function or 
>
>  # Author: Jori Mäntysalo // Added support for bar objects. 
>

That sounds fine, if that is what you are comfortable with.  *Removing* 
AUTHOR blocks is maybe not so good, though, unless there are specific 
reasons. 

-- 
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: Help SageMath-Android

[Eric Gourgoulhon]

Hi,

Le lundi 21 septembre 2015 21:45:28 UTC+2, Andrey Novoseltsev a écrit :
>
> Can someone with an Android system take a look at this? Is it device 
> specific or the same strange thing happens for everybody?
>

It works fine for me:
- Galaxy SIII mini with Android 4.1.2
- version() returns SageMath Version 6.9.beta6

Eric.


> On Monday, 21 September 2015 11:15:07 UTC-6, Matematica wrote:
>>
>> I installed SageMath 
>> https://play.google.com/store/apps/details?id=org.sagemath.droid on my 
>> phone Samsung mini2 
>> http://www.gsmarena.com/samsung_galaxy_mini_2_s6500-3883.php
>> I get the error from of the attached images. Thanks.
>>
>>
>>

-- 
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: SPKG Maintainers??

[Jori Mäntysalo]

On Mon, 21 Sep 2015, William Stein wrote:


If anything, I think we should systematically do vastly *more* to
clearly acknowledge and appreciate the code contributors to Sage.
They are by far the most important people to the existence of Sage.


As a code contributor I feel that core contributors should deserve more 
thanks, not me...


But what if I start to write my name to comments? Then they can later be 
semi-automatically extracted, if wanted? Like


def foo():
"""
Do foo.
"""
# Author: Jori Mäntysalo

for a new function or

# Author: Jori Mäntysalo // Added support for bar objects.

for an important enhancement.

--
Jori Mäntysalo