Re: [sage-devel] Re: sage on hacker news right now...

2016-02-14 Thread Ralf Stephan 
On Monday, February 15, 2016 at 1:11:18 AM UTC+1, William wrote:
>
> Combinatorics is definitely the strongest part of Sage.


Old school combinatorics perhaps. But see
http://unsexy-science.blogspot.de/2015/10/survey-sage-and-enumerative.html

-- 
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: sage on hacker news right now...

2016-02-14 Thread Daniel Krenn 
On 2016-02-15 06:25, Jori Mäntysalo wrote:
> On Sun, 14 Feb 2016, William Stein wrote:
> Shall we teach Python in the process of teaching Sage?

At least, I do in my first year's course.

D

-- 
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: sage on hacker news right now...

2016-02-14 Thread Daniel Krenn 
On 2016-02-15 07:56, Nathann Cohen wrote:
> -> We write to the guys who complain about the software and ask
> them to give us more concrete examples, so that we learn and fix them.
> The first one, the guy who mentions bad doc and useless functions
> surely has some stories to tell..
>
> If some of you like the idea, I really don't mind trying to ask the
> guys on the thread to tell their stories.

+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.

[sage-devel] For whoever cares about undergrads (or wants to help William earn more money)

2016-02-14 Thread Nathann Cohen 
Hello everybody,

This is triggered by the thread about 'not-so-nice' comments about Sage.

If you want to help william earn more $$ than he has now (by working
for free) or if you want to help Sage be more welcoming to novice
users (*), a first step would be to help this guy:

http://trac.sagemath.org/ticket/19953

Nathann

(*) Sadly you cannot do one without doing the other now. Well, I guess
I just won't do it then.

-- 
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: Regarding Flint in GSOC-2016

2016-02-14 Thread Ralf Stephan 
Hello Aaditya,
I can offer two ideas here:

The Pynac symbolic expression code mostly relies on differentiation
to get series expansions for expressions. This is the most general but
also the slowest approach. Your project would be to use Flint's polynomial
and series expansion functions to provide fast univariate power series
expansions for all basic and special functions in Pynac. Extending this
to multivariate would likely use giac as the fastest library at the moment,
or you could use stacked Flint polynomials as they do in nemo.

See https://github.com/pynac/pynac/wiki/%7C-power-series
http://trac.sagemath.org/ticket/14878
https://github.com/pynac/pynac/issues/116

Second, ring power series in Sage are missing nearly all member functions
that perform the same basic and special functions on its argument. Where
the result lies in nonsymbolic rings these could be computed via Flint's
functions. But we could handle e.g. `(1+x).sin()` as well by creating rings
over symbolic expressions.

http://trac.sagemath.org/ticket/20017

I'm willing to be mentor for either of these projects.

Regards,

-- 
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] interrupt.pyx on PyPi?

2016-02-14 Thread Jeroen Demeyer 

On 2016-02-01 10:40, 'Martin R. Albrecht' via sage-devel wrote:

In particular, I am *very* interested in turning Sage’s interrupt
handling into something that can be easily installed from PyPI.


This is now ready. The upstream package is hosted at 
https://github.com/sagemath/cysignals and the Sage ticket is 
http://trac.sagemath.org/ticket/20002


After the work which has been done, it would be very nice if somebody 
could review the Sage ticket #20002. Note that a review would mostly 
consist of checking the build and packaging. The "interrupt" part was 
just moved from Sage to an external package, so that doesn't need to be 
reviewed.


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: sage on hacker news right now...

2016-02-14 Thread Nathann Cohen 
> There's a big reddit discussion in which a lot of people say
> not-so-nice  things about Sage:
>
>  
> https://www.reddit.com/r/math/comments/45q7j1/sagemath_open_source_is_now_ready_to_compete_with/

What do you think about doing something smart?

-> We write to the guys who complain about the software and ask
them to give us more concrete examples, so that we learn and fix them.
The first one, the guy who mentions bad doc and useless functions
surely has some stories to tell..

For my part, I would really love it if those who know the code
involved could finally get rid of the  broken functions that
everybody inherit from element:

N, base_extend, base_ring, db, dump, dumps, is_zero, n,
numerical_approx, rename, reset, subs, substitute.

You will see them in combinatorial objects (there is no notion of .n()
there, e.g. Partition object), you will see them in places where no
base ring makes sense (PermutationGroupElement has a .base_ring), you
will see them in immutable objects (no 'rename'), you will see them
where no substitution makes sense.

Those functions raise exceptions when called, and do not make sense.

Additionally, it is important to remember for whoever implements
public methods in *very* generic classes that the method names must
also make sense in the context of subclasses, and that this is *very*
hard to ensure.

If some of you like the idea, I really don't mind trying to ask the
guys on the thread to tell their stories.

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

Re: [sage-devel] Re: sage on hacker news right now...

2016-02-14 Thread Jori Mäntysalo 

On Sun, 14 Feb 2016, William Stein wrote:


There's a big reddit discussion in which a lot of people say
not-so-nice  things about Sage:

https://www.reddit.com/r/math/comments/45q7j1/sagemath_open_source_is_now_ready_to_compete_with/


Shall we teach Python in the process of teaching Sage?

It should be clear that a language desinged for one purpose is (at least 
slightly) better for that purpose than any generice language like Python. 
OTOH Python is not that bad, so learning it may help CS students in other 
tasks too.


 * * *

What is first contact point between Sage and a student?

I think it may be a course of 1) linear algebra, 2) analysis, 3) group 
theory or 4) graph theory. So we should think about how easy it is for a 
student to 1) solve an equation of form Ax=b where A is 3x3 matrix, 2) 
compute derivates of sin(x^2) and (sin x)^2, 3) compute subgroups of S3 
and 4) check planarity of a given graph.


--
Jori Mäntysalo

Re: [sage-devel] Ultimatum: sage's trac and wiki

2016-02-14 Thread Thierry 
Hi,

On Thu, Feb 11, 2016 at 05:42:55PM -0800, William Stein wrote:
> Hi Sage Developers,
> 
> Can somebody *PLEASE* volunteer to move trac.sagemath.org and
> wiki.sagemath.org to a VM on GCE and maintain it for a while?

I already volonteered for this so i guess it is useless to answer again.

Actually, the following is currently done:
- contact people working on ask.sagemath.org and mmarco (who set up
  letsencrypt for the wiki), most seemed to be in as well (+dimpase now)
- ask for 2 fresh VM to my university:
  http://sagewiki.lipn.univ-paris13.fr/
  http://sagetrac.lipn.univ-paris13.fr/
- work on understanding the existing setup and try to reproduce a separate
  working environment on a personal VM (dpkg -l, debsums, diff, and all
  the like) with nonduplicated tools between services (e.g. apache+nginx,
  mysql+postgres,...), wiki is almost done, there are still things to
  understand for git/trac, i might have to ask some questions later.

Note that the wiki (trac as well) runs Ubuntu 12/04 and according to the
welcome message of the wiki:

Your current Hardware Enablement Stack (HWE) is no longer supported
since 2014-08-07. Security updates for critical parts (kernel and
graphics stack) of your system are no longer available.

As nothing is documented, it seems dangerous to just upgrade the
production VM on itself.

I think the current situation is the consequence of the kind of
"ultimatum" that is sent in such situation. While it looks easier to ask
students to do the job, or developpers to quick-and-dirty fix when some
problem happens, this costs actually a lot more time to the people that
eventually do the job. When ask.sagemath.org went down, i personally spent
a few days to just figure out how were the things configured. This kind of
archeology wastes way more time than a serious documented installation
from the beginning (at least it is more frustrating).
 
> I've setup a complete Google compute engine project specifically for
> this and I'll pay for this indefinitely.  The project has a machine on
> it with ssh keys that can ssh to the root account of trac.sagemath.org
> and wiki.sagemath.org.   I can easily add anybody with a Google
> account to this project (though I only want to add people that at
> least enable 2-factor auth, since you can waste a lot of money if you
> get hacked).
> 
> Why move to GCE from UW?   (1) the machines at UW are not backed up at
> all!   -- I assumed a certain student had them backed up, but nope; in
> fact, we can't even ssh into the VM host right now (obviously, I could
> get physical access).  (2) With GCE it's trivially to snapshot the
> whole VM's hundreds of times for very, very cheap, and things are very
> easy to scale, and every person who I add to the project has full
> access to everything through a web browser.  It's just vastly better.

I am not sure the arguments are related to the kind of hosting: it is more
about setting up scheduled backups! Having physical access to the machines
and having direct contact with the sysadmins can be very handy in some
circumstances too.

Moreover, the previous instance of ask.sagemath.org was hosted on GCE as
well, and i remember we finally had to move to a more permanent place at
Niles university (OSU), which is definitely more stable.

As in the "emergency" thread in may 2015, i asked to my lab and it is OK
to host those two servers, and they went just set up. I asked for similar
performance than the existing VM's, it seems they are lacking some RAM
right now but i guess we could ask for more when needed.

I did a quick backup (though not recurrent) of trac and wiki so that we
should not lose much in case of crash.

What is the deadline to leave UW for git/trac/wiki ? I will be almost
completely offline for the next 3 days, moreover i guess it is better to
have some time to do things collectively and documented, or we will face
the same issue again and again.

Ciao,
Thierry


> The other option is that we move *everything* to Github.   I think
> that would be way too disruptive to do right now.
> 
>  -- William
> 
> -- 
> 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.

-- 
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: sage on hacker news right now...

2016-02-14 Thread William Stein 
On Sun, Feb 14, 2016 at 4:00 PM, Travis Scrimshaw  wrote:
>
>
> On Sunday, February 14, 2016 at 5:32:34 PM UTC-6, William wrote:
>>
>> There's a big reddit discussion in which a lot of people say
>> not-so-nice  things about Sage:
>>
>>
>> https://www.reddit.com/r/math/comments/45q7j1/sagemath_open_source_is_now_ready_to_compete_with/
>>
>
> If someone doesn't like the documentation or feels it is lacking, they
> should contribute by giving changes and adding (thematic) tutorials... I
> thought that's one of the benefits of open-source software.

That's definitely 100% fine if our target audience is only research
mathematicians.  However, I aspire to create something that is very
valuable far beyond that target audience.   If I can make money from
SageMathCloud, then I'll be able to make it so sometimes people can
focus on writing better Sage documentation and exposition instead of
teaching sometimes ("teaching buyout").  At least that is my hope.
Actually, this should start happening soon -- if you know of anybody
who you think might be able to do something really good, please email
me.

> Also, yay for the combinatorics mention (and in my bias opinion, is the best
> place for doing combinatorics). Wikipedia currently only lists Sage as being
> the only general-purpose CAS which supports matroids.

Combinatorics is definitely the strongest part of Sage.  You guys have
a fantastic community and have accomplished a huge amount.   I wish
computational number theory wasn't so fragmented still.

>
> Best,
> Travis
>
> --
> 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.



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

[sage-devel] Re: sage on hacker news right now...

2016-02-14 Thread Travis Scrimshaw 


On Sunday, February 14, 2016 at 5:32:34 PM UTC-6, William wrote:
>
> There's a big reddit discussion in which a lot of people say 
> not-so-nice  things about Sage: 
>
>  
> https://www.reddit.com/r/math/comments/45q7j1/sagemath_open_source_is_now_ready_to_compete_with/
>  
>
>
If someone doesn't like the documentation or feels it is lacking, they 
should contribute by giving changes and adding (thematic) tutorials... I 
thought that's one of the benefits of open-source software.

Also, yay for the combinatorics mention (and in my bias opinion, is the 
best place for doing combinatorics). Wikipedia 
 currently only 
lists Sage as being the only general-purpose CAS which supports matroids.

Best,
Travis

-- 
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] Update on memory usage during docbuild

2016-02-14 Thread Travis Scrimshaw 
Hey all,
   Just to note, with 7.1.beta3, I was able to run make doc-clean && make 
doc using less than 6 GB of memory and 8 threads. So memory usage for 
docbuilding seems to have decreased significantly since last time I did a 
full docbuild with this many threads (I think it was in 6.9 betas).

Best,
Travis

-- 
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 on hacker news right now...

2016-02-14 Thread William Stein 
There's a big reddit discussion in which a lot of people say
not-so-nice  things about Sage:

 
https://www.reddit.com/r/math/comments/45q7j1/sagemath_open_source_is_now_ready_to_compete_with/

"Enjoy",

 -- William


On Sat, Feb 13, 2016 at 7:40 PM, William Stein  wrote:
> https://news.ycombinator.com/item?id=11078220
>
> --
> William (http://wstein.org)



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

[sage-devel] Regarding Flint in GSOC-2016

2016-02-14 Thread Aaditya Thakkar 
Hi,
 I am Aaditya Thakkar. I am an undergrad student from Dhirubhai 
Ambani Institute of Information and Communication Technology, India. 
 I have strated contributing to open source and I would like to 
continue this journy and also I am strongly willing to participate in GSOC 
this year. I have recently contributed to the Flint. I was working under 
the guidance of Mr. William Bill Hart and I recently got my commits merged 
for flint library. Also I have previously contributed to open source for 
Cesium Community and I have learnt many things by going through many 
mentoring organizations of the previous year. I would like to keep this 
going and for that I need someone's guidance.
  I got to know about that SageMath is also using Flint and I will 
be very happy to propose a project for it and work and contribute for the 
same, for that I need to know whether SageMath will be taking up projects 
of Flint or not. I have posted about the same on the sage-gsoc google-group 
also.
  Looking forward for your reply. :)

Regards,
Aaditya.

-- 
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] latest OSX 7.1 beta builds available for testing - reverse support request

2016-02-14 Thread Dima Pasechnik 
This is a reverse support request :-)

As there were many problems with binary builds of OSX 7.0 and early OSX 7.1 
betas, the latest (beta3)
are now uploaded on Sage mirrors (at least they are on a couple of UK and 
US mirrors I checked)
http://files.sagemath.org/osx/intel/index.html

Please test and report any problems.

-- 
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.