วันจันทร์ที่ 31 สิงหาคม พ.ศ. 2552

[sage-devel] Re: Anyone want an account on first release of Solaris 10 ?

Hi ,
I love to work with all software that help us in teaching mathematics using  Technology  for students in our schools,
I works as a trainer for  geogebra program in our schooles , give the teachers training for interactive mathematics class room using all technology software  , it is good , but I try all software posibiol in this fild for my study in PhD (Methodes and Curriculums In Mathematics )
 
I will  start to try sage too .
I will be in tuch soon
Maha Ahmed Ismail
Mathematics Expert
Egypt


 
On Mon, Aug 31, 2009 at 7:56 PM, Dr. David Kirkby <david.kirkby@onetel.net> wrote:

If anyone here builds software and wants to know if it works on the
first release of Solaris 10 on SPARC, let me know and I'll create an
account. I set up a quad 450 MHz processor Sun Ultra 80 with 4 GB RAM to
test Sage. Juanjo (the ECL developer) is using it to test ECL too.

I don't intend keeping the machine up more than a couple of weeks, due
to its rather high power consumption, so if you want to test something,
shout now.

Dave
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: sage fails to build with user built python in PATH

Hi Ondrej,

On Tue, Sep 1, 2009 at 4:01 AM, Ondrej Certik<ondrej@certik.cz> wrote:
>
> Hi,
>
> we got this bugreport in femhub:
>
> http://groups.google.com/group/femhub/browse_thread/thread/1bfad9ea3ec7df48
>
> and it turns out it is a Sage problem as well.
>
> Essentially if the user installs his own python and puts it into PATH,
> Sage picks this python instead (and then it fails of course, due to
> slight incompatibilities among python versions, like UTF). I don't
> have time right now to debug it.

Sage 4.1.1 is known to fail to compile on the following versions of SUSE:

* SUSE Linux Enterprise Server 10 SP1 ia64 with GCC 4.4.1. The build
process dies very early when it tries to build bzip2-1.0.5.

* openSUSE 11.1 x86_64 with GCC 4.4.1. In this case, the build fails
because cliquer-1.2 is known to fail to build on some 64-bit
platforms.

Other reported platforms under which Sage 4.1.1 fails to compile
include Mac OS X 10.5.8 compile in 64-bit mode (see #6681), Fedora 10
64-bit (fail due to cliquer, see #6746), and Xubuntu 9.04 32-bit on an
IBM Thinkpad T43 with Centrino 1.86 Ghz and 1.5 GB RAM; see this
sage-devel thread:

http://groups.google.com/group/sage-devel/browse_thread/thread/64ee480aaef342da

Where the compilation fails under SUSE, there is a system-wide
installation of Python and the PATH is not prepended with absolute
paths to any installation of Python. So if Sage 4.1.1 fails to build
because of cliquer, this is a 32- versus 64-bit issue in cliquer
itself.

--
Regards
Minh Van Nguyen

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: grant proposal season

William Stein wrote:
>
>
> On Tue, Aug 25, 2009 at 3:02 PM, David Joyner <wdjoyner@gmail.com
> <mailto:wdjoyner@gmail.com>> wrote:
>
> Things I think would be cool:
>
> 1) spreadsheet functionality,
>
>
> Hello,
>
> Does anybody know of an *open source* (so could be included in Sage)
> AJAX spreadsheet implementation? I.e., something kind of like Google
> docs spreadsheet or http://www.editgrid.com/, but open source. That
> is, something is like TinyMCE, but is a spreadsheet?
>


http://dhtmlx.com/docs/products/dhtmlxGrid/index.shtml (datagrid
javascript; GPL license for standard version)

http://www.simple-groupware.de/cms/Spreadsheet/Features (uses PHP on
server side---could this be replaced with python? I think this is GPL)


http://jqueryplugins.weebly.com/jquerysheet.html (based on jquery---I
assume this is open source)


http://code.google.com/p/jqueryspreadsheet/ (BSD licensed, also based on
jquery; I don't know how far along this is)

Do you need to pick a spreadsheet, or is it sufficient to demonstrate
that there are possibilities out there?

Jason

--
Jason Grout


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

Re: [codenode-devel] Re: [sage-devel] Re: grant proposal season

On Mon, Aug 31, 2009 at 6:14 PM, William Stein<wstein@gmail.com> wrote:
>
>
> On Mon, Aug 31, 2009 at 5:31 PM, David Joyner <wdjoyner@gmail.com> wrote:
>>
>> On Mon, Aug 31, 2009 at 8:19 PM, William Stein<wstein@gmail.com> wrote:
>> >
>> ...
>> >
>> > Hello,
>> >
>> > Does anybody know of an *open source* (so could be included in Sage)
>> > AJAX
>> > spreadsheet implementation?   I.e., something kind of like Google docs
>> > spreadsheet or http://www.editgrid.com/, but open source.   That is,
>> > something is like TinyMCE, but is a spreadsheet?
>>
>> Is this helpful?
>> http://ajax.phpmagazine.net/2006/11/simple_spreadsheet_open_source.html
>
> That looks interesting.  This is a better link:
>
> http://www.simple-groupware.de/cms/Spreadsheet/Home
>
> Thanks!
>
>>
>> http://www.zkoss.org/product/zkspreadsheet.dsp
>>
>
> Unfortunately, ZKspreadsheet is a Java program, not javascript/ajax.


Another idea is something like wolfram alpha:

http://www.wolframalpha.com/

it's on my todo to do something like "sympy gamma". And then something
for femhub (femhub.org) and PDEs, so this could be used with Sage too.
Besides being a cool thing (I really like wolfram alpha), my boss
actually came independently with the idea of being able to interact
with femhub (sage) without having to log in. So near the end of the
september, I plan to work on that.

If you could get funding so that someone else can do the job, that'd
be awesome. :)

Ondrej

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Factorial syntax

I also don't consider Fateman's comments flamebait.

This kind of syntactic-philosopy discussion comes up frequently on CA
mailing lists.

I've had an offline discussion about supporting the quantum mechanical
ket
operator with the syntax of "|state>". One possible implementation is
to
consider the vert and less-than symbols to be syntactic sugar for "("
and ")"
However, supporting |state> would involve Axiom parser changes.

There appear to be three target audiences, the programmers, the
mathematicians,
and the computational mathematicians. Sage has clearly been singing to
the
programmers. A postfix ! would appeal more to the mathematicians.

Fateman, being in the computational mathematician category has a clear
understanding of the range of tradeoffs. I suspect his blunt tone is
due to
the fact that, as he says, this has been debated since the 1970s.
However,
he's being quite clear about the range of possible philosophical
responses.

-Tim

On Aug 31, 1:38 pm, Robert Bradshaw <rober...@math.washington.edu>
wrote:
> On Mon, 31 Aug 2009, William Stein wrote:
> > On Mon, Aug 31, 2009 at 8:23 AM, rjf <fate...@gmail.com> wrote:
>
> >> You can either
> >> (a) adhere to Python syntax.
> >> (b) adhere to Python except for a few changes that are so subtle that
> >> no one would even notice.
> >> (c) adhere to Python except for a few more changes that you tell
> >> everyone about and hope that the discrepancy between Sage and Python
> >> (and Cython) will not be noticed by critics, and so you can continue
> >> to claim "Sage is just Python and everyone knows and loves Python".
> >> (d) use Python as a base, and add whatever you want, because you now
> >> have your own "pre"parser and can add, say, postfix "!".  And now you
> >> probably should tell your users that the Sage language is "python
> >> inspired".
> >> (e) use your own parser for your own made up language that looks like
> >> some mathematical notation.
>
> >> Maxima/Macsyma is in category (d) or (e):  it uses its own parser for
> >> Algol 60 as a base and added things to it, but also has a syntax-
> >> extension feature so that the parser can be changed by the user.  e.g.
> >> commands to add new post-fix operators etc.  (At run-time).
>
> >> One of the premises and alleged great advantages of Sage was, I
> >> thought, that it is "just Python", and you seem to be casually tossing
> >> that idea into the trash.
> >> (The idea that the user-language and the implementation-language for a
> >> computer algebra system has been a topic for discussion at least since
> >> the early 1970s, so the idea that all those clever people just missed
> >> the boat -- the solution is Python; Python does it all; seemed a bit
> >> simplistic.  So I'm not surprised at the slippage, to a "pre-parser"
> >> and Cython and now , oh, let's add a postfix "!".)
>
> > Wow, you are truly the master of flamebait!
>
> I don't consider this flamebait, I think it's a rather accurate (though
> blunt) assessment of the uneasieness some people have regarding the
> preparser.
>
> I'd say Sage is between (c) and (d)--just the fact that we're having this
> discussion means we're not willing to just go and add stuff to the
> preparser left and right.
>
> As for  the question at hand, I'm personally not convinced this is useful
> enough to merit another departure from pure Python. It also risks turning
> the valid Python expression "x!=120" into an invalid one (unless the rule
> is something like "! becomes factorial unless it's followed by one, but
> not two, equals signs...") And then people might expect "x!!" to be the
> double factorial instead of (x!)!.
>
> - Robert
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: sage in mobile wap browser

On Mon, 31 Aug 2009 at 06:32PM +0200, Harald Schilly wrote:
> I found this:
> http://math1.skku.ac.kr/wap_html
> (links on top just mean "example 1-4")
>
> Is this open source, or who is behind that? Never seen before ;)
> The base url is a korean (?) translated sage notebook ...

Jin-yeong Bak ("NoSyu") posted about it a while ago:

http://groups.google.com/group/sage-devel/browse_thread/thread/e6af6ca64c016d8c/

Dan

--
--- Dan Drake <drake at kaist dot edu>
----- KAIST Department of Mathematical Sciences
------- http://mathsci.kaist.ac.kr/~drake

[sage-devel] Re: MPIR build failure on first release of Solaris 10 (03/2005)

Juanjo wrote:
> On Aug 31, 6:37 pm, "Dr. David Kirkby" <david.kir...@onetel.net>
> wrote:
>> I just tried to build Sage (part of which is MPIR) on a Sun Ultra 80
>> SPARC) machine I set up for testing Sage on the first release of Solaris
>> 10.
>> It appears this was fixed in Solaris patch 123647-01 (August 2006),
>> though the latest version of that patch is 123647-03.
>>
>> http://sunsolve.sun.com/search/document.do?assetkey=1-21-123647-03-1
>>
>> which is dated Mon Aug 17th 2009.
>
> I think this is also blocking me from building the Boehm-Weiser
> garbage collector, and thus ECL, in the first version of Solaris 10:
> gcc uses gas instead of as and the resulting binary file has the wrong
> binary format. I did not experience this problem in the T2 cluster,
> AFAIR.

I never noticed before that gcc had switched from using the GNU to Sun
assembler in later Solaris releases. Not sure when that happened, but
you are correct, that the first release of Solaris 10 used the GNU
assembler.

I'll email you off list about the patches.

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?



On Mon, Aug 31, 2009 at 12:39 PM, Brian Granger <ellisonbg.net@gmail.com> wrote:

I would expect apache to come with any relatively recent Unix or
unix-like box. Perhaps it's not installed by default though. (It is on
Solaris).

I just checked and see the latest Apache is 6.6 MB, so not small, but
not particularly huge.

Apache really is a standard. I don't see it has any serious competition.


For deploying the type of web application that the Sage notebook is, there are definitely alternatives to apache.  As William has mentioned, there are many different aspect of the Sage notebook performance and scalability.  Then there is deployability - where apache is not that great even on linux (proper apache config is non-trivial).

I think the best solution is to abstract the Sage notebook web application away from the web server using something like WSGI.  Ideally, the Sage notebook could be deployed using any of a number of different web servers depending on a users needs.  Welding the notebook to a particular web server implementation is a second rate choice in many respects.

I very strongly agree with this.     My impression is that codenode has sort of done this, in that they use Django, which abstracts away and uses WSGI.

William
 

 
There is no way that the Sage notebook will ever *depend* on Apache.  That doesn't even make sense to suggest.

Whew!

Cheers,

Brian

 
William








--
William Stein
Associate Professor of Mathematics
University of Washington
http://wstein.org

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?

Dr. David Kirkby wrote:

>
> Apache really is a standard. I don't see it has any serious competition.


It seems like lighttpd (http://www.lighttpd.net/) is preferable to
apache in some instances.

According to wikipedia: "Lighttpd is used by some of the biggest
websites, including sites such as meebo. Wikimedia runs Lighttpd servers
[4][5][6][7] as does SourceForge.[7] Three of the most famous torrent
listing websites, The Pirate Bay, Mininova and isoHunt, which have more
than 1,000 hits per second, also use Lighttpd.[8] Lighttpd currently
holds fifth place on the Netcraft "Web Server Survey" (November
2008).[9] Lighttpd has not placed in the June 2009 survey except under
the "Market Share for Top Servers Across the Million Busiest Sites"
survey where it places sixth. [10]"


I agree that apache certainly seems to be the most common webserver.

Jason


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: MPIR build failure on first release of Solaris 10 (03/2005)

On Aug 31, 6:37 pm, "Dr. David Kirkby" <david.kir...@onetel.net>
wrote:
> I just tried to build Sage (part of which is MPIR) on a Sun Ultra 80
> SPARC) machine I set up for testing Sage on the first release of Solaris
> 10.
> It appears this was fixed in Solaris patch 123647-01 (August 2006),
> though the latest version of that patch is 123647-03.
>
> http://sunsolve.sun.com/search/document.do?assetkey=1-21-123647-03-1
>
> which is dated Mon Aug 17th 2009.

I think this is also blocking me from building the Boehm-Weiser
garbage collector, and thus ECL, in the first version of Solaris 10:
gcc uses gas instead of as and the resulting binary file has the wrong
binary format. I did not experience this problem in the T2 cluster,
AFAIR.
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

Re: [codenode-devel] Re: [sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?

On Mon, Aug 31, 2009 at 8:59 AM, William Stein<wstein@gmail.com> wrote:
>
>
> On Mon, Aug 31, 2009 at 8:45 AM, Thierry Dumont <tdumont@math.univ-lyon1.fr>
> wrote:
>>
>> William Stein a écrit :
>> >
>> >   The first thing I plan to do is consider switching from
>> > twisted to Django, as is done in codenode -- see http://codenode.org/ --
>> > hopefully, even sharing code with that project.
>>
>> Django is used (generally) with mod_python under Apache. Does it means
>> that the whole Sage will be served by Apache and mod_python? This would
>> (may be) improve performances, no ?
>>
>
> I doubt it would impact performance in practice.  Improving performance is
> going to require improvement to the rest of the server code.
>
> For Sage, the default would be that the notebook is still served by some
> Python library -- most likely Twisted of course -- but that at least the
> option of replacing use of Twisted by Apache would exist.
>
> Codenode guys -- since I speak from 0 experience (!) -- any comments on the
> relative performance of Apache versus Twisted?

For what we are trying to do, i.e. write an "online programming
notebook" (Sage notebook / Codenode),
the bottleneck is definitely going to be the processes that are
actually executing
the code, the communications from these processes back to the "Frontend",
and the additional data operations like saving user state.

Consider a "very small" load on a regular website, maybe 100 users at once doing
various things like reading static articles or checking out photos. In
the case of
the Sage Notebook or Codenode, that's *100 python/sage processes* -
Apache vs Twisted doesn't make much of a difference here.

A Twisted web server can easily handle 100 users, and so can Apache, it's
a moot point for what we are trying to accomplish, imho. There are other
optimizations that are far more important - like optimizing file
system / database calls.

To summarize, we have many more pertinent problems to solve than trying
to decided how a web server might improve performance.

Also, as William has mentioned, talking about all this in an anecdotal
way is mostly useless. We need well thought out benchmarks - which
is a challenging science in its self (there are lies, damn lies, and
benchmarks).


--

Also, Django is a web framework, meant to be served by a WSGI server.

With codenode, the default is to use Twisted, which *has* WSGI serving
functionality,
and it's very convenient to use Twisted because for most situations users
don't have to know anything about the fact that Twisted is doing the WSGI.
Surely this is desirable for the Sage notebook, as with codenode.

You can go to the extra effort to use Apache/Nginx to do the WSGI, but at
this moment with codenode it hasn't been needed for our usage because
Twisted works great. That said, as soon as possible, I want to test and
document all configurations so people can choose what they prefer.


-Alex

>
> William
>
>>
>> t.d.
>>
>>
>>
>
>
>
> --
> William Stein
> Associate Professor of Mathematics
> University of Washington
> http://wstein.org
>
> >
>

--
Alex Clemesha
clemesha.org

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: behavior of log for nonpositive integers

On Aug 28, 8:18 pm, Golam Mortuza Hossain <gmhoss...@gmail.com> wrote:
> Hi,
>
> On Fri, Aug 28, 2009 at 11:02 AM, kcrisman<kcris...@gmail.com> wrote:
>
> > I'm working on #6388,
>
> Thanks for taking the effort on this bug.
>
> > sage: a = Integer(-1)
> > sage: a.log()
> > ValueError about input being negative
> > sage: log(-1)
> > Same ValueError, since preparser turns -1 into Integer(-1)
> > sage: log(SR(-1))
> > I*pi
> > sage: log(CC(-1))
> > 3.14...*I
>
> > Would anything break if a.log() returned a symbolic answer as above?
> > Alternately, one could place a check in the global log() for
> > nonpositive integers and make them symbolic before moving on.  Or one
> > could leave the existing behavior and just document it better.
>
> Please return a log value for integer and float arguments that is
> consistent with rest of the Sage.  As you said, for CC and SR arguments,
> log() returns acceptable answer. So it would be desirable that we get compatible
> answers also for integer and floats. Currently, we get
> ---------
> sage: log(-1.0)
> NaN
> sage: log(-1 + 0*I)
> pi*I
> sage: log(-1)
> ...
> ValueError: self must be positive
> ---------

With no response in three days, I am assuming nothing bad would happen
to integer stuff. I sort of like log(-1) giving an error but as long
as log? makes it very clear why you get an answer (and which branch we
choose!) this is okay with me.

>
> > Something similar happens for input zero, except here SR(0) and CC(0)
> > give various forms of infinity while the request is for log(0) as an
> > answer.
>
> I guess, a consistent convention of "-oo" is perfectly acceptable for log(0).
> My main concern was that it is raising error instead of  returning
> at the least, unevaluated symbolic expression.

Okay, I like that better than sometimes -Infinity and sometimes log
(0). Hopefully I'll have time to finish this relatively soon.
- kcrisman


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?


I would expect apache to come with any relatively recent Unix or
unix-like box. Perhaps it's not installed by default though. (It is on
Solaris).

I just checked and see the latest Apache is 6.6 MB, so not small, but
not particularly huge.

Apache really is a standard. I don't see it has any serious competition.


For deploying the type of web application that the Sage notebook is, there are definitely alternatives to apache.  As William has mentioned, there are many different aspect of the Sage notebook performance and scalability.  Then there is deployability - where apache is not that great even on linux (proper apache config is non-trivial).

I think the best solution is to abstract the Sage notebook web application away from the web server using something like WSGI.  Ideally, the Sage notebook could be deployed using any of a number of different web servers depending on a users needs.  Welding the notebook to a particular web server implementation is a second rate choice in many respects.

 
There is no way that the Sage notebook will ever *depend* on Apache.  That doesn't even make sense to suggest.

Whew!

Cheers,

Brian

 
William




--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?



On Mon, Aug 31, 2009 at 9:40 AM, Tim Dumol <timdumol@gmail.com> wrote:

Codenode doesn't work with Sage currently, although Dorian Raymer is
working on the backend, and it seems it should be ready within a week.

It seems that Codenode can replace the Notebook eventually, assuming
there are ways to plug in a few templates (to include the Help, etc.).
It's well-modularized, so if there isn't a plug-in framework yet, it
should be possible (Django supports using middleware to preprocess
stuff, afaik).

In the meantime, we can stick with the Notebook and make changes to
make a transition to Codenode as smooth as possible. This probably
entails decoupling the components of the Notebook, and finding a way
to generate the documentation (live) that can be plugged into Codenode
easily.

How does that sound?

I definitely do not endorse this, at least until I know a lot more about Codenode and other options.  A month from now I will.   Regarding speed, there are a lot of parameters to consider, but in addition to "snappiness" for a single user, it's also important to consider how the notebook handles say N simultaneous users for various values of N.   It's also important that speed testing be do-able using precise tools rather than anecdote.

William
 

On Sep 1, 12:17 am, Robert Bradshaw <rober...@math.washington.edu>
wrote:
> On Aug 31, 2009, at 9:07 AM, Dr. David Kirkby wrote:
>
> > William Stein wrote:
>
> >> When we wrote the second version of the notebook, twisted.web2 was  
> >> *the*
> >> way to go, according to various people.  However, it turned out that
> >> using twisted.web2 was not the way to go, as the project was
> >> discontinued as an independent project.
>
> > Perhaps a silly questions, but is Apache the way to go?
>
> Whatever we choose, it should still work well as a stand-alone server  
> (unless we wanted to ship apache with Sage :) but it would be nice,  
> especially for large public servers, to use a platform that supported  
> some mod_* under apache as well.
>
> - Robert

As far as I can tell, Codenode doesn't support Apache yet either
( http://groups.google.com/group/codenode-devel/browse_thread/thread/176925e2b2e42279?hl=en
). Regarding the performance of Codenode, it seems to respond a bit
slower (at least for Python) right now, although I haven't actually
done any real measurements. I'm guessing it can be attributed to the
communication time between the server and engine.


 

--
William Stein
Associate Professor of Mathematics
University of Washington
http://wstein.org

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?



On Mon, Aug 31, 2009 at 9:44 AM, Dr. David Kirkby <david.kirkby@onetel.net> wrote:

Robert Bradshaw wrote:
> On Aug 31, 2009, at 9:07 AM, Dr. David Kirkby wrote:
>
>> William Stein wrote:
>>
>>> When we wrote the second version of the notebook, twisted.web2 was
>>> *the*
>>> way to go, according to various people.  However, it turned out that
>>> using twisted.web2 was not the way to go, as the project was
>>> discontinued as an independent project.
>> Perhaps a silly questions, but is Apache the way to go?
>
> Whatever we choose, it should still work well as a stand-alone server
> (unless we wanted to ship apache with Sage :) but it would be nice,
> especially for large public servers, to use a platform that supported
> some mod_* under apache as well.
>
> - Robert


I would expect apache to come with any relatively recent Unix or
unix-like box. Perhaps it's not installed by default though. (It is on
Solaris).

I just checked and see the latest Apache is 6.6 MB, so not small, but
not particularly huge.

Apache really is a standard. I don't see it has any serious competition.

There is no way that the Sage notebook will ever *depend* on Apache.  That doesn't even make sense to suggest.

William

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: permission problem when uploading worksheet

On Sat, Aug 29, 2009 at 05:37:20PM +0000, Willem Jan Palenstijn wrote:
>
> On Sat, Aug 29, 2009 at 10:25:54AM -0700, William Stein wrote:
> > On Sat, Aug 29, 2009 at 10:20 AM, Willem Jan Palenstijn <wjp@usecode.org>wrote:
> > > Ok, done. This is now http://trac.sagemath.org/sage_trac/ticket/6843 .
> >
> >
> > Awesome. Could you post an example broken worksheet to the ticket (or link
> > to one) to make debugging easier?
>
> Yes, of course; I added a basic '1+1' worksheet that exhibits the problem to
> the ticket.

... and now also a patch that should fix it. It runs a 'chmod -R go-rwx,+rwX *'
directly after untarring the sws. This has the effect of removing all
group/other permissions from all files, and then restoring those not masked by
the umask. (Except for the 'x' bit, which it only restores if it's not in the
umask, but is set for the user.)

This should ensure that permissions of an imported sws worksheet are the same
as those of a newly created one.

Any volunteers for a review? :-)

-Willem Jan

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: sage in mobile wap browser

That's very nice. Might be useful during exams. ;-)

Rob

On Aug 31, 9:32 am, Harald Schilly <harald.schi...@gmail.com> wrote:
> I found this:http://math1.skku.ac.kr/wap_html
> (links on top just mean "example 1-4")
>
> Is this open source, or who is behind that? Never seen before ;)
> The base url is a korean (?) translated sage notebook ...
>
> via:http://kdpu.1bbs.info/viewtopic.php?t=103
>
> H
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] sage fails to build with user built python in PATH

Hi,

we got this bugreport in femhub:

http://groups.google.com/group/femhub/browse_thread/thread/1bfad9ea3ec7df48

and it turns out it is a Sage problem as well.

Essentially if the user installs his own python and puts it into PATH,
Sage picks this python instead (and then it fails of course, due to
slight incompatibilities among python versions, like UTF). I don't
have time right now to debug it.

Ondrej

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: MPIR build failure on first release of Solaris 10 (03/2005)

I'm CCing this to mpir-devel and we will add a ticket to trac for
noting this in the manual, if it isn't already in there.

Bill.

On 31 Aug, 17:37, "Dr. David Kirkby" <david.kir...@onetel.net> wrote:
> I just tried to build Sage (part of which is MPIR) on a Sun Ultra 80
> SPARC) machine I set up for testing Sage on the first release of Solaris
> 10.
>
> $ cat /etc/release
>                           Solaris 10 3/05 s10_74L2a SPARC
>             Copyright 2005 Sun Microsystems, Inc.  All Rights Reserved.
>                          Use is subject to license terms.
>                              Assembled 22 January 2005
> $ uname -a
> SunOS goose 5.10 Generic sun4u sparc SUNW,Ultra-80
>
> One issue I hit is with mpir.
>
> libtool: link: `/usr/sfw/lib/libstdc++.la' is not a valid libtool archive
>
> sure enough, a look at the file
>
> $ ls -l /usr/sfw/lib/libstdc++.la
> -r-xr-xr-x   1 root     bin            0 Jan  8  2005
> /usr/sfw/lib/libstdc++.la
> $ file /usr/sfw/lib/libstdc++.la
> /usr/sfw/lib/libstdc++.la:      empty file
> $
>
> shows it is empty.
>
> A Google finds others have the issue too:
>
> "General Solaris 10 Discussion - libstdc++.la empty"http://forums.sun.com/thread.jspa?threadID=5073150
>
> Hence it is worth noting this in the MPIR documentation if it is not
> already done so. (Since this version of gcc is older than officially
> supported by Sage, it is not really an issue in Sage as such).
>
> It appears this was fixed in Solaris patch 123647-01 (August 2006),
> though the latest version of that patch is 123647-03.
>
> http://sunsolve.sun.com/search/document.do?assetkey=1-21-123647-03-1
>
> which is dated Mon Aug 17th 2009.
>
> Note one would need to be root to install this, though one could
> probably get around that by building ones own gcc.
>
> Dave
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Factorial syntax

On Mon, 31 Aug 2009, William Stein wrote:

> On Mon, Aug 31, 2009 at 8:23 AM, rjf <fateman@gmail.com> wrote:
>
>>
>> You can either
>> (a) adhere to Python syntax.
>> (b) adhere to Python except for a few changes that are so subtle that
>> no one would even notice.
>> (c) adhere to Python except for a few more changes that you tell
>> everyone about and hope that the discrepancy between Sage and Python
>> (and Cython) will not be noticed by critics, and so you can continue
>> to claim "Sage is just Python and everyone knows and loves Python".
>> (d) use Python as a base, and add whatever you want, because you now
>> have your own "pre"parser and can add, say, postfix "!". And now you
>> probably should tell your users that the Sage language is "python
>> inspired".
>> (e) use your own parser for your own made up language that looks like
>> some mathematical notation.
>>
>> Maxima/Macsyma is in category (d) or (e): it uses its own parser for
>> Algol 60 as a base and added things to it, but also has a syntax-
>> extension feature so that the parser can be changed by the user. e.g.
>> commands to add new post-fix operators etc. (At run-time).
>>
>> One of the premises and alleged great advantages of Sage was, I
>> thought, that it is "just Python", and you seem to be casually tossing
>> that idea into the trash.
>> (The idea that the user-language and the implementation-language for a
>> computer algebra system has been a topic for discussion at least since
>> the early 1970s, so the idea that all those clever people just missed
>> the boat -- the solution is Python; Python does it all; seemed a bit
>> simplistic. So I'm not surprised at the slippage, to a "pre-parser"
>> and Cython and now , oh, let's add a postfix "!".)
>>
>>
> Wow, you are truly the master of flamebait!

I don't consider this flamebait, I think it's a rather accurate (though
blunt) assessment of the uneasieness some people have regarding the
preparser.

I'd say Sage is between (c) and (d)--just the fact that we're having this
discussion means we're not willing to just go and add stuff to the
preparser left and right.

As for the question at hand, I'm personally not convinced this is useful
enough to merit another departure from pure Python. It also risks turning
the valid Python expression "x!=120" into an invalid one (unless the rule
is something like "! becomes factorial unless it's followed by one, but
not two, equals signs...") And then people might expect "x!!" to be the
double factorial instead of (x!)!.

- Robert


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Anyone want an account on first release of Solaris 10 ?

If anyone here builds software and wants to know if it works on the
first release of Solaris 10 on SPARC, let me know and I'll create an
account. I set up a quad 450 MHz processor Sun Ultra 80 with 4 GB RAM to
test Sage. Juanjo (the ECL developer) is using it to test ECL too.

I don't intend keeping the machine up more than a couple of weeks, due
to its rather high power consumption, so if you want to test something,
shout now.

Dave

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?

Robert Bradshaw wrote:
> On Aug 31, 2009, at 9:07 AM, Dr. David Kirkby wrote:
>
>> William Stein wrote:
>>
>>> When we wrote the second version of the notebook, twisted.web2 was
>>> *the*
>>> way to go, according to various people. However, it turned out that
>>> using twisted.web2 was not the way to go, as the project was
>>> discontinued as an independent project.
>> Perhaps a silly questions, but is Apache the way to go?
>
> Whatever we choose, it should still work well as a stand-alone server
> (unless we wanted to ship apache with Sage :) but it would be nice,
> especially for large public servers, to use a platform that supported
> some mod_* under apache as well.
>
> - Robert


I would expect apache to come with any relatively recent Unix or
unix-like box. Perhaps it's not installed by default though. (It is on
Solaris).

I just checked and see the latest Apache is 6.6 MB, so not small, but
not particularly huge.

Apache really is a standard. I don't see it has any serious competition.

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?

Codenode doesn't work with Sage currently, although Dorian Raymer is
working on the backend, and it seems it should be ready within a week.

It seems that Codenode can replace the Notebook eventually, assuming
there are ways to plug in a few templates (to include the Help, etc.).
It's well-modularized, so if there isn't a plug-in framework yet, it
should be possible (Django supports using middleware to preprocess
stuff, afaik).

In the meantime, we can stick with the Notebook and make changes to
make a transition to Codenode as smooth as possible. This probably
entails decoupling the components of the Notebook, and finding a way
to generate the documentation (live) that can be plugged into Codenode
easily.

How does that sound?

On Sep 1, 12:17 am, Robert Bradshaw <rober...@math.washington.edu>
wrote:
> On Aug 31, 2009, at 9:07 AM, Dr. David Kirkby wrote:
>
> > William Stein wrote:
>
> >> When we wrote the second version of the notebook, twisted.web2 was  
> >> *the*
> >> way to go, according to various people.  However, it turned out that
> >> using twisted.web2 was not the way to go, as the project was
> >> discontinued as an independent project.
>
> > Perhaps a silly questions, but is Apache the way to go?
>
> Whatever we choose, it should still work well as a stand-alone server  
> (unless we wanted to ship apache with Sage :) but it would be nice,  
> especially for large public servers, to use a platform that supported  
> some mod_* under apache as well.
>
> - Robert

As far as I can tell, Codenode doesn't support Apache yet either
( http://groups.google.com/group/codenode-devel/browse_thread/thread/176925e2b2e42279?hl=en
). Regarding the performance of Codenode, it seems to respond a bit
slower (at least for Python) right now, although I haven't actually
done any real measurements. I'm guessing it can be attributed to the
communication time between the server and engine.
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] MPIR build failure on first release of Solaris 10 (03/2005)

I just tried to build Sage (part of which is MPIR) on a Sun Ultra 80
SPARC) machine I set up for testing Sage on the first release of Solaris
10.

$ cat /etc/release
Solaris 10 3/05 s10_74L2a SPARC
Copyright 2005 Sun Microsystems, Inc. All Rights Reserved.
Use is subject to license terms.
Assembled 22 January 2005
$ uname -a
SunOS goose 5.10 Generic sun4u sparc SUNW,Ultra-80

One issue I hit is with mpir.

libtool: link: `/usr/sfw/lib/libstdc++.la' is not a valid libtool archive

sure enough, a look at the file

$ ls -l /usr/sfw/lib/libstdc++.la
-r-xr-xr-x 1 root bin 0 Jan 8 2005
/usr/sfw/lib/libstdc++.la
$ file /usr/sfw/lib/libstdc++.la
/usr/sfw/lib/libstdc++.la: empty file
$

shows it is empty.

A Google finds others have the issue too:

"General Solaris 10 Discussion - libstdc++.la empty"
http://forums.sun.com/thread.jspa?threadID=5073150

Hence it is worth noting this in the MPIR documentation if it is not
already done so. (Since this version of gcc is older than officially
supported by Sage, it is not really an issue in Sage as such).

It appears this was fixed in Solaris patch 123647-01 (August 2006),
though the latest version of that patch is 123647-03.

http://sunsolve.sun.com/search/document.do?assetkey=1-21-123647-03-1

which is dated Mon Aug 17th 2009.

Note one would need to be root to install this, though one could
probably get around that by building ones own gcc.

Dave

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] sage in mobile wap browser

I found this:
http://math1.skku.ac.kr/wap_html
(links on top just mean "example 1-4")

Is this open source, or who is behind that? Never seen before ;)
The base url is a korean (?) translated sage notebook ...

via:
http://kdpu.1bbs.info/viewtopic.php?t=103

H

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?

Dr. David Kirkby wrote:
> William Stein wrote:
>
>> When we wrote the second version of the notebook, twisted.web2 was *the*
>> way to go, according to various people. However, it turned out that
>> using twisted.web2 was not the way to go, as the project was
>> discontinued as an independent project.
>
> Perhaps a silly questions, but is Apache the way to go?
>

If we use Django, I don't think we need to restrict ourselves to Apache:

http://code.djangoproject.com/wiki/ServerArrangements

Jason

--
Jason Grout


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?

On Aug 31, 2009, at 9:07 AM, Dr. David Kirkby wrote:

> William Stein wrote:
>
>> When we wrote the second version of the notebook, twisted.web2 was
>> *the*
>> way to go, according to various people. However, it turned out that
>> using twisted.web2 was not the way to go, as the project was
>> discontinued as an independent project.
>
> Perhaps a silly questions, but is Apache the way to go?

Whatever we choose, it should still work well as a stand-alone server
(unless we wanted to ship apache with Sage :) but it would be nice,
especially for large public servers, to use a platform that supported
some mod_* under apache as well.

- Robert


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: -m64 is not a linker flag.

Bill Hart wrote:
> Just to add to what was said. The -m64 is *required* on OSX 64. If you
> don't include it, everything blows up. In fact you will get piles and
> piles of errors which all look very subtle and you will pull your hair
> out for months trying to figure out what is going wrong. All obvious
> to Michael, but I've been bitten by this and wasted hours of my life
> trying to figure out what had gone wrong.
>
> In particular you'll find piles and piles of unresolved symbols. It
> will complain you have compiled for the wrong architecture and the
> symbols will all have extra leading underscores. VERY frustrating.
>
> Bill.

Is this to the linker, or compiler?

Since when I did a 'man ld' on an OS X box, I can't find any -m64 option.

Dave

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?

William Stein wrote:

> When we wrote the second version of the notebook, twisted.web2 was *the*
> way to go, according to various people. However, it turned out that
> using twisted.web2 was not the way to go, as the project was
> discontinued as an independent project.

Perhaps a silly questions, but is Apache the way to go?

Dave

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Programs in Sage using SCons are really screwed up.

SCons is much newer than the ancient GNU autotools. It is quite
possible that various subtleties are yet to be sort out in SCons. It
may well be worth reporting this issue to the SCons people, as they
may well not know about it (or may be able to give advice on how to
avoid the problem).

Bill.

On 31 Aug, 15:39, "Dr. David Kirkby" <david.kir...@onetel.net> wrote:
> I have no idea if this is a fault of SCons, or in the many programs in
> Sage that use SCons, but the combination does not seem to work too well
> on Solaris.
>
> SCons seems determined to use the Sun compilers in some situations.
> Below the varaible CC was clearly defined to be /usr/sfw/bin/gcc. It is
> probably that is too old to build Sage, but that is not relavent here.
>
> Note how the C compiler 'cc' is called, despite the environment variable
> CC is set elsewhere. In this case 'cc' is not a command, as the Sun
> studio compiler is not installed. But still Sage tries to build with 'cc
> -KPIC' which is the Sun compiler name, with a Sun flag. The only issue
> is that the Sun compiler is not installed.
>
> This issues seems to crop up in many Sage packages using SCons. It would
> be really good if a SCons developer could take a look and hopefully sort
> out why the SCons+Sage+Solaris combo is working badly together.
>
> It seems SCons will try to use the Sun compiler, even if we don't want
> it to, even if it is not installed.
>
> This particular build is on an old Sun of mine, running the first
> release of Solaris 10. It has no optional software installed - I
> purposely left it unchanged, so see if Sage would build on the first
> release of Solaris 10.
>
> Packages which I know have various issues on Sage with SCons and Solaris
> are:
>
> * PolyBoRihttp://sagetrac.org/sage_trac/ticket/6582(just bad programming)http://sagetrac.org/sage_trac/ticket/6177
>
> * cliquerhttp://sagetrac.org/sage_trac/ticket/6852
>
> * modified sage library codehttp://sagetrac.org/sage_trac/ticket/6595
>
> For whatever reason, this combo is not working well. It is perhaps a
> fault of SCons, or perhaps badly written code using SCons, but for
> whatever reason, the combo is hassle.
>
> Here's the cliquer code, for which I have just created trac #6852
>
> x cliquer-1.2/.hg/dirstate, 147 bytes, 1 tape blocks
> x cliquer-1.2/.hg/undo.dirstate, 147 bytes, 1 tape blocks
> x cliquer-1.2/SConstruct, 301 bytes, 1 tape blocks
> Finished extraction
> ****************************************************
> Host system
> uname -a:
> SunOS goose 5.10 Generic sun4u sparc SUNW,Ultra-80
> ****************************************************
> ****************************************************
> CC Version
> /usr/sfw/bin/gcc -v
> Reading specs from /usr/sfw/lib/gcc/sparc-sun-solaris2.10/3.4.3/specs
> Configured with:
> /gates/sfw10/builds/sfw10-gate/usr/src/cmd/gcc/gcc-3.4.3/configure
> --prefix=/usr/sfw --with-as=/usr/sfw/bin/gas --with-gnu-as
> --with-ld=/usr/ccs/bin/ld --without-gnu-ld --enable-languages=c,c++
> --enable-shared
> Thread model: posix
> gcc version 3.4.3 (csl-sol210-3_4-branch+sol_rpath)
> ****************************************************
> scons: Reading SConscript files ...
> scons: done reading SConscript files.
> scons: Building targets ...
> cc -o src/so_cl.o -c -KPIC src/cl.c
> sh: cc: not found
> scons: *** [src/so_cl.o] Error 1
> scons: building terminated because of errors.
>
> real    0m11.561s
> user    0m9.580s
> sys     0m1.532s
> sage: An error occurred while installing cliquer-1.2
> Please email sage-develhttp://groups.google.com/group/sage-devel
> explaining the problem and send the relevant part of
> of /export/home/drkirkby/sage-4.1.1/install.log.  Describe your
> computer, operating system, etc.
> If you want to try to fix the problem, yourself *don't* just cd to
> /export/home/drkirkby/sage-4.1.1/spkg/build/cliquer-1.2 and type 'make'.
> Instead type "/export/home/drkirkby/sage-4.1.1/sage -sh"
> in order to set all environment variables correctly, then cd to
> /export/home/drkirkby/sage-4.1.1/spkg/build/cliquer-1.2
> (When you are done debugging, you can type "exit" to leave the
> subshell.)
> make[1]: *** [installed/cliquer-1.2] Error 1
> make[1]: Leaving directory `/export/home/drkirkby/sage-4.1.1/spkg'
>
> real    23m21.646s
> user    18m45.421s
> sys     4m33.002s
> Error building Sage.
> $ echo $CC
> /usr/sfw/bin/gcc
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: -m64 is not a linker flag.

Just to add to what was said. The -m64 is *required* on OSX 64. If you
don't include it, everything blows up. In fact you will get piles and
piles of errors which all look very subtle and you will pull your hair
out for months trying to figure out what is going wrong. All obvious
to Michael, but I've been bitten by this and wasted hours of my life
trying to figure out what had gone wrong.

In particular you'll find piles and piles of unresolved symbols. It
will complain you have compiled for the wrong architecture and the
symbols will all have extra leading underscores. VERY frustrating.

Bill.

On 30 Aug, 20:44, "Dr. David Kirkby" <david.kir...@onetel.net> wrote:
> Juanjo wrote:
> > On Aug 30, 5:51 pm, William Stein <wst...@gmail.com> wrote:
> >>> * It's hard to see why the options -O2 and -g are needed on Darwin, but
> >>> not on other platforms. Perhaps someone thought it was a good idea to
> >>> add debug support on OS X, but not on any other platform. Perhaps they
> >>> thought OS X is so slow, that it needs the code optimised, but other
> >>> platforms do not!
> >> Someone=somebody=Michael Abshoff in all cases here, so don't expect anybody
> >> to give any explanations.
>
> > This is just guessing, but when detecting GCC, the autoconf files set
> > CFLAGS="-O2 -g" so he was probably giving the defaults + what is
> > needed for 64 bits.
>
> > Juanjo
>
> Fair enough to add -m64 to CFLAGS, but not to LDFLAGS.
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?



On Mon, Aug 31, 2009 at 8:45 AM, Thierry Dumont <tdumont@math.univ-lyon1.fr> wrote:
William Stein a écrit :
>
>   The first thing I plan to do is consider switching from
> twisted to Django, as is done in codenode -- see http://codenode.org/ --
> hopefully, even sharing code with that project.

Django is used (generally) with mod_python under Apache. Does it means
that the whole Sage will be served by Apache and mod_python? This would
(may be) improve performances, no ?


I doubt it would impact performance in practice.  Improving performance is going to require improvement to the rest of the server code.

For Sage, the default would be that the notebook is still served by some Python library -- most likely Twisted of course -- but that at least the option of replacing use of Twisted by Apache would exist. 

Codenode guys -- since I speak from 0 experience (!) -- any comments on the relative performance of Apache versus Twisted?

William
 
t.d.






--
William Stein
Associate Professor of Mathematics
University of Washington
http://wstein.org

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Factorial syntax

We have not actually had any +1 votes for a new postfix ! operator, so
it's bit premature to criticise that decision be fore it hasn't been
made. (William's +1 was for a different suggestion, that Sage's
symbolic variables should not be allowed to be wider than python's.
Again, keeping to python conventions if at all possible.

John
(a veteran Algol 68 programmer, though not Algol 60!)

2009/8/31 rjf <fateman@gmail.com>:
>
> You can either
> (a) adhere to Python syntax.
> (b) adhere to Python except for a few changes that are so subtle that
> no one would even notice.
> (c) adhere to Python except for a few more changes that you tell
> everyone about and hope that the discrepancy between Sage and Python
> (and Cython) will not be noticed by critics, and so you can continue
> to claim "Sage is just Python and everyone knows and loves Python".
> (d) use Python as a base, and add whatever you want, because you now
> have your own "pre"parser and can add, say, postfix "!".  And now you
> probably should tell your users that the Sage language is "python
> inspired".
> (e) use your own parser for your own made up language that looks like
> some mathematical notation.
>
> Maxima/Macsyma is in category (d) or (e):  it uses its own parser for
> Algol 60 as a base and added things to it, but also has a syntax-
> extension feature so that the parser can be changed by the user.  e.g.
> commands to add new post-fix operators etc.  (At run-time).
>
> One of the premises and alleged great advantages of Sage was, I
> thought, that it is "just Python", and you seem to be casually tossing
> that idea into the trash.
> (The idea that the user-language and the implementation-language for a
> computer algebra system has been a topic for discussion at least since
> the early 1970s, so the idea that all those clever people just missed
> the boat -- the solution is Python; Python does it all; seemed a bit
> simplistic.  So I'm not surprised at the slippage, to a "pre-parser"
> and Cython and now , oh, let's add a postfix "!".)
>
> But what I think is appropriate here is an explicit formulation about
> what you are doing, either (a)....(e)  or something else, and the
> consequences.
>
> RJF
>
>
>
>
>
> On Aug 29, 10:13 am, William Stein <wst...@gmail.com> wrote:
>> On Sat, Aug 29, 2009 at 10:10 AM, Jason Grout
>> <jason-s...@creativetrax.com>wrote:
>>
>>
>>
>>
>>
>> > Fredrik Johansson wrote:
>> > > Hi,
>>
>> > > How about supporting n! as a shortcut for factorial(n)? This syntax is
>> > > very convenient and makes a huge difference for combinatorial
>> > > expressions with many factorials. M&M (Maple & Mathematica) allow this
>> > > notation, as do many scientific calculators.
>>
>> > > Although Python doesn't have any other postfix operators, I don't
>> > > think there's any ambiguity as ! is unused in Python except for the !=
>> > > operator. The worst I can come up with is x!=y, but given that =
>> > > denotes assignment and not equality, this still only has one possible
>> > > meaning.
>>
>> > In an attempt to find places that "!" could be used, I found the
>> > following interesting thing.
>>
>> > sage: var('x!')
>> > x!
>> > sage: _^2
>> > x!^2
>> > sage: x!
>> > ------------------------------------------------------------
>> >    File "<ipython console>", line 1
>> >      x!
>> >       ^
>> > SyntaxError: invalid syntax
>>
>> > Maybe Sage symbolic variables should be restricted to identifiers that
>> > are legal in python...
>>
>> Yes, definitely +1 to that.  I think this used to be the case at some point.
>>
>> William
> >
>

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?

begin:vcard
fn:Thierry Dumont
n:Dumont;Thierry
org;quoted-printable;quoted-printable:Universit=C3=A9 Lyon 1 & CNRS.;Institut Camille Jordan -- Math=C3=A9matiques / Mathematics.
adr:;;43 Bd du 11 Novembre.;Villeurbanne;;69621;France
email;internet:tdumont@math.univ-lyon1.fr
title;quoted-printable:Ing=C3=A9nieur de Recherche / Research Engineer.
tel;work:04 72 44 85 23.
tel;fax:04 72 44 80 53
x-mozilla-html:FALSE
url:http://math.univ-lyon1.fr/~tdumont
version:2.1
end:vcard

William Stein a écrit :
>
> The first thing I plan to do is consider switching from
> twisted to Django, as is done in codenode -- see http://codenode.org/ --
> hopefully, even sharing code with that project.

Django is used (generally) with mod_python under Apache. Does it means
that the whole Sage will be served by Apache and mod_python? This would
(may be) improve performances, no ?

t.d.


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Factorial syntax



On Mon, Aug 31, 2009 at 8:23 AM, rjf <fateman@gmail.com> wrote:

You can either
(a) adhere to Python syntax.
(b) adhere to Python except for a few changes that are so subtle that
no one would even notice.
(c) adhere to Python except for a few more changes that you tell
everyone about and hope that the discrepancy between Sage and Python
(and Cython) will not be noticed by critics, and so you can continue
to claim "Sage is just Python and everyone knows and loves Python".
(d) use Python as a base, and add whatever you want, because you now
have your own "pre"parser and can add, say, postfix "!".  And now you
probably should tell your users that the Sage language is "python
inspired".
(e) use your own parser for your own made up language that looks like
some mathematical notation.

Maxima/Macsyma is in category (d) or (e):  it uses its own parser for
Algol 60 as a base and added things to it, but also has a syntax-
extension feature so that the parser can be changed by the user.  e.g.
commands to add new post-fix operators etc.  (At run-time).

One of the premises and alleged great advantages of Sage was, I
thought, that it is "just Python", and you seem to be casually tossing
that idea into the trash.
(The idea that the user-language and the implementation-language for a
computer algebra system has been a topic for discussion at least since
the early 1970s, so the idea that all those clever people just missed
the boat -- the solution is Python; Python does it all; seemed a bit
simplistic.  So I'm not surprised at the slippage, to a "pre-parser"
and Cython and now , oh, let's add a postfix "!".)


Wow, you are truly the master of flamebait!

 -- William

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Factorial syntax

You can either
(a) adhere to Python syntax.
(b) adhere to Python except for a few changes that are so subtle that
no one would even notice.
(c) adhere to Python except for a few more changes that you tell
everyone about and hope that the discrepancy between Sage and Python
(and Cython) will not be noticed by critics, and so you can continue
to claim "Sage is just Python and everyone knows and loves Python".
(d) use Python as a base, and add whatever you want, because you now
have your own "pre"parser and can add, say, postfix "!". And now you
probably should tell your users that the Sage language is "python
inspired".
(e) use your own parser for your own made up language that looks like
some mathematical notation.

Maxima/Macsyma is in category (d) or (e): it uses its own parser for
Algol 60 as a base and added things to it, but also has a syntax-
extension feature so that the parser can be changed by the user. e.g.
commands to add new post-fix operators etc. (At run-time).

One of the premises and alleged great advantages of Sage was, I
thought, that it is "just Python", and you seem to be casually tossing
that idea into the trash.
(The idea that the user-language and the implementation-language for a
computer algebra system has been a topic for discussion at least since
the early 1970s, so the idea that all those clever people just missed
the boat -- the solution is Python; Python does it all; seemed a bit
simplistic. So I'm not surprised at the slippage, to a "pre-parser"
and Cython and now , oh, let's add a postfix "!".)

But what I think is appropriate here is an explicit formulation about
what you are doing, either (a)....(e) or something else, and the
consequences.

RJF

On Aug 29, 10:13 am, William Stein <wst...@gmail.com> wrote:
> On Sat, Aug 29, 2009 at 10:10 AM, Jason Grout
> <jason-s...@creativetrax.com>wrote:
>
>
>
>
>
> > Fredrik Johansson wrote:
> > > Hi,
>
> > > How about supporting n! as a shortcut for factorial(n)? This syntax is
> > > very convenient and makes a huge difference for combinatorial
> > > expressions with many factorials. M&M (Maple & Mathematica) allow this
> > > notation, as do many scientific calculators.
>
> > > Although Python doesn't have any other postfix operators, I don't
> > > think there's any ambiguity as ! is unused in Python except for the !=
> > > operator. The worst I can come up with is x!=y, but given that =
> > > denotes assignment and not equality, this still only has one possible
> > > meaning.
>
> > In an attempt to find places that "!" could be used, I found the
> > following interesting thing.
>
> > sage: var('x!')
> > x!
> > sage: _^2
> > x!^2
> > sage: x!
> > ------------------------------------------------------------
> >    File "<ipython console>", line 1
> >      x!
> >       ^
> > SyntaxError: invalid syntax
>
> > Maybe Sage symbolic variables should be restricted to identifiers that
> > are legal in python...
>
> Yes, definitely +1 to that.  I think this used to be the case at some point.
>
> William
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Is there any particular reason why Notebook uses Twisted.Web2 instead of Twisted.Web?



On Mon, Aug 31, 2009 at 4:46 AM, Tim Dumol <timdumol@gmail.com> wrote:

The Notebook currently uses Twisted.Web2 as its server. As stated <a
href="http://twistedmatrix.com/trac/wiki/TwistedWebPlan">here</a>,
Twisted.Web2 is being phased out and its useful features being merged
back into Twisted.Web. There doesn't seem to be any support for
Twisted.Web2 -- I cannot find proper documentation for Twisted.Web2,
aside from http://twistedmatrix.com/trac/browser/trunk/doc/web2. API
listings for Twisted.Web2 are gone in the latest version of API
documentation (http://twistedmatrix.com/documents/8.2.0/api/
twisted.html
).

Would it cause any problems if the Notebook were gradually migrated to
Twisted.Web?

When we wrote the second version of the notebook, twisted.web2 was *the* way to go, according to various people.  However, it turned out that using twisted.web2 was not the way to go, as the project was discontinued as an independent project. 

I'm not at all convinced that using twisted in any way (web or web2) is a good idea for the Sage notebook.   I plan to revisit this in late September.    The first thing I plan to do is consider switching from twisted to Django, as is done in codenode -- see http://codenode.org/ -- hopefully, even sharing code with that project.   Of course, twisted could still get used at a certain level behind the scenes, but the Sage notebook would then no longer explicitly use it.

So if you want to help as you describe above, perhaps you could accelerate this.  This involves:

  (1) getting familiar with Django, if you don't already know it.

  (2) reading through the current codenode codebase

  (3) then formulating a plan to replace server/notebook/twist.py with something based either directly on Django, or possibly using codenode in some way.

The above is what I would do, but if you do it first that would be spectacular.

William

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Programs in Sage using SCons are really screwed up.

I have no idea if this is a fault of SCons, or in the many programs in
Sage that use SCons, but the combination does not seem to work too well
on Solaris.

SCons seems determined to use the Sun compilers in some situations.
Below the varaible CC was clearly defined to be /usr/sfw/bin/gcc. It is
probably that is too old to build Sage, but that is not relavent here.

Note how the C compiler 'cc' is called, despite the environment variable
CC is set elsewhere. In this case 'cc' is not a command, as the Sun
studio compiler is not installed. But still Sage tries to build with 'cc
-KPIC' which is the Sun compiler name, with a Sun flag. The only issue
is that the Sun compiler is not installed.

This issues seems to crop up in many Sage packages using SCons. It would
be really good if a SCons developer could take a look and hopefully sort
out why the SCons+Sage+Solaris combo is working badly together.

It seems SCons will try to use the Sun compiler, even if we don't want
it to, even if it is not installed.

This particular build is on an old Sun of mine, running the first
release of Solaris 10. It has no optional software installed - I
purposely left it unchanged, so see if Sage would build on the first
release of Solaris 10.

Packages which I know have various issues on Sage with SCons and Solaris
are:

* PolyBoRi
http://sagetrac.org/sage_trac/ticket/6582 (just bad programming)
http://sagetrac.org/sage_trac/ticket/6177

* cliquer
http://sagetrac.org/sage_trac/ticket/6852

* modified sage library code
http://sagetrac.org/sage_trac/ticket/6595

For whatever reason, this combo is not working well. It is perhaps a
fault of SCons, or perhaps badly written code using SCons, but for
whatever reason, the combo is hassle.

Here's the cliquer code, for which I have just created trac #6852

x cliquer-1.2/.hg/dirstate, 147 bytes, 1 tape blocks
x cliquer-1.2/.hg/undo.dirstate, 147 bytes, 1 tape blocks
x cliquer-1.2/SConstruct, 301 bytes, 1 tape blocks
Finished extraction
****************************************************
Host system
uname -a:
SunOS goose 5.10 Generic sun4u sparc SUNW,Ultra-80
****************************************************
****************************************************
CC Version
/usr/sfw/bin/gcc -v
Reading specs from /usr/sfw/lib/gcc/sparc-sun-solaris2.10/3.4.3/specs
Configured with:
/gates/sfw10/builds/sfw10-gate/usr/src/cmd/gcc/gcc-3.4.3/configure
--prefix=/usr/sfw --with-as=/usr/sfw/bin/gas --with-gnu-as
--with-ld=/usr/ccs/bin/ld --without-gnu-ld --enable-languages=c,c++
--enable-shared
Thread model: posix
gcc version 3.4.3 (csl-sol210-3_4-branch+sol_rpath)
****************************************************
scons: Reading SConscript files ...
scons: done reading SConscript files.
scons: Building targets ...
cc -o src/so_cl.o -c -KPIC src/cl.c
sh: cc: not found
scons: *** [src/so_cl.o] Error 1
scons: building terminated because of errors.

real 0m11.561s
user 0m9.580s
sys 0m1.532s
sage: An error occurred while installing cliquer-1.2
Please email sage-devel http://groups.google.com/group/sage-devel
explaining the problem and send the relevant part of
of /export/home/drkirkby/sage-4.1.1/install.log. Describe your
computer, operating system, etc.
If you want to try to fix the problem, yourself *don't* just cd to
/export/home/drkirkby/sage-4.1.1/spkg/build/cliquer-1.2 and type 'make'.
Instead type "/export/home/drkirkby/sage-4.1.1/sage -sh"
in order to set all environment variables correctly, then cd to
/export/home/drkirkby/sage-4.1.1/spkg/build/cliquer-1.2
(When you are done debugging, you can type "exit" to leave the
subshell.)
make[1]: *** [installed/cliquer-1.2] Error 1
make[1]: Leaving directory `/export/home/drkirkby/sage-4.1.1/spkg'

real 23m21.646s
user 18m45.421s
sys 4m33.002s
Error building Sage.
$ echo $CC
/usr/sfw/bin/gcc

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

[sage-devel] Re: Formal sums... but not the ones we have !

Hi Nathann,


On Mon, Aug 31, 2009 at 5:59 AM, Nathann
Cohen<nathann.cohen@gmail.com> > Is there a way in Sage to define a
Formal sum ? Something like
> sum( a_i, i \in [0,...5] ) or even worse, sum(a_e, e\in g.edges()) for
> g a graph, etc.

For formal or symbolic sum, you need to make it a SFunction sub-class
of new symbolics. Unfortunately, it is yet to be implemented/ported
to the new symbolics and it would be great if you could do this,

See this ticket for related discussion

http://trac.sagemath.org/sage_trac/ticket/3587


I think, symbolic sum implementation would be very similar to
symbolic "integrate" implementation. Please have a look at the
patch

http://trac.sagemath.org/sage_trac/ticket/6465


> I insist it should be a "formal sum",  as it should not return sum
> ( [a_i for i in range(6) ] ) or sum( [a_e for e in g.edges() ] ) , but
> rather an abstract formula for this. It could be evaluated at any
> moment, giving different results if the a_i have changed or if the
> graph has changed meanwhile.


In the _eval_ method of the new class, you can decide on the
situation where you want to evaluate explicitly or leave it symbolic.
If you want to leave it symbolic then just "return None" .


Cheers,
Golam

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to sage-devel-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---