Scientific calculator in the notebook
Ted Kosan
for beginners because I am interested in encouraging high school
students to use SAGE. The approach I have been working on recently is
to embed a scientific calculator into the notebook that SAGE newbies
should be able to start using immediately with zero SAGE training.
The calculator is still at an early stage of development, but there is
enough running that people can start playing with it. Here is a
screenshot:
http://sage.math.washington.edu/home/tkosan/misc/mathrider_applet.png
And if people want to play with it, just open a worksheet in FireFox
and evaluate the following text in a cell:
html('<applet id="mathrider" code="org.mathrider.MathRider.class"
width="800" height="650"
codebase="http://sage.math.washington.edu/home/tkosan/mathrider/"
archive="mathrider.jar" MAYSCRIPT></applet>')
At this point it only works in FireFox but hopefully it will
eventually work in IE too.
The Calculator, 2D plotter, and 3D plotter are not talking to the SAGE
server yet, but the text areas in the Cell tab are. Type something
simple like "m = 7" into the Send text area and press <shift><enter>.
The code will be sent to the SAGE server for evaluation and a response
will be printed in the Receive text are. If one then goes to an empty
cell in the notebook and enters "print m", 7 will be printed.
At this point I am still trying to get my mind around the details of
the communications between the notebook and the server and I will
probably be asking questions about the communications protocol in the
near future.
Ted
boo...@u.washington.edu
On Fri, 7 Dec 2007, Ted Kosan wrote:
>
> I have been working on ways to make SAGE as easy to use as possible
> for beginners because I am interested in encouraging high school
> students to use SAGE. The approach I have been working on recently is
> to embed a scientific calculator into the notebook that SAGE newbies
> should be able to start using immediately with zero SAGE training.
+1 -- I've been wanting this feature for a while.
> At this point I am still trying to get my mind around the details of
> the communications between the notebook and the server and I will
> probably be asking questions about the communications protocol in the
> near future.
I wrote most of the client-side communications stuff, and it is relatively easy to explain, so if you have any questions, feel free to send them my way.
-- tom
William Stein
Tom,
By the way, do you have any thoughts about technical communication
issues with the server when using Java versus using Javascript to
implement the scientific calculator. I think Ted's calculator client
is a Java applet,
which might impact how it would communicate with the Sage server.
We should really really have a collection of simple example java applets that
communicate with a sage server, to explain how that can be done.
William
Nils Bruin
introduction to sage, you might want to consider including a side-
window where the sage commands that effect the same computation scroll
by. That way, one could use it as a "scientific calculator-to-sage"
translator and people might be able to pick up sage syntax while
typing in on a familiar calculator.
By the way, my experience is that students don't necessarily know how
to operate a standard scientific calculator anymore. They only use the
superfancy graphing calculators. For instance, to calculate sin(1),
they would actually type in [SIN] ( 1 ) [ENTER]
or [SIN] 1 [ENTER] rather than 1 [SIN].
Jason Grout
Ted, I think this is fantastic.
(Alex, William, or anyone else: I'd love to hear any corrections or
insights you may have about what I've said below.)
It looks like you're using the trick of asking Sage to evaluate some
text using cell id -1. This is a hack right now, as cell ids are
supposed to be nonnegative numbers. Currently, though, the output
directories are created and the right .py file is created and sent to
SAGE. Problems can crop up if you happen to have two requests going to
Sage simultaneously (one request overwrites the other since they're both
for cell -1). Also, Sage complains that cell -1 wasn't defined before
sending the text to it. I think there may be other places in the code
where the id is assumed to be nonnegative.
That said, Sage actually does send back correct a response (if the cell
hasn't been overwritten on a race condition).
Here are two ideas to deal with the issue:
1. Have a namespace associated with each cell. Each namespace would
correspond to a directory in the notebook directory, so that the cells
in namespace "calculator" would be stored in "calculator/code/" and
"calculator/data" directories. This would allow Ted to have his own
private cell id space so that he can create virtually a notebook inside
of the notebook, storing students' calculations in sort of a history
interface.
2. Have a flag saying that the cell is a "virtual" cell and is sort of a
one-off calculation. Any non-text results of the calculation are thrown
away. I would find this useful in interactive widgets, where I just
want to set the value of some variables based on some user interactions.
I don't care about the non-text input and I don't want to keep track
of cell ids. I just want to have Sage do a short calculation that
affects the state of Sage and check the return value to see if there was
an error.
Also, it would be nice to allow for requests to be prioritized, so that,
for example, Ted's calculator requests would always be queued right up,
while some other calculations would have to wait their turn (and
possibly be pushed back).
-Jason
Jason Grout
>
> Ted, I think this is fantastic.
>
> (Alex, William, or anyone else: I'd love to hear any corrections or
> insights you may have about what I've said below.)
I meant to say "Tom, Alex, William, or anyone else:"
Thanks, Tom, for your offer of help to us trying to understand the notebook.
-Jason
William Stein
All I have to say is that these are both (1) excellent ideas and (2) probably
not difficult for me to implement.
Also, Nils idea to have an option to show how scientific calculator code is
translated to Sage is great!
Finally, it would be really cool to have similar scientific calculators, but for
special subject areas, e.g., graph theory, combinatorics, elliptic curves, etc.
These would rock.
-- William
Ted Kosan
> By the way, do you have any thoughts about technical communication
> issues with the server when using Java versus using Javascript to
> implement the scientific calculator. I think Ted's calculator client
> is a Java applet,
> which might impact how it would communicate with the Sage server.
The applet uses JSObject to access the javascript environment of the
FireFox browser and it relies on the javascript notebook code in the
browser for all communications with the SAGE server:
http://java.sun.com/products/plugin/1.3/docs/jsobject.html
This approach worked well for eliminating the need for the applet to
log into the SAGE server and it automatically sets it inside the
context of an opened worksheet.
JSObject also enables the applet to do things with the worksheet like
bind a text area to a cell for enhanced editing capabilities (like
syntax highlighting, search and replace, etc.). It can also inject
javascript code into the browser as needed. On the javascript side,
it can call the applet's public methods.
> We should really really have a collection of simple example java applets that
> communicate with a sage server, to explain how that can be done.
As soon as we come up with a solid applet communications technique, I
can help with this.
Ted
Ted Kosan
> It looks like you're using the trick of asking Sage to evaluate some
> text using cell id -1. This is a hack right now, as cell ids are
> supposed to be nonnegative numbers. Currently, though, the output
> directories are created and the right .py file is created and sent to
> SAGE. Problems can crop up if you happen to have two requests going to
> Sage simultaneously (one request overwrites the other since they're both
> for cell -1). Also, Sage complains that cell -1 wasn't defined before
> sending the text to it. I think there may be other places in the code
> where the id is assumed to be nonnegative.
This is where I need to get my mind around the notebook code better.
What I am doing now is the following:
1) When the applet launches, it injects an applet_callback function
into the browser:
win = JSObject.getWindow(this);
win.eval("function applet_callback(status, response_text) {" +
"document.mathrider.response( status, response_text );"+
"}"
);
2) When the user enters code into the Send text area, the applet calls
the async_request function and gives the applet_callback function as
the callback (I have been experimenting with different values for
newcell and id to see what effect they have):
String code = sendTextArea.getText();
String encodedCode = URLEncoder.encode(code,"UTF8");
win.eval("async_request(worksheet_command('eval'), applet_callback,
'newcell=1&id=11&input="+ encodedCode +"');");
3) The response method in the applet is called by the applet_callback
function to handle the response:
public void response( String status, String response_text )
{
receiveTextArea.append("\nStatus: " + status + " Response: " +
response_text + "\n");
}//end method.
Eventually, I see different parts of the applet communicating with the
server concurrently. To enable multiple concurrent submissions to the
server, I was thinking about maybe having response objects in the
applet handle the replies.
In light of this approach, I am going to have to study your
enhancement proposals further to better understand how they would
work.
Ted
Ted Kosan
> Scientific calculator programs already abound.
What I like about a SAGE-based calculator is that it can be advertised
as being among the most powerful scientific calculators in the world.
People can be told that this calculator is like the wardrobe in the
Narnia series. When people use this calculator, its like falling into
a mathematical Narnia that is as rich and interesting as Narnia
itself. The other-worldly people who built the "calculator" (SAGE)
can be thought of as living inside of the calculator and anyone who
uses SAGE can enter that world and communicate with the world's
creators :-)
> As a gentle
> introduction to sage, you might want to consider including a side-
> window where the sage commands that effect the same computation scroll
> by. That way, one could use it as a "scientific calculator-to-sage"
> translator and people might be able to pick up sage syntax while
> typing in on a familiar calculator.
This is an excellent idea :-)
My goal with this specific calculator is to change it into something
that progressively exposes SAGE functionality until the user has
evolved to the point of being an actual SAGE user. For example, I
would like to change the memory functionality so that the results of
calculations are saved in SAGE variables. I would also like to allow
the user to bind SAGE scripts to the unused buttons of the calculator.
Ted
Ted Kosan
> Finally, it would be really cool to have similar scientific calculators, but for
> special subject areas, e.g., graph theory, combinatorics, elliptic curves, etc.
> These would rock.
I have a couple of thoughts related to this.
1) Not only can the applet inject javascript code into the browser,
but it can also inject SAGE code into a worksheet. When a given
applet launches, it can automatically inject code into the SAGE
environment on the server which can be used by the user to interact
with the applet.
For example, a graphing applet might inject an object called cgraph
into the worksheet and then the user might type the following into a
cell to view the graph on the applet:
cgraph(x^2, 0, 10)
cgraph will know how to communicate with the applet it came from and
each applet can even have its own unique protocol using this "proxy
object" technique
With approach, the applet code will be very loosely coupled with the
SAGE server code and applet development can occur without having to
coordinate with the SAGE developers very much.
2) It should also be possible for the SAGE server to inject a data
file into the applet that tells the applet what kind of GUI to create.
Also, JavaSE comes with a javascript interpreter. Therefore,
javascript can be injected into the applet from the server so people
can create their own applets using Javascript and then the scripts can
be saved in the notebook and used as needed. This might make it
easier for people to create their own custom calculators and other
applets.
Ted
Jason Grout
I like the direction you advocate (progessively exposing Sage). In
middle and high school, I was introduced to a lot of math by reading my
calculator manuals (well, at least it was a lot to me then!). Pressing
a button on the calculator made me curious about what in the world
standard deviation was, which led me to the manual, which fortunately
listed the formula and a terse description of the quantities involved.
Then I had to figure out the formula. Then came the "why?" questions.
It was very exciting to me in a school that didn't teach calculus and
had no real computer classes.
But then again, I'm a geek :). But seriously, I think I would have
loved learning about abstract algebra and category theory and graph
theory, for example, if I had been exposed to Sage in high school. And
I probably would have learned how to program in python then, instead of
trying to wrap my head around C++!
-Jason
Nils Bruin
> 1) Not only can the applet inject javascript code into the browser,
> but it can also inject SAGE code into a worksheet. When a given
> applet launches, it can automatically inject code into the SAGE
> environment on the server which can be used by the user to interact
> with the applet.
are some tasks that are more conveniently done interactively: Pick a
viewpoint for a 3d picture, select the viewport for a graph, select
colors in a graph, etc. Each of those options can be set on the
commandline too. It would be wonderful to have applets that let you
select these things and then give you back the command line for later
use (so that you can generate the picture through another toolchain
that produces a print-quality version, for instance, or inside a
script that generates many similar graphs which need the same
parameters set). If the interactive tools always have the option of
"giving back" to the command line, you would really have the best of
both worlds.
Carl Witty
> William wrote:
> > By the way, do you have any thoughts about technical communication
> > issues with the server when using Java versus using Javascript to
> > implement the scientific calculator. I think Ted's calculator client
> > is a Java applet,
> > which might impact how it would communicate with the Sage server.
>
> The applet uses JSObject to access the javascript environment of the
> FireFox browser and it relies on the javascript notebook code in the
> browser for all communications with the SAGE server:
>
> http://java.sun.com/products/plugin/1.3/docs/jsobject.html
>
> This approach worked well for eliminating the need for the applet to
> log into the SAGE server and it automatically sets it inside the
> context of an opened worksheet.
from Java, it doesn't have to log in separately... the cookies from
the browser's login are automatically used by the Java
communications. At least, it works for me, and this document
indicates that it should be portable (unless we care about Netscape
4 :):
http://java.sun.com/j2se/1.4.2/docs/guide/plugin/developer_guide/cookie_support.html
Carl
Ted Kosan
> If the applet uses java.net.HTTPUrlConnection to talk to the server
> from Java, it doesn't have to log in separately... the cookies from
> the browser's login are automatically used by the Java
> communications. At least, it works for me, and this document
> indicates that it should be portable (unless we care about Netscape
> 4 :):
> http://java.sun.com/j2se/1.4.2/docs/guide/plugin/developer_guide/cookie_support.html
I also experimented with using the AsyncHttpRequest class:
Currently, however, the applet is being served from a different server
than the SAGE server it is communicating with. Doesn't this prevent
it from contacting the SAGE server directly unless it is signed?
Ted
Carl Witty
a file $SAGE_ROOT/local/java/foo.jar, then it is available through the
Sage server as https://localhost:8000/java/foo.jar (assuming default
settings).
Carl