mds-core — Medium traffic list for developers.

[mds] personal info

[Gabe J. <joh...@cs...>]

hey guys, i've been putting together a web site for the MDS, and
I want to have a 'people' section in there, unless anybody has
objections. so let me know what you want up by sending me any of
these things:

Name (duh), an email address you don't mind having on the web,
your homepage, and a short (<200 chars) description of you.

This section also has the standard thanks to those who fund us,
the amath dept and L3D. Eric, do we have a name to attach to the
grant money we've been using? A link to that would be good, if
there is one. Actually, that's why i wanted to have this section
to begin with, and thought it would look silly if it just
consisted of ass kissing.

-gabe

[mds] the tree works, it is very cool

[Gabe J. <joh...@cs...>]

If you haven't taken a look at the state of things lately, you
might want to check it out now. I fixed the tree problem, and it
works beautifully. I tested it on one of the nice machines in
ECCR 143, and it simply smoked. It felt just like a normal
windows program. The URL that you may want to look at is:

http://mds.colorado.edu/x305.html

By the way, colin, (if you're still reading this!) the problem
with the tree was that the panel we were putting the JTree into
has a default layout manager of FlowLayout, which doesn't let
things grow. The only thing that was needed to get it to work was
setting the layout to BorderLayout (which will let the center 
component grow to its hearts content) and then adding it to the
center. So, FYI, now you know. You may sleep comfortably, now.

-gabe

[mds] announcing the mds-announce list

[Gabe J. <joh...@cs...>]

This message is an invitation to join the mds-announce mailing
list. I'm sending this out to anybody who has at any point in
time expressed some level of interest in this project. Instead of
pestering people that potentially don't really care, I've set up
this mailing list for people to subscribe to if they are
interested in knowing about future developments and releases of
the MDS. 

This will be a very low traffic list--in fact, the only person
that can post to it is myself--I expect it will only be used once
every few months or so. If you would like to join this list,
visit the URL:

http://lists.sourceforge.net/mailman/listinfo/mds-announce

... and enter your email address. Reply to the message it sends
out, and you're set.

Thanks everybody. Those of you who don't sign up will be bothered
no more!

-gabe

[mds] MDS status

[Gabe J. <joh...@cs...>]

Hi Dr. Curry. I'm attaching a document (HTML format) that Eric
and I have written. It is a status report that you might find
informative. It details where we are now and where we might go in
the future. If you would want to discuss this, we can set up a
time to get together. My schedule is fairly open. I'll paste it
in here, and you can pick a time that works for you.

Monday:    
            10 - 11      Digital Logic II
            1 - 2        Theory of Computation
Tuesday:    
            2 - 4        Digital Lab
Wednesday:
            10 - 11      Digital Logic II
            1 - 2        Theory of Computation
Thursday:
            2 - 4        Digital Lab
Friday:
            10 - 11      Digital Logic II
            1 - 2        Theory of Computation
            2 - 3        Office Hours for Java II

-gabe

Re: [mds] MDS maiden voyage

[Gabe J. <joh...@cs...>]

"Eric S. Wright" wrote:
> 
> Would you send me the staroffice version of this?  I will try to distill
> it so that curry can have his summary too.  I'll run it by you when I am
> done.

here you go.

-gabe

Re: [mds] MDS maiden voyage

[Eric S. W. <ew...@ne...>]

Would you send me the staroffice version of this?  I will try to distill
it so that curry can have his summary too.  I'll run it by you when I am
done.
-Eric
Gabe Johnson wrote:
> 
> I've just written up a "here we are, here's where we may go" sort
> of thingy, and I've put it up on mds.colorado.edu. I'd like
> comments on this, so if you have nothing better to do with your
> time, you might want to wonder over and take a look. It's about 4
> pages single spaced, but it's fairly fluffy so it shouldn't be
> that hard to digest. I put a single hook at the top so you can
> comment on it if you like. (and help test the theory to a small
> degree, even though hooking doesn't work as it should)
> 
> In order to run it you will need to get the Java 1.2 plugin for
> your browser. If you go to the file, it will help you out with
> this. Also, if you have any MDS classes in your system's
> CLASSPATH, that will probably interfere with things, so unset
> that.
> 
> Also, if you have problems, let me know what sort of setup you
> have (browser, OS, plugin version, astrological sign, etc)
> 
> the file is located at http://mds.colorado.edu/status_nov_1.html
> 
> have fun.
> 
> -gabe
> _______________________________________________
> mds-core mailing list
> mds...@li...
> http://lists.sourceforge.net/mailman/listinfo/mds-core

-- 
*-------------------------------------------------------------*
|Eric S. Wright		| University of Colorado at Boulder   |	
|ew...@co...	| Department of Applied Mathematics   |
|phone:	(303) 735-4165	| Campus Box 526                      |
|fax:	(303) 492-4066	| Boulder, CO 80309                   |
|-------------------------------------------------------------|
|	http://amath.colorado.edu/appm/student/ewright        |
*-------------------------------------------------------------*

[mds] your usernames and passwords

[Gabe J. <joh...@cs...>]

I forgot a minor detail... your user names and passwords.

For the most part your username is whatever your normal user name
is, and your password is just your last name. This is mighty
secure, eh? Oh, EDS, your username is 'eds'. 

blah.
-gabe

[mds] MDS maiden voyage

[Gabe J. <joh...@cs...>]

I've just written up a "here we are, here's where we may go" sort
of thingy, and I've put it up on mds.colorado.edu. I'd like
comments on this, so if you have nothing better to do with your
time, you might want to wonder over and take a look. It's about 4
pages single spaced, but it's fairly fluffy so it shouldn't be
that hard to digest. I put a single hook at the top so you can
comment on it if you like. (and help test the theory to a small
degree, even though hooking doesn't work as it should)

In order to run it you will need to get the Java 1.2 plugin for
your browser. If you go to the file, it will help you out with
this. Also, if you have any MDS classes in your system's
CLASSPATH, that will probably interfere with things, so unset
that.

Also, if you have problems, let me know what sort of setup you
have (browser, OS, plugin version, astrological sign, etc)

the file is located at http://mds.colorado.edu/status_nov_1.html

have fun.

-gabe

[mds] MDS demo Thursday (tomorrow) in the L3D lab at 4:15

[Gabe J. <joh...@cs...>]

Hello everybody. 

I am going to give a presentation on some of my recent work in the
L3D lab tomorrow (Thursday, October 26) at 4:15. 

The system, aka the "Mathematical Discussion System", or MDS for
short, is a tool for enabling people to embed conversations
directly into an online textbook. There is more information in our
initial project proposal, which was written by myself and Eric
Wright, an applied math doctoral student. The URL for this can be
found on our horribly out of date web site:

http://mds.sourceforge.net/archives/initial_proposal.html

The presentation is mainly for the benefit of Dr. Curry (who
helped fund the effort and now understandably wants to make sure
that we did something useful with the money), but everybody is
welcome to attend. Also, Leysia and I may conspire to set up an
L3D meeting on the subject at some point in the near future.

--
-gabe
http://www.cs.colorado.edu/~johnsogg/

[mds] Sockets work

[Gabe J. <joh...@cs...>]

Hey guys. 

I feel obligated to let you know where I am with things. You saw
my desparate plea for help to the netscape newsgroup (that got no
response), and I've given up on using RMI. So my plan for now is
to re-implement the client/server communication for using Sockets.
I have written a very simple test program to make sure this is
possible, and as far as I can tell, this works. I'm going to keep
testing it at every possible point, so I can make sure it doesn't
spontaneously stop working (like RMI bailed on us). 

I'm guessing that this will take tonight and tomorrow. Fortunately
I no longer have a 9am class, so I can stay up and code. I might
get done tonight... if I do I will let you know, and you can go
pound on it.

-gabe

Re: [mds] Re: can't use RMI to connect to the server the code came from

[Colin T. <Colin.Tucker@Colorado.EDU>]

Gabe-

what version of Java are you using to compile your code with?  maybe you
should try using the oldest one you can.  just a random idea...

cya
Colin.

[mds] Re: can't use RMI to connect to the server the code came from

[Eric S. <Eric.Scharff@Colorado.EDU>]

Is this problem unique to Netscape?  If not, sending a full exception with
line numbers would be really helpful.

-Eric

[mds] can't use RMI to connect to the server the code came from

[Gabe J. <joh...@cs...>]

I have a fairly simple client/server system that should be
communicating via RMI. We have apache running on mds.colorado.edu,
which hands out a jar 'mds.jar' to clients. The applet runs okay
until there is any RMI usage. I have a debugging statement that
System.outs the server it's looking for,
rmi://mds.colorado.edu/server, which is correct. However,
communicator (4.74 on Linux) throws an exception:

java.rmi.UnexpectedException: Unexpected exception; nested
exception is: 
	netscape.security.AppletSecurityException: security.Couldn't
connect to '127.0.0.1' with origin from 'mds.colorado.edu'.

I'm actually not entirely sure how to interperet this. I'm
assuming that it realizes that the code came from
mds.colorado.edu, but that it is being asked to talk to the
localhost.

It behaves correctly when the server and client are running on the
same host, and it behaves correctly through appletviewer.

Any ideas? If you want to look, the apache URL at the moment is
http://mds.colorado.edu/testing/ and our code is at
http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/mds/?cvsroot=mds .

--
-gabe
http://www.cs.colorado.edu/~johnsogg/

[mds] i need mysql_uncomp.jar

[Gabe J. <joh...@cs...>]

Does anybody have the file 'mysql_uncomp.jar'? If you have it,
please email it to me. That's the database driver that we've been
using.

-gabe

[mds] xml

[Rian B. <ria...@co...>]

Hey guys,

well the parsing has gone reasonably well until this point. I can take a
xml file with MathMl specfic markup for a matrix, parse it and have an
applet display it. The problem comes wihen the applet tries to display a
matrix which contains text data (as opposed to an empty one) It display
correctly breifly and then spews garbage all over the applet panel. And
empty matrix displays and edits just fine. I have no clue how to debug
this thing, since it is a mysterious awt issue, related to text. I am
open to ideas, including just serializing the damn things.

Rian

--
*********************************************************
*Rian Bogle                                             *
*Department of Applied Mathematics                      *
*University of Colorado                                 *
*-------------------------------------------------------*
*Work: (303) 492-3751                                   *
*Web: http://amath.colorado.edu/appm/student/boglerc/   *
*Email: Rian.Bogle@Colorado.edu                         *
*********************************************************

Re: [mds] chapter 1

[Eric W. <ew...@ba...>]

11 is fine.  I will be in the newton lab around that time.

-Eric

Colin Tucker wrote:

> Eric--
>
> okay, when would be a good time for you tommorow?  11AM would work well
> for me, but I'm flexible.  I've been thinking quite a lot about what the
> hooks are going to look like- I think it's very important since they're so
> prominent in the overall user experience.
>
> cya
> Colin.
>
> > Colin, let's get together monday and talk about hooks for this thing.
> >
> > -Eric
> > --
> >   ------------------------------------------------------------------------
> >         Eric
> >         Wright  Department of Applied Mathematics
> >
> >         Office: Grandview site (N.E. corner of 15th and
> >  [Image]        grandview)                                      [Image]
> >         Phone:  (303) 735-4165
> >         E-mail: ew...@ba...
> >         WWW:   http://amath.colorado.edu/appm/student/ewright
> >
> >
> >
>
> _______________________________________________
> mds-core mailing list
> mds...@li...
> http://lists.sourceforge.net/mailman/listinfo/mds-core

--
  ------------------------------------------------------------------------
        Eric Wright Department of Applied Mathematics
        Office:     Grandview site (N.E. corner of 15th and grandview)
 [Image]Phone:      (303) 735-4165                                     [Image]
        E-mail:     ew...@ba...
        WWW:        http://amath.colorado.edu/appm/student/ewright

Re: [mds] chapter 1

[Colin T. <Colin.Tucker@Colorado.EDU>]

Eric--

okay, when would be a good time for you tommorow?  11AM would work well
for me, but I'm flexible.  I've been thinking quite a lot about what the
hooks are going to look like- I think it's very important since they're so
prominent in the overall user experience.  

cya
Colin.

> Colin, let's get together monday and talk about hooks for this thing.
> 
> -Eric
> --
>   ------------------------------------------------------------------------
>         Eric
>         Wright  Department of Applied Mathematics
> 
>         Office: Grandview site (N.E. corner of 15th and
>  [Image]        grandview)                                      [Image]
>         Phone:  (303) 735-4165
>         E-mail: ew...@ba...
>         WWW:    http://amath.colorado.edu/appm/student/ewright
> 
> 
> 

Re: [mds] compilation errors in the symbol stuff

[rian <ria...@co...>]

Gabe,

the MdsDiv is a dead class and should not be in the current tree but
rather in the attic. You probably just have it leftover from an old
checkout.

As far as i know all the symbol.* compiles.

Each symbol now writes out its content in xml markup. Just call
toString(). I'll be adding the parsing later tonight.

later,

Rian

[mds] chapter 1

[Eric W. <ew...@ba...>]

Hello,

Chapter 1 is now pretty much done.
You can view it at :

http://laguerre.colorado.edu/MDS/linear

Could some of you take a look at the applets to see if there are any
glaring mathematical errors?  I will do the same, but I am notoriously
prone for missing my own obvious errors.

Colin, let's get together monday and talk about hooks for this thing.

-Eric
--
  ------------------------------------------------------------------------
        Eric
        Wright  Department of Applied Mathematics

        Office: Grandview site (N.E. corner of 15th and
 [Image]        grandview)                                      [Image]
        Phone:  (303) 735-4165
        E-mail: ew...@ba...
        WWW:    http://amath.colorado.edu/appm/student/ewright

Re: [mds] (almost) friday update

[Colin T. <Colin.Tucker@Colorado.EDU>]

Eric - can you send me the lastest version of your text please?
 
> When you design the hook, keep two things in mind:
> 
> 1) It should be conspicuous enough that there's no question about it:
> clicking on this thing will let you read messages or will cause you to
> start writing a message. Right now the hook I have in node4.html only
> sends you to read-mode, but we need to have a dual-purpose hook: read
> and write. Look at the js for that, and create an appropriate js
> function for writing. I'll deal with the Enviro.

where is this wonderous thing known only as "node4.html"?

when you talk about dual purpose are you meaning an image map, two
images, or something else? 

> 2) Eventually we will liven things up so that the web server will
> magically make the images look different when there's more or less
> content. Maybe leaving some whitespace in the image would be a good
> idea. I dunno, do what you can do.

cool, I'm going to rope an unknowing but artistic roommate into designing
the hook with me since I generally suck at that kind of thing ;)

cya

Colin.

Re: [mds] (almost) friday update

[Gabe J. <joh...@cs...>]

> I was and still am tempted to convert it to a tree-table because I
think
> that would look cooler, but it's tricky to do, and I couldn't find a
> decent example of one anywhere.  I plan to come back to this in my
own
> time though.
>
> I'm in Newton right now for the afternoon, I'm going to document my
code,
> go back and look at the javascript and add some platform/browser
checks
> that spawn warning dialogs.  Also I'll design some new hooks, and
start
> putting them in the text.
>
> exactly what version of Netscape & IE does our applet require now?

For now, your tree component is fine. I think your mission at this
point should be to work with Eric on hookifying the text. He's got
some applet-ized version of the first chapter, so you need to get that
from him, and whatever other changes he's made. Like I said when we
were chatting in Newton the other day, I think our hooks should be a
bit more sparse. I think that a hook at the top of every section would
do the trick

When you design the hook, keep two things in mind:

1) It should be conspicuous enough that there's no question about it:
clicking on this thing will let you read messages or will cause you to
start writing a message. Right now the hook I have in node4.html only
sends you to read-mode, but we need to have a dual-purpose hook: read
and write. Look at the js for that, and create an appropriate js
function for writing. I'll deal with the Enviro.

2) Eventually we will liven things up so that the web server will
magically make the images look different when there's more or less
content. Maybe leaving some whitespace in the image would be a good
idea. I dunno, do what you can do.

Only now, at the end, do you understand the power of the dark side of
the force.

Re: [mds] friday update

[Gabe J. <joh...@cs...>]

> The example applets for chapter 1 are done.  I still need to check
them
> for mathematical errors, but they all look and work correctly.
While
> writing them, however, I noticed that VisualMatrix and its
derivatives do
> not play nicely with GridLayout.  GridLayout tries to stretch the
> components to equal sizes.  This seems to conflict with the way that
they
> resize themselves.  It causes an infinite loop of paint() calls if
you
> have two matrices with different preferred sizes in the Grid.  No
matter,
> I just stuck the matrices in their own FlowLayout Panels befor
putting
> them in grid cells.  This seemed to work.  Just something to be
aware of.

Maybe we could change the MdsMathComp so that it includes this
two-leveled container thing. Sort of like how many swing components
have a 'content pane' where you're supposed to stick things.

> Gabe, when would you like to start talking about hooks?

See my other message. I'm putting Colin on hook detail, you need to
make sure he's using the right version of the text (applets and
whatnot included).

> Oh.  I will try to get a web server up and running this weekend.  No
> promises though.  I have a spectral solver to finish writing too.
I'm
> becoming a *&%*%&^$ code monkey.

If you need help with this (I've set up web servers more times than I
can count) give me a call. I will either be at home (720) 565-1127 or
in the lab.

Are we all down for lunch on Tuesday? I think we should try and go for
beers or something in the PM.

[mds] (almost) friday update

[Colin T. <Colin.Tucker@Colorado.EDU>]

Well my week was spent living, breathing, thinking and dreaming in JTrees. 
Apart from some documentation and other little shit, I'm done with the
MdsTreePanel.java (woohoo!)

I was and still am tempted to convert it to a tree-table because I think
that would look cooler, but it's tricky to do, and I couldn't find a
decent example of one anywhere.  I plan to come back to this in my own
time though.

I'm in Newton right now for the afternoon, I'm going to document my code,
go back and look at the javascript and add some platform/browser checks
that spawn warning dialogs.  Also I'll design some new hooks, and start
putting them in the text.

exactly what version of Netscape & IE does our applet require now? 

later
Col.

(720) 304-7792

[mds] friday update

[Wright E. <ew...@ba...>]

Howdy folks,

The example applets for chapter 1 are done.  I still need to check them
for mathematical errors, but they all look and work correctly.  While
writing them, however, I noticed that VisualMatrix and its derivatives do
not play nicely with GridLayout.  GridLayout tries to stretch the
components to equal sizes.  This seems to conflict with the way that they
resize themselves.  It causes an infinite loop of paint() calls if you
have two matrices with different preferred sizes in the Grid.  No matter,
I just stuck the matrices in their own FlowLayout Panels befor putting
them in grid cells.  This seemed to work.  Just something to be aware of.

I am going to pull a night session at my office today (saturday).  I
should be able to finish the drill applet so that chapter 1 will have all
of its applets complete.  I will also try to fix any mathematical errors
in the example applets.

Gabe, when would you like to start talking about hooks?

Oh.  I will try to get a web server up and running this weekend.  No
promises though.  I have a spectral solver to finish writing too.  I'm
becoming a *&%*%&^$ code monkey.

-Eric

------------------------------------------------------------------
Eric Wright	Department of Applied Mathematics
Office		Grandview site (nw corner of 15th and grandview)
Phone		(303) 735-4165
E-mail		ew...@ba...
WWW		http://amath.colorado.edu/appm/student/ewright
------------------------------------------------------------------

[mds] compilation errors in the symbol stuff

[Gabe J. <joh...@cs...>]

Rian, i'll let you handle this. I think all you need to do is
make that final method non-final.

client/symbols/MdsDiv.java:14: The method void
paint(java.awt.Graphics) declared in class
mds.client.symbols.MdsDiv cannot override the final method of the
same signature declared in class mds.client.symbols.MdsMathComp. 
Final methods cannot be overridden.
        public void paint(java.awt.Graphics g){
                    ^
client/symbols/MdsDiv.java:30: The method java.awt.Dimension
getMinimumSize() declared in class mds.client.symbols.MdsDiv
cannot override the final method of the same signature declared
in class mds.client.symbols.MdsMathComp.  Final methods cannot be
overridden.
        public java.awt.Dimension getMinimumSize(){
                                  ^
client/symbols/MdsDiv.java:58: The method void
paint(java.awt.Graphics) declared in inner class
mds.client.symbols.MdsDiv. Mid cannot override the final method
of the same signature declared in class
mds.client.symbols.MdsMathComp.  Final methods cannot be
overridden.
                public void paint(java.awt.Graphics g){
                            ^
3 errors
make: *** [client/symbols/MdsDiv.class] Error 1  


-- 
-gabe
http://www.cs.colorado.edu/~johnsogg/