Hi,
The current slots/commands in basic category is part turtle geometry and part cartesian geometry. I recast the viewer to include only turtle geometry primitives in basic to make etoys easier for children who are yet to develop intuitive understanding of co-ordinates. Why present two systems when one is sufficient?
I would have liked to make "pen down false/true" to "pen down/up" but I am yet to figure out a simple way to get Boolean slots to support arbitrary strings/symbols instead of just true/false values.
throwing it open for discussions, Subbu
Subbu,
The current slots/commands in basic category is part turtle geometry and part cartesian geometry. I recast the viewer to include only turtle geometry primitives in basic to make etoys easier for children who are yet to develop intuitive understanding of co-ordinates. Why present two systems when one is sufficient?
Even if we follow this line of thought, you would still like to keep "heading" in the basic category, yes?
Ideally, we could have easier version(s) of position vector representation in the basic category, but for various reasons (like hiding scary-looking "@" operator), we sort of substitute the vector representation of location with x and y. This has some merit: the first thing the end-user does after painting/instantiating an object is to pick it up by a mouse and move. If the viewer is opened, the x and y values in the viewer will change and give good indication that the viewer is actually a "view" on the object. (And, by typing the numbers, smart kids can tell that the relationship is bi-directional.)
In future systems, the end-user should be able to reorganize the categories for each object. This way, a teacher can set up a nicer material, for example. You have a point, but but for now, I think current affair is "ok".
In future systems, various different notation/representation of vectors should be really nice, too. What kind of notation would be good for children?
I would have liked to make "pen down false/true" to "pen down/up" but I am yet to figure out a simple way to get Boolean slots to support arbitrary strings/symbols instead of just true/false values.
Hmm, you might have to define a new DataType for it...
throwing it open for discussions,
Thank you!
-- Yoshiki
On Thursday 01 November 2007 11:37 am, Yoshiki Ohshima wrote:
Subbu,
The current slots/commands in basic category is part turtle geometry and part cartesian geometry. I recast the viewer to include only turtle geometry primitives in basic to make etoys easier for children who are yet to develop intuitive understanding of co-ordinates. Why present two systems when one is sufficient?
Even if we follow this line of thought, you would still like to keep "heading" in the basic category, yes?
I am not so sure. Three reasons. Firstly, heading counter is not really needed to "play turtle". 5-7yr olds are quite sensitive to small changes and the changing counters could divert their attention from the etoy. Secondly, heading is based on an absolute reference. Body systonic movements are always relative and there is no fixed "north". Older children who understand absolute measures will intuitively reach out into geometry for the heading counter. Thirdly, "north" is forward only for top-view (as in children playing with toys on the floor). When sketching on screen, children tend to draw side-view or front-view and then the "forward" movement is disconcerting (cf. Alan the mouse in Squeaky tales video).
... the first thing the end-user does after painting/instantiating an object is to pick it up by a mouse and move.
Seymour Papert invented the "turtle" to get children to steer it indirectly through instructions and not through direct manipulation. Picking up etoys and moving them around helps build hand-eye co-ordination but I believe opening up the etoy and controlling its behavior through scripts is what stimulates deep learning thru body systonic explorations.
In future systems, the end-user should be able to reorganize the categories for each object. This way, a teacher can set up a nicer material, for example.
Excellent! The "tear off" could also be used for a new slot instead of the "v" button. Both variables and behaviors could be added by tear-off->define->categorize sequence
You have a point, but but for now, I think current affair is "ok".
It serves 9-12yrs well. I was trying to reach down to younger kids. The Basic tab in the catalog does not have an etoy. An low cost change would be to add a turtle sketch with pen down and black color nib, so children can start using it to create pen trails right away.
In future systems, various different notation/representation of vectors should be really nice, too. What kind of notation would be good for children?
The arrow serves well to indicate both direction (head) and length (stem).
I would have liked to make "pen down false/true" to "pen down/up" but I am yet to figure out a simple way to get Boolean slots to support arbitrary strings/symbols instead of just true/false values.
Hmm, you might have to define a new DataType for it...
That wouldn't be extensible. I am trying to get Boolean slots to display any two names - up/down, inside/outside, dead/alive etc. so that the tile names reflect real world behaviors and don't look contrived.
Subbu
Subbu,
I am not so sure. Three reasons. Firstly, heading counter is not really needed to "play turtle". 5-7yr olds are quite sensitive to small changes and the changing counters could divert their attention from the etoy.
You could drop the bi-directional connection idea for the younger learners, yes. But can end up with something a bit different from Etoys.
Secondly, heading is based on an absolute reference. Body systonic movements are always relative and there is no fixed "north". Older children who understand absolute measures will intuitively reach out into geometry for the heading counter.
However, the turtle does move and rotate on screen currently. Unless we can really fix the turtle on screen and use the turtle local coordinate system throughout and move the rest of the world (i.e., "turn left" makes the world rotates right, and "forward" means the world slides backward). If so, I could agree with the idea to take out the global coordinate values.
Thirdly, "north" is forward only for top-view (as in children playing with toys on the floor). When sketching on screen, children tend to draw side-view or front-view and then the "forward" movement is disconcerting (cf. Alan the mouse in Squeaky tales video).
Yes, that is a brilliant part of Logo. But I think this is a weak argument to remove heading from basic.
... the first thing the end-user does after painting/instantiating an object is to pick it up by a mouse and move.
Seymour Papert invented the "turtle" to get children to steer it indirectly through instructions and not through direct manipulation. Picking up etoys and moving them around helps build hand-eye co-ordination but I believe opening up the etoy and controlling its behavior through scripts is what stimulates deep learning thru body systonic explorations.
Well, I don't disagree. But a typical Etoys session begins with drawing a painting, so how to merge the idea you described into that flow is not trivial (like removing heading and x, y would not.)
You have a point, but but for now, I think current affair is "ok".
It serves 9-12yrs well. I was trying to reach down to younger kids. The Basic tab in the catalog does not have an etoy. An low cost change would be to add a turtle sketch with pen down and black color nib, so children can start using it to create pen trails right away.
Actually, we can have a project like that. We can in fact a lot of projects to do different tasks and include them. (Even like calculator, painting, camera manipulation, etc.)
In future systems, various different notation/representation of vectors should be really nice, too. What kind of notation would be good for children?
The arrow serves well to indicate both direction (head) and length (stem).
That is one. And how to use it in scripts is something interesting to think about.
I would have liked to make "pen down false/true" to "pen down/up" but I am yet to figure out a simple way to get Boolean slots to support arbitrary strings/symbols instead of just true/false values.
Hmm, you might have to define a new DataType for it...
That wouldn't be extensible. I am trying to get Boolean slots to display any two names - up/down, inside/outside, dead/alive etc. so that the tile names reflect real world behaviors and don't look contrived.
Yes. DataType and friends is the classes to modify to provide that.
Thank you!
-- Yoshiki
On Saturday 03 November 2007 4:11 am, Yoshiki Ohshima wrote:
Subbu,
I am not so sure. Three reasons. Firstly, heading counter is not really needed to "play turtle". 5-7yr olds are quite sensitive to small changes and the changing counters could divert their attention from the etoy.
You could drop the bi-directional connection idea for the younger learners, yes. But can end up with something a bit different from Etoys.
Once a young learner gets used to the idea of control and starts exploring other properties, (s)he will encounter it in geometry. heading is not gone, it continues to be in geometry category. So how does it change Etoys?
Secondly, heading is based on an absolute reference. Body systonic movements are always relative and there is no fixed "north". Older children who understand absolute measures will intuitively reach out into geometry for the heading counter.
However, the turtle does move and rotate on screen currently. Unless we can really fix the turtle on screen and use the turtle local coordinate system throughout and move the rest of the world (i.e., "turn left" makes the world rotates right, and "forward" means the world slides backward). If so, I could agree with the idea to take out the global coordinate values.
I am not suggesting that we change the co-ord system. Currently Etoys 101 introduces both turtle and cartesian co-ords (which was fine for its initial test group of 9-12yr olds). All I am suggesting is to keep only simple turtle co-ords for Etoys 101 (5-7 yr olds) and bring in cartesian co-ords (geometry) in Etoys 201. Recall the case of Deborah (Mindstorms, page 118).
Thirdly, "north" is forward only for top-view (as in children playing with toys on the floor). When sketching on screen, children tend to draw side-view or front-view and then the "forward" movement is disconcerting (cf. Alan the mouse in Squeaky tales video).
Yes, that is a brilliant part of Logo. But I think this is a weak argument to remove heading from basic.
The observation is from watching 6-9yr olds (not all of whom know English) using Squeak. I am having way too much fun with kids in this age group :-). Getting a sketch to draw a triangle, square etc. doesn't need heading. It is more of a distraction in basic category. heading makes sense only to kids old enough to understand compass points and negative numbers.
a typical Etoys session begins with drawing a painting, so how to merge the idea you described into that flow is not trivial (like removing heading and x, y would not.)
It still does. Kids still get to sketch first and script a circle. Young kids continue their fun with shapes and trails with their car while older compass aware kids can build wheels and control the car with that. My suggestion is to remove these only from the basic category. The cartesian co-ords will be available in geometry.
That wouldn't be extensible. I am trying to get Boolean slots to display any two names - up/down, inside/outside, dead/alive etc. so that the tile names reflect real world behaviors and don't look contrived.
Yes. DataType and friends is the classes to modify to provide that.
Thanks for the pointer.
Subbu
etoys-dev@lists.squeakfoundation.org
-
subbukk -
Yoshiki Ohshima