[Cuis-dev] Cuis, Dynabooks, Teaching and Learning

[Gerald Klix]

This better be a private answer.

On 2021-05-05 17:49, Hilaire Fernandes via Cuis-dev wrote:
> These are great requirements you are posting! I even think a dedicated 
> hardware is needed.
Yes something really sturdy!

My fifth child, she is 15, has an iPad provided by her school. As far I 
can, she handles it
well. (The have GeoGbra installed).

My wife works as a speech threapist (Logpädin)
at a Swiss school. She had to do remote therapy
during the look-down. Even in Switzerland not
every child has a suitable device and I am
talking just video-conferences (Please, do
not ask about the situation in Germany, were we live, I will start to 
rant ...).

My suggestion was provide them with cheap
Raspberry Pies.
Her answer was: "That would never
survive the first week or come
back in one piece".

Perhaps we should port the Opensmalltalk VM
to Nintendo's Switch. Rugged hardware
and children will value it more.
> 
> That's a big project indeed. I don't know much about founding hunting, I 
> guess it is complicate and time consuming.
I have considerably less experience with that
topic than you have. I just have about 30 years
of experience with software projects (small ones
and big ones, with more than 200 developers) as
a freelancer.

My personal situation is such, that it is hard
to acquire projects/contracts for me.
I am tempted to do that Haver project as full
time job, but that needs funding.

And to be a success it needs a showcase!

Dr. Geo would bee a really cool showcase,
OTH I don't want to offer substantial
help, when I am not sure that I can deliver.

I wonder if some a crowdfunding campaign
on Kickstarter, Indiegogo or CrowdSupply
would help.

If there is a 10% chance to start
a crowdfunding campaign for Haver or Dr. Geo
I will happily invest a week or two.

> 
> In the other hand I am still astonished to see every day how poor is the 
> impact of computer in teaching (in k12 and lower grades). In the other 
> hand I am gratefull the use of computer in education is poor because 
> regarding the existing option it will be a nighmare. Still most of the 
> society use computer on a daily basis but the option for the education 
> is poor. In my opinion, a perfect computer in education will be 
> ubiquitous : it will /look/ like a book when the learner needs a book, 
> it will /look/ like an exercises book when the learner need to 
> exercises, it will /look/ like a notebook when she needs to take 
> handwritten note, it will /look/ like a binder when she needs to access 
> her lessons, etc. All contents would be dynamic of course.
That's a lovely idea.

My wife uses her SurfaceBook with a second pen
for the pupils/patients,
this works much better than any laptop or
finger. So yes, an electronic book with pen will
work well.

OTH I see no way to equip every child
with this 3000€ device.

Perhaps it would be better if this dynabook is
one big brick of epoxy resin, like the Apollo Guidance Computer, 
otherwise these things will break every third day.

BTW: I am thinking about this Nintendo Switch
idea for more than 3 month.
I wonder if they would tolerate educational
Software in there AppShop.
> 
> And of course it be lightweight, very and replace this horrible 10kg bag 
> too many young learners needs to carry on.
Yes, but in this 15kg bag, the iPad is the least
problem. So even a 1kg device will do wonders,
if it is sturdy enough.
> 
> Hilaire
> 
> Le 30/04/2021 à 22:43, Gerald Klix a écrit :
>>
>> That got me really thinking.
>> I once had similar dreams.
--- snip ---
>>
>> - A simpler module system than mine,
>>   for this use case explicit imports are
>>   probably best.
I will resume working on Friday.

Since I have to modify or sub-class Cuis' compiler for that purpose, I 
was side-tracked
by an old idea of mine,
implementing access-levels like the Swiftlanguage has:
https://www.appcoda.com/swift-access-levels/

This means the compiler has to replace
the selectors of message sends with other
(private) selectors.

This in turn is an interesting feature,
if you want to replace English message selectors
with French ones. I think you see, were I am
driving at: I will help to implement language
specific scripting for Dr. Geo.

In addition to that:
I am pretty sure, the (horrible) method I am modifying 
(Parser>>#messagePart:repeat:)
can be modified so that the language obeys
ordinary arithmetic precedence rules.
Of course, it would not be Smalltalk
any more, but a big help for Dr. Geo
users. And there is no harm done, because
you can replace the compiler-class
on a class basis. Only the classes
accessible for scripting will have
the modified compiler.

It just needs a tiny operator precedence
sub-grammar for binary message sends:
https://en.wikipedia.org/wiki/Operator-precedence_grammar#Operator-precedence_languages

Enough of my ramblings.
I hope you got an impression of what I am up to.


Best Regards,

Gerald