[Cuis-dev] About complexity in the face of the user

[ken.dickey at whidbey.com]

On 2024-03-18 09:49, Douglas Brebner via Cuis-dev wrote:
> On 17/03/2024 21:29, ken.dickey--- via Cuis-dev wrote:
>> There was a Scratch done some time back for the RaspBerry Pi called 
>> NuScratch, so we should be able to get the Smalltalk source code:
> 
> Doesn't Tim Rowledge maintain an up to date port of Scratch to current
> Squeak for the Pi?
> 
> Maybe he'd be the best person to ask about it.

Yes.  I am looking at NuScratch20210213-64.image on a RasPi 5.  Tim is 
the one.

The docs indicate "When an error occurs in a script, if you are not in 
fill screen mode, or have turned error catching off, Squeak displays 
popups instead of the block turning red".

ScratchStageMorph>>stepProcess catches errors which is checked by 
BlockMorph>>updateCachedFeedbackForm but I am still reading sources and 
puzzling things out.

The expectation seems to be "correct by construction", with immediate 
feedback by a block turning Red and a beep, but I have yet to find a 
debugger.

ScratchFrameMorph has a #UserErrorCatcher but I have yet to determine 
its use.

Still exploring the code..

A small DrGeo construction language with a simple evaluator could be 
presented with blocks or a simplified Smalltalk subset debugger.

The question for us is, what is the pattern of DrGeo scripting and what 
breakage do we have to deal with.

Some exploration and prototyping required.  Hey Cuis is great for this!  
:)
-KenD