[Its.an.education.project] An OLPC Development Model

[david at lang.hm]

On Fri, 9 May 2008, Bert Freudenberg wrote:

> On 09.05.2008, at 09:56, david at lang.hm wrote:
>
>> which is why I fail to see the big point of Sugar.
>> [...]
>> a perfect example was the suggeation to make the sugarized
>> activities use
>> a standard file picker call so that it could go to the journal on
>> the XO
>> machine, or to a normal file selecteor window on other desktops.
>
> Your example indicates that you indeed fail to see the "big point" of
> Sugar. The point is to not have a document-centric environment, but an
> activity-centric one. Verbs rather than nouns. Yes it gets
> philosophical here. And I'm not the best to explain it.
>
> Maybe an analogy helps. Many developers fail to see the "big point" of
> object-oriented programming. For them, it's just that structs have
> function pointers now, so what's the big deal? But that misses the
> point completely, oo is all about decoupling and encapsulating
> concepts, it's a philosophy rather than an implementation technique.
> Or maybe the analogy does not help, depending on which camp one is in.
>
> Back to your example: even if all the world thinks applications with
> file dialogs are "normal" that does not imply it has to be that way.
>
> "Don't be misled by the enormous flow of money into bad defacto
> standards for unsophisticated buyers using poor adaptations of
> incomplete ideas." -- Alan Kay
>
> We do want to create something better than the status-quo. We may fail
> for a gazillion of reasons, but we're trying anyway. Children deserve
> the best, our future is in their hands.

Bert,
   if you try and say that the entire world is wrong in how it writes 
software, and only software specificly written for the Sugar environment 
should be available to the children, you are doing them a great 
dis-service.

it's fine to produce an alternate approach, but to bet-the-business on 
that approach with no fallback is betting that you know better then the 
rest of the world. it's possible that you are right, but not very likly.

however if you allow for compatibility you have a fallback.

many people have pointed out the limitations of the journal approach, and 
problems with not naming activites and files. yes it's easier to get 
started if you don't have to deal with confusing matters like directories, 
but as more documents are created a flat namespace for them will get 
overwelmed (be it a time-based journal, or a single layer home directory)

you are optimizing for the beginner so much that once they have used the 
system for a short time it will no longer be suitable for them.

dom't make the training wheels for beginners so ridgid that the kids can't 
remove them as they learn more.

As for your arguments about object oriented programming vs functional 
programming, opject orientation has it's place, but there have been a lot 
of evils foisted on us over the years under the banner of Object 
Orientation. just becouse an idea has merit doesn't meant that any 
implementation of that idea is automaticaly good.

David Lang