[etoys-dev] Sugar vs legacy OS

[Bert Freudenberg]

On 29.10.2009, at 09:03, Walter Bender wrote:

> On Thu, Oct 29, 2009 at 8:57 AM, Bert Freudenberg <bert at freudenbergs.de 
> > wrote:
>> Begin forwarded message:
>>
>>> From: Bert Freudenberg <bert at freudenbergs.de>
>>> Date: 29. Oktober 2009 08:39:11 GMT-04:00
>>> To: Hilaire Fernandes <hilaire.fernandes at edu.ge.ch>
>>> Cc: OLPC Devel <devel at laptop.org>
>>> Subject: Re: Sharing files among several XO
>>>
>>>
>>> On 29.10.2009, at 02:47, Hilaire Fernandes wrote:
>>>
>>>> Under Etoys, If I keep down the mouse button and choose "Save under
>>>> another name" (I am not sure about the exact message), and save the
>>>> project with another name, then in the journal the previous  
>>>> instance
>>>> of the project is replaced with this new one. I will check again
>>>> tonight.
>>>>
>>>> Note: Pay attention I am not just using the Keep button
>>>
>>> Why?!
>>>
>>> Simply stop Etoys to save to the Journal, overwriting the previous
>>> entry.
>>>
>>> Simply click the Keep button to create a new entry in the Journal.
>>>
>>> Simply edit the name in the Etoys toolbar to rename. Or rename in  
>>> the
>>> Journal.
>>>
>>> The hidden save menu options let you save to a file or upload to the
>>> Squeakland website. They do not save to the Journal. They are hidden
>>> for a reason.
>>>
>>> - Bert -
>>
>> If anybody has ideas on how to avoid misconceptions like above,  
>> let's hear
>> them.
>
> I think that if the Sugar sharing and versioning were working better,
> then this problem/misconception would never have arisen. 0.84/86
> addresses many of the sharing issues. We'll be discussing versioning
> in Bolzano as part of our 0.88 plans. The good news is that OLPC seems
> to be motivated to get 0.86 onto the XO 1.5 machines.
>
> -walter

How would that address the differences between using Etoys on a  
machine with a user-managed file system and Etoys on Sugar with  
automatic saving?

Are you suggesting that when running under Sugar, we should disable  
Etoys' ability to work with the filesystem directly? What Hilaire ran  
into is that he accidentally used the lower-level filesystem interface  
instead of the higher-level Journal interface.

- Bert -