Scratch

@Hardmath123

See also post #3859 et #3862

Taneb wrote:

ALSO also, the .deb installer doesn't work for me. I'm on Ubuntu 11.04, which may explain it, but still.

what error?
made on 11.10, works for me

Taneb wrote:

"Internal Error: The file could not be opened"

Hmm, I wonder if there was a problem with my download.  The one on the web site is 2391793 bytes long; is that what it should be? (@roijac)

@Jens
This is a typical example where a preudo byte-code would be an advantage.
No dobt that you will see that in due time !!

xly wrote:

@Jens
When I decided to rewrite my fractals examples from Byob to Snap! I thought it was a simple formality. In  fact it took me hours. The trick is that with Snap!  - contrarily to Byob and Logo - you need to omit the final STOP of the IF part of the Script. See the result :
http://www.xleroy.net/ByobTuto/New/Snap!/ccurve.gif

show off 

Taneb wrote:

Well, that gave me a different error message. "The package is of bad quality: The installation of a package which violates the quality standards isn't allowed. This could cause serious problems on your computer. Please contact the person or organisation who provided this package and include the details beneath"

try rename it to byob-3-1-1 (lower-case)
is it working now?

bharvey wrote:

Hardmath123 wrote:

By any chance, could you give me the code for an individual block? I'd like to try making some Snap! blocks.

Load Snap! in your browser, then View Source, then click on objects.js to get the latest greatest.  But bear in mind that a lot of Control blocks that are currently handled as special cases will be regular blocks by the time we're done.

And the code?

EDIT: Got it. Here's a block:

Process.prototype.reportExp = function (a, b) {
    return Math.pow(parseFloat(a), parseFloat(b));
};

blocks.push(block(
        'reporter',
        'operators',
        '%n ^ %n',
        'reportExp'
    ));

Last edited by Hardmath123 (2011-10-23 09:14:14)

roijac wrote:

rename it to byob-3-1-1 (lower-case)

I re-downloaded and put it on the BYOB page again.

^ Cool! 

Last edited by shadow_7283 (2011-10-23 11:27:42)

Hardmath123 wrote:

I'm trying to make a block installer with a javascript:() bookmarklet which asks for the two codes. How do I edit the "function() =" with the 'blockspecs'? The code, I assume, can be added as a simple (global) function.

Maybe I'm missing the point about bookmarklets.  What is the end goal you're trying to accomplish that the proposed callJS won't do?

The reason for this plan is that we hope to have a minimum of primitive blocks, extended by both our own and user-provided libraries that define other things as custom blocks.  We've even talked about moving some of the redundant Scratch blocks (e.g., CHANGE <var> BY <delta> ) into a library.  So, we want users to be able to customize the language with Snap!-level programming, rather than JS-level programming.  But there will be a lot of JS capabilities that couldn't be done entirely in Snap!, so we want one ring to... no, I mean, one block that will give access to the entire zoo of JS libraries.

But, for instance, we don't want to encourage users to import JS control structures (e.g., FOR loops); if it can be done in Snap!, then it should be (and it should run fast enough not to discourage users from relying on custom blocks).  Not that you shouldn't program in JS if you want, but if that's what you want you should fire up Emacs and write a .js file.

rubiks_cube_guy238 wrote:

Instead of this, why not add a block like (JavaScript[alert()]) that can be put into a call or run block. It looks nicer, and it adds one block instead of two (or three if a launch block is planned)

I think your version will have to be two shapes, too; no matter how you handle the JS side of the interface, the Snap! side of the interface distinguishes blocks that report values from ones that don't.

The reason to make it JS apply (call one function) rather than eval (run some piece of code) is that we are trying to provide access to low-level (e.g., hardware interface) primitives, rather than to encourage people to write whole scripts in JS instead of in blocks.

Of course people could do that anyway by writing a JS function in a private library and then using our apply block to call it, but nevertheless we think that the two possibilities encourage different programming styles.

bharvey wrote:

one ring to...

...rule them all? 

Of course, what else did you think I meant?? 

was bored 

Last edited by roijac (2011-10-24 11:17:00)