Scratch

asampal wrote:

Hi,

I've got a couple of questions - one about BYOB and the other about Snap!

In BYOB, if I import the tools project, I get the error "Unknown class VariableDeclarationBlockMorph" when I try to save out the project that I imported in. Is there a workaround for this issue?

With Snap!, I find the UI to be quite sluggish, e.g. scrolling the list of blocks is really slow even with Firefox 6 and hardware acceleration. Is this something that remains to be optimized? As it is, I prefer the standalone BYOB for its responsiveness. BTW, will the standalone Smalltalk version of BYOB be abandoned altogether. Could you at least look at putting out an updated version 3.1.1 based on Pharo with the Cog VM?

Cheers,
Adrian

I'm gonna have to agree with you in terms of snap! being sluggish, even on my intel i7 2600k (top of the range) I've only been trying with firefox though, maybe chrome is faster or something?

rookwood101 wrote:

asampal wrote:

Hi,

I've got a couple of questions - one about BYOB and the other about Snap!

In BYOB, if I import the tools project, I get the error "Unknown class VariableDeclarationBlockMorph" when I try to save out the project that I imported in. Is there a workaround for this issue?

With Snap!, I find the UI to be quite sluggish, e.g. scrolling the list of blocks is really slow even with Firefox 6 and hardware acceleration. Is this something that remains to be optimized? As it is, I prefer the standalone BYOB for its responsiveness. BTW, will the standalone Smalltalk version of BYOB be abandoned altogether. Could you at least look at putting out an updated version 3.1.1 based on Pharo with the Cog VM?

Cheers,
Adrian

I'm gonna have to agree with you in terms of snap! being sluggish, even on my intel i7 2600k (top of the range) I've only been trying with firefox though, maybe chrome is faster or something?

It does seem to be faster in google chrome, although still not really fast.

Smalltalk on Javascript. This might have been very useful some time ago, perhaps somewhat useful or at least interesting even now. 

http://amber-lang.net/

bharvey wrote:

joefarebrother wrote:

or.. implement firs class colours like in panther!

On the list.  But remember, no major new features in 4.0, just getting the 3.1 stuff working in JS.

I thought it was html

itsme213 wrote:

Smalltalk on Javascript. (...)
http://amber-lang.net/

Hi, Sophie.
I love the Amber project! Looks like they're checking out Snap!'s Morphic GUI:  http://irclogger.com/.amber-lang/2011-09-20

bharvey wrote:

titaniumbones wrote:

The issue is, how to detect if a sprite X is touching another sprite of 'kind' Y.

I think this should do it:

http://cs.berkeley.edu/~bh/touch-kind.png

Here, "ant" is the name of the parent of all the ants (type Y), and "myself" is the object that wants to know who's touching it (object X).  The OBJECT (MYSELF) block is providing an extra input to ASK that fills the input slot in the TOUCHING? block.  You can't just say TOUCHING? (OBJECT (MYSELF)) because then "myself" would be the object being asked, rather than the one doing the asking.  But the second ASK doesn't take an extra input (so you click the left arrow to get rid of the slot).

ASK and KEEP are in the tools project or ToolSprite.

Disclaimer: I haven't actually tried this. 

P.S.  If I were doing this myself I'd use the grey-border feature instead of explicit THE BLOCK blocks, but people seem to find it more understandable this way. 

P.P.S.  As I look back on this, it's needlessly inefficient; you could just ask the parent for its ATTRIBUTE(CHILDREN) and make the TOUCHING? test on them.  But I don't have time to rewrite it just now.

P.P.P.S.  And you're right, there should be a FIND FIRST in the tools collection.  I'll do that for 4.0.

thanks for this brian!  Sorry I took a while to respond. In the end Mika worked around this problem using messaging before we had a chance to come back to the forum.

It's really interesting working with a smart 7th-grader on BYOB.  Last year I would have said that BYOB is too complicated for her -- or at least, that the advanced features mostly just fly over her head.  But after doing this project she already has an intuitive grasp, for instance, of what the gray border syntax means, and of when it makes sense to build a new block.  Some of the lambda-y features are still hard to grasp, but by exploring and copying the toolsprite scripts she's slowly learning more; and I'd say she's in most respects more comfortable than I am with the language now. 

So:  nice work guys!  I wish there was some way for me to tell the Scratch folks how much richer Mika's programming experience has been with BYOB!

Snap! New features : broadcast, wait, glide
@Jens Do you  intend to implement into Snap! the sophisticated ways of doing of Scratch as to keep (more or less) identical speed and delays, irrespectively of the processor speed ?



ps - Jens, sorry to unveil your new features !!

xly wrote:

ps - Jens, sorry to unveil your new features !!

it's wonderful that you're testing them!
Other new features are:

- LAUNCH (thread forking)
- error catching
- implicit parameters (empty slots in lambdas)

Enjoy!

I'm assuming you guys are working on some kind of array system, such as lists for snap? As this is a feature I use a lot in scratch.

xly wrote:

Snap! New features : broadcast, wait, glide
@Jens Do you  intend to implement into Snap! the sophisticated ways of doing of Scratch as to keep (more or less) identical speed and delays, irrespectively of the processor speed ?

http://www.xleroy.net/ByobTuto/New/broadcast.gif

ps - Jens, sorry to unveil your new features !!

Oooooh, cool new colors and drawing of blocks. Snap will be awesome. But I don't quite get it. Broadcast, wait and glide were in BYOB 1, 2 and 3. Are you recreating all of the blocks?

Jens wrote:

yup, I'm re-writing BYOB 3.1 completely in JavaScript.

Would you be able to give me a template code for me to edit? I really want to learn JavaScript more, and this would be really helpful!

@ Jens
See in the attached example that THE BLOCK input variables are inverted :

[CALL [THEBLOCK [ blank / blank]] WITH INPUTS 3 2 ]gives 0.666666 instead of 1.5