Scratch

bharvey · 2012-10-04 11:49:58

xly wrote:
This list-of-list can be modified modify and then reloaded with DEFINE, and you get a new block with the same name

Yes. You have to DEFINE again. I thought maybe you were suggesting that a change to the list would automatically be reflected in the procedure.

But in any case, all Logo procedures are global, and they're dynamically scoped, so they're not a precise analog to Snap! scripts.

bharvey · 2012-10-04 11:52:31

Hardmath123 wrote:
How about a list-of-list structure of lambdas?

That's pretty clearly going to be the external representation of the text of a script. Alas, the cost of having a visual language is that it's not the internal representation, so we have to invent conversion functions to get self-reflective code.

Hardmath123 · 2012-10-04 12:15:12

bharvey wrote:
Hardmath123 wrote:
How about a list-of-list structure of lambdas?
That's pretty clearly going to be the external representation of the text of a script. Alas, the cost of having a visual language is that it's not the internal representation, so we have to invent conversion functions to get self-reflective code.

Is it just me, or are your posts steady increasing in concentration of words I don't grok? tongue I'm gonna puzzle over that one during the bio test tomorrow.

bharvey · 2012-10-04 12:18:11

Hardmath123 wrote:
words I don't grok?

Sorry. "Self-reflective" means that there is a means in the language for manipulating programs in the language, e.g., inspecting the instructions in a script.

As for dynamic and lexical scope, search for them in the earlier messages in this thread. smile

joefarebrother · 2012-10-04 13:08:59

Hardmath123 wrote:
Scripts aren't mutable (yet), right?

Well you can't mutate the script itself, but a script can mutate it's own lexical environment.

xly · 2012-10-04 13:31:22

@bharvey
"But in any case, all Logo procedures are global, and they're dynamically scoped, so they're not a precise analog to Snap! scripts."

My contribution was certainly "off-topic". Incidentally my layman point-of-view. is that the form of a Logo procedure defined as a list-of-list has some similarities with a Java definition of a Class : name + variables list + constructor.

Hardmath123 · 2012-10-04 21:01:39

bharvey wrote:
Hardmath123 wrote:
words I don't grok?
Sorry. "Self-reflective" means that there is a means in the language for manipulating programs in the language, e.g., inspecting the instructions in a script.

As for dynamic and lexical scope, search for them in the earlier messages in this thread.

No problem, and thanks. smile

nXIII · 2012-10-04 22:06:12

bharvey wrote:
Hardmath123 wrote:
How about a list-of-list structure of lambdas?
That's pretty clearly going to be the external representation of the text of a script. Alas, the cost of having a visual language is that it's not the internal representation, so we have to invent conversion functions to get self-reflective code.

What about making ITEM OF polymorphic? ITEM n OF a stack would return the nth block in that stack and ITEM n OF a block would return its n-1th argument for n > 1 and the block's prototype for n = 1.

bharvey · 2012-10-05 01:42:51

nXIII wrote:
What about making ITEM OF polymorphic?

Ah, I see, so you don't have to do an explicit conversion to list form. Interesting. I want a polymorphic ITEM OF anyway, for bitmaps.

joefarebrother · 2012-10-08 16:14:40

bharvey wrote:
nXIII wrote:
What about making ITEM OF polymorphic?
Ah, I see, so you don't have to do an explicit conversion to list form. Interesting. I want a polymorphic ITEM OF anyway, for bitmaps.

And if ITEM OF was also polymorphic to deal with strings so it is a synonym for LETTER OF then that would be even better.

nXIII · 2012-10-08 16:47:35

joefarebrother wrote:
bharvey wrote:
nXIII wrote:
What about making ITEM OF polymorphic?
Ah, I see, so you don't have to do an explicit conversion to list form. Interesting. I want a polymorphic ITEM OF anyway, for bitmaps.
And if ITEM OF was also polymorphic to deal with strings so it is a synonym for LETTER OF then that would be even better.

Alternatively, it could treat strings like sentences.

funelephant · 2012-10-08 17:23:43

How hard was it to create the online player?

bharvey · 2012-10-08 21:54:42

nXIII wrote:
Alternatively, it could treat strings like sentences.

To me this is a reason not to have ITEM OF work on strings, but instead to provide both LETTER OF and WORD OF reporters. ITEM OF is ambiguous for a string.

bharvey · 2012-10-09 00:24:52

funelephant wrote:
How hard was it to create the online player?

Are you asking about Scratch, BYOB, or Snap!?

Hardmath123 · 2012-10-09 02:37:04

bharvey wrote:
funelephant wrote:
How hard was it to create the online player?
Are you asking about Scratch, BYOB, or Snap!?

Or JSScratch or sb2.js?

Caviac12 · 2012-10-10 17:30:59

I have a very quick, stupid question, but I haven't been able to find an answer anywhere and couldn't find a way just by searching the program. How do I play my game outside of the program?

bharvey · 2012-10-10 18:25:14

Caviac12 wrote:
I have a very quick, stupid question, but I haven't been able to find an answer anywhere and couldn't find a way just by searching the program. How do I play my game outside of the program?

In BYOB, you mean? There's a "Compile this project" option in the Share menu that will make a standalone executable. (It actually still contains all of BYOB, but when run it starts with your project loaded in presentation mode.)

joefarebrother · 2012-10-11 02:13:50

bharvey wrote:
Caviac12 wrote:
I have a very quick, stupid question, but I haven't been able to find an answer anywhere and couldn't find a way just by searching the program. How do I play my game outside of the program?
In BYOB, you mean? There's a "Compile this project" option in the Share menu that will make a standalone executable. (It actually still contains all of BYOB, but when run it starts with your project loaded in presentation mode.)

It doesn't see, to work for me.

jji7skyline · 2012-10-11 02:16:29

I made a page on the Scratch Wiki a while ago. The best way to do it is to use ToJar.

bharvey · 2012-10-11 22:07:04

joefarebrother wrote:
It doesn't see, to work for me.

What OS? And how does it fail?

joefarebrother · 2012-10-12 15:04:27

bharvey wrote:
joefarebrother wrote:
It doesn't see, to work for me.
What OS? And how does it fail?

Windows 7.

It doesn't seem to do anything, it just opens 2 command line consoles for a few seconds, then closes them, then opens my 'My Documents' folder in windows explorer. I can't find any new files that might be the compiled copy.

Last edited by joefarebrother (2012-10-15 15:52:09)

Jens · 2012-10-15 14:52:34

Folks, I've just completed the full German localization for Snap! (along with translated drop-down options and other previously missing strings). Except for the About-Dialog everything should now be localized and still working. While I'm writing up instructions on how to contribute translations for LTR languages I'd appreciate some rigorous testing smile

Thanks!

TRocket · 2012-10-15 15:00:02

Jens wrote:
Folks, I've just completed the full German localization for Snap! (along with translated drop-down options and other previously missing strings). Except for the About-Dialog everything should now be localized and still working. While I'm writing up instructions on how to contribute translations for LTR languages I'd appreciate some rigorous testing

Thanks!

when deleting a custom block the dialogue text is in English...

nXIII · 2012-10-15 15:47:28

A number of unrelated bug reports:

—

Preferences: The "Clicking sound" preference shows the state of the "Input sliders" preference.

—

UX: Clicking past the end of a line in an editable TextMorph should place the insertion point at the end of that line, not the beginning of the next one; sometimes it is even a few characters into the line:

—

UX: Scrollbars and buttons should regain their "active" state rather than their "hover" state when:
1) the mouse is pressed on a button
2) the mouse leaves the button's bounds
3) the mouse reenters the button's bounds

—

UX: Dialogs should not be draggable by their buttons.

—

UX: Input slots should be deselected when they are blurred:

—

Evaluator: When errors are thrown, threads do not stop and use up the full atomic execution time every frame.

—

Editor: Dragging a reporter block onto a ring to replace it deletes the ring and does not put it back in the editor.

—

UX: Dialog shadows should not be identical to drag-and-drop shadows.

—

Blocks: Pop down inputs are not correctly autosized in the palette:

—

UX: The palette should reset to motion when a new project is created.

—

Blocks: The boolean ring should either:
- not allow drops of non-boolean reporters; or
- reshape when a non-boolean reporter is dropped on it

—

Evaluator: This makes no sense:

—

Rendering: Empty speech bubbles' pointers go off the edge of their bubbles:

—

Editor: This should not be possible:

—

Evaluator: Evaluating this lags Snap a huge amount:

It should either throw an error or give an empty string.

—

Evaluator: I disagree:

—

UX: Shift+arrow keys should not alter the anchor point in text editors.

—

Evaluator:

NaN doesn't appear anywhere else, and:

—

Evaluator: We are not JavaScript.

—

Blocks:

—

Blocks: Error-outlined blocks have weird shadows:

—

Blocks:

(Create a new block with spec "test" -> click label fragment -> click "Delete" -> click "OK")

—

Blocks: The default variable names in SCRIPT VARIABLES are not unique (i.e. they are "a" through "z" followed by an infinite stream of "foo"s):

—

UX: The bottom-right pixel of checkboxes should not be really light:

—

Evaluator: This isn't necessarily a bug, but it's unintuitive:

changes "a" regardless of the order of the LAUNCH blocks.

—

Evaluator: This can hide bugs:

—