Scratch

What if it isn't a script variable? The same thing occurs with global and local sprite variables...

2.995.005 more bugfixes.

Jens wrote:

this release fixes the overlapping list watcher bug reported by shadow (thanks, your're awesome!), and features a huge speed-up in nested lists (don't get all too excited, because it doesn't affect ttt). I'm also attaching a BYOBbed version of the Scratch sample project MouseRecorder, which demonstrates (among other things):

   - custom blocks
   - lists in lists
   - making a sprite go to coordinates
   - first class procedures
   - nested sprites

BTW - nice sample project! Though definitely a bit slow for recording 

Why are you guys in such a rush to make the final version? For the Scratch Conference? I think you guys should take your time to make it as shiny as possible :3

Hey Lucario621, did you try 2.995.005? It's been made real fast for recording nested lists...

bharvey wrote:

There will be some form of procedure definition in 2.0, they say, but, for example, one idea they had tossed around was to limit it to commands with no inputs, to keep it simple.

I'm referring back to your contact with the "Scratch Team".
I'm just to point out that the Logo language offers - since 30 years at least - the same "sophisticated features" as offered by Byob : custom-designed functions, block editor, list of lists, and even some more like building a function inside a function. ( DEFINE and TEXT primitives if I remember correctly). These have not prevented to use Logo at all ages and educational levels without problem. Hundreds of tutorial, samples ,books have been written to teach and use Logo. Scratch should at least implement the equivalent of these largely used and known features.

For example MAP and ASSOC applications can be easily implemented by using some couples of simple primitives like FIRST, ALLBUTFIRST etc. These MAP/ASSOC primitives (or blocks) are very useful to create educational tools like quizz, dictionaries, sentence building, grammar exersizes etc

I understand that the Team is not prepared to adopt all the Byob concepts, but they should consider , at least, BYOB features already used in Logo language.

nXIII wrote:

If I finish it in time, do you want my (true) scripting language parser/evaluator?

As a sample project, you mean?  Sounds great!  Yes!

xly wrote:

the Logo language offers - since 30 years at least - the same "sophisticated features" as offered by Byob... These have not prevented to use Logo at all ages and educational levels without problem... Scratch should at least implement the equivalent of these largely used and known features.

First of all, I want to emphasize that the Scratch Team are the same MIT team that developed Logo, or at least the intellectual children of the Logo developers.  Mitch Resnick is a Logo pretty-old-timer.  So it's not that they're unaware of this history!

I really don't want to be the one to talk about why the S.T. are resistant (but not antagonistic!) to BYOB; when I try, I don't always get it right and then Mitch gets annoyed at me.     But I think it's more like "we want to move slowly" than "we think this is too hard."

When I was working at Atari a million years ago I heard a talk by one of their star game designers who pointed out that if the video games of (then) today had been introduced back at the beginning, kids would have found them impossible to understand, with bazillions of unfamiliar details and very fast action, but now, even kids new to video gaming can play them, because a set of conventions about how video games work has become part of the general culture, and so enough kids are around who already get it to teach the newcomers.

In the same spirit, I can see some merit in the idea that the ideas in BYOB, introduced all at once, might be too foreign to the culture of what kids already know and expect.  We have some counterexamples hanging out in this forum, but how many?  Two dozen?  Out of half a million Scratchers.  I believe that if these ideas turn up in Scratch, more than two dozen will get it right away, and most of the active users, the ones who upload projects, or at least the ones who upload serious projects, will get it after a while.  But I believe this just out of general optimism, not based on any real field testing -- what little testing I've done indicates that almost everyone needs a fair amount of teaching to help them.

Now, you're basically right that it was all in Logo.  (Not quite right, because Logo is dynamically scoped and has first class expressions rather than first class procedures -- you had to pass around procedure names as data.  Al Cuoco, the world's best math teacher, used to bug me all the time about how he needed lexical scope.)

But if you look at that history more carefully, every Logo kid got simple custom commands with no inputs, or maybe one numeric input; a much smaller number wrote operations at all, let alone recursive operations, let alone higher order operations.  So you can imagine someone back then saying "we could simplify Logo by leaving out these capabilities that hardly anyone uses."  In fact you don't have to imagine it!  You may recall Radio Shack's TRS-80 "Logo," or Turtle Tracks, or a bunch of other alleged Logo equivalents that weren't.  Here in the UK (that's where I am right now) there were big fights because the government picked one of those lesser turtle graphics programs as the standard for use in schools.  That was sad, but I don't think we can point to any educational disasters that are really rooted in that particular choice (rather than in the way schools are altogether, but don't get me started).

So, I don't know.  Maybe in the long run it would be better if Scratch 2.0 dipped its toes into first-class data (let's say with first-class lists but not procedures), and then Scratch 2.1 or 2.2 or 3.0 went the rest of the way.  I don't think so, but I'm not sure I'm right.  It certainly would be a pain in the neck for those of us teaching college students, but that's not the S.T.'s job.  We can muddle along with a separate BYOB for a while, if we have to.  (We certainly have to right now, because Scratch 2.0 is not imminent!)

2.995.006

Jens wrote:

this release fixes a minor (but annoying) bug that would truncate leading zeros in temporary variables (thanks, Stefano!). It also contains a small fix regarding the STOP BLOCK block, although there appear to be some cases in which something wrong still seems to be happening.

There are two small new features:

- an additional option in the EDIT menu offering "Scope Contrast", so you can tell late bound variable setters from early bound ones
- a little Windows screensaver that will be installed by the Setup program in the final distro

Note: The Screensaver only works if you have at some point installed BYOB via the Windows setup program (because it requires registry entries made by the installer).

To test the screensaver

    1. copy the .scr file into your Windows\system32 directory
    2. add a folder named "Projects" (without quotation marks) to your BYOB installation folder
    3. add one or more . ypr or .sb files to the Projects directory
    4. fire up your Windows screen settings and choose the BYOB screensaver

the Screensaver starts BYOB in presentation mode with a random project from the Projects folder. If you want to always run the same project (your own special screensaver) be sure to just put that one into the Projects folder (you can put other projects into subfolders, in this case they won't be used as screensavers). Like a "compiled" standalone project the screensaver exits back to Windows once the user presses Esc or exits from presentation mode. The Screensaver also makes sures that only a single instance of itself always runs.

Enjoy!

Wow cool! A screensaver option! Awesome! 

bharvey wrote:

I think deadlines are good for a project.  If we waited until it's perfect, we'd never get it out the door.  There are a million little things I'd like to change (e.g., some dialogs have RETURN key = click OK, but others don't) and half a million big ones (first class sprites!).

But, there's always 3.1.   

I think this version looks very polished! I haven't found any bugs at all (of course I haven't really been looking for them    )

Hey, there's still time to add psuedo-first-class sprites. All we need is that (my name) reporter!

bharvey wrote:

ScratchReallyROCKS wrote:

I've pretty much hit the limits of Scratch - hard. I'm trying to do things that it just can't do.

but

if you make Scratch too complicated, you will lose your original and intended audience.

That's the dilemma.  I don't know if you intended it this way, but you present arguments on both sides.

I think the BYOB blocks themselves are not a problem.  What worries the S.T., maybe correctly, is that the collection of online projects will use custom blocks a lot, if they're in the language, and so the projects will become hard to understand.

OTOH, I find the existing Scratch projects, the interesting ones, hard to understand already!  They have scripts that go on for miles.  And they're structured that way often for lack of custom blocks, which would allow huge simplifications.

Yes, I did intend it that way. Also, like you said, I don't think the custom blocks are the problem, but the lists-in-lists and blocks-in-lists and 'script variables (a) >' may be harder to understand.

I still think BYOB should be implemented in Scratch 2.0 though, I was just stating why it might not be.

bharvey wrote:

Jens wrote:

Sorry for spamming you with frequent releases. I'm considering this release to be the "final" beta. Now I'll start going through all the other files once again and building the full installer.

Yeah, like that'll happen