Stirring the soup

[tperry2x]

Wow: those vector graphics can be easily rotated: I wonder if that could be implemented in OXT
-
No jaggies on rotate!!!!!!!!

I think you'll find these images are simply over-sized and scaled back as default. When you enlarge them past a certain threshold, you'll probably get pixellation galore.

Movies are not vector graphics and always have a fixed resolution.

[tperry2x]

But as a simplistic no-code solution to the problem of creating something, it looks like a lot of work went into it at some point. Not to take away from that. It was very obviously produced at a time when everything was following Apple's aqua-scheme, and arguably things were presented in a clearer manner.

Some elements of this are great, so an interesting plaything at the very least.

[richmond62]

Some elements of this are great, so an interesting plaything at the very least.

Yup: it is a plaything insofar as one cannot get at the underlying programming language [which I believe you could in earlier recensions].

What I am interested in are any objects and/or functionality we could use in OXT.

Of course as:

1. HyperStudio is commercial software,

and

2. The underlying code is written in LOGO,

we CANNOT steal things/aspects of HyperStudio.

BUT, there is nothing stopping us writing xTalk code to imitate the functionality of what we like.

As, a large part of my stake over here is the result of my obsession to send SCRATCH and imitators off to 'Siberia', but also not drop
10-11 year old into the cauldron of Python, I am on the lookout for things that can make the LC/OXT interface slightly better for self-study by the Primary school brigade (and more inviting) than it is at the moment.

It was very obviously produced at a time when everything was following Apple's aqua-scheme, and arguably things were presented in a clearer manner.

That's a HEAVY sentence that deserves some unpacking and analysis.

It was very obviously produced at a time when everything was following Apple's aqua-scheme

1. Intend to try to get the Windows version up and running on WINE this weekend. . . watch this space.

It will be interesting to see, after the Mac version,how 'Windowsy' that is.

2. Frankly I feel that Apple have sunk since they gave up their Aqua-scheme.

and arguably things were presented in a clearer manner

I do think the Properties palettes knock the socks off those provided by LC 9.6.3. . . . and I was NOT convinced by their change whenever it was:
-

Props.jpg (304.99 KiB) Viewed 1847 times

-
From the left: LC 6.1, LC 8.1.10, HyperStudio 5.2.

HyperStudio's Properties palettes do not, obviously, owe anything to Mac Aqua.

[richmond62]

Way-back-when, my MSc thesis basically consisted of a decision tree for Primary teachers (21 year obsession), and, had the thing moved from a prototype to a full-blown GUI to strap on what was then called Runtime Revolution [now in a period of post-revolutionary stagnation], it would have been a more text-driven version of what HyperStudio 5.2 offers (and, frankly,my graphics were utter crap).

Now:

1. I certainly do NOT want OXT to become another HyperStudio because I want its users to learn how to code/program, which they cannot with HyperStudio.

2. I do want a better front-end to OXT [and I do NOT mean what LC are, supposedly,preparing for their final release of LC 10, as this will NOT make anything easier for learners], and aspects of the HyperStudio can be imitated/copied/leveraged to that end.

3. HyperStudio's 'HUGified' Mac version is charming, but I would far rather aim for a complete cross-platform consistency for the IDE.

It is pretty pointless if Primary sprogs learn how to use OXT on, say, a Linux distro and then find, on their Mac and Windows laptops at home the OXT interface is sufficiently different for them that it gets in the way of their homework.

These aesthetic considerations may seem a bit 'way off' when considering functionality, but they are 2 sides of the same coin.

[tperry2x]

These aesthetic considerations may seem a bit 'way off' when considering functionality, but they are 2 sides of the same coin.

It's all valid, as if we don't get the look-and-feel right, then this can switch people off straight away.
I've been fooling around with a revised "App Browser" / "Project Browser", and what's become apparent is how intrinsically linked it all is. One wrong move and nothing will load at startup.

Looking through the various stacks, it's also clear just how much legacy stuff there is in place, and comments throughout amounting to "...don't know why this happens..." Not a great sign.

I almost think it'd be worth stripping the syntax-parsing parts out, just so we have something that can interpret & run xtalk script, and start with a complete rewrite of the interface. As you mention, I don't mean with "Modern interface guidelines" as IMHO these suck, to put it bluntly - or whatever you could say Livecode or OXT currently is. A mishmash of UI ideas.

If I were starting afresh with an interface, I'd have options available for how it could be presented. Perhaps there would be a "Classic" mode where it looks a bit more like hyperstudio in some ways, a "Modern" version where people can indulge themselves all they like with widgets and sideways pop-out tool options, or even a "Minimalist / Distraction free" mode where it's more designed for a code focussed view.

But in giving people the choice of perhaps three different interface layouts, the underlying code that makes it all tick can remain largely the same. Just the appearance theme does the work, but behind the scenes, it's all handled the same.

Purely because if we are trying to get this to appeal to everyone, use cases will vary and they might not all want the same thing from it. A huge task, and not something that I have time or necessarily the skills to accomplish.
Coupled with the limitations of the engine (the one we have access to) not running on Arm, it's another hurdle. Not to say hurdles are insurmountable though.

[tperry2x]

It is pretty pointless if Primary sprogs learn how to use OXT on, say, a Linux distro and then find, on their Mac and Windows laptops at home the OXT interface is sufficiently different for them that it gets in the way of their homework.

True, but if you consider the current interface




The interface isn't that different between all three currently.
Yes, it might not have the initial welcoming appeal or handholding approach of hyperstudio, but I get the feeling that they were aiming for a different target market. Probably someone who wanted rapid results and had used a tool like realbasic in the past.

[richmond62]

Richmond's imitation game is in full swing:
-

Screen Shot 2023-10-11 at 17.33.12.png (482.44 KiB) Viewed 1818 times

-
Mind you: HyperStudio's Aqua buttons look far, far better than what LC 963 could offer.

I might fire up the G4 Mac Mini and take a look at what RevMedia4 can offer me in terms of buttons.
-

Screen Shot 2023-10-11 at 17.40.57.png (343.1 KiB) Viewed 1817 times

-
Ah 'Marginally' better.

[richmond62]

Back to messing around with HyperStudio in my lunch hour:
-

Screen Shot 2023-10-16 at 2.19.53 pm.png (81.83 KiB) Viewed 1798 times

-
If HyperStudio can do #3 [Web App, without any plug-in] how come 'some people' have been getting 'things' wrong repeatedly?

I find numbers 1, 3 and 4 extremely appealing.

One thing that does make me sad, is HyperStudio 5.2 is (apparently) unscriptable, one cannot export snapshots of buttons and so on:
-

Screen Shot 2023-10-16 at 2.26.00 pm.png (10.42 KiB) Viewed 1798 times

-
In fact the HyperStudio bunch remind me, in their use of the word 'future' of another group of people:
-

Screen Shot 2023-10-16 at 2.33.48 pm.png (24.34 KiB) Viewed 1797 times

-
Promises, promises, promises which are rarely kept.

[FourthWorld]

In fact the HyperStudio bunch remind me, in there use of the word 'future' of another group of people:
-
Screen Shot 2023-10-16 at 2.33.48 pm.png
-
Promises, promises, promises which are rarely kept.

It's very difficult to do both scripting and point-and-click programming well. It's hard enough to completely nail either one of those, but attempting both creates the round-trip problem: will the system be able to parse user code well enough to allow point-and-click alteration? Is the point-and-click complete enough to even support the possibility?

Back when SuperCard premiered I interviewed Bill Appleton for a local computer mag. His previous title was Course Builder, which had done rather well, and he was delightfully surprised to see it used to make games and other things far outside its envisioned scope of courseware. With all that going for it, why make a scripting tool like SuperCard?

"Expressiveness", he said. There's a point with point-and-click tools (these days referred to as though they're new - despite 30+years of VPL history - by using the fashionable "no-code" moniker) where the system you're making becomes hard to express with graphical objects. Even just navigation, he said, shows this, where a card-to-card flow diagram becomes an unreadable spider web. If the GUI is unreadably low value anyway, why not drop it altogether? The cognitive abstraction of systems design can only be concrete in very limited form; if we embrace it as inevitable in sophisticated works we might as well go whole hog and allow for the expressiveness, the flexibility, of scripting.

After all, he reminded me, Pinyin exists for the same reason, to overcome the limitations of trying to handle everything you want to express with symbols rather than phonetics.

So I can understand why HyperStudio never got around to adding a scripting language. They had a great tool, well done for the specific needs of its audience. Adding scripting adds far more than that one feature, and broadens the audience in ways that might initially seem attractive but which risks not excelling at completely capturing the needs of any given audience segment.

[richmond62]

The situation is quite the reverse: HyperStudio had a scripting language: LOGO, but between versions 4.5 and 5.0 it vanished.

In fact early HyperStudio was extremely similar to HyperCard except for 2 things:

The underlying scripting language was LOGO.

Because of LOGO all the turtle graphics stuff was there automatically.

Roger Wagner left HS after version 3.0 or so, and then it ended up at MacKiev, and somewhere down the line it went in what we might call the 'opposite direction' to Runtime Revolution. As HS 5.2 seems to have been developed at least 12 years ago it might be going the way of SuperCard insofar as, on Mac, it does not gave a 64-bit version.

To my mind its loss of its scripting language and its not being updated is a very real pity, even (as it is cross-platform between Mac and Windows) if only because a bit of competition might stimulate LC in some positive ways.

[richmond62]

My interest in HyperStudio is not, however, really about whether it is scriptable or not [tho' see earlier image which suggests the current (err, 2012) developers have been toying with the idea of making it scriptable again], but in its "educational outreach" in terms of some pretty groovy help palettes: palettes that might be good stuff as far as using an OXT version as a teaching tool at primary level.

[richmond62]

And so, to HyperStudio 4.5 running on MacOS 9 on 'Classic' on MacOS 'Siberian Tiger'.

My interest in this is to have a rethink [or, not really wishing to reinvent the wheel, see how HyperStudio 'did' the HC/MC/RR/LC thing differently], particularly in view of using OXT for teaching young learners the principles of computer programming without falling into something I personally regard as a mistake: the SCRATCH universe.
-

Picture 1.png (4.78 KiB) Viewed 1748 times

-
The Toolbar looks as cruddy and as antiquated as HC/MC, so no gold there.
-

Picture 2.png (20.61 KiB) Viewed 1748 times

-
So, now we get a load more hand-holding as we start to make a program.
-

Picture 3.png (26.78 KiB) Viewed 1748 times

-
The Properties palette for a button comes up AS one pops a button on a card.

This seems sensible for learners as it guides them through and shows them things they might not otherwise spot.

[richmond62]

Don't let the bitmapped jaggies tickle your nose, that is NOT the point.

Try, and believe me its bloody hard, to think of yourself as an 8,9,10 year old in a confusing world sitting down in front of an xTalk IDE for the first time.

A bright boy of 11 pointed out to me 15 minutes ago when I told him I was 61 that I was 50 when he was born: cripes, not even slightly funny.
-

Picture 4.png (31.38 KiB) Viewed 1748 times

-
What the @#$%^&* 'CLOWNINGWAY' is, I just don't know: assume it is some cutesy built-in font.

[Mind you that dog proves that Roger Wagner was ever so slightly into kitsch.]
-

Picture 5.png (23.29 KiB) Viewed 1748 times

-
Nice and easy to add icons to buttons.
-

Picture 6.png (16.48 KiB) Viewed 1748 times

[richmond62]

Obviously you can get at these properties palettes at later stage should you want to change things.
-

Picture 7.png (14.95 KiB) Viewed 1748 times

-
I haven't worked out yet whether on choosing an action it is included in the button's script or not.

Note the 'Use HyperLogo' option.
-

Picture 8.png (15.89 KiB) Viewed 1747 times

-

Picture 9.png (15.58 KiB) Viewed 1746 times

[richmond62]

Scripting is a little bit hidden [unlike HyperStudio 5,2 where it has GONE completely]:
-

Picture 15.png (24.68 KiB) Viewed 1745 times

-

Picture 16.png (70.13 KiB) Viewed 1745 times

-
A fairly bare-bones scriptEditor, BUT with a library of inbuilt, premade action scripts.

[richmond62]

I am not going to bother to examine HyperLogo because the focus here should be on the 'xTalk' language (and I don't like Logo very much).

I do feel that a "beginner's setting" with OXT very similar to what I have just seen with HyperStudio 4.5 (that can be turned OFF in the Preferences stack) would:

1. Make things easier for learners (at all ages).

2. Allow learners to be a lot less dependent on teachers.

3. Be a 'selling' point for OXT (especially as the 'other lot' provide nothing of that sort).

[richmond62]

Choosing an action is NOT included in the button's script, which is counter-productive and will NOT help learners with the underlying language:

But this is nice:
-

Picture 1.png (6.89 KiB) Viewed 1743 times

-
But this is extremely ugly, and using slang is a right turn off:
-

Picture 2.png (9.5 KiB) Viewed 1743 times

-
"Rog" gets his kitschy register wrong again.

Pandering to kids' "street talk" is a load of rubbish, as those children will despise adults' fakery and those adults will get it wrong anyway: the whole point, after all, of kids' street talk is to demarcate cultural territory off from adults.

[tperry2x]

Some good ideas, and I can see where you are going with the handholding approach for a younger users of OXT.
I do wonder though if creating a 'beginners' stack (better term needed), would be easier and could contain all the handholding required in getting them up and running. This then does not need to involve internal code changes to OXT, just a stack that can be added to OXT lite and the RC versions to build in tutorials of this sort?

[richmond62]

just a stack that can be added to OXT lite

That sounds a marvellous idea.

[richmond62]

We could, also, have a sort of secondary Tools stack attached to the IDE with all that guided handholding as substacks.

The Start Center can offer end-users a series of choices: Educational, Standard, and Advanced.