OpenXTalk Forums

I am cross-posting this:

https://forums.livecode.com/viewtopic.php?f=9&t=28059

Mainly because of my comment in my 'revival' posting:

"I am really writing this message to revive this thread as I feel that LiveCode [the company] has lost the plot
in terms of pushing its product [err . . . LiveCode] as a rapid prototyping tool."

OXT should be the go-to tool for prototyping . . . after all xTalk is perfect for prototyping
and with you-know-who pricing themselves out of a prototyping market . . .

Thw whole thread bears reading right through.

Any opinions on how LC stacks up to Figma, XD, Bubble, QuantUX, or Proto.io?

Not in my case as I have never used them.

However, one of the advantages about using xTalk for prototyping is that, if one wants, one can
convert that prototype into the finished product in xTalk itself as xTalk is not just a prototyping
tool.

Figma is usable in a collaborative setting: however, if that means developers can edit the same thing at the same time, or
merely access code and edit it when it suits them, is not clear.

https://www.figma.com/
- -
It would be perfectly possible [whether this has been tried or not, I don't know.] to keep xTalk stacks in some
shared space such as a DropBox account (GIT hub?) where they can be accessed by various people, and with
proper implementation of versioning, they can be collaborative.

richmond62 wrote: ↑Tue Mar 28, 2023 7:48 am Not in my case as I have never used them.

Business advice given to others will be more compelling where it includes effort to survey the range of solutions prospects are currently using.

Quite possibly, but as my business is involved in running and advertising an ESL school in Bulgaria, that comes fairly far down my list of priorities.

To which I could add that, when it comes to building apps I do not engage in that for business:

My Devawriter program is open source, and my other programming is involved in providing ancillary support
programs for my ESL school; something on which my business does not depend on.

richmond62 wrote: ↑Tue Mar 28, 2023 4:11 pm Quite possibly, but as my business is involved in running and advertising an ESL school in Bulgaria, that comes fairly far down my list of priorities.

Of course. What anyone does in their own environment is literally their own business.

Let that be what it is: your experience for your specific needs.

The constant droning about what others "should" be doing is as tiresome as it is uninformed.

I am not droning about 'should', but I am suggesting 'could'.

"Uninformed" . . . not completely . . . and certainly NOT uninformed in the field of CALL software.

"Uninformed" . . . Yes, almost completely in terms of the circles that you, FourthWorld, move in.

You maybe uninformed about CALL software and the market for that that is almost endless.

The FACT that CALL software can be developed relatively painlessly in xTalk by Language teachers who
do not have either the time or inclination to get into the niceties of 'real' languages such as Python or Java
is something the ONLY commercial contender in the xTalk field singularly fails to address.

Beyond CALL (Computer Assisted Language Learning) there is CAI (Computer Assisted Instruction) and so on,
in the Education field . . . untouched upon and unmentioned . . .

If my continual mention of this is 'droning', sobeit: maybe one fine day, before everything has vanished down the rabbit hole,
some sleepy bunch of people who might leverage this for profit, might wake up and take notice.

Certainly the late lamented Brahmanathaswami was well aware of xTalk's use for CAI, as that is what his programming consisted of:
I know as I did quite a lot of work on his Yamas and Niyamas project for him.
- - -
My own major production: Devawriter, is used just exactly for computer assisted learning of Indic writing systems in
a number of universities.
-

AND, to return to Prototyping . . .

I am now in the unenviable position that teenagers who are expected to dream up off-the-cuff apps in C# and Python
at high schools where the teachers know and mention NOTHING about any sort of prototyping whatsoever are coming
to me and taking up my lunch break asking me how to set up a semi-functional prototype in xTalk for them on the basis
of this sort of instruction from their teachers:

"Design and Program a Point of Sale app for a handheld device."

"Design and Program an app for a handheld device to tell English-speaking foreigners how to navigate themselves about a big city
in Bulgaria."

Apart from my personal feeling that those teachers are the ones who need to go back to school . . .

Given a few quick tips, those teenagers can generally have a 'half-cock' prototype up-and-running in xTalk in about 90 minutes:
saving them hours and hours in class and homework time: so they can show that to their teachers and their teachers can say "yes, a bit
like that."

Mind you 3 teachers in the last month have said something that roughly translates as, "Lor, Luv a Duck, how in fudge did you
manage that so quickly?"

Were Ministries of Education aware of this capacity of xTalk, and its benefits, they might be prepared to invest money
in that direction.

AND so, to wire frame software . . . my choice [perhaps inevitably] being Pencil, and open source, cross-platform this:

https://pencil.evolus.vn/

AND, Yes, you can do all sorts of things that are neither no better or no worse than you can do with a box of crayons and
some paper: the ONLY obvious advantage being that you don't have to squidge your picture under the scanner and your
co-worker can b*gger up your design with their version of Pencil.

SO? Why does Richmond think that xTalk is a better deal than this?

1. You can do ALL that Pencil has to offer on a stack.

[actually, as far as I can see, Pencil, only allows one to import SVG images . . . which is seriously limiting]

2. You can then move from a wireframe to some sort of functioning prototype.

3. AND, if the programming requirements fall inwith what xTalk is capable of, one can develop ones' functioning
prototype into a finished application.

richmond62 wrote: ↑Mon Mar 27, 2023 10:35 am I am cross-posting this:

https://forums.livecode.com/viewtopic.php?f=9&t=28059

Mainly because of my comment in my 'revival' posting:

"I am really writing this message to revive this thread as I feel that LiveCode [the company] has lost the plot
in terms of pushing its product [err . . . LiveCode] as a rapid prototyping tool."

OXT should be the go-to tool for prototyping . . . after all xTalk is perfect for prototyping
and with you-know-who pricing themselves out of a prototyping market . . .

Thw whole thread bears reading right through.

Coincidentally I'm now taking somewhat of a deeper dive into the graphics and layout features in the IDE, which I had occasionally noticed before are fairly broken, but now I'm actually trying to fix some of the problems, and additionally I want to add new export features.

For example in 1.963 currently the edit mode for polygons shapes does not turn on from the toggle menu, which never toggles. It does work fine if you 'set the editMode of graphic X to polygon' from the msg box.
I've written a method for converting Graphic controls shapes ('the effective points of...') to basic SVG path strings for use with various widgets, a method for SVGIcon to rasterized Image Data, and another method for exporting SVGPath strings back to valid .svg files you can open in InkWell, Illustrator, etc. for editing.

It seems to me that multimedia stuff among other things have long taken a back seat to Database stuff or Unicode text features or porting the engine to web, which of course are useful too, but maybe the reason people don't use these other tools in the IDE so much, which I imagine would useful in prototyping, are because they've been broken for a long time or the features to better facilitate easier use for prototyping (like export features) don't exist (yet). I imagine there may have been a concern about exporting would making it easier to switch a project to another Development environment or programming language, but maybe that's just my imagination. At any rate there is no such concern here, and in fact I think it would be awesome to add dual-directional transpiling capabilities.

Here's some current IDE to-do items:
Reshape graphic toggling needs fix
Rotates menu item only works on graphics controls, with but with image controls it is broken (displays menu and dialog but then the image doesn't actually get rotated).
'Rotate' and other graphics items don't show in the context menu,
The IDE contextual menu in general has limited usefulness in its current state.
Add 'copy points to SVGPaths' IDE and contextual menu item
Add Export>... to main menu for exporting content to various other formats (like SVG Icon to SVG Files, HTMLText to HTML, Export All Images, Export All Sounds, etc. etc.)

That's super, although I would beg you, on bended electronic knee, to release builds for W, M & L just as things are at present.

Oh, and while I am here: I am trying to lay my sweaty paws on ALL the SVGs in LibreOffice Draw to import them into a palette: as they are Open Source they should be up for grabs, but I am far too lazy to export each one individually from an OO Draw document.

Have you any pointers?

richmond62 wrote: ↑Wed Mar 29, 2023 4:14 pm That's super, although I would beg you, on bended electronic knee, to release builds for W, M & L just as things are at present.

Oh, and while I am here: I am trying to lay my sweaty paws on ALL the SVGs in LibreOffice Draw to import them into a palette: as they are Open Source they should be up for grabs, but I am far too lazy to export each one individually from an OO Draw document.

Have you any pointers?

Not sure what you mean since I haven't used Libre/OpenOffice in a long time. Is this a clip art collection that is included? If the art is in SVG format then that can fairly easily be imported and used with the SVGDrawing library. SVG is basically a markup for the ancient LOGO drawing commands, and in the SVGpath strings the letters jive up: M = moveto, L = lineto, C = curveto, etc., its not that hard to parse the SVG XML. I believe the OpenOffice odoc format is XML based as well.

As for releasing new packages, I've just been too busy, working on related things whenever I can. I'm still finding things that need fixing, I just found some 'branding' in the script editor templates sub stack that I'd missed for example.

I gave a quick search in the LibreOffice GitHub repos for SVG and didn't turn up a folder called 'shapes' with SVG data in it, the first pages of hits were all used in the Help Docs. But if you can figure out where those 'shapes' are coming from as far as source SVG files, I have some xTalk scripts that can parse SVG in bulk that could be used at least as a starting point. Some of those shapes look like Material Icons, which I already made an LCB library to load those (and many other SVG symbol 'webfonts') into IconSVG 'fonts' (arrays actually) for use with the various widgets that support that IDE library.

richmond62 wrote: ↑Wed Mar 29, 2023 10:01 am "Design and Program a Point of Sale app for a handheld device."

That reads like an purposefully vague prompt that I would use to test an AI like ChatGPT with. Which I think is something that xTalk may be uniquely well suited for using that with, because xTalk script is already rather 'natural language' oriented and can practically read like pseudo-code / English already (particularly if you're good at naming variables so that they fit int with that idea).

I'm interested in that open source GUI prototyping tool you posted. I'm going to install that and LibreOffice tonight if I can stay awake long enough to.

LibreOffice has some C++ and JS libraries that could be useful for our things here, down the road. Like for adding support for Import/Export of MS.DOC DOCX '.PUB(lisher), SVG, .ODOC, and various other formats.

"shared space such as a DropBox account (GIT hub?) where they can be accessed by various people, and with
proper implementation of versioning, they can be collaborative."

Sure people can use cloud storage or GitHub to collaborate on Stacks, but not so much collaborating simultaneously / realtime/live or whiteboard style (which really was the original, circa 1960s, idea for this 'Hypertext' network, wasn't it?).

GitHub has some nice things for collaborating, tracking ideas, bugs/issues, etc., building documentation, etc.

I think the macOS version of Pencil is pretty bad, laggy, and crashed on me.

I have access to Adobe XD, I'll have to give it a whirl sometime.

OpenXTalkPaul wrote: ↑Thu Mar 30, 2023 2:37 am GitHub has some nice things for collaborating, tracking ideas, bugs/issues, etc., building documentation, etc.

Yes, Github is collaborative, but it's over-complicated and to me at least, is a complete mystery.
That's why I'm here using OpenXTalk, and not XCode / VisualStudio or Bluefish.

Just a link to download, unzip and make executable and run. That's what we need. As far as shared projects go, this can be a menu inside OpenXTalk that shows community projects. It could read from github or whereever, but fetch from github and present it in an OpenXTalk way perhaps?

I don't see what's wrong with keeping it in the tutorials / sample stacks section.
(apart from it being slow to load)

tperry2x wrote: ↑Fri Mar 31, 2023 5:11 pm

OpenXTalkPaul wrote: ↑Thu Mar 30, 2023 2:37 am GitHub has some nice things for collaborating, tracking ideas, bugs/issues, etc., building documentation, etc.

Yes, Github is collaborative, but it's over-complicated and to me at least, is a complete mystery.
That's why I'm here using OpenXTalk, and not XCode / VisualStudio or Bluefish.

Just a link to download, unzip and make executable and run. That's what we need. As far as shared projects go, this can be a menu inside OpenXTalk that shows community projects. It could read from github or whereever, but fetch from github and present it in an OpenXTalk way perhaps?

I don't see what's wrong with keeping it in the tutorials / sample stacks section.
(apart from it being slow to load)

Fetch from GitHub / Present in OpenXTalk IDE is basically what I've done with the Extension Manager "Store", and I made that with easy to maintain set-up files with simple tab-separated values file, collected by and managed from a stack with a DataGrid, then if there's peripheral files in the same directory with same root name, but with a resource extension such as .svg (icon/logo) or .png / .jpg (screenshots) then those are displayed in the browsing client stack. The client stack is currently also hosted on GitHub as a HTML5/WebStack, running in Browser Widget in the IDE (similar to the Dictionary stack), but maybe a copy should be installed locally with the IDE for speed.

For sharing community stacks I think we could keep that in a separate sample stacks area like it is now, for submitting stacks for that people could do either pull request on GitHub or upload to a section here for peer review and adding to the collection.

If you use the GitHub Desktop app, then GitHub can sort of just be treated as if it were cloud storage, that syncs your project folder on you local drive to a copy on your GitHub repos, if its an image or a text based file (script only stack) then you can compare differences before you sync, if it's a binary file ('GUI' regular stack) or an otherwise not supported by GitHub format then you don't get to compare first but it will still sync.
The difference from personal cloud storage is that other people can maintain copies of the project folder too, modify files and then ask to merge their changes to the main project.
But there's other benefits, like hosting project related web pages, wiki, issues, bug reporting / feature requests, etc. which can mostly all be done from Github Desktop or from the GitHub web UI.