OpenXTalk Forums

So I've been working on an Enscripten/GitHub based Extension "Store". Besides delivering Extensions (Widgets/Libraries) my plan is that it will eventually replace the user sample stacks (revOnline stack).

I'm wrapping up some loose ends for release (such as removing bunches of back-up-type extra files). There's still lots of ideas and a few IDE annoyances that I wanted to fix but just couldn't get to (for example Palette/IDE Windows behavior when they're selected from the revMenuBar and its window is already opened but are minimized or in a state of occlusion, should be brought to front rather then closed). Of course there will always be more work needing to be done, but there needs to be a release so you have to stop somewhere in order to do that.

MacOS Ventura is out today, and there were reports of problems with using LC CE with early betas of it a few months back. So I will be installing Ventura (macOS 13) on my (Intel-based) iMac tonight so that I can see for myself how well (or not) OXT IDE runs on Ventura official release.

Of course there are new macOS things now (such as 'Stage Director' window manager) that ONLY function on the Apple Silicon hardware, so if someone feels like testing a CLEAN install of LC CE 9.6.3 on that hardware and report results here, that would be very helpful!

I'm not at home yet to start test on Ventura release, but I see nothing in the AppKit changes or other release notes that should prevent our xTalk Engine from functioning properly.
https://developer.apple.com/documentati ... r-macos-13

Stage Manager works perfectly well on my 2018 Intel Mac Mini.

But, as I don't like it, I have turned it off.

Oh, and . . . LC Community 9.6.3 works fine, and as nothing else seems forthcoming . . .

richmond62 wrote: ↑Tue Oct 25, 2022 6:12 pm Oh, and . . . LC Community 9.6.3 works fine, and as nothing else seems forthcoming . . .

Something else is coming forthcoming, I know, I know, I've been saying that for months but it is what it is. Maybe I'll hit the MegaMillions Powerball Lottery and then be able to quit my job and work on OXT full time, but I'm not holding my breath.

Thanks for reporting about 9.6.3 on Ventura. I did get Ventura installed on our 2019 Intel iMac but didn't have time to test OXT IDE. I had suspected that the problems with the early Ventura betas would've been early beta OS related problems. So that's great, no holdup there.

I know, I know, I've been saying that for months but it is what it is.

The problem is that everyone else, beyond the lone nut-case (me), has pushed off elsewhere,
and if this is to become the seed of a sustainable community I am not entirely sure how interest
is to be revived.

There is always the risk that operations like this:
- -
are going to seduce an awful lot of would-be supporters . . .

https://www.omnis.net/community-edition/

Mind you:

"Note that some features and class types are not available in the Community Edition, such as those relevant to developing desktop applications."

emphasis is mine

from my point of view, at least, that is utter bollocks.

As is this:

"Omnis Studio will run anywhere you like – so you can develop your application on one platform, Windows or macOS,
and deploy it to any other platform or device, including all desktops, tablets, and mobiles – while using the same set of
code for development, without any modifications."

as 'any' does NOT include Linux (except for Google's screwed up version: Android).

Why the fudge would Omnis push a FREE version that excludes desktop apps and includes
web-based stuff, rather than the other way round?

Wrote 'abusive' email to OMNIS:

"1. Why does the OMNIS Free version allow web-based and Mobile-based app development, and NOT desktop development, when, surely, the way to 'drag clients in' would be the other way round.

I am considering transitioning my highly complex desktop software from LiveCode to another development IDE, but am not going to pay for something until I have had an adequate chance to assess whether this is doable.

2. "any other platform", is, of course untrue, and does not help me much when about 250of my user-base is on dektops running Linux distros.Does OMNIS offer nothing in this respect?"

and got this reply:

"Thank you very much for your support request.

Regarding 1: The permanently free version of Omnis Studio (Community Edition) is the entry-level model of our products. Therefore, not all features of the paid version are represented here.
If you would like to test Omnis Studio to the full extent, we also offer this possibility. You can simply download a free trial version (https://www.omnis.net/developers/free-trial/).

Regarding 2: It is also possible to support Linux by developing a web application. This should then also be accessible from Linux without any problems."

I pointed out that their answer to my point 2 was disingenuous as I, at least, have end-users who have extremely dicky and intermittent internet
possibilities.

OBVIOUSLY, as my software (Both for ancient Indic and ancient Slavic writing systems) is FREE
(or, as they coyly say, "Donationware" = a few cents once in a blue moon), there is NO way, however 'sexy'
OMNIS may be for desktop development, that I can consider transitioning (quite apart from the mental gymnastics,
at 60, involved in getting my head round whatever OMNIS's underlying language is (which is OPAQUE to say the least)
to the standard that I have got after about 29 years with HyperCard, MetaCard, RunRev and so forth)
from "the programming 'thang' we know and love".

Just waiting for 'someone' to chime in that it is 'bad form' to publish personal emails in a public
space . . .

. . . tough cheese, boyo: as the contents of these emails redound on how we think about any sort of future
of LC/OXT/whatever.

How much the OMNIS people conceive of their FREE version as a truly FREE version, and how much they conceive of
it as a baited hook for commercial versions is another question that deserves a bit of rumination.

One more IDE bug fix is making it into the Extension management scripts today.

Extension Libraries can be written in xTalk script, there's no inherent need to use Extension Builder (really only if needed to use Foreign Function Interface for platform native stuff ) however it turns out that Script libraries did not show any included "samples" stacks in the Extension Manager list menus, although the "samples" folder was correctly included in the extension package (which is actually just a zip file with an .lce extension). This is now fixed. There have been a bunch of other fixes and changes to Extensions management in particular. For example previously demo "samples" stacks for Builder extensions (but not Script Extensions) did show in the lists menus, but it would not open the stacks when selected, that was fixed a while back. I've built entirely new 'Extension Store' that is served up from GitHub pages at zero financial costs to the community.

I need to make release notes.
This screenshot has the file tagged with lc extension (renamed while debugging) but it also correctly shows .oxtstack or .oxtscript files as well.

It would be helpful to have other contributors, only a handful of people have contributed a few things, or directly gave permission to use their xTalk work(s). I guess that also 'is what it is'. I'm just going to keep plugging along whenever I can.

richmond62 wrote: ↑Wed Oct 26, 2022 7:58 am There is always the risk that operations like this:
-
SShot 2022-10-26 at 10.56.46.png
-
are going to seduce an awful lot of would-be supporters . . .

https://www.omnis.net/community-edition/

Looks identical to other IDEs targeting business users that don't want to have to learn much to code some database using app.

I can't easily find what language they're mainly using, it seems to be JavaScript now, I know it was Pascal (which I'm kinda 'meh' on Pascal) in the past.

It has some sort of proprietary license.
If I loved Pascal I would be using Lazarus and FPC (which we can still use libraries from via Builder FFI), which I might point out only has a small community around it after some 20 years of work on the FOSS version and multiple corporations funding pascal development at various times in the past.

There's exactly one reason for me to not drift off myself to go get into using some other language (probably JS, Python, and/or Lua) with MUCH larger communities around them, and that reason is this 'they aren't xTalk'!