uliwitness wrote: ↑Fri May 02, 2025 11:53 am
Takes some getting used to.
Indeed it does.
I appreciate your thoughts, as always, and would not dispute matters of taste, especially where our preferences are the same.
There was a time when I held out hope that the xTalk Way would lead the charge in 21st century app development. But I see the world's preferences are not my own, and my question relates to why that may be.
Syntax aside, the two elements of the xTalk Way you mentioned are design choices, with a large and growing body of examples in the world outside xTalks.
Auto save has been around for a while, these days nearly ubiquitous on macOS as a HIG mandate. xTalk engines make it easier to do, but in ways that seem to me tradeoffs, as all things are. Separating data from the UI in which it's displayed became important to me way back when I as building HyperCard stacks for clients; Adding features to the UI meant having to impose an export/import scheme on the user experience with each version. I believe the modern separation of data, view, and the code that coordinates them has evolved to become the norm for useful reasons.
And as Paul noted, the web engine has had native packagers available for some time. Though the number of native desktop apps made with the web engine is relatively small, today it far surpasses apps delivered with xTalk engines. And on mobile it's so common to use the web engine packaged in native standalone format that we see it used by a very large number of the leading apps, and a good many besides. React Native, Electron, and others are still not as easy to use as xTalk standalone builders, but that's largely a UX priority for dev tools makers rather than an inherent limitation of the runtime engine.
--
For myself, this line of inquiry is part of a personal process for establishing workflows and methods for new works going forward.
I began my career in a world where we were given a powerful, scriptable rich media engine pre-installed on every GUI device (tho such devices in mainstream '87 were limited to Macs).
Today we have a world where we're given a powerful, scriptable rich media engine pre-installed on every GUI device (and such devices are plentiful, in a wide range of form factors).
While it eventually became useful for HC to make standalones, one of the great things about HC was the ease of sharing apps: they were small files, which could be shared easily with the confidence that the user already had the large engine needed to run them.
Today we have an Internet which makes delivery even simpler: we don't have to send the app at all, just a single line of text, the URL, describing its location.
MetaCard made Internet stack delivery trivial, but despite my best efforts to help popularize it as a general delivery option, it never really caught on beyond being an efficient way for standalones to update their components.
We never saw a general purpose stack browser with any popularity, despite some good efforts from Ken Ray and others in that direction.
--
That Internet component is perhaps defining of our modern world.
At the beginning of my computing journey, single-user single-computer apps were the norm, as they would remain until the Internet's privatization in '96.
Now nearly three decades out, I can hardly imagine trying to get thru the day if I had to manually sync my data across a phone, a tablet, a laptop, and three desktop computers. The prospect would be more daunting considering each of these runs a different OS.
For better and sometimes for worse, our world is now an interconnected flow across people and machines, with compute and data resources available and interoperable nearly everywhere except when I'm backpacking in remote regions of the Mojave beyond cell range.
This "flowy" quality of how we use data today has reshaped expectations beyond the silos of yesteryear.
We can still make single-user single-computer apps as needed (I've been considering one for filling out archeology site record forms), but as the Internet's availability has grown such use cases are ever fewer.
--
This has become quite the ramble; please forgive the length. I just have some apps to make, and while I see how the world has evolved, I still find a voice from my younger days poking and prodding me to try to justify doing new work in the paradigm I know best.
For legacy apps, keeping them with the engine they're written for makes business sense.
But for new works I wander the desert, tasting the waters at each oasis I come across, observing that these are the places where flora and fauna still thrive, so different from the rocky barrens where I record archeological artifacts.
