So they had to delay Xojo 2025r3 because there was so much work needed to make it compatible with macOS 26, but somehow had time to create this?
Priorities, ya know? Gotta add the shiny buzz word of the day before bug fixes. It really is an indictment of the Rapid Release Model. Instead of big new features once a year and then a bunch of smaller bug fix releases we get new, and mostly untested, features every release. Sure some bugs get fixed every release but it sure seems like new ones are created at about the same rate.
To be honest, I’ve already done quite a bit with the AI APIs. If anyone needed more than 30 minutes for that, I sincerely hope they’re either an intern or a marketing manager.
Why as a developer I have to thoroughly test clients code before committing to “updating” (I wont call it upgrading) to a new version
Some use one version for some projects and different ones for different projects because while the version might fix bug X Y and Z it introduces new ones in a different place (just found a truly brilliant one in the version we’re using for Windows builds. Shells on Windows that do NOT end when the main app that started them is quit … even when explicitly closed & nil’d. YAY !)
And that to me is an indictment of a lack of rigorous mechanisms for finding, fixing, and testing to insure bugs stay fixed
I know how we used to do that but that seems to have fallen by the way side
Again
You used the word Rapid in combination with Xojo? The only thing I’ve ever experienced with Xojo that could even remotely be called fast was the charge on my credit card. Literally everything else - without a single exception - is running on the slowest suite imaginable, and then slowed down by another factor of ten.
Okay, they’ve now added another feature to the beta process that — at least apparently — works well and actually counts as a “major” feature. It also doesn’t seem to be dramatically buggy or implemented in the typical “Xojo way” of being unusable. So this could actually be a genuinely useful new feature that does what people expected it to do.
That would indeed be a really positive development.
Even though, of course, they’re still hopelessly lagging behind in everything else. Well, actually, they usually only start running once all the other participants have already crossed the finish line and the award ceremony is over.
I don’t know what this feature is or whether you can presently discuss it, but if it really works well and is stable, then I’d guess it didn’t come from Xojo. Maybe one of the plugin developers? That’s the only way I could see Xojo publishing a new feature that works well in the initial release.
This new feature has a lot of potential, i think. It is not really a secret, i guess. The roadmap should give a clue. Geoff demoed some things of the roadmap at the last MBS Xojo conference in Germany, Christian does have video about this demo on his site.
Let me say it so: you can dream that this will be happened but I guess it will be more like Xojo Android. At the end it will be so like it was always with them. I have no Idea why but they act always in this wise ![]()
What ever they try to do in case of AI: this train left before a few years. And XoJo is not a language which will bring the success you’ll need. It is sad that they have lost so many chances they had to get to a much better performance in the market. One they lost when using Kotlin but instead of using Kotlin Multiplatform mobile for both, iOS and Android they developed it for Android. Makes two projects from it. Even when and if they start to use one project base: they will internally be stacked to this double sided solution. But what do I know about it?
Which video ?
The roadmap lists a pile of things
And the most interesting one is libraries
But, from what I’ve seen, + the long list of feature requests & public bug reports already lead me to expect the result may be underwhelming
And the other top 2 are supposedly in the works and again, as usual, of questionable utility & quality at this time - or so I’ve been lead to believe
LIke this (and this is JUST public cases and only the first screen of them - there are more)
I think the expectation is that Libraries will be “like plugins” and somehow adopt capabilities & compilation options that give them better access to the OS and speed like optimized C+ compilers give to plugins
Neither of which is likely to be true
At best they are spitting out LLVM bit code that then gets compiled & linked into your project
Its still Xojo code at the end of the day
So any access to the OS is via declares - not direct API access like in a C++ plugin etc
And somethings will be devilishly hard to make into a Xojo library - things that require asynchronous callbacks - that are already damned hard if not impossible to do well in Xojo
There’s no magic here
Its a bit better than encrypted code, maybe a bit more secure, but its is, when all is said and done, Xojo code
I’m often wondering why we can’t publicly discuss stuffs added in beta, while the reports in the tracker site are actually public and, well, everyone with a standard account can see them being discussed there. ![]()
Because being in testers channel is something special where people shall feel “VIP” cause they are allowed to test the newest Ideas of the development. That’s in most cases while the Leaders are organizing the entire stuff so crazy that at the end there is always an Alpha status (Android, iOS) or a beta status (Web and Desktop). Nothing in clear production ready status. And while people are believing that they are something special they need to renew all times cause: wioth renewing they get to the point that they are still a member of this club.
To be honest, I think the reason only Pro users have access is simply because the Xojo team doesn’t have the time to test anything themselves — and in the Xojo world, it’s completely normal that users work for Xojo for free. I never saw myself as a VIP, but rather as an exploited customer whose work no one appreciates — or even acknowledges.
Sorry, I wasn’t clear. I know why testers aren’t allowed to talk outside of the testers channel.
What I meant is, it’s worthless to disallow this in the forum if the same information can be found in the issues.
Being a member of Testers is a “special benefit” of being a Pro license holder
So they restrict discussion to those types of members & that forum
Never made sense to me either even when I worked there
Then when concerns are raised in the Testers forum, the response is that they don’t represent typical users.
Oh yes I recall THAT as well
They are not interested in what the users say. If testers or not. When I was member of testers forum I realized that this “testing” is worthless while it works always as a testing of an ALPHA state software which is never really working like it should or like it was to expect when reading their promises. None of this packages runs like it should. None. And testing is not helping. You can test, send them the bugs you’ll find. After: nothing happens. Mostly there are only the showstoppers which they take care. After the app stops crashing they start deployment. So for what testing? Beside the fact that I am - like a few others - not in testers forum following the fact that I am banned from forum for speaking truths.
Which one is the “feeling smug” emoji? ![]()

