I think it stopped when the people there sopped having experience building in other tools and comparing where they are with Xojo to those other toolsets
If they regularly worked with Xcode to build a Mac app in obj-C/Swift they’d see those differences
Same in VS or other tools on Windows & Linux
Those tools and their frameworks have come a long way in the time that Xojo has been around
But horse changes and evolution are unobvious to the Xojo team
This is similar to DB Kit. I asked for that back when I was trying to help Geoff bring FileMaker devs to Xojo. When I suggested it, I expected it to be build into Xojo, not some kit that was external to Xojo.
It’s so goofy.
A well done ORM already existed - Active Record
They should have gotten permission to use that as the basis for DBKit and build it right into the IDE
Or one of the other ORMS that exist
Xojo’s latest marketing trick: boasting and bragging about a capability of Xojo (DBKit, Android support) and keeping that stuff in beta state eternally, so that nobody can accuse Xojo of delivering crap software. ´Oh, you know, it’s in beta, you shouldn’t have used that in production’.
Android is definitely still beta at best
DBKit is not something I’d use having read the code
I’d use ActiveRecord
And the easiest way to do that is with ArGen where you point it at your database & it generates code you can copy & paste into a Xojo project
Always makes me feel good to see that ActiveRecord and ARGen are still being referenced and used. It’s not without its flaws and drawbacks but we used it effectively, for many consulting projects, to do DB apps quickly and easily. That was our bread and butter back in the day so it made sense to make the tooling to make it as automated and easy as possible.
Its still easier than writing it all by hand and its pretty flexible
We have changed it somewhat But we still use it heavily.
Norm wrote some stuff for us that looks at he Active Record classes and auto sets up the DB!
well and then were those Devs, who developed their own Active Record Classes, improved them over the years making them robust and stable and were confronted with API 2.0
Jobs talking about Apple and advertising
And he flat out says we spend a lot on advertising but you’d never know it
How did it turn out? I still use 2021R3, so I’m kind of curious 
meh
same old same old
fixed some bugs
added some new features with attendant new bugs
pushed a bunch of cases from 2024r4 out to the future - again
Quite honestly for ME 2024r.4 isnt useful - broke things that worked and added things that dont work
Neither is 2024r4.1
Waiting to see if 2024r4.2 helps at all but it only has 2 cases assigned to it so unlikely
Well, in the las couple of years xojo has being made actual improvements. But they have a LOT to cach on. 2024r4 Has many interesting things that I really wanted but the quality in general is not good.
For desktop, most of them are still not better or not as good as the “hacks”, custom controls or plugins I use on 2019r2.1. On the other hand, the few things that indeed are better (mostly IDE improvements) are tempting but still not really worth a new license.
For web, I still use 2018r3 but I was toying with 2024r4 and it is the first version of web 2.0 that didnt felt like a beta. Maybe the next Web proyect could be worth the upgrade.
My biggest issue is the things they’ve improved + the bugs added dont make it possible for me to move projects
I’ve got one that opens & compiles in 2024r3 and doesnt compile at all in 2024r4
Its an android project and I have exactly 0 clues why it no longer compiles
IMHO they really need to separate bug fix releases from feature addition releases so this situation can be avoided
Tried to make that case when I worked there and you can tell how successful I was 
We’ve all been trying to make that case ever since the Rapid Release Model was implemented…
Normal builds are like people gathering together to lift large heavy bricks to put into place in a building.
Rapid release is like making those large heavy bricks smaller, stacking them on a single pallet so it’s larger and heavier than before, but instead of carrying it, you use the people as wheels, rolling it over them, placing new people in front when space becomes available as the burnt out, used people spin out from behind the pallet.
And when you get to the building, you realize theres no one left to place the bricks. 
Thats truly a funny way to describe it
Rapid release is about cash flow. It has nothing to do with development. Any comment that says it has to do with development is just rationalization/marketing To help solve very real problem of paying for engineers Who, like everybody else has monthly bills to pay.
It is really stupid to have many releases installed and carefully controll what proyect runs on which release…
Too bad that some CEO had the idea that devs love running after the shinny release notes instead of having a stable tool. 
I only tried Android in xojo once. The “compiler” looked like a bunch of transpilling hacks pointing to a SLOW and ridiculously BLOATED custom runtime designed for a decade old devices. Not really something I want to waste my time on. B4X has me cover on Android iOS and Web services.
The cash will flow as long as there are developers to expend as wheels. 