Tried to use AI with XoJo compared with Java and IntelliJ

1

I just tried XoJo with AI. It hallucinates between API 1 and 2 and can’t find the right solution. It makes crucial mistakes and is only a bit usable.

With Java I am using AI as a second programmer at my Side. No Boilerplate writing anymore, refactoring like a charm, working with embedded API’s bluetoothLE API and much more without any problem.

For example: had to change an App from sqlite to Postgres. Was working automatically even with moving processes from code to database while that is possible here. A few corrections, testing, ready and sold.

When will XoJo be able to serve such AI usability?

1 Like
2

I’m the same with C and Vala - AI has enough reference material to have been trained fairly well. Experienced developers can use it as a tool to aid them in programming whilst reviewing the code as you would from a junior programmer.

For most mainstream languages AI is fairly capable but for Xojo it’s just a mess - but what is the incentive to train a model on Xojo?

1 Like
3

When Xojo first said they wanted to do this it only made sense IF they trained an AI using one API or the other exclusively

But they havent done
And even in their latest version they provide “Jade AI” or whatever that is really just a basic UI for chatGPT (not sure if they support others)
Its not their own model they trained using JUST API 2 or anything like it
THAT’s is, IMHO, the only way general AI’s like Claude, chatGPT, codex etc will be useful

But, now they can check of the “AI help” box on the bean counter list and say “Yep done” and move on to something else

But y’all knew this already

1 Like
4

It is based on Claude Sonnet 4 and you need to buy a token before you can use it. At the end it is so that XoJo can’t train the AI and that there are not enought users to train it. Using Claude Sonnet for XoVo is extremely productive.

5

I attended an AI meeting well over a year ago and the comment that the speaker said was something to the effect of ‘if you think you need to train your model you’ve already failed.’ The problem for Xojo is that Geoff said he was training it. Fail #1. Fail #2 is the lack of data for the models to look at to get proper statistics. Fail #3 is that there is so much history out there of API 1 code and not enough of API 2.

I just don’t see how Xojo will EVER work properly with AI. Even if all a model ever did was look at the Xojo forums it would still be confused. Even years later users still mix and match API 1 and 2 all the time. So the Xojo model just gets more muddied every day.

Remember, it’s a great big prediction routine because it really doesn’t understand anything. So if the input data is limited and quality is shit then the prediction model will be shit. They’d need to double the number of users, and open source projects all moved to API 2 and then wait 10 years before the models would get good. They don’t have that long.

4 Likes
6

Yes Bob, that’s the outcome of a CEO which believes that he can control everything. Let’s say it so: he can control his forum. He can control his company. But he can’t control the real world. He should know that.

Working on projects today got real speed. Software which was costing three quarters of a year for a two man team is done in three weeks by one person with AI in Java and I guess also in C#, C++, Python and so on. There is a difference. Millions of users on one site, a few users on the other.

People believed and told me that XoJo could compete with them. And believed it. Today they can see that XoJo isn’t niche, it is a one way dead end. The code is dead property. that makes it impossible to work in the today times. It is a workflow from the day before yesterday. It is a dead end.

The Webapps are still critical, crashing and there are tons of missed functionality. On mobile XoJo is not really usable and has leaks of elementary functionality for the iOS platform and Alpha status for Android. I only say: Tables without columns. Who the heck does that?

The Screen Formats they used for mobile are far away from the reality of the market today. Also not a good Idea. They did not even try to learn.

And yes, when it dies the code is worthless. That is one of the fundamental parts why it is not a good Idea to write commercial software with it. From my point of view: if Jetbrains dies I can two minutes later work with netbeans of with Eclipse. No problem. But with XoJo? No chance at all.

Let’s celebrate the funeral and that’s it. I tested this for a customer to show him that it Is what it is: a dead end.

7

I was building small apps within a day and serious apps with Xojo within three weeks without that AI crap years ago while it took 10x longer in C# or C++ so not sure what you’re saying here besides the fact you found yet another Xojo thing to wine about. Please dump Xojo from your computer and get a life…

1 Like
8

I builded small Apps already within a one and a half hour. Where is the problem? That is really simple. For example a data analyzer which looks in a therapy database of a medical device and analyzes the therapy results to be able to sort the clinical data and clinical conditions.

Also I wrote an App for a CNC company which needed to convert their CNC programs for a new machine type with a Heidenhain Controller instead of a fanuc controller. They had production programs for 920 parts they produce regular over the year as serial parts. The converter was ready within three hours.

Also, that was last month, I was building an application for Serial number tracking in medical devices production. Automated assembly serial number tracking in their manufacturing line. Four days. Including the API connectors for seven different machine lines they use.

Forget this “Only XoJo can do it” dream. It is not tue at all.

That made me laughing. It is a joke. There is really nothing what you can write in XoJo in three weeks what will not possible to build in C++ or C# in the same time. Sorry. That is another surreal Idea of somebody which has no Idea what can be done with them. Let me say it so: I can do any job you try to do in XoJo for Desktop, Web or mobile in half of the time you need without AI in Java language. Why? Because I have the ecosystem around which makes my life easier. I have a really assisting IDE which allows much more productivity. I have a programming language which has much more capabilities. I can’t stand it anymore that people trying to tell me that the master language XoJo can do that faster.

Also: the resulting code is slower when written in XoJo. Simple. Tested even with 2026 Release 1 and XoJo 2019R3.2. The 2026 Release spits ut a result which is a bit faster. But still can’t compete with Java 25 and with a much bigger distance not with C++. Sorry but that is something you may have to live with.

And speaking about that: I have always to listen: for XoJo I have a drag and drop UI editor. Please. For JavaFX I have SceneBuilder, for Java Swing I have JFromDesigner, for Vaadin I have the Vaadin Designer and for mobile (CodenameOne) I have the Cn1 UI Designer. The simple difference between this worlds is: the functionality. For Desktop the XoJo one is acceptable but for Web and mobile it is crap. Show me iOS Buttons with different background colors. And an Android App with a Table with more than one Column. Without using declares. Without using Plugins. While I also do not need Plugins for that. Nor declares.

I saw programmers which are slow while not good programmers in C++, C#, Java, Python, RUST. Yes, people like this can be found. When I hire engineers for my company I look especially for this capabilities. Productivity is important for me to exist cause at the end of the month I have to pay everybody working for me his money.

So this competition is not really working. You can’t win this. While it is not possible for you to present evidences for this 10 times faster programming than a professional C++ programmer of Java programmer. Man, stop dreaming. It is not the reality you speak about.

9

Who cares what language you use??? If it works for you great!

So tired of these “my language is better than yours” arguments.

1 Like
10

That’s not the point. You are right with what you say. But telling: it is not possible to write in C# or C++ while this will cost ten times more time than programming it with XoJo is nonsense. I said all time: if you are comfortable with XoJo and you don’t hit problems you don’t find a solution for: as long as it is available on the market: super. The Idea behind XoJo is really good. No question. And yes, when you can realize every project you have to realize with it: good.

11

That’s exactly what I mean! In contrary to what thorstenstueker claims, I am NOT constantly saying Xojo is the best, I’m only defending it when he found new ways to start a rant against it. For me and my goals, Xojo has been an almost perfect programming language. I get it that it’s not for all people out there and that there are people that hate it because of the path the company chose, same goes for Microsoft, Apple and Google. Some love it, some hate it and there will always be debate.

I don’t care that C++ can produce faster running code or that Java is better, I can make a working app in Xojo withing hours and up until now, I always managed to create what I wanted to create. I only need it for Mac (and sometimes Windows) so no crappy web or phone environment needed. But it seems that thorstenstueker is only out there to wine about Xojo. It’s getting long in the tooth…

1 Like
12

I already stated that, but here I am again:

When I have a memory hole (forgot how to do something), I ask Gemini (free) about that, an most of the time I say “Yes, that is that”.

Easier to ask than search in the Xojo Documentation…

Worst, I asked in TOF what is the Xojo 2926r1.1 new Feature Print that I read the entry in the RN and do ot found in the Documentation.

I still do not know what it do (and I sleep like a baby)…

Regards

13

The entire point was: you said that with XoJo Apps could be written ten times faster than with C++ and C#. That is definitely not the case. And so arguing starts. Not ranting. Arguments. truths. And not little jokes like the one with the ten times faster development and jobs that people could not program with C++ or C# in month but you can do with XoJo in a week. Try to compare programmers on the same level. I am always open for a fair competition.

14

It’s not about competition. It’s about letting people enjoy their favorite thing and not hitting on that all the time, trying to make it look like a pile of crap. In your opinion Xojo is a piece of shit, fair enough. In my opinion C++ and Java are pieces of shit, but do you see me opening topics about it all the time?

It’s like said before:

15

Can’t resist.. you’ld write anything 10x faster in any language if you spent the time coding, instead of writing novels about how bad Xojo is for your needs.

The relationship is over. You have a new girl/boyfriend. Why can you not move on?

1 Like
16

So what the heck is your problem, critics on XoJo forbidden? Na, your problem is simply: stuck with it.

I have still a relationship while I have a few customers needing rewrites of their XoJo applications. The most of them while there is a tiny little edge which nobody could get working. They have speeded months or years of development in it.

So may be you are so happy. Why are you here? This is the forum with people left XoJo which are allowed to speak about the problems and the crap. So if you don’t wanna read critics I have a good Idea for you: go back to the XoJo User Forum of XoJo, inc. which bans everybody which says something against them. We have a few here. Man. And for you as an information: I did not mentioned that I am ten times faster. Yes, you could not resist. For you it is more important to protect this clean and nice world of XoJo. For me not. For me it is important that people have a real chance to get the problems coming out before they invest a year of development with the end: bugs are not always fixed and showstoppers are normal in XoJo world. And THIS is the important part.

What you think about me: I am not really interested. I don’t care to speak truths. You don’t care for me. For me the truths care.

17

Why are you here? This is the forum with people left XoJo which are allowed to speak about the problems and the crap

Is it though? That’s not how I remember it. It wasn’t supposed to be the ‘bash Xojo’ and nothing else forum.

image
image1604×622 54.8 KB

I came here to learn about alternatives, and to benefit from discussions with people I consider knowledgable experts. (Not much of that happening any more)

But this unrelenting anger is just depressing.

1 Like
18

Xojo F’d many of us over and many of us are STILL bitter.

19

Good old times when he was founding inn. He left before long time. Cause many things are changed between 2020 and now. And may be that you understand or not. I can be here even if I am not a XoJo fanboy. And news for you. I doubt that Norman Palardy, Dave S., Hal Gumbert and many others are still XoJo fans. Or let me say it so: mostly they are also not users anymore.

I am helping my customers with their XoJo Apps they wrote. I am helping them to move to another language when they reach borders without solutions. I am actively supporting XoJo users with their problems. So may be you might get that it is not so funny to see that customers lost much money cause their development ends up at point x.

And no, I am not ranting against the language itself. I am ranting not against the Idea cause the Idea behind XoJo is perfect. I doubt that you understand this but..

20

Oh for goodness sake, Had enough of you now.