It is worth checking out this thread on Xojo’s Forum.
… and this one
… and this one:
… and this one:
That’s 3 serious bugs reported in the first 10 threads I looked at.
I’m a big fan of the canvas - it’s the bugs that make Xojo unacceptable.
Why I chose Xojo?
Please keep in mind that I’ve only been using Xojo for a year now! Prior to Xojo, my computer programming background has been in C++, C#, PL/SQL, VB.NET, and PowerBuilder.
There are (RAD and Cross-Platform support) primary reasons why I decided to begin using Xojo and becoming a Pro Plus subscriber:
- RAD (Rapid Application Development
- Over the decades, the .NET Framework (while extensive in capabilities and cross-platform support) has become far too bloated and no longer a RAD solution.
- Cross-Platform (iOS, macOS, Windows, Android, and Linux
- I really like this about Xojo (coupled with RAD) (.i.e. One code base)
I fear he is in for a rude awakening expecting ONE code base …
Hi Markus,
I do not think the second discussion is Xojo fault. It is IMHO a documentation miss far more than the IDE fault.
Also, sometimes, it is because of the user lack of knowledge, sometimes it is the developer who want to fight against the OS his application will run on.
Sometimes Xojo is not the good RAD for the job.
At last, sometimes it is the reader who do not understand (or the developer’s design):
Along the nearly last 30 years (1998-04 to now), some people here or there ask for a window with a red border. I never understand why until some days ago while I was looking at the Dialysis screen machine:
Standard border: everything is OK,
Orange border: a simple warning (a period ends for example, when the dialysis ends),
Red border: Blood pressure is too low (9 or 8 or lower).
Developer design:
Instead of working against the HIG and add a non standard colored border, a simpler and more visual User Interface is to add a color in the window’s backdrop: pink, orange, yellow, etc.
I hope my idea is not against the HIG.
Now, Xojo is … what it is. Its only advantage is is fast start to develop (vs the market).
No need to tell about how it sucks.
He will get it, but in how many years after the rude awakening ? 
The moment he actually tries to have one code base on iOS and Android 
That he thinks he can have it across all those OSes tells me he literally hasn’t tried to do it yet
Xojo claims you can do it and they should know, right?
lets just say THEY know that’s not true
you cannot write s single code base that works across desktop, iOS, android, and web
its hard enough to write across all the target OSes for JUST desktop
I’ll point you to the Xojo IDE in evidence
Totally agree with you! My question was sarcastic.
Xojo Inc knows better of course, but their marketing as I recall made the claim or at least strongly implied that you could.
absolutely agree they used to imply you could
I picked up on that topic too and it’s sweet what Christian Wheel has done there!
On the contrary of what you read here: Xojo is a very good programming environment, as long as you realize it’s limits. It’s not stellar in cross platform, although I managed to get good results with Mac (developed) software also perfectly running on Windows. As long as you limit yourself between Windows and Mac, you can develop cross platform with not that many problems.
I’m a hobby programmer that mostly develops on/for Mac and I get very good results with it. I have around 10 apps under my belt varying from simple small apps, an iTunes/MediaMonkey rivaling music app to a multi-user CRM system driving a company with 15 employees. That specific CRM app was developed in RealBasic in 2004 and is after 22 years STILL running strong with regular updates.
This specific forum topic by Christian Wheel helped me discovering new ways to accomplish things by adding external libraries, like the BASS library Christian used in his impressive app. I still get amazed on a regular basis of what I can build with Xojo. That might be because I’m not a professional software guy and I guess there’s plenty of software out there that does a better job than Xojo, but I AM HAPPY USING XOJO. 
That’s entirely fair. If it works well for your purposes, you can and should use it. Although I’d encourage you to be open to alternatives like Objo that wouldn’t be entirely new to you, and where the learning curve isn’t very steep. And it’s free – you only need a license to compile a distributable version of your app. If you can get by with running your app within Objo Studio and you don’t need premium features like the Profiler, then it’s a pretty good deal; the price is right 
Like you, I also was able to coax two small Xojo GUI apps written originally for MacOS to come to life under Windows and using a different DB for the backing store. It was more painful than it needed to be and that plus the bug backlog and general attitude toward users (especially pro users) made me unwilling to argue for leaving those apps in Xojo. In fact I told my client I was using Xojo to get them up and running quickly with existing code I was bringing with me, but I never presented it as a permanent solution.
I could have argued to keep it the way it was, but as they say, once burned, twice shy: I had been through this with Microsoft and FoxPro in the early to mid 90s. “Don’t worry, the Mac OS9 and Xenix versions will still be available” vs gone in a year – you could get any of several versions (32 vs 32s vs 64 bit) so long as it was a Windows version. “FoxPro is part of Visual Studio” until it wasn’t, etc. I just am not going to bat with a skeptical client for a vendor that I have no reason to trust, AND isn’t a politically safe recommendation. Where would they find Xojo devs to maintain it after I’m gone, for example? It seems too much like an attempt on my part to create job security for myself.
And therein lies the problem that Xojo is facing. If you’re looking for Xojo developers to hire or work on your project as a consultant there are very, very few options. As a potential Xojo developer if you look at reviews and opinions you’ll find a ton of negative feedback at worst, and qualified positive feedback at best.
Many of us left Xojo after years of begging for better stability, features that we absolutely needed, and consistency. What we didn’t ask for was API 2, a project file that couldn’t handle desktop, web, RPi, and mobile projects all in one file that shares code between all of the targets.
At this point in time I would pay for a Claude license and use it to develop in whatever the best tool is for the platforms I’m targeting. Using AI is not perfect. You have to monitor and check its work but it’s so good at some things that I can’t imagine NOT using it for development work.
From what I can see, LLMs just move the problems around to different places. Ironically, forcing a return to waterfall development in many respects. Yes if you plan and specify the living daylights out of a project, the code gen part might end up going fast enough to overcome the original objections to a heavily spec’d waterfall project: the requirements change too fast, the environment is too dynamic.
IDK. Right now for example I have a rather complex subsystem that the government just blew up by making unannounced changes with no notice to a daily data feed that has been stable since, oh, about 2008. Most of the work is rediscovering exactly what steps are executed and why and figuring out all the new tables and changes to how it’s processed. The coding is the least part of it anyway, so I’m faced with doing the coding myself vs letting a stochastic parrot do it and then figuring out what it missed. Given that, I just find the value proposition of the LLM to be underwhelming. Particularly when it is clearly guessing (and often, wrong) when it offers to just write a single loop of code for me. Often that doesn’t even qualify as “slop”, it’s just plain wrong.
It’s not that I haven’t found LLMs useful for anything. They aren’t too bad if you (VERY clearly) present some obscure configuration or deployment issue or error, in finding a way out of the thicket. That’s a very focused and deterministic problem and it’s usually something that represents a portion of my responsibilities that I find consists mostly of drudgery. It usually saves research time. Although in once case the solution the LLM executed was wrong in some small detail that blew up the whole project – fortunately it was easy to spot and fix.
But I have no desire to become an orchestral conductor, frantically waving my hands trying to get a particular nuanced performance from a swarm of 'bots when I have worked mostly solo for over 40 years now learning how to translate my thoughts to code.
Hence why the postings to Xojo own “find a consultant” forum have literally dried up
There’s been 2 in about a year
INN’s founding was because of the disappointment with Xojo, their lack of openness and much other frustration with them
It also means that the pool of experienced developers that Xojo has to pick from to just maintain the product is dwindling. You simply can’t alienate and drive away the serious developers and claim dogfooding as a core tenet at the same time. It’s just not sustainable.