Today started with helping someone out with a new project in another cross platform tool. Headache would be an understatement. The installation took a significant amount of time, only to then install more later. The build time on even a simple hello world was lengthy and then the resulting app seemed slow. After a morning of pain, they were convinced to try Xojo.
Machine wiped fresh, Xojo 2021 r1.1 was downloaded and installed quickly - great!
First a quick example app to show something good - Web Push (an example I remembered being great in 2017 when I moved to Xojo from VB) - errors - something about sessions being dropped even though they were still active. Web uploader example - the least said about that the better. Are these examples not tested for each release as a basic UAT/QA? (Windows 10 latest released build)
Defeated, machine wiped again, Xojo 2018 r3 installed and finally at the end of the day we can write something that works.
Either today I was just unlucky or cross platform development tools in 2021 are designed to make developers wince?
The results of your test answer that speak by themselves
2018 r3 is the one I also use for web. Really miss the performance improvement of 2019, but those have API 2 and not a lot of improvement on web 1 so not worth the money.
And after this, here they come the posts about “what are you talking about, its is working for me”, or the others like “it is working”, just dont use those features, give them a break, a not working functionality is not htat bad
I still use 2019 R3.1 to maintain productive applications. But I also keep a copy of one of the productive projects as a test project on 2020, and now 2021. What changed with 2020, is the introduction of Web 2.0. It is a very different beast, to say the least. In some areas, it is better than Web 1.0 (the web framework that existed prior to 2020). In other respects, it is not yet on par. It really depends on your use case whether you want to develop a productive application with the Web 2.0 framework.
Now, a totally biased and objective opinion: I never do an unscripted, unrehearsed demo of anything, and I always have plan B available. I learned the hard way supporting pre-sales colleagues on a number of major ERP demos. What can go wrong, will go wrong. So, nothing “on the fly”. Nothing unscripted. Even a program that I know is working. Because what can go wrong, will. And also what cannot go wrong, will. Think of an empty ink cartridge on the demo printer. Think of somebody just made a change to the hotel firewall, that you tested last night. Think of some well intentioned colleague who fixed errors in your data. (when that was precisely the the subject of the demo!)… And then some. So, never an unscripted demo, always a plan B - Powerpoint slides that show what should have happened if the colleague did not fix my data, for example. Even trying to think of every way things could go wrong in a demo, it still happened that we had surprises. Demos are evil.
I often maintain projects in older versions of Xojo but I’d never start a new project in an old version, particularly a completely different web framework. I’d be too concerned about the technical debt that would be built up.