Hi @Per_Ronny_Westereng , I’m constantly using the Web and Services aspect of Xojo for some serious in-house production systems as well as public-facing HMI/SCADA for controlling remote lab equipment. Feel free to message me directly if you ever want to chat about this stuff - I love to give back to our awesome Xojo community
I would say that Web use is down from a peak in 2019, but still healthy based on my numbers. I have quite a few customers who have deployed large Web 2 applications that I’ve actually seen, both intranet and internet.
The big hurdle for a lot of people that I’ve talked to who are not doing development in Web 2.0 is the conversion. A fair number say that their app is “done” and they don’t want to move it to Web 2.0 until there is a compelling or catastrophic reason to do so, and that far outnumbers the folks who’ve told me “We’re just not doing Web anymore” or “We’re moving to a different development environment.”
Granted, my numbers are only representative of those of my customers who I’ve actually discussed it with and those Xojo users in general who reach out to me, but you’d probably be surprised how often I talk to users about things like this.
Yes, I’m one of those who developed a large application with the first framework, and didn’t switch to the second framework. I’ve also noticed the decline in interest on the forum for the web framework.
I suppose that those who need to develop for the web today will use the new framework without any problem, but I have the impression that a certain dynamic has been broken. That’s a shame. With the first framework, Xojo had a certain originality in the way to design a web app too, really like a desktop app (and only in that way, so the concept was clear).
I have been using the web framework for ERP + MES development since 2016. In 2021 I switched everything to 2.0 and it was not so catastrophic because I had already rewritten all the controls via WebSDK(Listbox, dialog, data fields, toolbars, buttons). Not being interested in WebListbox bugs etc. I concentrated on reporting all the performance bugs, memory leaks, container drawing problems etc. Thanks to the good work they are doing I think that with 2022r2/3 the web framework is becoming stable enough to be able to create reliable projects.
I don’t understand why for many the conversion is so difficult, for sure it is tedious to change the names of Open->Opening events however just start fixing the project on Web1.0 using custom classes that rename the events. A small initial effort to then be able to access all the improvements of 2.0.
Thank you for sharing your experience. There are many differences between the two webSdk?
PS : Sometimes it’s not just the conversion itself that’s difficult, it’s mostly a matter of cost and development priority. When development schedules are already filled to the brim for months and years, and customers are already urgently waiting for new features (and that bring in money, unlike the change of framework), it’s hard to integrate in the schedules a quite important development and testing time without any real benefit. Especially since there will always be small changes in the application’s interface when changing frameworks, which means that this will generate new questions from customers and users, and therefore an additional cost, etc. It’s an expensive spiral with little benefit.
Yes, they changed a lot of things even in the WebSDK in fact I had to rewrite the controls to go from 1.0 to 2.0. I kept most of the JS code I had written good though.
I fully understand that it’s a cost that seems to add up to nothing. Unfortunately, the way of computing evolves every year and one must always update and stay current. Refactoring and maintenance activities must always be scheduled in the development plan of any software. True, in this case it seems like a forced upgrade that does not strictly depend on your business logic, but it can bring big benefits now (with new features), future in terms of technological innovation or for security updates etc. For example, if in a few months a security flaw is discovered in some library that is used by xojo for your webapp, by the time you convert your app from 1.0 to 2.0 or change programming language you have already missed the train. It’s fine that if a piece of software works you shouldn’t break it, but it can’t stay that way for life either.
Obviously this is my personal thought! everyone is free to do what he wants
edit: Before you try to convert your app maybe look in “Issues” if there are any bugs that might break your app. If so vote for them so that the more people request such resolution the sooner they are fixed. This would allow the 2.0 framework to get better and better.
I am currently using Web 2.0 for the development of a larger SAAS application (in development for over a year). Currently in the RC phase and after some optimizations with load balancing I am very satisfied. So far I had only developed a few dashboards and admin interfaces to gain experience.
I think that it has also become quieter because those who use web 2.0 intensively have gained experience (and deal with the still open problems) and all those who tried once quickly probably jumped off again.
I hope that web 2.0 gets the attention it deserves from xojo. @Geoff_Perlman@Ricardo_Cruz The concept and also the performance (RAD) has really potential.
Indie developers are also complaining about less and less interaction on Twitter and LinkedIn. I guess the Summer is near the corner.
I personally think CSS and JS could be the main pain points for people willing to migrate from Web 1. We will have to address that ASAP. The good thing I’m seeing, with the new issue tracker, is bugs are being reported there instead, which is much more effective for everyone.
@Ricardo_Cruz I personally don’t think that’s the main pain point - I can hire a JS and CSS person really cheaply to do what I want, rather then me doing it, as they’re a dime a dozen.
Migration from Web 1.0 is not that. It’s more than that - if someone has a decent sized project.
My next migration will be to another language - I have used Xojo Web as a POC / MVP but to my mind is not truly scalable. Make absolutely no mistake Xojo Web has served its purpose for me as a proving ground for my platform and so I am very happy with it (now have a few very large names that will be using my solution in various forms) but if I’m going to spend x on upgrading to Web 2.0 with the constraints that it imposes, I’d rather spend 2x on migrating to an enterprise level scalable platform, without those constraints, with the benefits (including the ecosystem) that it provides.
Not expecting an answer but confirming I am a happy Xojo Web user - got out of it what I wanted, so mission accomplished
@SteveP, performance issues will be addressed. Many of them has been already fixed in 2021, not sure if you’ve tested Web 2 again since the initial versions.
A good amount of the questions I see on this forum are about struggling with styling. Is the only one?, probably not, but it’s still a big pain point.
Of course you can have your own reasons, and I’m glad to hear the success of your project is forcing you to make some decisions (switching to another language won’t make them to magically disappear). If you change your mind, the whole community is here to help. Feel free to open threads about how to address performance issues, open issue reports or just reach us in a DM, if it’s confidential.
I am a chemist not a “pro”, but in every job I have ever had I created apps for use at work both in and outside my department.
Over the years I used a number of different languages, but in the last 20ish years it has been only using RB/RS/Xojo.
Back when Web 1 one was first introduced, though it was still in it’s infancy, I created an in-house app with it.
For me the attraction was all about being like desktop and not needing to know web technologies.
Things changed at work that made web apps impractical… but that changed again about 8 months ago and i had an idea for an app that would better have been a web app than a desktop app.
But the basic change in philosophy in web 2, not knowing web technologies and being too busy with my job to learn to learn them, as well as remembering the growing pains of Web 1, (along with Web 2 having to be API 2) made me decide not to do it.
I don’t know how many “citizen developers” out there are like me, but I suspect the number is not zero.
Xojo web is being used on a large scale application for specific users. We managed to get it to work with a really good response time and fast load time, this can all be done pretty simple. Even our buiding and deploying is as easy as Pressing build and it deploys to the www.
The performance is much improved when using a good load balancing system. Also xojo could improve much more here. Note that the amount of sockets is an important factor to improve app speed.
Xojo web use is not down, at least not here. I think it’s much more UP and devs have less and less issues with it these days…
To be clear, the philosophy in web 2 is only subtly changed from web 1. In web 1 we were trying to make the look and feel as much like the desktop as possible. The feeling was that web apps in general (not those made in Xojo) looked and felt very different from desktop apps because the functionality to make them similar didn’t exist when web apps first came along but by the time we started on web 1, it did. Thus if we could make web apps look and feel like desktop apps, that would be a good thing. It turned out that the look and feel of web apps had been around long enough that people had grown accustomed to it. They no longer saw it as a deficit. It perhaps wasn’t a strength either but it wasn’t a deficit. Thus the desktop look and feel of web 1 was not a strength and could be a liability. With Web 2 the look and feel is designed to be more like any other web app you might find. But this is the ONLY place where there’s any philosophical divergence.
Our vision for the way in which you develop web apps in Xojo and the knowledge required to do so are unchanged. We want you to be able to apply as much as possible from what you have learned building desktop or mobile apps to building web apps. We will continue to improve Xojo on all fronts in this regard. Of course that’s not an overnight process but I feel we are moving in the right direction.