I think the degree of dog-fooding isn’t the root of the topic we should be focused on. Really what we want is for Web to get dramatically better and it’s struggling pretty hard for how long we’ve had to wait for it and for how long it’s been out - borderline on the unacceptable level - which is also my guess why Xojo hasn’t really chimed in at this point - it’s an embarrassing thing to admit. It’s especially hard when you’re rooting for something and for it to continuously miss the mark and drag on for so long in trying to do - so let’s maybe pivot to what Xojo can do to further dog-food the Web project and what features and areas are struggling the most. I actually have less concern about scaling and more concern about responsive design being baked in. There’s also just a ton of language features that would be nice to get parity from other languages inline-delegates(lambdas), Object wrappers on the primitives, automated interops support, etc…
I’m not a Xojo web user but it seems that things are moving forwards since @Ricardo_Cruz started. He’s been pretty active on the forums and on the issue tracker.
Every time this topic comes up – we’re at, what, once per month now? – I say the same thing. I have a lot in Web 2.0 and I have a lot of customers using Web 2.0. Some of these customers have some really massive Web 2.0 projects that are either in late stage development or already released and performing well.
Sure there are still kinks to work out, but it gets better with every release. To say that Xojo Web 2.0 can’t handle a large project, in my mind, exposes a serious lack of imagination and tenacity. I see what it can do every day, and I’m really excited to see where it will be in one or two years from now.
As for me, I’m glad they went with GitLab for the issue tracking.
Me too. I don’t use Xojo for web at all and don’t see myself ever having the need for it. But given how clunky the old Feedback app was, I’m happy they’re using a modern, off the shelf tracker so they can concentrate on other stuff.
GitLab is light years better than Feedback. I see no point in reinventing the wheel when there are perfectly good alternatives out there.
Agreed. Dog-fooding would only make sense for an app that they needed for their own purposes that no one else could provide. In this case, the GitLab stuff not only exists and fulfills their needs, but it’s free, too (I assume). What’s not to like?
I am in agreement with all that Gavin has said here, and I also didn’t like the optics of abandoning it and made that clear at the time. Now, though, I’m glad. Bringing Ricardo up to speed on Web 2.0 itself is enough, but it could have been a long time before he was able to devote a significant enough portion of his efforts to Web Feedback, and I’m happy to have the functionality and usability of GitLab right now.
Regardless of the features we would like to be in Web 2, the biggest challenge that I have had to grapple with is a lack of feature/control events parity with Web 1. The struggle to recreate in Web 2 what I could create in Web 1 has been real and often very painful. There is no way to actually convert a project and have it work, but that is not nearly as important as having controls missing events that they used to have that users relied on and the missing basic styling tools that once made designing an app simple without having to rely on finessing CSS. I’m rehashing old news here but this has been the most frustrating 2 years in the 20 years I’ve been using Xojo. It feels like I will likely retire from programming before it is ever a viable Web Application tool again.
lack of imagination and tenacity is not the problem.
The problem is why should someone had to put a lot of imagination, tenacity and TIME doing hacks and workwrounds instead of Xojo fixing the bugs…
Well, Xojo is fixing the bugs in Web 2.0.
In my opinion dogfooding is not only building an application with your own tooling, it’s also using your application intensively every day. For example the old desktop Feedback application. It showed what could be made with Xojo but in day to day use is was not the most fantastic and stable ticketing program I ever had. That said, I don’t see Eddies Electronics as eating your own dogfood.
I am not and expert in Web-technology but I am afraid it’s more than just bugfixing what WE needs. Advancing insight and new developments and techniques can make a project already obsolete before you release. It happened to me more than once. Perfect work takes too long if you don’t have enough recourses.
At the risk of looking like an old muppet, I still remember the early days of Web 1.00. Any new tool needs a while to be closer to perfect.
Yes Web 1.0 certainly had issues to start with and it took a while for it to mature. I lost many hours due to unexpected issues of events not firing when expected or firing when not expected. I also got unexpected issues due to my application working differently in different browsers and application crashes were common. I had to come up with many workarounds. It is no different in Web 2.0 which will certainly be better than Web 1.0 and will be a great platform for future web development.