Rappstar wrote:
USCoregonian wrote:
nightfend wrote:
The whole Zwift experience as it exists now is a hodge-podge of code built on top of the original coding that the triathlete guy (Jon Mayfield) built years back and showed on Slowtwitch. No one expected the popularity the program has had. The company did not start with a proper development core, and they have just been tacking more and more features onto the old code to try to keep up with the popularity and feature requests. This is why you get weird aspects to using the program like having to quit out of the program to change worlds or events.
Are they building a new version from scratch? I hope so, because I don't see how they keep advancing the program without that.
The entire platform is a buggy, sloppy mess and building hardware makes software and infrastructure look easy. I can't imagine they turn out anything of interest that isn't OEMed.
As someone who actually works with the code on a daily basis, I am well aware of the bugs and the tech debt that we have. However, I also see what comes into firebase. And I also have my own experience. For the first six months I rode zwift - before I worked there, I always dual-recorded to a head unit, because no way I trusted a video game to not crash and lose my ride. But zwift never did. And I said to Jon, I probably wouldn't be working here if that happened with any sort of regularity. And, to this day, with over 20,000km ridden, I've literally never had zwift lose a ride of mine. Now, I know it happens. Believe me. I've written code that has caused user crashes. And I hate that. Because I know how that would make me feel. But I also know that we are - literally - well over 99.95% crash free in terms of activities. I play plenty of games from major studios that don't even come close to that.
Most of what I think people are frustrated with is actually bad UX. Like having to quit the game to change routes. That's annoying. But I would - and do, daily - draw a very hard line between UX issues and bugs.
It's also a significant misstatement to refer to Jon Mayfield as "the triathlete guy." Jon has actually never done a triathlon, though he plans to. What Jon has done for almost 30 years is build video games. And most of that time has been spent as someone who has built and worked within game engines. The original Zwift code was built by someone who - at that time - had 20 years of experience building AAA games and - critically - the foundational platforms they are based upon.
Likewise, on the hardware side, the hardware team is led by someone with decades of experience building fitness hardware. Surrounded by a team of engineers and industrial designers from that industry.
I truly do understand everyone's frustrations with Zwift and also realize that it's easy to make hyperbolic potshots on an internet forum. But I'd also say that Zwift has not raised in excess of $500M by being a "buggy, sloppy mess." And there are *parts* of the code that I do hope we rebuild from scratch - and we have and we are. But we also deliver a pretty damn good experience to a ton of users every day, and I'm proud of that. And I fully expect the same from our hardware when it arrives.
Thanks for chiming in. I do agree that it is actually fairly stable, as I can't recall any chrashes myself. Just as it is not really bugs that annoy people rather than UI/UX things.
A common seen one is why is it hard to implement distance to go when riding a route, when all the code is there including lead in(untill X route actually starts), from the races. There you have the distance etc. That is a UI issue I presume?
If UX is then only seeing last 30 days PRs instead of all time or both, or having to exit to change world or start a new route so be it.
I just think people like to zwift it is awesome and want the full experience and for 99'95% of users tdf races, active steering etc. Comes lower on the priority list so why is the UI things not addressed first.