Got the .40 update, and it didn’t fix much.
- Scrolling in the Custom Workout list still does not work.
- Editing the name of a custom workout has the cursor red.
- Still can’t adjust the numbers in any of the interval boxes as arrow keys bolt out of the box.
Looks like the only thing that might have been fixed is the ‘Zwifters Nearby’ isn’t an over caffeinated twitching hot mess.
I think there were a couple of other issues but I just did major surgery on my Neo Smart Bike, and want to ride, so off to TR again I guess.
And what exactly was the aim of the update that started all this? (I’m tired and cranky, but can I go back two updates?)
They fixed a lot of bugs, just not the ones you care about. It’s reasonable to be disappointed but I appreciate the focus on bug fixes. Clearly they didn’t prioritize workouts.
So you came to my post to insult me? I’m glad the bugs you care about were fixed. Happy you are happy…
I was told by support ‘fixes would be in this next update’. Forgive me for being disappointed they weren’t there.
I thought your disappointment was reasonable, because workouts are very important. And your comment about the only thing fixed being the riders-nearby list is completely incorrect. They sped up their release cycle and fixed a lot of public-facing bugs in 1.40. It’s a positive development and I hope they follow up with more releases focused on fixing the stuff they broke. If they let users opt into beta testing (or postpone if they don’t want to volunteer) and roll back if they run into problems that would be a huge improvement as well. The phased rollout process makes it seem like that should be possible since running latest is clearly not required by the platform.
In a past life, I was a programmer and realize that updates can be complex and take a lot longer than a ‘user’ might think they should. I can’t count on one hand the number of times someone said ‘this shouldn’t take long’ and I had to be the rain on their parade, or delay it a while.
But I also realize the value of testing updates, and running as many different types of possible user interactions that I could think of. So the number of errors surprised me, and the time it’s taking to squash them is too. Did they do any testing before releasing the previous update? Possibly. You can only test what you think a user might do/see, but the in-ride Zwifter list? I’d think someone who tried the update would have encountered that?
But whatever… Some of us will have to wait for the next updates. Custom workouts aren’t important for most Zwifters, but it is for ‘many’ Zwifters. People can argue that I should use TrainerRoad for workouts if I’m so obsessed, but TR doesn’t (to my knowledge) allow riders to make their own rides, plus I like the scenery of Zwift, and want the self-induced pain and flexibility of TR. I want it ALL!! And being a user now, I want it last week…
I was going to volunteer to be a beta tester (I’ve done it several times before for other software vendors). But testing only tests what people think is important. On big projects the beta team is often instructed to focus on certain features just to make sure they were hammered enough. It can be fun, being a beta tester. Like when everything goes sideways: ‘Well that was interesting…’
I have done a bit of programming, and when testing it myself, I found that I often “test for success”, meaning I use the application the way I intended it to be used.
I had my partner use the latest build, and he started clicking buttons that needed some input that wasn’t provided, causing errors. I told him he was doing it wrong. He said, “If you don’t want me to click it, don’t let me click it.”
My point being, that testing usually works in the lab (WITL), since the testing is only focusing on what was “fixed”.
That’s why third party testers are needed, even for a minor update. Yes, people that wrote a program should almost definitely be the LAST people to do testing. The maxim I was familiar with is ‘nothing tests better than shipping’, as in putting it in front of users, and by then, if it’s silly and massive faults, the developer looks inept and sloppy. (Like the old days of Microsoft where updates occasionally broke older fixes)
A really funny story on beta testing: The military was developing a ‘smart weapon’ that could detect enemy tanks hiding in battle field environments. Well, the manufacturer was having near 100% scores in it correctly identifying areas where enemy was present, so they went to ‘live tests’, and the likely billion dollar wonder failed miserably. It was no better than taking a wild guess. Well, other people looked at the issue and found that, in the in-house testing all the images shown where enemy weaponry were present were also darker, dimmer images (no sun, heavy clouds) and the ‘clean’ images were bright and sunny. Yeah, the system was detecting the brightness of the image and indicating that the darker images contained ‘badness’ and scoring 100%. I read that and laughed myself nearly off the couch.
Wow, pairing is amazing. I usually pair up the bike and then do the HRM when I finally get on, and it usually took several attempts and a bit of time, but now the pairing of the HRM seems almost instantaneous. Well done…
And it is a galaxy in the night sky on the loop heading into Tempus. Awesome…