Devs at GARMIN, do you test software before releasing it?

philipshambrook said:

FlowState said:

Yet Garmin has a worse reputation for software quality and usability compared to certain other companies

Says who? 

- DC Rainmaker, in 2019. Wonder why he would bother to post "Garmin’s Biggest Competitor Is Their Own Software Instability" back then, and follow up a while later by saying Garmin has really improved their QA process, if there were no problems.

- People on the internet, for years and years.

As far as usability goes, most runners I know never touch the Connect app and only use Strava. Ppl I've talked said that Connect is too complicated. It's also fairly obvious that Garmin will never catch up to Strava in terms of social features (so they don't even bother.) Anecdotally, something which should be simple like changing gear (e.g. running shoes) takes many more clicks in Garmin compared to Strava, and is much less intuitive. I wrote a post about it here.

Garmin is pretty open about the fact that the current beta (in the app and website) is meant to simplify the experience for users who are unhappy with existing look and feel. Wonder why they would bother if everything is already great? DCR wrote a whole post about how the changes in the beta are much needed, btw.

Also, as a Connect IQ app developer, I can see lots of bugs and design issues that regular users don't see. There's a lot of stuff that recurs over and over again, due to poor design decisions and obvious lack of testing.

Some examples:

- Internal dates are *sometimes* off by about 20 years, because of an internal confusion between two different standards of timekeeping (ISO standard vs Garmin standard.) The worst part is this bug has been fixed on some devices and not others, so 3rd party app code has to guess / figure out whether a given date has the bug or not.

- Running cadence reported by the CIQ API is *sometimes* halved. (e.g. The CIQ API may report 90 steps per minute instead of 180.) This is clearly bc of internal confusion between bike cadence (which is revolutions per minute -- for one foot) vs running cadence (which is steps per minute -- for both feet). The dumbest thing is that in one case, the current/instant cadence range was correct, but the average cadence range was halved.

Both of these problems have been around for almost a decade (pretty much as long as CIQ has existed). These problems are "well known" to those of us who've been on the dev forums for years, but do you think Garmin bothers to document them so devs can avoid these bugs in their apps? Nope. They're happy for existing 3rd party devs to provide support (free labor) to new devs. Same as garmin customers are often unpaid beta testers (whether or not they actually enrolled in the literal beta.)

Hopefully it's clear that the nature of those bugs isn't simply "oops, we messed up but no worries it can and will be fixed", but more like "we have fundamental issues that will never be fixed."

From my POV, Garmin is either unable or unwilling to get software right, whether it's customer-facing sofware, or a 3rd-party developer framework.

Can anyone imagine Microsoft, Apple or Google releasing a software development framework where you have to guess whether internal dates are relative to 1970 (the universal standard) or 1989 (the Garmin standard), and they don't even both to let you know?

Yeah I get that they have a lot more resources than Garmin.

I could also go on and on about the various issues in the forum platform. Garmin didn't develop this platform but they chose it, which says a lot. When I look at some of Garmins direct competitors, they use modern forum platforms which beat this one in terms of speed/responsiveness, lack of bugs, usability on mobile, and usability in general. Not to mention the fact that Garmin broke a lot of inbound and outbound links when they migrated forum platforms years ago. Even now, they have broken links on their own developer blogs which point back to the old forums. That's right, Garmin broke part of their own developer website when they migrated their forums, and it's been that way for years.

(Not that it matters, but I begged them to use on of the modern forum platforms - like nodebb or discourse -- when they switched over years ago, and they chose this instead. Suunto uses nodebb - check it out, it's like night and day: https://forum.suunto.com/.)

Anyone ever use a forum platform or social media site/app that breaks when you try to quote an emoji? This one does.

How about intermittent overlapping text?

Again Garmin didn't develop these forums, but it's kinda funny how the overall experience perfectly reflects Garmin all the same.

Then there's the fact that the Garmin "Ideas" page is a just a one way form which submits your feature request to a black hole at Garmin's HQ with no possibility for feedback from Garmin or community discussion. Not like Spotify, which actually has a public ideas page where the community votes on ideas and discusses them, and the company gives feedback. To me that's the difference between a company that doesn't care and a company that at least pretends to care.

It's just a general pattern that's pretty obvious to me. Can't speak for anyone else.