Not sure what's going on with beta development or development in general at Garmin...

I've been an Edge user since 2010 and a systems engineer my entire career.  I have not been able to imagine what goes on in the Edge development group. Of course, despite their shortcomings, they are making money hand over fist.

Same here. There are so many regressions with the updates, and I'm still battling the unreliable Bluetooth connection every time I use the unit. This is in contrast to my Forerunner watch which is pretty solid.

My guess is that the team working on Edge software is too small and overworked and that management feels like they cannot prioritize reliability because they need to meet deadlines.

And the codebase is probably not set up in a way that allows for quick, automated regression testing.

This may sound condescending, but I prefer a glasses/helmet mirror to any electronics, to "watch my six".

Anyway, similar to other comments, I'm also an electronics and computer guy with a love/hate relationship with Garmin products. The more I use OsmAnd+, the more everything else sucks for navigating. I'm really thinking of switching over to some other bike computer to adorn my handlebars, track my rides, and help me find my way.

Nah, not condescending at all. I gave my brother my bar end mirrors. I didnt like them. The mirror was too small so all my attention was focused on a little mirror taking my eyes off the road. The varia WAS perfect. Big screen (E1040), audio and visual alerts…really quick glances at the display and i knew what was around me at all times. I pulled in more information with the varia than I ever could with a mirror in the same timeframe. There are other devices making their way to market. As soon as someone make a comparable radar, ill leave this failing company. Like you said, there are other better or equal bike computers out there already. If they keep messing up my edge 1040 Ill switch, warn others of my experience and never look back. All this is a result of pure mismanagement. All C level management related to exercise/training software and the ceo should be fired for letting it get this broken. 

My fenix 7 has been pretty good with updates. Although, I’ll wait a very long time before updating my watch given my experience with garmin updates in general. 

I tried bar-end mirrors, and found them useless. I've been using glasses/helmet mirrors for years, and I feel naked riding without one.

Some people prefer the bar-end mirrors, but I've found that glasses/helmet mirrors have much better field of view, they're more stable, less prone to vibration, and you can look around by turning your head, which becomes instinctive, kind of like normal looking around. To "look around" with bar-end mirrors, you have to turn the bars.

Some people prefer bar-end mirrors; I have no idea why.

Fully agree with the sentiment. However, as a product manager at a software development company myself, I do have to admit it's far from trivial to have good coverage from regression testing, let alone automated regression testing. It's not like no one is able to connect via Bluetooth, or that everyone experiences high battery drain. There's a great number of variables at play here, with hundreds of potentially connected devices, via either different Bluetooth versions or via ANT+, with those devices all running different software versions, people with one or multiple (out of thousands of available) CIQ apps installed, having updated directly from the previous release, or having skipped one or several releases, having updated via phone, WiFi, or Garmin Express, etc. The list of potentially impactful variables goes on and on. It's not as simple as connecting a Bluetooth HR monitor with the new build and checking whether it works or not.
Having said that, Garmin is a large company that should be able to hire great architects, developers and test engineers. Their software stability has improved compared to the past (I had a FR610 watch for which you could literally flip a coin on whether it would sync or not) while providing far more functionality delivered via faster development cycles. Maybe that is part of the problem though: being forced to keep developing new functionality by the marketing department, who (rightfully so) see customers looking at what companies like Hammerhead are doing with interest.

A focus on quality first would be beneficial from my perspective, but while we complain about stability issues, we also complain when Hammerhead or other competitors have features that aren't implemented by Garmin quick enough. Unfortunately, research has shown time and again that customers on average value quick development cycles for new functionality more than stability. So companies like Garmin, who value their market share, then prioritize accordingly. Selling a rock-solidly stable Edge 1000 (functionality-wise) only gets you so far in 2024 as a company needing to make a profit.

The varia has never, repeat never, been sold as a safety device. Only an aid. They even say that in the manuals.

If you were a developer in safety systems you would understand and appreciate the difference. 

25+ years in the software business, 20+ years a garmin user.  For their sports equipment, a beta release is really an alpha, a production release a beta.

It's not that they don't know how to do it, as their equipment for flying has excellent and well tested software, it's that they choose not to.

When Edge units were a couple of hundred dollars you could forgive them, but they are pushing $750 now.

This is super easy to understand when you try to experience every update that keeps causing technical troubles.  Look, I and many of the users are still suffering sync and signal drop issues between my Shimano Ultegra DI2 gear and rubbish 1040 solar.