Why I'm returning my 965 after one week.

You have a defective unit. Get it replaced and join the bug-filled joy of owning the 965, it's worth it, I promise!

Its most likely just defective, I have used mine to record about 15 hours of cardio in "run" mode in a week and battery needs to ba charged after that, 5 hours straight of running takes the battery from 100% to aout 78%

What are your settings?  I get 3 days of battery life with a fraction of the running you're doing.

Disable PulsOX and than see battery life.

I returned it and I'm looking at alternatives.  Thankfully, my 255 is doing well and I disabled automatic software updates so tgat Garmin can't introduce bugs into my working watch.

I've had none of those issues and i've had a 965 since release date as well as a 955 745 630 620 and many since the 201. I've found the Support Chat helpful in resolving most issue. Some software bugs are irritating , but non serious enough to make me change ecosystems.  I guess we use the device differently and sometimes experience different issues. I would have got a replacement device and given that a whirl.

I have a 955 and my device has spontaneously crashed and rebooted more than once during basketball activities. (It's happened enough times to pass threshold of "freak occurrence", and it's happened on more than one firmware version.)

When I got a 935 (near release date), I immediately found several navigation related bugs, and spent a bunch of time going back and forth with product support to give them bug reports, recreation procedures, etc. I felt like an unpaid beta tester. (These were actual bugs that Garmin eventually fixed, to their credit, not just stuff that annoyed me.)

Sometimes it gets to the point where you just ignore certain things bc it doesn't seem worth reporting all the bugs.

TedW said:

The watch displays big font numbers and vibrates continuously until I hit stop.

S.M.R. said:

I've had none of those issues and i've had a 965 since release date

This sounds like a variant of the random alert problem that more than one person has posted about, where an alert with a huge time or distance is displayed immediately when an activity is started. Is it possible those other ppl all have defective units? If so, that's almost as bad as a pure software bug (imo).

Obviously this kind of stuff isn't happening to everyone (or even the majority of ppl), otherwise Garmins would probably be returned en masse and DCR would be blogging about it.

Doesn't mean it isn't annoying. If I were a first-time Garmin customer and I ran into some of this stuff, it might make me think twice about the brand.

I think everyone has their own threshold for the type, amount and severity of bugs they're willing to tolerate. If my watch started crashing during runs (my primary use for the device), it would def be a problem for me.

I agree it's possible that some of this stuff might be resolved with a new device, but I've also seen ppl complain that they've found a problem which is clearly related to software, but all Garmin does is offer to replace hardware.

I will say Garmin is pretty good about replacing hardware in several cases (such as when an accessory like an HRM strap breaks.) As far as software goes? Well Garmin has the rep that it has.

You don't remember the crash with the FR620 after ca 1 km in the Run profile which turned out to be a bug between watch and (any make of) Heart Rate strap? That was a Doozy! At least back then they released quick bug fixes, within a week or ten days normally. And let's not speak of the FR630 touchscreen. A disaster by any name - though better, according to Ray Maker, than the FR610 touch bezel...

But the Fenix branch also had a lot of crashes, especially when navigating. I remember (on a 5X) having run about two hours, closing in on home, when I decided to take an alternate street from the programmed route... The watch was in 'follow map' (?) mode - ie rerouting - and simply blew up. Erasing all the FirstBeat data of the run with the crash.

I reported that crash (with debug dumps) to the Fenix team - in 2017 they had a direct e-mail to the developers - but of course never heard back from them. After some time reporting bugs I simply quit. Even when experiencing a crippling, potentially bricking, crash I didn't utter a word. This pertained to connecting the Fenix 5X to a computer, for example to copy over the activity .fit files. When disconnecting the watch from the USB port, the watch erased every file it had (including the maps!) and recreated the basic file system from its hidden partition.

This happened intermittently. You never knew. I kept an updated backup of the whole filesystem for these occasions. About a year after my first sight of this bug, they published a firmware which had a specific fix for it. In-between I had seen people post about these strange erasures on the forum (the old forum, which they consigned to the garbagebin when 'updating' to the present forum).

So, I would never use any rose-tinted glasses when looking back on Garmin firmwares. My foray into a Coros Pace 2 has been a much better experience. Rock solid firmwares. What they do less well is the supporting analysis. Eg every single run, no matter how light, they claim at least 1.x Anaerobic training effect for me. That is a regression from ca april 2023 when they rewrote the backend and the on watch analysis. Before this, the training effect was about the same as Garmin FirstBeat output. I went back and forth with support (who mostly was thrilled that the watch identified the same Lactate Threshold as a recent blood test session had done), but they didn't really understand that the Anaerobic numbers were totally bogus. And since support are in effect gatekeepers, it has never been corrected.

So, I bought the mature FR955 and run with both watches. They do complement each other. Especially since Garmin denies us separate Alerts in conjunction with Workouts. Coros solved that problem in 2022...

Edit: changed Anaeronic to Anaerobic (bugfix ;-) Phone is a xxx to type on...

soptunna said:

So, I would never use any rose-tinted glasses when looking back on Garmin firmwares

On the contrary, I see the exact same bugs (or classes of bugs) recur repeatedly over the years. It’s one to say things “well everyone has bugs”, but it’s another thing for a bug to be fixed at one point, but come back later (maybe in a slightly different form), as if nothing was learned.

One example is the “half cadence” bug in fenix devices which seems to mostly affect CIQ apps, but sometimes native apps as well.

Another example is the problem where the the pan/zoom buttons on the map page have a black background (which wastes a lot of real estate), when it was previously translucent. Somehow this bug first appeared on 945 LTE (and was fixed), but now the same bug is apparent on FR955 and similar devices (probably FR255, 265, 965 and Fenix 7/7 Pro). It’s even visible on Fenix 8 device renders for online stores, although it’s not visible in review videos.

From the 3rd-party Connect IQ app development side, I’ve personally seen cases where:

- a device configuration bug was reported and acknowledged (with a promise to be fixed), but it was just forgotten about for years until it was reported again. To be clear, they just forgot to fix the bug until they were reminded years later

- someone noticed something strange about the Enduro 2 device configuration, which sparked a whole discussion with developers and the CIQ team about certain changes which would affect how Enduro 2 would be supported going forward. There were open questions which were never adequately resolved and it was just left at that. But a few weeks later, devs noticed that it wasn’t possible to release Connect IQ apps for Enduro 2 anymore. This situation persisted for a couple of months. Finally, the Connect IQ team to basically revert most of the changes that they had previously discussed. 

The exchange went like this:

Devs: ”Why doesn’t Enduro 2 support the latest CIQ?”

CIQ team: “No worries, the Enduro 2 part number is going away. The CIQ SDK will be updated with this change”

Devs: “But how will we release Enduro 2 apps in the future?”

CIQ team: …

[weeks pass]

Devs: “Uh, we can’t release Enduro 2 apps anymore. It’s been like this for 2 months”

CIQ team: “Hmmm”

[back and forth discussion about how to fix this]

CIQ team: “Oh it looks like the Enduro 2 firmware was supposed to be updated so it has the same part number as Fenix 7X, but it never happened. Once users get the firmware update, everything will be fine”

Devs: “But doesn’t the store need to be updated too? Enduro 2 still has its own part number on the store”

CIQ team: …

CIQ team: “Ok, we’re reverting the original changes to the CIQ SDK”

So to be clear:

- devs identified a problem with the Enduro 2 SDK configuration and filed a bug report. The CIQ team said don’t worry we’re making changes to the SDK bc the device is changing

- devs questioned whether these changes would break stuff but there was no reply

- a short time later, devs noticed that in fact, things were broken (it was no longer possible to release apps for Enduro 2)

- the CIQ team responded that the device team forgot to change the device firmware, but no problem, once the device was changed, everything would work

- devs questioned whether this change would in fact make everything work, as there would need to be changes to the store as well

- the CIQ team finally conceded that in order for the store to work (without changes) they would have to undo the original changes mentioned at the start of the discussion

IOW, in this case, devs correctly noticed (twice) that changes which the CIQ team had planned to make would not work (or would potentially cause problems). And despite explaining that certain changes were coming on the device side (and assuring devs that everything would be ok), it turned out the device team just forgot to make those changes. 

You would think that:

- the CIQ team would be able to anticipate these kinds of problems in advance, considering that devs were able to

- given the CIQ team told devs that changes were coming to the device, they would double-check with the device team to make sure those changes actually happened (since changes had been made to the SDK which were dependent on the device changes.) But no, the device team forgot to make the changes and nobody else at Garmin double-checked. The only ppl who noticed were CIQ devs who wanted to release apps for the device (Enduro 2.)

Reminds me of frustrating times at a previous job where I was able to predict certain problems years in advance, but management did not care.

You break me up! Crazy stuff. And: 'One example is the “half cadence” bug'. Still cropping up!? When doing a complex datafield back in 2015 for my FR630 I had to do a '2 * cadence' to correct. Don't remember if the Fenix 5X needed the same correction.

At times I've carried the notion that the Garmin dev teams are too small, but your example from the interactions between 3:rd party - CIQ - Device teams feels more akin to largish structural problems.