I have a Garmin "smart" watch (with every app notification etc disabled) and I love the fact that I can do almost two weeks of exercises (ride, walk, gym) without needing to charge it. The bike computers are also solid. But sadly the UX of the software on these leaves a bunch to be desired, and I've been bitten by many software and firmware bugs in the last years... Including months for which HRM would randomly and persistently drop it's value from say whatever the real value (say 145 for argument sake) to 80.
I know all of the wrist watches experience this issue, but this was extreme like drop from 145->80 for like 60+ min then rapidly shopt back up. Not like a small couple min blip.
This was a near the top end model at the time, and after complaining Garmin support owned up that this was a firmware bug impact all sensors of that generation and it would take 2+ months to fix (took like 5).
But they did send me a HRM for free and I've been using that. So I am grateful that and using it since. But for short rides (like 90 min or less) I don't always remember to think to bring the HRM.
Prior to that I had two lower end Garmin watches, and despite having theoretically lower end HR sensors they did not experience such bugs or drop outs (an unexpected blip every once in a while).
But I think the main point still stands, their software/firmware/UX has not moved in relation with the hardware. Next time I'm in the market I will be consider all the options. Feels like Coros and others have come a long way.
Prob the biggest thing keeping me in their ecosystem is multi sport (variations of bike riding types -- I do all), hiking, strength training, erg, winter sports. But even there the list of strength exercises has not been updated in like a decade.
Look at coros. Currently wearing Nomad - getting around a month of runtime on a charge WITH notifications enabled (not too many tho, only important ones). And UX is great too imho. (Not affiliated, just a happy customer)
Pre tariffs you could get a whole Carbon bike frame from CN (including stem, handlebar, fork) for $500.
And while I'll give that there was variety between vendors, some of them were doing sophisticated layups with multiple grades of wave that withstood a torture tests far beyond the ANSI bike tests.
In fact you could argue they were overbuilt compared to Western brands (at a cost of 100 - 150g) since they didn't want to deal with warranty claims.
Truly amazing, cheap, good quality mass market carbon products.
> In fact you could argue they were overbuilt compared to Western brands (at a cost of 100 - 150g) since they didn't want to deal with warranty claims.
That's the single strongest advertising for a product I've seen in a while.
As a customer, I also don't want to deal with warranty claims. Overbuilt = as it should be. Excessive optimization ("value engineering") = bad, waste of resources, and producer of trash.
(Excessive optimization also limits what you can do with a product, and is potentially unsafe, as normal use could exceed structural or operational limits of the product, breaking it, and potentially hurting the user.)
(Okay, I feel like I'm gearing for a long rant about sorry state of physical goods on the market, so I'll just stop now.)
My knowledge here is rather dated, and I'm unfamiliar with the minio code base... But ~ a decade ago Intel invested a fair amount of effort into optimizing the oss Erasure Coding libraries for Intel CPUs.
Back then CPUs had less cores, RS coding was relatively more expensive and certainly CPU bound on then new NVMe flash devices
It's possible, even likely this is the result of that
Each of the CPUs tested are from various generations and have different TDPs. It's hard to benchmark 2-cpus if they're not in the same class TDP wise or from the same generation etc.
Politicians are not in step with the general electorate, they are instep with party elites and primary votes. The primaries tend to select most "ideologically" aligned candidates and by the time you're in the general election the politicians on both sides are far away from the median voter.
Oh boy. We had somerhing like 5 our of 8 drives fail all at the same time. All of them were affected models bought at begining of summer and failed a couple months later.
It was a pretty maddening thing to debug and figure out where the issue was (servers, rack, drives, RAID do controllers). 2 different machines 2 and 3 drives. Week later we found the Intel bulletin about the issue.
New large generation (Millenials) coming to the housing market. Low interest rates driving folks to refinance. New housing is flat YoY if you look at "House Starts" statistics.
FICO score for mortgages is not the same score as the modern FIDO score. The mortgage score is often what's called FICO v2 and the modern one is something like v10. They've even branded their different old versions as Insurance score, Auto score, Mortgage score and those industries kind of stick with it.
To illustrate why this matters. My modern FICO score is something like high 700s, low 800s (depending on credit data vendor). My V2 score is like 690.
How did that happen? Turns out when I moved out from my last house (5 years ago) and canceled my internet with Time Warner they failed to charge me $31 which was always set to auto-play. I actually settled 5 months after the move when TW sent me to collections. The shady collection agency promised to remove it off the record if I paid (they didn't). Instead, it shows up as a 5 year old, $31 late payment, paid in full.
How did I find out, while looking to refinance at these current rates. I'm getting this sorted out now. Both TransUnion and Experian got it fixed in a few business day.... fucking Equifax cannot get their shit together 3 weeks later.
For me it's a hassle, annoyance and wasted time. But as you can see it can impact real people and the score methodology is pretty dumb.
It's insane that a 5 year old, paid in full debt for $31, that's not even my fault drags my credit score down ~100 points (that's what it is once corrected) and prevents from getting a refi. Doesn't matter that all my other credit cards are always paid in full, no late payments on mortgage, car, insurance ... which all add up to several magnitudes more then $31 over the 5 years.
Absolutely, but if an institution is taking some form of deposit in euros and promises to make predictable payments in euros then being long euros is not a directional bet for that institution.
Of course the picture gets a little murkier once you start considering inflation caused by exchange rate fluctuations. But that is probably not the main concern in a large currency area with an independent central bank.
We must have different ideas of what fx fluctuations are, +/-500 bps over a decade sure… not +/-500bps on -30% decline line over a decade vs a major pair…
Yeah it has to be the assumption, but at the same time it makes one wonder why EUR/USD hedging is increasingly expensive in the first place… though if one pulled up a monthly chart going back 10 years and their head wasn't tucked between their legs, its easy to see why.
I have a Garmin "smart" watch (with every app notification etc disabled) and I love the fact that I can do almost two weeks of exercises (ride, walk, gym) without needing to charge it. The bike computers are also solid. But sadly the UX of the software on these leaves a bunch to be desired, and I've been bitten by many software and firmware bugs in the last years... Including months for which HRM would randomly and persistently drop it's value from say whatever the real value (say 145 for argument sake) to 80.