It doesnt really feel “slow” to me but just buggy and like its trying to do too many things at once.

Going to college doesnt allow you to buy a house at 30 either lol

These are lab conditions, so in reality the degradation is probably quite a bit worse. In practice your battery will reach temperatures ranging from 10-50°C (or much more if you leave it in the sun) for example. High heat especially can cause battery aging to progress at 2-3 times the normal rate. Its also not evident from this graph if these cycles were going from 25-100 and then immediately back down to 25% or if the charge was stuck to 100% for 6h like it is on a real phone that gets charged over night.

5 years is IMHO good enough for a daily phone.

Its better than the current situation, but fundamentally phone hardware hasnt really changed in the last 10 years. There isnt really any reason to not use one phone even longer than that. The Fairphone 6 is promising 8 years of software updates for example.

Sure there are phones that have enough capacity overhead to still be usable despite a 20% reduction, but thats not most phones.
Also they just stated “No” in response to a question to which the answer is unequivocally “Yes” and that pissed me off.

Charging limit features exist for a reason and not just in phones. EVs have the same problem with charging over night. The best thing to do is not to drain it too low or charge it too high. The orange line (65-75) is probably not realistically doable, but the purple (45-75) should be doable for most people that arent constantly hanging on their phones.

As you can see going low is not as bad as going high because 25-75 is much better than 50-100 even tho the difference is 50% in both cases.

https://www.ifixit.com/News/90983/why-charging-your-gadgets-over-80-is-such-a-bad-idea
https://www.roulonselectrique.ca/en/toutes-les-actualites/recharger-son-vehicule-electrique-a-80-ou-a-100/
https://batteryuniversity.com/article/bu-808-how-to-prolong-lithium-based-batteries

If you are wasteful idiot yes.

Thats just 14h tho, not too bad.

Yeah me neither. I buy it in glass bottles, but considering the fact that canadians have milk bags i dont doubt that this exists too.

Microsoft has and will continue to do Trumps bidding when it comes to the ICC, so this is a very good but also very late decision.

Ah i see. With full fat non homogenized milk you always have a big chunk of separated out pure fat/cream sitting on top, but i guess that not it in this case? If in doubt just taste test it, a few droplets of spoiled milk wont harm you.

What exactly am i supposed to see here? Just looks like normal milk to me?

This sounds like an effective plan to me. With enough numbers it will work eventually.

I find the “AI” part undescriptive too. Its mostly just a fronted for a bunch of LLM APIs right? If you want to target idiots then use “AI” in the name, if you want to target IT people use “LLM”.

AI/LLM Center? Multi AI/LLM?

The nane sounds like an AI version of PornHub to me.

I see. I havent had a Google account for like 8 years, so i didnt notice the difference.

This isnt new. Its been happening for years. There is a post about this on lemmy every few weeks. It just doesnt happen consistently, so people always think that they “discovered” this for the first time.

They have a page about that: https://watchy.sqfmi.com/docs/battery-life

With only time keeping, Watchy should have a battery life of 5-7 days, while with fetching data over WiFi, it should last between 2-3 days. These numbers can be extended through further optimizations (e.g. sleep during off hours, waking up only on motion/tilt, etc.).

By default it wakes up the CPU and updates the display every 60 seconds, but you could totally change this to make it only update on demand (the most efficient would be by button press) which should extend the battery life waaay beyond 7 days.

I havent used it yet, so cant really say much, but ive been looking at watchy. https://watchy.sqfmi.com/

ESP32 based with a epaper display

Doesnt overwolf rely on ads?

That seems like something that wouldnt hold up in court if tested. If you make money with it, thats a different issue (Wettbewerbsverbot), but if you publish free and open source, then this wouldnt fall under that to my understanding.