dacryn

justify the overhead? But you are using flask and sqlalchemy?

You made the wrong choice I would guess. Django does not have 'overhead' to justify in that scenario.

I have been on android for 10+ years and never has Meet gotten in my way or even launched itself.

It's totally uncomparable. Also google is not actively pushing it's customers to use it to kill their competitors

cities skylines and age of empires 2 sadly ....

not superrich, but I avoid foie gras because of the horrible animal welfare component. If you get it to a price palatable for michelin 1 star restaurants, it s definitely going to get adopted

But old iPads already work for yeeeears.

My work ipad pro is 4 years old, and I can't be bothered to replace it (I can upgrade for free at work, at the 'cost' of having to migrate my apps and data etc...)

My personal ipad mini 3 has been travelling with me until last year, as ebook reader. Sure it was heavily handicapped in the sense that I did not get app updates or any new apps really. But I still have my books and goodreader app, and still had VLC, and until very recently also netflix. That thing was 9 years old and I only retired it because work unlocked my work ipad so I only travel with 1 ipad now.

that is low cost ...

I don't know what you are referring too, maybe it's aliexpress level and they skimped on wire thickness. Most of the cost is the pure copper cable here, which should indeed cost 100+ for a 16ft cable, if you don't want it to melt when it's hot outside.

We were creating many complicated yet great user experience frontends before React was even born, using server side template engines.

React makes development at scale of these things easier, and makes a website really a website (clicks are just triggering javascript instead of the classic post/get cycle that fetches new html). But it is completely unnecessary to build any level of GUI

good performance and putting in extra effort should be rewarded over those who just do their job at the minimum effort levels.

That's a fundamental problem with teacher pay, there are no bonusses, there are hardly incentives or room for raises. It's all just tenure based. Once you're in, you're in.

no it doesn't, at least not in my case. Any settings to do?

why do they want 34? I pay 12

it would be so easy to just implement an opt-out function, but Microsoft doesn't want to.

There doesn't have to be a tradeoff if you make it optional.

it's how microsoft does it's business. They have to lead by example.

What message would it send if they can develop this with 5 devs, but the customers they sell their advice and licenses to needs a 1000?

So they have to dogfood, leading to all of these bullshit bloated apps

whilst similar, an antitrust investigation is different than this one. Microsoft should be on both lists.

the memory is inherent to the gpu architecture. You cannot just add VRAM and expect no other bottlenecks to pop up. Yes they can reduce the VRAM to create budget models and save a bit here and there. But adding VRAM to a top model is a tricky endeavour

if you can handicap a competitor for 15 million to save your 13 billion investment, go for it

There is a reason EU is investigating this, and this is exactly it. They are sick of the Microsoft extinguish aspect

synapse is such a disaster. I don't know any example of a decent size succesfull deployment that. It grinds to a halt and charges you a lot of money for it.

Snowflake is an amazing Agile database that has a great ecosystem. Teradata still remains king if you want this type of cloud datawarehouse, and becomes price competitive with Snowflake if you actually use it intensively (transactions are cheaper on teradata than snowflake)

But in the end, a good managed postgresql is probably enough for 80% of the clients of Synaps/Snowflake anyway. It's just that CTO's are starting to lose technical knowledge and are more politicians nowadays

It just caused me to stop paying for Spotify. Spotify is clearly better, but Youtube music is really 'good enough'

There are two modes of watching TV in my opinion: intentional and casual

Intentional: look for a certain show/movie, e.g. the weekly episode of House MD, Lost, Friends, HBO, ... -> This is the streaming war between Netflix, HBO, Disney, ...

Casual: You turn on the tv with the aim to just browse around and relax, and get something accidentally interesting/amusing -> Youtube is king and more or less the only option here (and a bit of Nebula)

The two together is the traditional TV

you are lucky, where I work WSL2 is blocked for security reasons

WSL 1 is allowed luckily, but it's kinda meh

the answer is not to hide their unlawful data scraping to save their business. The solution is to actually follow the rules

what's wrong with a text/whatsapp/email/push notification/messenger/...?

many systems exist where you have to respond yes to acknowledge receipt. They can manually call the 1% that didn't respond.

Robocall technology isn't free either.

at that point, they can send you a text, an email, a letter, ...

robocalls are never necessary

such a shame it's done in C# and is using components that do not exist in Mono

If written in something more open, it would have more chance of pickup and using it on other weird platforms would become a thing

This happens very rarely, and is usually of the same family. You replace an Airbus 320 with another airbus, not a boeing. Simply because those airlines don't have both on them in their portfolio.

Now if its specifically to avoid a certain type of Boeing 737 max when the airline already flies boeing, yeah thay might happen indeed.

The only cases of switching families is long haul, like Emirates replaced an Airbus a380 with a 777 once on a trip I took. Still caused massive issues because the boeing is smaller, so they don't like doing this.

an API does not mean there are no hidden dependencies, and that is often the biggest failure.

If an API call is not stateless, or requires to be chained with other calls, you're in for a world of pain in the long run

you are limiting computation to discrete computation.

Biology is analog, which is a whole different ballgame compared Turing completeness. And extremely hard to replicate with digital paradigms

I'd argue backprop is still handy just to learn the basics

It doesn't have to be production ready of course, but spending 3-4 hours to write it out in code, debug a few steps, ... are useful in my opinion.

or at least watch the karpathy video and try to follow along

it happily chugs along at 1gbit for me, that's the max I can test.

I'd be surprised if it doesnt support it. Its google after all. Despite their shitty customer support, their engineers dont mess around

I have started to just create physical photo books a few years ago.

It started as a gift to my girlfriend, but now every year we take our best pictures of the year and put it in a photobook. Cant be deleted that way. All the other photos are less important indeed.

I store about 20gig of important photos on my laptop (weddings, events, ... the really important stuff) and everything else is just dumped into google photos. We'll see how that works out in 20 years. But I feel at ease

the problem is that many people dont use it as a backup. In a backup scenario, you are immune to these things, because you still have your primary copy. It's annoying yes, but in a backup you can always afford to lose the backup with no problem.

But many people treat these solutions as an extension of their hard drive. A physical analogy is putting your books into a Shurguard/self storage. Thats not a backup, that's just storage. You should make a photocopy instead