HN user
Some of what you’re saying is likely true but quality is something that will never be optional. There is quantitative data suggest that LLMs perform better on higher quality code base which makes sense. Code that is well organized and coherent will make better use of the context window.
Software refactoring and tech debt is absolutely where LLMs will shine as is evidenced by the Bun rewrite.
Quality is different than something that feels “special” which every Ferrari I’ve sat in or drove has in spades.
Whether or not it’s well put together is another topic entirely.
I don’t understand why anyone is jumping to conclusions about anything before anyone has driven it.
A Ferrari is about driving, and while it wouldn’t surprise me that the driving experience is generally the same as most EVs I’m unwilling to dismiss this based on looks alone.
Large Apache ecosystem (Spark, Flink, Pinot) is completely missing and third-party SDKS (looking at you AWS) almost ALWAYS have worse SDKs. The java Kinesis consumer and producer libraries are amazing, the ones for C# are simple wrappers around the AWS APIs which means there's a few foot guns waiting for developers to run into, even if they should know better.
The speed is really cool but the fact that your rules are written as rust code meaning that new rules need a new binary. That might be fine but just wanted to point it out to anyone who's interested.
I haven't gotten this notification yet
A huge number of people are convinced that OpenAI and Anthropic are selling inference tokens at a loss despite the fact that there's no evidence this is true and a lot of evidence that it isn't.
I think it’s fairly obvious that Anthropic is lighting cash on fire and focusing on whether or not they’re losing money per token on inference is missing the forest for the trees.
Tokens become less valuable when the models aren’t continuously trained and we have zero idea what Anthropic is paying for training.
Have you never run a team of software engineers as a lead? Agentic coding comes naturally to a lot of people because that's PRECISELY what you do when you're leading a team, herding multiple brains to point them in the same direction so when you combine all their work it becomes something that is greater than the sum of it's parts.
Lots of the complains about agents sound identical to things I've heard and even said myself about junior engineers.
That said, there's always going to need to be people who can reach below the abstraction and agentic coding loops deprive you of the ability to get those reps in.
Nice straw man.
Nobody is talking about banning anything, we’re talking specifically about holding social media companies accountable for marketing to children a product that is knowingly addictive and potentially harmful to their health.
Part of the issue with social media is that no reasonable parent lets their 12 year old watch porn or drink but Instagram and ticktock are on a lot more 12 year old’s phone’s than you realize. Social media has network effects and creates tremendous social pressure to not make your kid “different” when half the classroom is sharing TikToks.
I’m not conservative in the slightest but I see no reason to treat social media any differently than alcohol, tobacco or gambling. Available without restriction to adults but limited to children under a certain age.
I think people are overreacting about what amounts to a patent. There's no evidence if BMW will actually use this and if they do how.
I'm not defending BMW here but there was a similar freak-out a few years ago about heated seats requiring monthly payments that ended up being a giant nothing burger.
There’s a million reasonable situations where this pattern could arise because of you want to encapsulate a domain behind a micro service.
Take the simplest case of a CRM system a service provides search/segmentation and CRUD on top of customer lists. I can think of a million ways other services could use that data.
I don’t think renting or buying makes a difference in communities. I had a much better relationship with my neighbors when I rented vs now where I own my own home.
I chalk up the difference to the fact that a lot of people my age couldn’t afford to live where I moved so my neighborhood is full of wealthier older people who I have nothing in common with.
I agree with you 100% as someone who's developed on and run many Kubernetes clusters over the years. Picking a choosing what to use can be daunting.
That said, the path to ease of use usually involves making sound decisions ahead of time for users and assuming 99% will stay on that paved path. This is how frameworks like Spring and Rails because ubiquitous.
I am a graduate of Khoury and find this very disappointing. I wasn’t a huge fan of learning Scheme/Racket but after looking back during my career I’m grateful for the education I got there.
This feels like a step backwards.
I have no idea why anyone would use Cognito unless they don’t care about availability.
Almost every other SaaS vendor supports multi-region active-active and Cognito does not.
I haven't looked at Cloud Run pricing but running Kubernetes in the cloud is pretty cheap these days and my experience with solutions like Cloud Run in the past is that they end up becoming expensive.
Kubernetes can be as complex or as expensive as you'd like but it's also fairly possible to run a pretty bulletproof simple Kube cluster.
It varies state by state in the US but the level of insurance you're required to buy for operating a vehicle is incredibly low. Generally, the level of insurance you must buy is dictated by the terms of financing because the only collateral the bank has is the vehicle itself.
If I were in the market for a CyberTruck (I'm not) I'd be thinking long and hard about what I'm getting myself into.
As someone who worked at a very large SaaS company this is a good recommendation if the vast majority of your customers come from large enterprises with competent IT departments.
The problem is when you work with smaller shops that don’t have IT departments or worse bad IT departments you’re going to pay a fortune in support costs.
Use an open source identity provider or pay someone to do it for you.
I was also the SAML person at one point in my career and I 100% agree. I used to laugh at all the HN criticisms of JWT because of how much of a nightmare SAML is.
I personally love sports betting and I’m glad that it’s legal to do and I don’t have to send money to the Caribbean to do it. For me $10 is enough to get me interested in a game and I don’t gamble compulsively.
The cat is out of the bag with sports betting, any teenager can open up a Bovada account with no verification.
I’m happy to talk about advertising and reasonable regulation but banning sports betting at this point seems silly.
I’m not saying I agree with the comment above you but Kubernetes upgrades and keeping all your addons/vpc stuff up to date can be a never ending slog of one-way upgrades that, when they go wrong, can cause big issues.
The Tesla subreddit collective proactively bans people who comment in any sub-Reddit they don’t like. I dropped a drive by comment on a subreddit that was Tesla related but apparently had the wrong opinion so I was banned from a number of other subreddits unless I deleted my post and apologized.
That and the fact that in the US the car is tuned to only turn on the range extender when the battery is nearly dead and it can’t always keep up.
Thankfully it’s trivial to change.
I can tell you that I worked at a large company that was moving pieces like identity and secrets management to SaaS or open source solutions years after rolling their own and failing to keep up. I'm well versed in built vs buy vs opensource.
For the record, Hashicorp vault was 10x the cost of our fully managed Identity system for a million+ unique logins a month.
The high price of Hashicorp doomed them in the end. I don’t know what their financials looked like internally but the quotes we got from sales reps got them laughed out of the room.
Working with deeply nested data structures in Go is still frustrating, it’s one place where the Java still wins thanks to the streams API.
I just starting using .NET professionally very recently and coming from Go and Java a lot of the .NET platform is just downright pleasant.
You can’t get the old behavior back with a flag as of 1.17 and if you could I wouldn’t be complaining.
Also, the reason CN is deprecated has nothing to do with security but the maximum length of the field in the spec. Chrome ignores it but every cert I’ve seen recently still includes it for legacy compatibility.
It’s not a huge dealbreaker but it forced me to go make 3rd parties regenerate their certs before I could upgrade my version of go.
I believe it’s more widely used than Go
This isn’t true at all and I’ve complained about it before but the go maintainers introduced a big breaking change in the x.509 standard library that caused a lot of scrambling.
https://github.com/golang/go/issues/39568
As far as I know every other language standard library still allows CN to be used to validate the hostname when the SAN field doesn’t exist.