nemothekid

>The article then ends before saying if the new setup solved their problems.

This reads to me like a recruiting post; thats why they are open if it solved their problems. I'm not a fan of the way they framed it, it seems they have made their architecture decision and now need help implementing this.

I'd imagine if you are the kind of person that wants to work on large scale data query problems this is an invite for you to apply. That said, if you are that person it might not be ideal to join after all the fun decisions have been made.

>Caching with Swatinem/rust-cache. It helps, but there’s a trap. A workflow run can only restore caches from its own ref or the default branch, and GitHub evicts caches unused for 7 days.

>A self-hosted runner for the heavy builds. This was by far the biggest win, 90 minutes down to 40. It also means I now maintain a build server.

WarpBuild (https://www.warpbuild.com/)[1] solved this for me.

1. You can cache into your own s3 bucket, so you manage eviction policy

2. They manage the infrastructure, or you can link in your AWS account. I have builds that run on spot instances that are requested when a build starts.

Then for the caching issue, I just have a scheduled job that reruns a build at midnight on the master branch.

[1] Not affiliated, found them through Google several months ago

I think the point is it is hard to claim that this exploit is the "end of the world", if it is trivially defeated by the default setup. Almost every deployment of php is going to be fronted by apache or nginx, and the default client_max_body_size for nginx is 1m.

Maybe I'm finally old, but I don't see how it's surprising that Raleway is at all successful. Heroku built the same business ~15 years ago and had plenty of "enterprise" clients. I don't really understand if there's that much more value in Raleway vs Vercel, and Vercel claims to have OpenAI and Meta as clients.

For me, what was strange is there were no write-ups on how the vulnerability actually worked. It really does feel like a lot of people went "Claude found a bug, and Claude says it's dangerous" but no one actually verified how exploitable it actually is? Not to say its worth ignoring, but compared to Heartbleed (where exploits had to be written by hand) there are a lot fewer active PoCs.

If you cannot make the effort to communicate to me about your project, I don't know why I should bother reading it.

Frameworks in the past didn't always have custom designed web pages, but at least one could understand if the author spent time actually building documentation.

If Claude is writing all the documentation, that has freed up "resource constraints" that you can use to actually make sure your site is legible. If you can't pass that bar, I don't know why anyone should seriously consider your framework - just like if someone had created a framework without documentation in the pre-ai days.

I think my new tell for low effort sites is if they default to the standard Claude Design language. This site has all of it

* Unnecessary badges

* text-uppercase section headers

* Randomly italicized words

Man I recently executed a migration from Aurora to Planetscale. I seriously considered PingCap but decided against it because I thought they cloud pricing was too complex. I wonder if my personal information is in the leak.

I haven't checked in a while, but I believe nginx does not support http/2 with proxy_pass.

>What would I gain from using a passkey?

The number one gain I see with passkeys is that they are phishing resistant. The biggest issue I've had with TOTP keys is they aren't phishing resistant.

Forcing people to use passkeys gives the authentication layer a stronger means to ensure that people are logging into the correct domain. A password manager accomplishes this, but it's very difficult to enforce. (I've heard a number of variants of "John Smith sent me an important document and when I went to open it, the password manager didn't work so I just copied password")

Rust + Axum has sort of supported this with a middleware (https://github.com/lazear/axum-aws-lambda). Essentially you write your API as normal, and then when you build for lambda you enable a feature that does the special handler for you.

This + Function URL + Cloudfront, and you end up with an API entirely hosted in lambda. I have a backend that does is hosted like this that does pretty complex work (image/video uploads, image conversions, on top of the standard auth, database connections, etc). It even renders a frontend with HTMX/Alpine.

Even with all the bot traffic I receive, at almost 15 million requests per month the cost is under $10/mo (median latency is roughly 90ms). Cold starts are a non-issue with the Rust runtime.

Instead of a database, the LLM will expose an api endpoint and build a database on demand?

That's interesting. Maybe to decrease latency the LLM could "cache" it's build of it's database and reuse in between instances. It could host this artifact on a "hub" of git trees and then any new use cases that come up, can be added to this git tree. Then it can possibly be reused in different use cases.

There are plenty of businesses with customers that don’t sell products. The PMF distinction is useful because many startups are actually specialized consultants/agencies, not products.

There are plenty of companies (successful ones) that I’d argue don’t have PMF, but have plenty of customers (ex Palantir)

Looking at the frontend design examples; why do these models seem to love the "01 - UPPERCASE TEXT" motif. It's everywhere now (see https://try.cloudflare.com/, which has '01 · QUICK TUNNELS', but no "02" anywhere).

>If software developers want their users to manage their own ordering schemes or use customization features, make it easy and convenient to do

This is akin to telling parents that if they want to their children to clean their rooms, you should make it easy and convenient to do. Obviously this is non-sensical: if you make a mess it might be hard and inconvenient work. Thats why kids don't do it!

I find some tech criticisms to be simply inane; the tech companies are not going to absorb the costs of educating users or fixing unworkable strategies. Not because Bezos needs to buy another yacht, but because the companies that try to do that all die before they get big. There is a giant graveyard of startups who have tried to be the "responsible and patient parent", and they have all died. Users don't want to put up with it, and if given the option will run to the company that just handles everything for them with a subscription fee or reselling their data.

If Jev can maintain its cost and speed with multi-modal input I think it will be well worth the hype.

In the US, Nike/Adidas were #1 and #2. Thats all I meant - I haven't seen any change in Adidas, if anything they are even more popular now and are probably the #1 sneaker brand I see in San Francisco. The Samba is incredibly popular, it's as salient to me as the "air force 1" used to be.

Nike's destruction over the past 5 years was really surprised and until I read up about the whole "direct to consumer" fiasco it really helped me understand another trend I was a bit confused about.

About 2-3 years, it felt like, out of nowhere everyone was wearing Hoka and On brand sneakers. In my entire life it felt like the Nike/Adidas duopoly was unbreakable, and then it felt suddenly that Hoka was everywhere.

While Hoka probably executed extremely well, I imagine it was an incredibly lucky position Hoka found themselves in when all those retailers had to replace Nike with something. Maybe even allbirds would be in a better position today had they launched 3-4 years later.

Edit: I just looked it up and it seems both Hoka and On were founded before Allbirds. Seems allbirds imploded (~2021), just as Nike was ceding all their retail space (also around 2021). Had they had another year, maybe they wouldn't have had to pivot into an AI datacenter company.

As you get larger there is a fire everyday and ever more competing priorities on which fires get put out.

How much money does movie maker bring Microsoft? Now which executive would want to prioritize movie maker so much that they would directionally alter the company to fix this? Consider the issue that the execs had was fixing all the issues involved a huge number of teams.

Discredit to Bill here but I honestly think he was the only one who had the power to fix this. A Jobs-like figure is the only person with the organizational power and freedom to drive details like this to completion. Anyone else would end up in a power struggle

If the correctness of my program depends on the ordering of data (for operations that would otherwise be commutative), that seems like that should be something explicit rather than implicit.

I'm not sure if I agree that "hidden setting actually keeps your data correct" is something that should be the default.

The problem in practice with CCSV files, is they are no longer human editable, nor will tools like Excel automatically parsed them (maybe, I haven't checked in ~10 years, however I doubt CSV parsing has improved at all).

If the files aren't human editable anymore, why stick with CSV at all?

>If you don't believe us, maybe you'd believe TigerBeetle which is also single threaded by design.

If SpaceTime was as rigorously developed as TigerBeetle, with TigerStyle, that would have been worth reading into as well. I don't think this can compare with TigerBeetle just because both are single threaded. TigerBeetle achieves it's numbers with a novel architecture

If Opus 5 can do it, there will probably be a Chinese OSS model that can do it before the end of the year.

This makes sense for a verifiable task like a regex engine - but those are clearly defined and I doubt an LLM will have a meaningful speed increase over current human regex engines.

Anything else kind of just feels like "make GTA 6, make no mistakes". The ability for you to write fast and correct software is really going to depend on well you can specify your problem, and having the ability to specify correct programs in such detail an LLM can reasonably performance tune it is a big enough reason for me to believe there will still be slow software

The SpacetimeDB launch video came up on my YouTube feed - and I was surprised the video didn't go at all into how it was implemented. I assumed it was proprietary magic, but then I was surprised to find it was opensource. That confused me more - to get the kind of semantics they were talking about I had assumed that it was pretty novel and if it was open source they should be leading with that.

Kind of bummed to see its essentially 2015-era React Flux in Rust around a mutex.

>There’s no reason why this can’t be replicated elsewhere. The key is to make the programming language reasonably robust with at least 80% of common non-UI dev needs built in and put the remaining 20% and UI bits into a small family of well-supported, community-embraced, preferably first party libraries.

The big reason is money & time. Maybe less so today if you are happy with an AI generated stdlib, but I don't think it's fair to compare Rust (built by Mozilla, who's primary product is an open source web browser) to Swift/AppKit (built by Apple, 4.6T marketcap at the time of writing).

When Go was released it had an amazing stdlib, but that was because Google was funding it.

>Guess we all to suffer with braindead supercomputers in our pockets

Complete non-sequitur. You don't have to suffer, you can just buy a different platform. Android has had alternate stores for decades.

This is concern trolling. If you cared about having an open platform, you could just not buy apple. But that's not what you care about - you want to be able to push your shitty software onto other people's supercomputers without Apple getting in the way.

>open platforms instead of being at the mercy of a private corporations protection racket as if it's the 19th century

But your gripe is not about open platforms. Your gripe is that Apple iOS is not an open platform. Those are two different things, there are open platforms out there.

My complaint is that iOS as a closed platform is a feature, and Apple has been a good steward for that platform that has made it a stable computing platform for the general public. Corporations (cosplaying as "poor indie developers"), are simply frustrated that Apple isn't allowing them to abuse their users for extra dollars.

These same "developers" who pretend to care so deeply about open platforms don't even distribute their software on Linux. The concern trolling about open platforms is a farce. Picking on epic games again - you can't play Fortnite on Linux. It's "open platforms for me, but not for thee"

Forget the advertising angle, lets look at payments. Apple forces you to use their payment processor and they charge a larger fee. What does the consumer get? They get a single control panel for all their subscriptions that they can cancel at any time. "Developers" hate this - not because they want to charge you less, but so that they can make you jump through 50 hoops in order to cancel.

If I have to choose between two evil corporations, Apple and the greater tech industry that shit on the web, I'm picking Apple. If I care about open platforms, I'll move to an open platform, not lobby the government for Apple to open theirs.

Your approach is fine, if that is something important to use then use linux. What I dislike is taking a platform where the sandbox is a feature, and crying you aren't allowed to shit in the sandbox.

>All in all, still extremely anti-consumer.

One thing that ticks me off about this conversation is they are always framed as anti-consumer, instead of anti-developer. It's Tim Sweeney, Mark Zuckerberg, and the ad networks who have the biggest gripe with these policies, not consumers. Consumers are the ones who bear the brunt of the cost when it comes to developers who desperately need to exercise their right to capture every and any piece of data the OS provides.

>If I can download and run arbitrary code on my Mac--even if I have to jump through scary warnings--why should I not be able to do so on my phone?

We saw how developers used that "freedom" - now all almost all software runs in a remote server and is delivered through a sandbox (and even then developers try as hard as possible to poison that as well).