It's very, very far from a CRUD app or "just updating a DB". GUI-heavy apps are notoriously hard to get right. Any kind of "editable canvas" makes it 10x harder. Online collaboration is hard, so that's another 10x—there are known solutions, but it's an entire sidequest you have to pour a massive amount of effort into.

Custom text editing and rendering is really hard to do well.

Making everything smooth and performant to the point it's best-in-class while still adding new features is... remarkable.

(Speaking as someone who's writing a spreadsheet and slideshow editor myself...among other things)

"Having CRUD operations" and "Just a CRUD app" are very different things.

The custom text rendering bit alone should have been a good cue for this distinction.

That was the point. It is a CRUD app but it is not just a CRUD app.

I'm trying to imagine a scenario with a non-trivial app that is missing a create, read, update or delete operation. I'm coming up with so few examples that I have to imagine the colloquial use of CRUD app means just CRUD operations.

When people say "a CRUD app" they mean "an app that is mostly just CRUD"

This might have been the response trcf22 needed.

But sure, I’m being too pedantic here I suppose.

You should look into figma. Its one of the few marvels of software engineering made in recent times.

If you want to know how tough realtime editing is, try making a simple collaborative drawing tool or something. Or an online 2 player text adventure game

Theres a reason tutorials for those arent common

I wouldn’t necessarily say so. I guess that’s what they are trying to « pulse » people and « learn » from you instead of just providing decent unbiased answers.

In Europe, most companies and Gov are pushing for either mistral or os models.

Most dev, which, if I understand it correctly, are pretty much the only customers willing to pay +100$ a month, will change in a matter of minutes if a better model kicks in.

And they loose money on pretty much all usage.

To me a company like Antropics which mostly focus on a target audience + does research on bias, equity and such (very leading research but still) has a much better moat.

Not so different from matrix isn’t? I’ve got no idea how it works but does it also destroys all forms of electronics in the area?

How different is it from going to a public library?

Almost every country in the world apart from the US pays authors for library lending.

I wasn't sure how it works where I live, so I looked it up and apparently in Germany (according to Wikipedia) public libraries pay 3-4c per checkout to a central private body which redistributes it somehow.

So unless the book is checked out a thousand times over and its lifetime, buying it still dominates overall.

The libraries also bought the book originally.

Now I’d love to know more about that model!

After a quick check on Vercel stories, it seems all payments were discarded or mistakes in the first place.

Does it really happen to really have to pay such a bill? Do you need to tweet about it to be reimbursed?

> Do you need to tweet about it to be reimbursed?

This is what scares me, is social media the only way to get things sorted out nowadays? What if I don't have a large following nor an account in the first place, do I have to stomach the bill?

This is exactly what happened to me during Covid... I had a flight that got cancelled at the beginning of the pandemic since the country closed the orders (essentially). A year after, still on lock downs and et al, I wanted to enquire about a refund, for months I got not answer, until I caught wind that people using Twitter were actually getting results. Now, I don’t use social media at all, so I had to create a Twitter account, twit about my case et voila! 30 mins after I got a response and they send me a PM with a case number... Not even going to mention the airline, but it is infuriating...

I can't imagine them sending it to collections. What kind of recourse would a company like Vercel have if you don't pay it?

They can sue you for the bill plus the legal costs.

Someone at a community group I'm in messed up playing with Azure through their free for non-profits offering^. We were out about 1.2k€. Not huge but huge for us.

Encouraged by comments on HN over the years I had them ask support to kindly to wave it. After repeating the request a few times they eventually reduced their bill to <100€ but refused to wave it entirely.

So even without shaming on social media, But it probably depends. It's worth at least asking.

^The deal changed about six months ago.

It's waive, not wave

Relying on the mercy of a support agent that may be having a bad day is a poor strategy

No, at least in enterprise consulting for these kind of hosting, usually there is a contact person on the support team that one can reach directly.

However these projects are measured in ways that make Oracle licenses rounding errors.

Which naturally creates market segmentation on who gets tier 1 treatment and everyone else.

Once you're in a contract + TAM territory, pricing works very differently. Also, temporary experiments and usage overruns become an interesting experience where the company may just forget to bill you a few thousands $ just because nobody looked at the setup recently. Very different situation to a retail user getting unexpected extra usage.

I mean if developer got charged with 100k, more often than not the bank would decline that first maybe if you didn't have that high credit limit

but what happen if this happen to corporate account and somewhere resource get leaked???

multi billions dollar company probably just shrug it off as opex and call it a day

Great job! Would it be possible to know what was the cost of training such a model?

From their report:

> Once a production environment has been set up, we estimate that the model can be realistically trained in approximately 90 days on 4096 GPUs, accounting for overheads. If we assume 560 W power usage per Grace-Hopper module in this period, below the set power limit of 660 W, we can estimate 5 GWh power usage for the compute of the pretraining run.

Would you have the same strategy if you were to do it again?

Do you have LLMs in mind in your SEO strategy (like really long articles bragging about how good your app are in a non human readable way?

Would love to have your thoughts on that!

Thanks for the question.

As I noted it took 7 years to actually get any substantial Google traffic – if I was to start again I would potentially change some of what I've written to be more effective or focus on different topics earlier.

However my approach is always to write what will be useful and interesting for humans, if Google search also finds it relevant then that's a bonus. I never write just for machines... as I think that's a long term losing strategy.

Ok thanks!!

Exactly! If you can get some exposure as a « specialist », build a network or just learn a ton of new skills (marketing, accounting, PR, devops) it tends to be a win/win. That’s what I’m currently doing and by no mean would I have better myself as much in any other way.

If you enjoy Charlie’s, you will definitely enjoy Kahneman’s Thinking Fast and Slow, especially the part about being an « expert »: a few talks in empty classrooms in a famous Uni, a radio show nobody knows and voila, you get some cred!

Yes! Just find any page-turner like The Name of the wind and have fun!

Or a nice biography on someone you find interesting.