veggieroll

Robots.txt is irrelevant after hiQ Labs v. LinkedIn (2019)

Wilhoit’s law:

> There must be in-groups whom the law protects but does not bind, alongside out-groups whom the law binds but does not protect.

If spending is so important why are the most expensive districts the worst. Shouldn’t that money have helped them? And yet year after year, they remain the worst despite $20k+ per student.

I think the adminstration's plan to execute on this is basically garden leave. They tell the "retiring" employees to stop working, but keep them on the payroll so they keep getting paid.

This administration has been playing a lot of games with "budgeted" vs. "delayed" vs. "actively being worked on" (or not). So this isn't really that different than the abrupt cancellations or delays or re-org'ing of funded and legislatively mandated work.

The main difference is the uncertainty. IMO anyone would be incredibly foolish to accept a deal from a random email with such limited info on the exact terms of what happens in edge cases like you describe: shutdown, budget shenanigans, actual official RIF, etc.

This is exactly what [Tarantool](https://www.tarantool.io/en/doc/latest/platform/app/lua_tuto...) does. It's a shame I think that it never got much popular attention outside of Russia. I'm not sure why.

This was my first prompt after downloading too and I got the same thing. Just spinning again and again based on it's gut instinct that there must be 2 R's in strawberry, despite the counting always being correct. It just won't accept that the word is spelled that way and it's logic is correct.

> We do email only deals, we do LinkedIn message only deals and we have a self-serve funnel.

That's awesome. And, I think that's what a lot of people here are talking about: options.

If your customers largely wants sales calls and sales engineers, that's great. But don't ignore the (smaller) population that does prefer no-calls.

> whose customers will not approve a deal by email only

That's kind of begging the question. Part of the point of the thread here is that it seems B2B is leaving some money on the table by refusing to do deals with less interaction. For example, this CTO's comment [1].

A good comparison I think is AWS and other cloud offerings. Of course, past a certain spend, you will start getting calls and have the opportunity to negotiate better rates, SLA's, etc. But, you can just go and spend 5-6 figures per month without any human interaction.

Another good example I think is 5 figure (or maybe even 6 figure) equipment purchases. You can definitely go ham buying most professional equipment without having to talk to someone over the phone. Think servers, GPU's, storage, cameras, etc.

It'd be great if more services offered this.

[1]: https://news.ycombinator.com/item?id=42728008

What does the cost look like for running OpenHands yourself? From your docs, it looks like you recommend Sonnet @ $3 / million tokens. But I could imagine this can add up quickly if you are sending large portions of the repository at a time as context.

Maybe that's true for some people. But there's a lot of frustration being shown here and elsewhere that proves there is a demographic of people who really don't want this.

Cars are similar I think. Sure maybe some people need help. But there's is huge demand for a one-click, no-negotiation car buying "experience" (or lack of experience rather).

My conspiracy theory is that this has more to do with Salespeople and established sales channels (dealers) not being able to understand this both because their job depends on it and because they are naturally people-persons. So it feels intuitive to them and they have trouble understanding/accepting that many other are not.

The compute limit is dead on arrival, because models are becoming more capable with less training anyways. (See DeepSeek, Phi-4)

> Please don't post comments saying that HN is turning into Reddit. It's a semi-noob illusion, as old as the hills. [0]

But also, if you want people to stop mocking "Open" AI, then maybe they should stop being such a mockable caricature of themselves.

[0]: https://news.ycombinator.com/newsguidelines.html

IMO like with most other out-of-the-box LLM frameworks, the value is in looking at their prompts and then doing it yourself.

[1] https://github.com/microsoft/graphrag/tree/main/graphrag/pro...

I wonder if you can heat up a nice hotpot from that torpedo juice.

> might need more money to fuel their addiction

This makes a lot more sense to me than blackmail.

I'd quibble though that this overlooks that quite a lot (possibly even most) drug users are able to lead productive lives without having to resort to illegal activity to fund their drug use.

The other thing this doesn't cover is that there are many alcoholics that fall into the same trap. Though there again, most alcoholics are able to lead productive lives despite their addiction.

And through personal experience, I unfortunately know quite a few alcoholics that work in defense with various levels of clearance.

I don't do drugs; I barely even drink alcohol. But, I think drug testing is a violation of privacy and none of a company's business. This isn't 1950, nobody cares if you're doing drugs on your personal time as long as you're sober for work.

Yes, you need to drug test for clearance. But I'm saying that's a job I'm not applying to unless I have no other choice. It goes straight to the bottom of the pile for me.

Almost everything listed in this article is a downstream effect or side issue. Yes journalism is expensive and hard work, but that's not why people aren't reading/watching. People aren't reading because they don't trust mainstream news to deliver true facts. Journalists have consistently lost that trust over time and need to reckon with why that is.

> Royal Jelly

Is this a Spelunky reference? If so, I love it.

Absolutely. In this case, CLA isn't an issue. But I bring it up, because it seems to me that this article is a consequence of the prevalence of CLA's in projects advised by professional legal advice.

This issue of "how can we ensure this code is under copyright terms XYZ" seems to have been popularly solved by CLA's. And thus a pall has been cast over the traditional open source dynamic, where there'd normally be no question that a fork with a license in the repo is obviously licensed under the same terms.

I've heard this on legal podcasts as well. Where lawyers spend like 30+ minutes talking about whether a license in the project root counts. To the traditional free software / open source movement, this seems really silly, because that's how software projects worked for almost the entire existence of free software / open source software until the last 5-ish years.

The problem is that licensing under the same terms is not enough for most corporate open source stewards. They want copyright assignment, meaning you give them all of your rights.

I did this (3 elections in a row from 2018-2020), and .... now I don't vote, because it became clear to me that the process is not trustworthy.

5 or 20 workers (and their families) who don't get jerked around for stuff way outside their control is an absolute win.

Right now, there is no actual downside for executives. Just less upside. Did they earn XX Million this year or XXX Million. Some tangible downside would be nice.

I mean, heck, why aren't they fired? And really, it's more then middle management where that'd make a huge difference. If bad performance led to actual shakeups in the entrenched middle management, we might actually see business practices change rather than continue on through the established fiefdoms and petty corporate politics.

The prevailing wages are comically low though for most jobs that I've seen (like half of the real reasonable salary for the job we're offering). And employers have the option of 2 data sources so that they can selectively pick the lower of the 2 depending on the jobs they typically apply for.

Yeah. My first choice for approval is admins because some managers are better than others on this for sure. But that really is only possible because I build for a fairly small user base.

Audit logging is definitely another must for me. I literally won't make an app without it. The app I'm working on now had a case where a PM promised up and down that edit history wasn't needed for a particular section, and lo-and-behold 2 years later it turns out that actually we do need the history.

One thing on my list for CRUD software is an „I know what I’m doing” button that bypasses validation. Usually it’s behind manager or admin approval.

It’s pretty tricky to get right and is very case by case in what it specifically means. But it’s been critical to handling edge cases.

For example, in 99% of cases, you can’t progress to stage X without doing Y. But in this case „I know what I’m doing”.

Though this puts a lot of pressure on downstream processes to handle missing data. Fortunately that hasn’t been an issue too much because validation grows and changes over time and must naturally deal with records that were created under prior validation when their state was originally valid.

Genius-level risk taking

I switched to Go years ago because I know that code I wrote a decade ago is still going to work. And that guarantee is even more ironclad in recent years with modules and vendoring.

With a Go project, I can leave it to sit for 3+ years and then pick it back up and add a feature without any issues. I've never had that with a Python project. (and this isn't even about the 2-to-3 situation, I just mean minor version to minor version and packages)

The best cooking YouTube channel I've found are Food Wishes and J. Kenji López-Alt. Both of these channels focus on making a specific dish in each video, but they explain the why of the ingredients, cooking techniques, prep skills. Especially from a food science perspective, their knowledge was so deep and they clearly have a lot of experience in restaurants.

This helped me understand what role each ingredient and cooking step actually does. Similarly, they have great advice on efficiency so that cooking doesn't take as long or as much work.

[1] https://www.youtube.com/@foodwishes

[2] https://www.youtube.com/@JKenjiLopezAlt

As a developer, doing one or two support cases per day has been one of the best things I have done.

It really helps me understand what they’re doing better. It short circuits a lot of bureaucracy. And it’s great when you can push out a change for someone an hour after they wanted it for simple things. So they know that you can make changes fast to address their problems.

And it’s just nice to get to know the users and be friendly. We get to chatting. And it just makes everyone more chill knowing that the developers care.

I bought a Framework 13 DIY Edition [1] a few years ago have been using that as my main personal laptop since then. Recently, I had to make some repairs due to an accident and it's awesome that all the parts are available still and it was dead simple to repair.

I also loved building my Prusa printer. That took much longer to build (like probably 5 hours for me). But it was really cool to learn all the parts and feel really comfortable with how it works.

[1] https://frame.work

[2] https://www.prusa3d.com/category/3d-printers/