HNHacker News
TopNewBestAskShowJobs

yebyen

2,755 karma · joined October 9, 2012

[ my public key: https://keybase.io/yebyen; my proof: https://keybase.io/yebyen/sigs/Q4xYtntxlQX51bnJaCqx5ERBtsA-G0clfn1VJVEY2F8 ]
submissionscomments
yebyen··on My talk with Kris Nova about being homeless. See you on the other side, friend
I'm having a real hard time with this, not as hard as the people that new her personally I'm sure. But here's why Nova's passing is still so hard for me personally.

I grew up having my social life on the internet with strangers I don't meet. That is mostly it for me. I don't go to churches. I don't make close friends at work, generally. Nova is someone I knew through these same circles but didn't really interact with more than a handful of times. She made an outsized impact on me.

She was a pillar of the community. I feel strange mourning her as if I knew her personally. We never spoke outside of Twitch, Discord, chirp-box, Github. Not because I didn't want to know her, mainly because I am weird and didn't want to make things awkward, and I hope she appreciated not being bothered. I carried Hacking Capitalism with me all the way to Detroit to get it signed, and carried it around the conference hoping I'd get a chance to say Hi and it wouldn't be weird.

That chance never came and now I definitely have the regret.

But I do feel like I knew her better than I know most of y'all, who I call friends. Nova made an impact on me, as well as many other people, and she will be missed.

yebyen··on Tailscale vs. Narrowlink
It isn't usable if your goal is to take the open source project and turn it into a component of your own commercial product that is for sale and is closed-source.

But I think more than half of the people in this thread have the opposite opinion (that if it isn't open source, it isn't usable) and I tend to agree with that.

yebyen··on Tailscale vs. Narrowlink
Ah, looks like there are several! And they're all third-(third-)party...

> https://headscale.net/web-ui/

yebyen··on Tailscale vs. Narrowlink
Reverse nit, there is a UI for headscale:

https://github.com/gurucomputing/headscale-ui

Granted, that's also not tailscale.

yebyen··on OpenTerraform – an MPL fork of Terraform after HashiCorp's license change
You might be thinking of the "obviousness test" for patents. If an invention is a combination of two things which are established prior art, and there is a suggestion, teaching, or motivation which any person could discover through straightforward analysis of the existing prior art, then it's not a patentable innovation.

There being one right way to solve a problem, doesn't really change whether a bit is copied or not. If you have a bug and there's one right way to solve the issue, then you don't need to copy the original. You can reproduce the bug and solve the issue, and any similarity between the independent solutions is coincidental.

If you take the source code and accept the new license terms, I think you'll be bound by the terms. I am not a lawyer and I haven't read the full text of the BUSL-1.1 but I will now, since that's probably the best way to understand what the license does or does not do.

yebyen··on OpenTerraform – an MPL fork of Terraform after HashiCorp's license change
I think appropriating code against the license would be a license violation. Whether it's a bug or a feature doesn't really make any difference. If it's not licensed for copying, then you can't copy it. You can't re-license it without permission.

That's why they are saying it's not an "open source" license anymore – because it's not. Source available means just that. You can read it (say for the purpose of figuring out some annoying bug so you can contribute the fix upstream) and you can probably build and run it, so long as your use is according to the terms granted in the license. But you can't redistribute it, any copy you take probably isn't yours, and it's only usable according to the terms of the license.

yebyen··on Plunder and Urbit
> maintain the tradition (mistakenly) established by Urbit

No, your traditions are a mistake. (How do you like it?)

yebyen··on Plunder and Urbit
They literally kicked him out of town. If there was any person you could pick that best represents the whole Urbit community now, it would be anyone but this person.

And that's not because I am beefing with him or have any personal beef with him. He's literally not there anymore. He went away, years ago.

yebyen··on Plunder and Urbit
You may laugh, but a few years ago all those guys who sold us (some of? us - me anyway) on Kubernetes evidently got bored with it, and now they're all building a Hippo Factory[1]. And it's actually really good. This is the current timeline!

[1] https://docs.hippofactory.dev/

yebyen··on Compilation of claims/reports of LK-99 replication efforts
https://www.kjmagnetics.com/blog.asp?p=electromagnetic-levit...
yebyen··on WASI: WebAssembly System Interface
I implemented my first real-world project in WASM for OSS Summit, and everything you said about WASI echoes my real-world experience. It's not a set of functions to call. It's a terrific bundle of constraints (enabling constraints) on how to interact with the world outside of WASM, that embracing will force you to make your problem smaller until it's well-defined and testable, unit-ized even.

You're not building the old monolith the same old way in WASM, at least not if WASI is the main tool in the toolbox. And you're also probably not building the whole thing in WASM, (whatever it is that you had as original goal to build as an enterprise.) I think we'll have legacy code that needs to integrate with WASM for a time, and the time might be forever.

As a Rubyist, from this perspective I have been studying WASM and I have to admit I was really disappointed when I first began to understand the limitations we have in the current crop of tools – I now firmly believe the next generation of WASM tools will do it better.

Sweeping enhancements that are going to change it all again. Dynamic loading will hopefully bring us the capability for Ruby gems with C extensions to join the vfs assembly. (But now, as an end user and not a deep systems implementor, this is the part of the conversation where I begin to lose the plot, and going out of my depth...)

yebyen··on Apple stole $100s from me, Customer Service says there's nothing they can do
OK, but if it has value (utility) and it has scarcity, then it's going to get a cash price. My favorite internet argument about Web3 was the time that I had two people on Twitter telling me that "scarcity" was a problem with my imagination.

If I could just imagine a world with no scarcity, then it would be so! Shame for this tremendous failure of my imagination. It's clear enough that I'm the problem now.

Well friend, I'm here to tell you that it isn't so. There will always be some kind of resource scarcity, and as long as there is money, people will be putting dollar amounts on our million dollar art projects. Whether or not we think they are worth the money, is not at issue. Whether someone will pay the money, is not at issue.

If there isn't a "finitely deep supply" for long-term holders to protect somehow, and the "cost to create more supply isn't negligible" then what cost is there to burn your pseudonymous identity? Just mint another one. It has to have some cost, or spammers will just push this button continuously and ruin it for all us good people. It only takes a few bad apples, (but there is really nothing worse than infinitely many procedurally generated bad apples!)

It requires someone (it can be a shared job) to take ownership of these barrels, and occasionally remove some dangerous bad apples. There can be enough barrels for everyone (even the bad apples! they can get a barrel to themselves...) even if a few are spoiled, but there cannot be infinitely many barrels with infinitely many apples and still make a good community platform. It has to have boundaries. This is just IMHO.

These are the table stakes for a successful distributed community platform to grow without limits. (It has to have limits!)

yebyen··on Apple stole $100s from me, Customer Service says there's nothing they can do
Well I mean, you don't actually need to pay for it, you only need to "make them cost something." (You know, like a JPEG? Nobody really pays for these, do they?)

It's just a bit of friction, think of it like I'm giving you an egg. You need to carry the egg around with you and keep it safe, I'm not giving you another one. OK, maybe just one more. (But do try not to break it!)

But there are egg-as-a-service managers so you don't really have to carry the egg, just pay a small fee and someone will handle it for you according to the protocol. They are actually giving the managed accounts away now, if you wanted to remove all that cost and hosting friction and see what "superconducting feels like" J/K

The point about the identities is that they shouldn't be easily forged. If enough people do have them, if they're actually buying them, they won't have bought just one. They'll have them in bulk, they'll hold surplus, and they'll be careful about who they invite. You'll probably get one at no cost. Unless of course it's really becoming popular, and there is real scarcity. If Urbit is still around long enough for there to be any question of real scarcity, that will have been something else!

Just don't expect to purchase another one on any street-corner if you burned yours by sending spam, (then well it's better to have been a productive spam that made you at least back in value as much as the burned ID cost!)

Nobody really pays for this now. It's all million dollar art projects, and I think it's like the old days when you just went on IRC and who knows what you'll find!

I'm really glad to see you're willing to debate this in spite of having a strong opinion front-loaded, and I'm sure you're right about most people that heard of Urbit so far, ...but I think you'll find that many people will have only heard the good news and intentionally shied away from the controversy. We humans don't like conflict. No need to create it fresh from the bones of our ancestors, I say.

Or most people in reality probably haven't heard of Urbit yet at all.

yebyen··on Apple stole $100s from me, Customer Service says there's nothing they can do
I'm glad you've heard of it!

There are lots more authors now. I'm not sure why you're so hung up on this one. (He was literally forced out of town. So don't hold him against us.)

I was also pretty sure that story would be a good one to explain why it does have a chance. Ever heard of Elinor Ostrom? (I'll assume you didn't hear of her, but you might have heard of Garrett Hardin who came up with Tragedy of the Commons, which I learned was one of the most widely cited "wrong arguments" in the history of publishing papers...)

I'm not a scholar in the field, but it was my understanding that retaining the capability for "driving someone especially bad out of town" was literally the thing that made it possible for "tragedy of the commons" to have been found an invalid construct as foregone conclusion or an inevitable outcome of working with communal shared resources.

They did studies of the tight-knit communities in Africa where they had managed to avoid the tragedy of the commons for many generations, and I'm sure I'm drastically over-simplifying the conclusion, someone here can explain it better than I can...

But so, we've got that going for us, as well. We can kick people out without "losing their shit for them" on the way.

yebyen··on Apple stole $100s from me, Customer Service says there's nothing they can do
How do you make decentralized identities that are spam-proof? You have to make them cost something, that's how. That's what this argument misses, and why you are all wrong about this IMHO.
yebyen··on Apple stole $100s from me, Customer Service says there's nothing they can do
I still think Urbit has a chance.
yebyen··on Simple Sabotage Field Manual – How to Destroy Your Organizations
I bet a lot of people already know Technology Connections, but I like to bring up the first video of his that I ever watched at a time like this - that's right, the one about heat pumps!

He goes on this rant towards the end about Midwestern Values and I had been living in Indiana for about 5 years at this point, and nobody ever explained it to me so concisely and bitingly accurate what that perspective was. The short version in a picture of the story is the old man going up two flights on a rickety old ladder to check the roof, because he's been using that ladder for 30 years and it's "perfectly fine." You should get a new ladder, or you'll probably kill yourself!

The ladder is not perfectly fine, and each year he goes on using it the risk gets bigger. But we don't strive for perfection because we're used to "making do" with "good enough" tools. I could never make these my core values. As a professional, I need the freedom to bring my own tools, and as a full-time employee I'm going to need the right tools provided on the job site (because it's not in my contract to provide my own tools!)

As an app dev for non-profit corp whose primary business is not app development, I couldn't hack it here. I still live here, but I work for a foreign company now, on the open source project that I wish we could have adopted to make my life as an app dev a bit easier, or at least a bit more livable.

Your perspective on this classic document (that gets reposted at least once every year) reminds me of this struggle of my own.

yebyen··on Amazon’s electric Rivian vans have delivered 150M packages
That's not an optimization, lol, and it's not an answer. If the 20 minute recharge was built into a break, with a fast charger, that's going to add an extra 120 miles onto the delivery route range. IDK how many miles Amazon routes are expected to drive in an 8-10 hour shift, but given that the top range of the vehicles is about 350-400 miles with the biggest battery, that's almost an extra 35% of range.

If the trucks go far enough, and the drivers don't need breaks, then say that I guess. Drivers don't want to do their job is not an engineering problem, whereas delivering more packages per shift could be one.

yebyen··on Amazon’s electric Rivian vans have delivered 150M packages
> Teams that find a way to not do the break wind up winning

Teams that don't take the break will be inhibited by the limited range of their un-recharged vehicles. Are you suggesting that some people will find out how to swap their vehicles, and keep the shift going? (Or add an extra length of hose to the fuel line, for a virtually bigger tank? That type of trick isn't going to work well with battery vehicles...)

How is it you propose that someone is going to get more work done in less time, when their total range / route length is reduced by the extra effective distance they would have been able to continue with 15 minutes of fast charging added in, at the one or two points where it makes sense to add a stop in their shifts?

yebyen··on Wearable device that records your voice for legal defense
As a patient I am so offended by the idea that getting a complete and accurate record of one's own medical history and communications might actually just be a big hinderance that really is only gonna make someone's insurance premiums go up.
yebyen··on fe: A tiny, embeddable language implemented in ANSI C
There's an awful lot of languages with an .fe code file extension, aren't there:

https://ferite.sourceforge.net/

yebyen··on I'm afraid I can't do that: Prompt refusal in generative language models
Yeah, I was about to ask it "OK, how long should I wait, or do you need me to provide more input?" then I remembered, I'm not talking to a being. There's no reason to ask it questions like that. Keep talking about the problem, it's trained to talk about a solution, (don't spoil the illusion! Just let it pretend it saw the PDF file, and see how far this goes.)

I still think things might have gone differently if I provided it with the PDF as base64 encoded content. That's me, still not understanding what type of machine is a GPT/LLM, isn't it. It can write a parser but it can't test and run the parser. But it can pretend we did, and it can sure apologize when I call it out for lying!

yebyen··on I'm afraid I can't do that: Prompt refusal in generative language models
> very convincing imitation of being able to do so

So that's really what was happening! Then it's even more impressive how close it came to parsing the PDF file correctly, (or I guess I gave it more direct hints than I remembered.) Thanks for the link!

yebyen··on I'm afraid I can't do that: Prompt refusal in generative language models
I was working with GPT the other day on a new (not programming) problem for the first time, and I asked it to download a Google Drive spreadsheet and take a look at it. Which I was quite surprised to get a flat response that it couldn't do that, since just the other day, I had been working through a programming problem with GPT 3.5 (and I've never paid for it) when there was a point in the dialogue that I was asking it to generate a parser, and I suggested it would be really handy if there was some way I could provide it with example data for the parser to be tested against.

Wouldn't you know, it said this:

> I'm sorry, but as a text-based AI model, I cannot directly receive or process files or attachments. However, I can guide you on how to provide a sample PDF and assist you further with the script.

> If you'd like to share an example PDF, you can upload it to a file-sharing service (such as Dropbox, Google Drive, or any other file-hosting platform) and provide a link to the file. I can then guide you on how to modify the script based on the specific structure and content of the PDF.

Can it download the PDF, or is it faking? I've told it that it is a competent Ruby programmer as a first prompt, and this is the only reason that I can think of why I'd get such different results. Is it drawing a line between "reading the structure" and "processing a set of data files" or is it really just faking me out? The next prompt after I shared a link:

> After reviewing the PDF, I'll provide an updated version of the script that should work well with the given example. Please allow me some time to analyze the PDF and make the necessary adjustments.

... OK, should I just wait here then? I thought about it for a minute, and then I provided it with the header row for the output CSV, which was the only thing I really needed it to get from the PDF file. We moved on after that, and I do not know if it actually ever read the PDF, or if it was pulling my leg supportively, "yeah I can't really download a file, but sure send me the file, I am a great Ruby programmer as you said before, so surely I will know what to do with it friend-o!"

Is it really just the presence or absence of the reassuring "you are a programmer and you know what you're doing" prompt at the beginning of my conversation with it, that made the bot behave differently in both circumstances? I was told that it would be important to do things like this. What's really happening here, I wonder? I would love to understand a bit better!

We did get to the end of the problem, and no the parser did not work reliably (but it worked much better than I was honestly expecting, also I did not try with GPT4 so maybe that would have been the difference...)

yebyen··on Our right to challenge junk patents is under threat
LOL you're right, we could have had a civil conversation and productive dialog if I had just included that one word in my first post! My bad.
yebyen··on Our right to challenge junk patents is under threat
If you want to go ahead and remove again the specific context of my question/assertion after I re-added them back together, to emphasize how important that context was, go ahead and do it. I said what I said, in the order that I said it, on purpose.

I put the context back there together literally in the comment right above this one (and it was there in my top comment, too) because that was the context that I felt was most important, but now you've cut it right back out again. Are you sure?

Humalog as we know it today was first released in the 90's and it's still patented (or, at least Eli Lilly is the only company that can produce it, the specific "it" that we mean when we say Humalog and not "biosimilar to Humalog".) What are the actual enhancements that have been released in 2018-2019 to warrant the granting of a new patent for Eli Lilly's Humalog? Or is it out of patent? What's a biosimilar?

(Why can't anyone make Humalog other than companies that own patents on insulin, if Humalog has been out of patent since the mid 2010's?)

yebyen··on Our right to challenge junk patents is under threat
> First of all, you made a general statement and I demonstrated its falsity

This is an extremely disingenuous reading of our conversation. I made a specific statement about a particular instance (which you dismissed as cherry-picking), and I asked a question at the same time. Which you reduced overall to a "general statement" while dismissing the specific example that was really important to the context of what point I was trying to make:

> James Collip and Charles Best [donated] the insulin patent "so that everyone could have it" and we still don't have universal low-cost insulin over 100 years later, and prices continue to skyrocket

> Is there any evidence that drug patents can actually enter the public domain productively at their end of their expiration period?

There's an impedance mismatch between "nearly all insurance companies require a generic to be used" and "drug patents should be clearly in the public interest" when "drugs have to be profitable or they are not made by companies anymore."

Why not engage a bit more directly with the original example, if you're going to accuse me of generalizing now? Let's talk about insulin, I'm sure I picked it for a reason.

Is there a generic insulin that is chemically identical to a name brand insulin with the same efficacy? No, there apparently is not [1] – I won't claim to be fully informed on these issues, but it seems pretty clear from reading that the issue cannot simply be reduced to "lots of generics exist, so you are mistaken."

Why can't we have low cost insulin from an entity that operates like the post office, that exists as a public benefit corporation without any profit motive?

Does it really make sense that people who have insurance that is not "top tier" will not have access to the better (still patented) drugs, if they are really better? Are they better, or are they really just "evergreening" patents? If insurance typically only covers generics, then it's really problematic that there is no generic form of this lifesaving drug, and the monopoly persists more than 80 years after the patent should have expired on the original invention!

The deal was supposed to be 20 years. What went wrong, does it "go wrong" a lot, and does it "go wrong" more or less frequently when human lives are at stake? These aren't unfair questions, and if Teva (or any company other than Eli Lilly) made a generic form of Humalog insulin then I probably would not ask them at all. But only Eli Lilly does, only Eli Lilly can.

The same company apparently makes the so-called "generic" insulin that also makes the top-tier patented "Humalog" so maybe there are too many regulatory hurdles or something else is wrong. (I thought that the generics were supposed to represent the competition, that comes to save the day after patents expire...)

If we're central planning, wouldn't we want to promote a resource arrangement such that the best drugs with the most promise and the greatest life-saving properties will go to the most people? Is it really so off-base to conjecture that a resource arrangement that actively does something else is "really evil?"

[1]: https://www.healthline.com/diabetesmine/why-is-there-no-gene...

yebyen··on Our right to challenge junk patents is under threat
I picked the one that was donated on purpose by its creators though, and it's still making profits for the drug companies 100 years later. So why do we need patents, then?

From the page you linked:

> While these drugs are nearing the end of their patent term, the release of generics may be delayed due to litigation, acquisitions, or other exclusivities. As a result, some expected generics from previous years may still be in progress.

The point about the PBMs is that apparently, drugs have to be profitable or they are not made anymore. So even the generic companies need to position themselves in a way where their products are profitable, not only for them but for the insurance (or they will be dropped by formularies, and nobody's insurance will cover it.)

If patents are the only way for firms to protect the profitability of their drugs then we are doomed.

I picked one, sure, and you provided a list as counter-example. But the FDA also ships its own list... of off-patent approved drugs without an approved generic (and the current publication is a full 16 pages of drug names):

https://www.fda.gov/drugs/abbreviated-new-drug-application-a...

If the pharma companies can make a change to the formula so that the new formula of a drug is no longer covered by the original patent, they can get a new patent. I'm not an expert, but this is how I've understood the problem through reading about it. Yes, we have generics, but what doctor will prescribe "worse-pirin" when there's a better aspirin on the market now? What insurance company will cover it?

So, with 20 years of head start and a deep war chest, the companies you mentioned can all iterate, stop research on the previous drug, put up legal roadblocks in the way of their competition, put out a new drug which cannot be copied, while it costs a lot of money (and time) for competition to get their generic form of the drug tested and approved. It's not as automatic as the term "patent cliff" might suggest.

I'm just saying that the intent of patents in general is for innovations to enter the public domain after the period of exclusivity on the idea that enables their creators to enjoy profitability and a legally protected (temporary) monopoly on the idea. It seems like a fair question to ask, is that really what's happening? (What percent of time does it happen like that, what percent of time does it go down some other way?)

yebyen··on Our right to challenge junk patents is under threat
I don't know much about it honestly, but this article that talks about PBMs is very interesting:

https://publichealth.jhu.edu/2023/eli-lilly-lowers-insulin-p...

> “Even though they're dropping their list price by 70% and capping their cost to patients at $35, it’s important for us to make this very, very clear—the company is still making a profit,”

...

> During a 2019 congressional hearing, insulin makers admitted that there’s pressure to keep list prices high so that PBMs can also profit.

> PBMs favor the drugs that will yield them higher profits. By lowering prices, Lilly runs the risk of being excluded from formularies, meaning the insurance companies will no longer cover the product for their patients.

yebyen··on Our right to challenge junk patents is under threat
"Obviously." I like how this is the argument for drug patents, and if you don't just simply agree with this, the only real other option is to accuse the person arguing that it "obviously" is the right decision of arguing in bad faith.

...because it's an argument in bad faith? I too don't see why drug patents are "obvious" or clear case that it is helpful enough to be in the public interest to maintain drug patents. Is there any evidence that drug patents can actually enter the public domain productively at their end of their expiration period?

I mean, if James Collip and Charles Best sold the insulin patent to the University of Toronto for a mere $1 each, "so that everyone could have it" and we still don't have universal low-cost insulin over 100 years later, and prices continue to skyrocket, I'm more inclined to believe that for-profit medicine is just evil.

← PreviousPage 3 of 34Next →