jasode

It isn't about being superior, mysterious, or clever. The following "nobody understands" type of quips and thought experiments are different ways of saying our intuitions don't help us understand it :

- "God does not play dice with the universe.", Albert Einstein

- "Shut up and calculate", N. David Mermin

- "Schrödinger's cat" being both alive and dead

- various interpretations/analogies explaining quantum physics leading to absurdities or contradictions

Yes, humanity has harnessed quantum mechanics with mathematic frameworks and repeatable experiments to make useful things... but that's a different layer of "understanding" than the deeper levels in the above quotes.

"Nobody understands what gravity is." ... Wait, if nobody "understands" gravity, how did we send people to the moon?!?

Because those are different usages of the word "understand".

The public libraries in major cities like the New York Public Library have sizable collections. The smaller libraries in the outer suburbs have a much smaller selection. Yes, there's also inter-library loan system for the suburbs to pull from the big city library but there are restrictions such as only having 3 items at a time.

The neighborhood public libraries are wonderful resources but there can be limitations (quantity of titles) that prevent it from being a sustainable substitute for streaming for someone that wants to watch ~2 hours of content every day. My friend tried the public library and it was only a temporary reprieve from the ads on streaming platforms. The library simply didn't have enough content for her to watch.

Physical media is only cheaper if the user is ok with re-watching the same videos repeatedly. The higher cost of the DVD/Blu-ray discs can be amortized over multiple repeat viewings. (Good example is a 5-year old kid wanting to re-watch Frozen or Shrek cartoons again for the 100th time. A DVD player is cheaper than a Disney+ subscription in that scenario.)

Digital streaming subscriptions are cheaper if the user almost always prefers watching new shows instead of re-runs. That's the viewing habit of the vast majority of subscribers. Trying to only buy physical media to get a constant diet of new shows -- even buying the cheaper used discs from thrift stores -- will cost more than a subscription. Also, the physical discs are not immediately playable on an iPad/smartphone so requires extra hassle of ripping/transcoding that most normies don't want to do. Also, buying harddrives for the NAS server to store the rips isn't cheap either.

I'm all too aware of people preferring brand new shows because I've ripped several hundred DVDs to a stack of terabyte harddrives that nobody ever watches. Instead, they'd rather see what's new on Netflix, HBO Max, Amazon Prime Video, Apple TV+ ...

EDIT to add an update of a previous comment about my friend's defiance against ads added to Amazon Prime Videos by borrowing DVDs from the public library. She lost that battle because she finished all the DVDs she wanted to watch so had to go back to Amazon Videos for new content. She now endures the Amazon's ads. https://news.ycombinator.com/item?id=42388626

She really tried to fight ads by avoiding the platform but couldn't sustain it because she ran out of free DVDs to checkout from the library. This means buying DVDs to get something new to watch if trying to stick with physical media and unfortunately, that costs more than streaming platforms. She deals with the ad interruptions by multitasking on an iPad when they come on.

Putting ~170 million .com records into a RDBMS on a 90s computer is not enough infrastructure to serve ~746 billion DNS queries a day : https://www.verisign.com/about-us/

They don't use off-the-shelf Redis/PostgreSQL/MySQL and they have more uptime reliability than Github, AWS us-east-1, Anthropic Claude, etc.

Verisign may still overcharge for what they do but nevertheless, the computing infrastructure isn't trivial.

No agenda. I've been a Amazon customer for almost 30 years and not trying to run a smear campaign. I buy from them every week. I'm just explaining why I've revised my thinking about the "shopping hack" of buying 3-and-return-2 that I used to think was harmless and supposedly a "win/win" for both me and Amazon. My intuition thought Amazon would always restock and resell the other 2 was wrong. Maybe they will but there's a good chance they won't and I'm not interested in being the cause of more environmental pollution. So I try to be more conscientious and avoid frivolously returning things.

Several reasons combine to cause the "illogical" disposal of undamaged returned goods:

- delayed vendor payments (NET90) and how Amazon can deduct the cost of customer returns. That the hidden "perverse incentive" that reduces Amazon's "skin in the game" and motivation to resell it. They're no longer financially on the hook to pay wholesale costs for those returns.

- vendor calculates it's not worth the cost of re-inspecting and re-stocking the merchandise. If the NEW item is now OPEN BOX and must be priced much lower, it negatively affects the ROI calculation even further.

- vendor decides that the extra risk of customers tainting the returned items and causing the next customer to leave a 1 start review (which lowers future sales) is not worth it

The above factors are especially relevant to clothes. I'm ok buying and then returning a Bosch power drill for $200 because I'm confident that the reverse logistics will go through a liquidator instead of the landfill. However, I don't have the same confidence for a $50 shirt. It's the ~$5 vendor's cost not the $25 wholesale cost that must be used as basis for calculating ROI of risk-vs-resale and we have to be realistic that will often favor the landfill.

The vendor agreements say if the returns are Amazon's fault, they will not deduct from remittances. If you're a vendor/seller, you use the Amazon Vendor Portal to look at the account status of invoices, purchase orders, inventory, returns, payments, deductions, etc: https://vendorcentral.amazon.com/

When a customer returns something, they can pick from a menu of reasons such as "not as described", "changed my mind", or "wrong size", etc. For those reasons, the amount is always deducted from the seller's account because none of those reasons are Amazon's fault. (Again, Amazon didn't spend the ~$25 for $50 shirt that they're desperate to recover as you're claiming.) In many cases, the seller doesn't want to pay extra to ship the returned items back so they just let Amazon throw it away. That's how a high volume of perfectly resellable items end up in the garbage -- especially before the bad publicity in 2019. Amazon has reduced the environmental pollution of customer returns by giving more options to sellers to restock and resell but it's still not totally eliminated.

EDIT to reply: >Under first-party, Amazon is still purchasing the goods wholesale. Amazon absolutely spent the $25 for the $50 shirt. That is just how it works.

In a 1P relationship (not 3P), Amazon will deduct the wholesale $25 amount caused by customer returns from the next payment to the vendor/supplier. That's how it works. Yes, Amazon acting as 1P buys wholesale -- but they can also deduct the wholesale amounts owed to the supplier for customer returns that are not their fault.

I can't copy&paste the Amazon 1P vendor agreement but here's a similar one from Costco and notice the verbiage:

>14. CUSTOMER RETURNS. Costco may pursuant to Section 13.1 reject or revoke acceptance of and return to Supplier any Product returned by Costco’s customers. For any Products returned by customers, Costco reserves the right to return such Products to Supplier or to salvage, donate, dispose of, re-use, refurbish or recycle such Products at Supplier’s sole cost and expense.

Similar wording for the big retailers like Walmart, Target, etc.

>Nobody on either side is throwing away perfectly good, sellable merchandise just for the fun of it.

You keep repeating this but the investigative reports with GPS trackers put inside new items immediately returned to amazon to trace where they go and hidden cameras at landfills and industrial shredding facilities that receive truckloads of Amazon customer returns shows otherwise. The volume of it has been reduced somewhat, but not eliminated. Amazon doesn't throwaway all resellable items, but they still throw away a lot of it.

No, not a consignment model. A "chargeback model".

>, but this doesn't change the fact that they still have to pay suppliers.

Not for the returned items. The returned items are deducted from the next unpaid vendor invoice. (The "vendor chargeback" or "vendor clawback" in industry parlance). All the big retailers with leverage such as Walmart/Amazon/Target/HomeDepot/BestBuy etc do this. It makes the financial hit from customer returns mostly the vendors' problem instead of Amazon's.

>, has a little to do with the fact that it has been returned (except for things that are so damaged

The new law in France was trying to stop Amazon from destroying undamaged products because of investigative reports. Again, Amazon didn't care about sending undamaged returned merchandise to the landfill by the truckloads because of the chargebacks above. In other words, Amazon never "bought" the items the customers returned.

If it was expensive items like electronics, they would ship it back to the supplier or a liquidator for resale but that often didn't happen for clothes because it wasn't worth it the extra costs to the supplier. That's why a lot of it went to the landfills.

The key you're missing is Amazon didn't have to spend that ~$25 in the first place.

I'm guessing your assumption of Amazon's out-of-pocket cost is the source of your insistence that Amazon is determined to restock and re-sale instead of throwaway. That mental model of how big retailers operate is incorrect.

The giant retailers like Amazon/Walmart/Target don't pay upfront for merchandise. Instead, they use a complicated "chargeback model". They just deduct customer returns from the ongoing rolling invoices "net 90" etc.

https://www.google.com/search?q=amazon+walmart+target+%22net...

So, Amazon doesn't really "lose money" for customer returns if they don't re-stock & resell. It's the manufacturers and sellers that take the financial hit. (Which would be the much lower ~$5 instead of $25.) The above chargeback model is why Amazon was so nonchalant about throwing away perfectly good items that were returned by customers. Amazon wasn't incentivized to bother trying to re-sell it. Instead of taking a risk with a tainted customer return, just mathematically deduct it from the next vendor's payment. No ~$25 wholesale cost needs to be "recovered" because Amazon never spent it.

>That's not how businesses operate.

It is (or was) how they operated. That isn't FUD. Canada, France, Germany did investigative reports with hidden cameras and GPS trackers to discover that billions in Amazon merchandise was being dump at landfills and/or burned at incinerators. (What Amazon euphemistically calls "energy recovery"). This includes lots of expensive items that were more than ~$4 such as new coffee machines. The public relations backlash at Amazon destroying so much product was so bad that France passed a new law targeted at Amazon to stop doing it: https://www.google.com/search?q=amazon+france+passes+new+law...

That's when Amazon started a more earnest "re-sale or donate" and reduce the destruction of customer returns.

To clarify, I'm already aware Amazon re-sells returns. I often buy "Used Like New" electronics from Amazon Resale which are obviously customer returns that Amazon re-boxed. However, the category of clothes is different where it customer returns are often thrown out.

>What Amazon sends to landfills is when items (mostly third-party seller items) aren't selling and the seller doesn't want to keep paying for warehousing or to have them shipped elsewhere so Amazon sends them to landfills.

No, that's not the full story. What's also happening is that a seller has different options for Amazon's reverse logistics and they have different costs.

E.g. pay $1.50 for Amazon to inspect and resell or pay a lower $0.52 to throw it away. A lot of sellers just pay 1/3rd less and let Amazon throw it out.

That possibility of me inadvertently adding to landfills is enough to stop me from buying 3 sizes and returning 2. Maybe in some circumstances the other 2 returned items won't be thrown out -- but I can't know for sure so I won't do it. I'll continue to buy clothes from Amazon but only do so when I'm 100% sure I'm going to keep it no matter what.

Deep link to a video comment about online returned clothes: https://www.youtube.com/watch?v=66qOop6J8Q8&t=2m56s

The food and groceries I already knew about. However, Amazon throwing out returned clothes was a surprise because I've seen clothing retailers like Macy's just put the returned items right back on the rack for re-sale. When you go to a shore store, they don't throw out every pair of shoes that customers tried on and didn't buy. So I just thought Amazon just did the same thing at their warehouse. I've also returned car wax to an OReilly's Auto store and I watched the clerk just put the bottle right back on the shelf for the next customer to buy.

It was disconcerting to find out I was adding to the landfill just because I returned clothes to Amazon. Nowadays, I don't order clothes from them unless I'm 100% sure I won't be returning it.

I used to take advantage of Amazon's return policy by deliberately buying 3 sizes of the same item (1 below and 1 above my expected size) because clothing manufacturers are not consistent with their sizing. I can then try them all at home to choose best size and return the other 2 that don't fit. I did this for pants, shoes, etc. This really seemed great because I hate going to the store and changing clothes in fitting rooms.

However, I stopped doing this many years ago because I found out that some online retailers like Amazon do not restock the returned items for re-sale and instead, throw it away so it eventually ends up in a landfill. In those cases, the sellers just eat the cost of unsold goods that got thrown out.

It seemed illogical to throw away perfectly good clothes that are still NEW but somebody explained that Amazon does this because they're paranoid about bodily fluids etc tainting the returns. So to play it safe, they just throw away the returns.

We never do impulse buys for frivolous throwaway consumerism junk but fast ~2 day delivery is still very helpful because... Life Happens ... such as stuff unexpectedly breaking and you want to have a replacement to quickly fix it.

Example of 2 events just in the last month: (1) garage door screw coupler broke. It's the piece that connects the motor shaft to the screw drive. You cannot buy this type of oddball part from Home Depot or Lowes. Amazon had it and delivered it the next day so I was able to repair the door immediately. The alternative of waiting a week+ for slower delivery means the door has to be disengaged from the track so it can be manually raised and lowered. This is not acceptable because I don't have the specific key to the garage locking mechanism from the outside. Why don't I have a key?!? Because life happens. I guess I could manually lock it from the inside and leave the car outside the garage for a week but I didn't like the idea of leaving a catalytic converter exposed to drive-by theft.

(2) was my Skechers work boots had the sole separated from the shoe. I bought the same model shoes from Amazon for $59.99 and had it in 2 days. I could drive to the local Skechers store but they charge $79.99. So either wait a week+ for slow delivery or pay $20 extra plus round trip gas money to buy at the local store.

Amazon quick delivery acts as a JIT (Just In Time) inventory system to take care of unexpected things that happen in life. That's a more sane alternative than holding a bunch of inventory of random parts for things that might go wrong.

Btw, Amazon often offers extra reward points credited to your account if you voluntarily choose slow delivery instead of 2 days. This is what we do the majority of the time. I assume Amazon does this pseudo "cashback" incentive as a way to consolidate as many customer orders as they can to optimize their delivery routes and reduce costs.

I keep retrying Firefox over multiple years and it's still not quite equal to Chrome for Youtube 4k playback ...

2021: https://news.ycombinator.com/item?id=28783904

2024: https://news.ycombinator.com/item?id=42176450

2026 still talks about the problem: https://go.tabbit.ai/blog/firefox-youtube-4k-stuttering

And a recent Youtube comment (2026-September) notices that Firefox doesn't run as well as Chrome on his modern high-end laptop with 64GB RAM: https://www.youtube.com/watch?v=vIE_PkuFeng&lc=UgxfGeBL6sopr...

Sorry for not being clear. The Crestron/Savant panels are not typically purchased separately as 1 item. They're installed by certified system integrators as part of a total package where they wire up the whole house (often new construction).

There have been rumblings and predictions that Apple could expand HomeKit's capabilities to eventually be a real alternative to Crestron/Savant. The previous iterations of HomePod and Apple TV acting as a home hub really wasn't a credible threat but a new product like HomeView is a step closer to the "professional smarthome automation" space. E.g. Apple offering a wall-mounted PoE option is evidence of that.

One of the motivations is to have a "master controller and dashboard" for the (smart-) home that's conveniently always attached and connected to the house itself.

Consider that many houses already have wall-mounted control panels in the main hallway or kitchen for the burglar alarm or thermostat. Replace that with an all-in-one smarthome touchscreen that configures the same security system as well as everything else ... HVAC, lights, blinds, door cam video feed, etc.

Examples of that concept from other home automation companies such as Crestron, Savant, Control4, etc:

https://www.google.com/search?q=Crestron+home+touchscreen+ho...

https://www.google.com/search?q=Savant+touchscreen+home&tbm=...

https://www.google.com/search?q=Control4+touchscreen+home&tb...

Apple's product marketing pages for HomeView at the moment don't heavily promote the idea of customized programming for home automations (HomeKit API) to compete with Crestron/Savant but many observers think Apple wants to expand into that space. (https://www.apple.com/home-app/)

A lot of homeowners don't want to be locked in with Crestron/Savant. Yes, Apple is itself also a "walled garden" but it's a more palatable lock-in situation than the ridiculous re-programming fees that Crestron dealers charge for every little change.

It's a 1-man operation so it may have not have been financially viable to architect the app as server-based.

- server-based : must invest a lot more money in server farms and extra disk storage, or pay high AWS cloud fees. E.g. if a million users do a blur or denoise filter, all that cpu processing has to happen on the servers, and massive disk space to hold the intermediate files, and extra bandwidth costs to send the changed bytes back to the client.

- client-based : just ship Javascript blobs to end users' web browsers because the blur/denoise/etc filters happen on the desktop.

Also, this type of pixel-editing software still needs a ton of client-side Javascript to behave like a Photoshop clone because users want to see interactive changes as they dynamically slide the blur/noise/etc settings. Round-tripping that with extra server latency is not a fluid UI experience.

We can't confidently replay the past and say that starting it as server-side app from the very beginning means he'd have the same $million in revenue today. Instead, the extra server costs and UI jankiness could have doomed the project.

Less than 7 figures (~1 million) per year -- not per month -- based on previous comment from 2021: https://news.ycombinator.com/item?id=26769141

A later 2023 interview updated it to ~$200k/month (~2.4 million/year) : https://web.archive.org/web/20240606073354/https://saastrapp...

To not lose sight of the discussion subtleties, the gp was saying that in this 1st scenario, Bob still didn't "solve it (totally) on his own" because he still depended on the previous work of others to build on. Likewise, we can say Andrew Wiles "solved Fermat's Last Theorem" but Wiles acknowledges that seeing Ken Ribet's proof of epsilon conjecture was a breakthrough he used.

>In another, Carbon-based Charlie breaks into Bob’s lab, copies enough of Bob’s notes to understand Bob’s work, and provides the final insight to solve the open question. In this world, it’s fair to say that Charlie and Bob both contributed to solving the open question, but I think also fair to say that Charlie didn’t have the capability to solve it on his own.

Again, using gp's framing, Bob also didn't have the capability to solve it on his own. By omitting the previous papers and prior works that Bob built on, it makes your hypothetical scenario incomplete when judging Bob vs Charlie.

We don't have an objective standard of how much the "standing on the shoulders of giants" applies to each breakthrough. There was a blog post (might have been Terence Tao) that said society unfortunately awards the fame to the person who solves the last step of a proof and forgets about the people who solved the intermediate steps that led up to it.

I'm using "normies" to describe people who are not developers and they would not install something like OpenClaw on a Mac mini.

E.g. the OpenAI's 2-minute marketing video for Dot trying to show non-developers using AI assistants to get tasks done: https://www.youtube.com/watch?v=uXspbC2srEQ

A man wants his website changed, and a woman wants her presentation slides updated. "Serious engineers" don't need cutesy re-incarnated mascots of Microsoft's Bob or MS Office Clippy paperclip assistants. OpenAI dots doesn't have to win over HN techies who are comfortable with CLI tools.

Take off your "engineering" hat and put on your "normie" hat...

OpenAI's Dots, Meta's Muse, xAI's Grok Bot, etc are trying to target non-technical people and give them "24/7 AI personal assistants". Apple is also pursuing this angle by adding more features to the new Siri (partnership with Google-Gemini).

Perspective of product and market : we need our AI to be more deeply intertwined with customers lives instead of just being on-demand chatbot Q&A sessions. The "sticky" product to create this customer relationship is "AI personal assistants". Instead of just providing "answers" (chatbots) -- provide "completed tasks" (AI assistant).

Perspective of technical predecessors and concepts overlap : OpenClaw and Hermes self-hosted software on Mac minis and Linux boxes as "personal assistants" is geared for techies instead of normies. Instead, repackage those types of tools for normal people in cloud VMs to be hosts for "long-lived agents". Zero install required.

The above scenario of fixed time regardless of DST/TZ changes is what I tried to call "cultural/political time". In other comments, I called it "appointment time".

What you call "calendar time", others will call it "time with calculated UTC offset". (Which then leads to more meta discussion of "no... calendar time is not UTC offset because ..." )

Both examples of our ambiguous labels causing more confusion is prime example of the industry not converging on good names to make devs aware of the difference.

>I think practically what matters for developers are three things:

That categorization is fine but is still obscuring the key issue: many developers think they can collapse all of your 3 types into one simple strategy of "always store it as UTC"

I happen to call them "scientific datetime" vs "cultural/political datetime". However, the software dev industry has not converged on a standard vocabulary to delineate the 2 types which is unfortunate because that means programmers are unaware that the difference exists. Concepts are more top-of-mind when there are good names to label them.

If a programmer doesn't understand how the 2 datetimes behave differently, they will create software bugs as I've outlined before: https://news.ycombinator.com/item?id=39418897

There's the meme of "store UTC everywhere" (maybe perceived as correct because of superficial similarity to "use UTF-8 everywhere") ... but storing datetimes as UTC is only unambiguous for historical events such as timestamps of activity stored in server logs.

But future datetimes can have ambiguous edge cases which causes the split into 2 different types.

That hybrid setup can deliver very satisfactory sound but for some guitarists, they also want real tube power amps because they behave differently than solid-state power amps.

The tube preamps give the distortion fuzz but the tube power amps also provide the extra dynamics that makes the whole response under the fingers more spongy, velvety, etc. Connecting the preamp outputs of a Mesa Boogie or Soldano into a solid-state amp has a different feel than tube power amps. The solid-state amp is definitely usable... but it's different.

That said, I sold all my tube amps and just use the stock digital emulations of guitar amps in Logic Pro. I got rid of almost all the analog gear that was heavy and required maintenance (e.g. new replacement tubes cost $200) because I didn't want it weighing me down as I got older.

GCC 11 (2021) std::visit was slower than virtual dispatch.

GCC 12 (2022) optimized std::visit so it can be faster than virtual dispatch.

https://shubhankar-gambhir.github.io/posts/your-stdlib-imple...

Apple doesn't see individual transactions when Apple Pay is used at retail stores' tap-to-pay terminals. The secret card payment token is sent from the phone to the credit-card's issuing bank and bypasses Apple servers. In this way, using Apple Pay is more secure and private than plastic cards because the real card number details remains hidden from the merchant.

The iPhone does contact Apple servers to add a new card to the digital wallet. Apple servers then contacts the issuing bank to get the secret token the bank generates and then puts it in the digital wallet. Conceivably, the "add a new card to digital wallet" could also have been done without Apple in the middle but it would require a much more convoluted, less secure, and more user-hostile workflow to do it. (e.g. the end user would have to know what bank endpoint to contact, manually enter the long and cryptic digits of the secret token, or maybe scan a QR code on a computer screen that's vulnerable to interception and phishing.)

When people say software is "expired", "obsolete", "legacy", "limited shelf life", etc ... they're not talking about whether somebody somewhere in the world has the ability or idiosyncratic desire to continue running that old software.

Instead, it means the software's ability to be practical and usable in relation to the rest of the current computing ecosystem. E.g. In this perspective, Win95 MSPAINT is "expired/obsolete" because it can't open .PNG and .JPEG files. In today's world, .PNG/.JPEG files are much more common than the .BMP files that MSPAINT 95 can open.

I still have an old 16-bit Microsoft C/C++ v 7.0 compiler from 1992. Did it expire?!? Well, not literally. There's no ticking time bomb in it that checks the system clock and refuses to run. Instead, the issue is that it doesn't have the newer C99/C++98/C++11 features and I can't use it to compile today's ffmpeg source code. That's what people mean by "expired/obsolete".

Appliances with mandatory wi-fi is a bad trend. The mid-tier dishwasher from Bosch locks the Eco(nomy) energy saver cycle, delayed start, etc functionality behind a smartphone app. This requires connecting the dishwasher to wi-fi and using the smartphone to create a Bosch online account. That's what the whole Jeff Geerling complaint video was all about.

https://www.youtube.com/watch?v=5M_hmwBBPnc&t=4m21s

A person's perspective on the older dev tools being "superior" to today's newer tools depends on the things they want to build.

I had jobs working on Foxbase/FoxPro for DOS and migrated those apps to FoxPro for Windows, and Visual Basic 3.0 to 6.0 ... and I don't look back at those with the same positive nostalgia.

Classic VB was "simple" but every project at each company had a different hodge podge of 3rd-party VBX/OCX add-on controls to do anything non-trivial. E.g. 3rd-party calendar date picker, report writer, etc. VB didn't have a decent datagrid so everybody bought a 3rd-party one. In contrast, C# WinForms standard controls included a usable datagrid with data binding out of the box.

In a FoxPro for Windows project I was working on, we needed to open .zip files. FoxPro didn't have that functionality so the company paid an extra $300 for the FoxPro Library Construction Kit[1] and I hand-coded a .zip file parser in C Language to be called by FoxPro code. Instead of FoxPro dev work being simple, it's complex! Today's Python or C# can trivially open .zip files with builtin modules and functions so devs don't waste time on writing low-level C code or use 3rd-party libs.

With old 1990s tools, you can't do "simple" tasks such as getting a UTC timestamp, use a regex expression, or retrieve a web page with an http call, etc without calling Win32 API functions or 3rd-party tools.

If one wants to do things beyond the limitations of classic VB/FoxPro, the newer dev tools like Python/C# etc absolutely do make things easier and faster.

[1] https://www.scribd.com/document/384159949/Ms-Foxpro-Develope...

https://stackoverflow.com/questions/53401654/why-was-stdbit-...