I worked on the censorship and government reporting (sending all logs) infrastructure for Akamai China CDN. I'm glad to see it get shut down. Happy to answer questions.
Thanks for the disclosure. Don’t feel too bad, those stuff you helped built may not cause much damage as you feared.
CDN used to be at the front line of blocking content and surveillance citizens, nowadays that happens mostly thru social networks.
Unlike other countries, Chinese citizens are ultra online, and mostly concentrated on two platforms Webo and Wechat. Most other online services all require authentication via either of these two, plus cellphone number. So for govt it’s very simple to block anything or see any identity, CDN hardly play any major role anymore.
1. How was working with China requests and logging, differing from working with other nation states?
2. Was there full services brought up only for China specific needs? What would they take care or?
3. How would any blocks work? allowlist or denylist? Was takedown immediate, or was it working with the customer/client and getting them to take it down within SLA?
1. At the time it was the only nation state that had specialty infrastructure except for maybe the US.
2. There were specific infrastructure changes made for blocking and sending logs inside mainland china.
3. The CDN node would deny access to specific urls uploaded by the Chinese partner company. I don't remember the SLA. The SLA for reporting visited URLs was 15m IIRC
A friend mentioned this a few weeks ago during a conversation with someone in the industry. Apparently, it’s because the China government directly mandated it, so Akamai had no choice but to transfer their business to domestic China providers. For example, PSN is now being handled by Tencent.
My enterprise sized day job is an Akamai customer. Nothing in China directly so we won't be directly impacted.
As a hypothetical, if we got told that we had to switch providers to stay in a region, we'd need to rebuild pipelines, EdgeWorkers, edge caching rules, origin routing configurations and probably more I'm not aware of. Plus testing all of those changes in a non-breaking way across the entire enterprise. Along with all the normal business delivery priorities.
It'd probably take a solid year for us to fully execute it.
This. People tend not to realize how sprawling enterprise software stacks tend to be, how many implicit dependencies have to be untangled, etc. Even simple things can take years and complicated things often just don’t get done at all.
Yes, dealing with the mess that is your software stack, the mess that is your corporate structure, and the mess that is your change management process means that things a couple dudes in your startup could accomplish in a week would take a year at a crusty enterprise.
There's also a finance process. Akamai deals mostly with enterprise customers, which means step 1 may be technical validation, but step 2 is negotiating an appropriate contract with another provider, which may take weeks on its own without a clear go/no-go answer in the meantime.
It's actually the opposite. I thought the same thing before working in big enterprise though so I definitely understand how you could think that.
In reality everything takes 10x longer because things are done in a very thorough way and typically with significant redundancy (high availability). The code bases are typically shite and personally I'd rather eat nails than work on them, but they are reasonably well tested and changes are typically done very conservatively. Big enterprise devs are also really good at not breaking production. As much as I detest that environment, I do think startups in general could learn a great deal about not breaking production from the big enterprise people.
What I meant to be dependent on a single essential service that much with that much difficulty to overcome. Smells like lots of trust and hope (of things will stay as they are for long) put into the architecture. It is nice there is a workaround, and there is a will for workaround, in this situation for instance.
No, quite the opposite. Big companies likely also have other big(ger) companies as partners/customers, all of which want stableness and see things keep working. Therefore companies need careful planning, execution and testing to ensure there is minimal disruption.
Startups and a certain company can move fast and break things. But not everyone can do this.
It’s the difference between all hands on deck for 6 months and a reasonable pace over 18 months.
If you are just running a single website with DNS fronting that’s not an issue.
But large customers tend to have more advanced connectivity on L3/4 and asymmetric routing.
Then there is the CDN part itself are you only using the basic auto caching? That’s not a problem but if you manually manage it then all that needs to be converted as well and there is no guarantee that the partner API would be compatible or even have the same functionality as your current CDN.
It's specifically because Akamai wants to preserve its reputation amongst enterprise customers paying $$$ that it's giving such a big delay. And I can predict it won't be enough for many.
No doubt there be a lot of companies who don't finish until nearly the end. Barring legal reasons, I'm guessing there will actually be an extension because enough corps won't be ready at that point. Also would expect Akamai to offer extended support beyond that date (for a significant cost) on a customer by customer basis.
CDNs are much more than dumb http1.1 compliant content caches these days. And every CDN has a huge number of integrations and functionality. All of which have a different impleme tation and behavior for different providers. Its probably a good analog to an infrastructure (“cloud”) migration. And even harder to test, validate, and switch the actual service provider as they _are_ “the front end.”
Because Akamai is substantially more than a CDN (arguably, CDN is now a smaller – albeit not small – part of their business): it is also certificate management, WAF, web app/API protection, IAM, edge DNS, edge workers, complex CDN rules, analytics and a whole bunch of other stuff.
Enterprise customers also typically and mostly use Akamai in non-CDN scenarios so they will have a hard time migrating off it if a need be, especially if they have invested heavily in Akamai.
Akamai is also a CDN that [works for domain fronting][1], which is an amazingly reliable way to get around the Chinese Firewall via [Collateral Freedom][2].
Tor's meek used Azure for domain fronting.
Akamai likely faced some pressure from the govt because of this.
I remember the days when jsDelivr suddenly lost its ICP licence, and therefore got an immediate ban in China Mainland. It used to be a really reliable CDN service for both China and abroad. Then it suddenly failed.
PS: China now allows foreign entities to run CDNs, ISPs and IDCs in specific cities, starting from Apr 2024. This could be because something else.
I see a lot of US organizations decoupling stuff from China, especially if it's fairly peripheral, such as pulling back from running events there. But there's also a lot of more fundamental second-sourcing and the like.
Well depends on the organizations. Some of the organizations aren't actually pulling away from China, as they're just getting outcompeted in terms of price/quality by local companies. To my understanding, government is trying to convince people that "Chinese products can also be luxury", and some people are switching.
Or it could be part of general trend of splitting off internet and creating a firewalled chinese internet. PRC is explicitly favoring this outcome for decades now with explicit incentives via legal, financial and social routes.
They already did that ages ago. Any web tech company that wants access to Chinese market needs to play ball with PRC (just like what Apple and Microsoft does). Average Chinese person really doesn't care about global web, and the ones that really do, figure a way out through VPNs. I'm not Chinese, so I might be totally wrong, but that is my perception of talking to expats or people who still in the country.
Organizations dealt with certain "weirdnesses" (the great Firewall) for a long time. But my sense is that over the past few years there's been an increasing sense of pulling back from all this.
It is going both ways. This is less about censorship and more about having a full local and independent supply chain. The two are going to war unless one of them collapses first or the US cedes Asia to China.
The writing has been on the wall for over a decade, at this point. The only companies that "risk it" with China are the ones that rely on borderline (or in some cases, literal) slave labor to maintain their margins. If you don't have an outstanding manufacturing investment to honor, it's a net-negative reliance in many cases.
This is not remotely accurate. Plenty of large tech companies still play in China. The difficulties are very high but if you are big enough the huge market size it’s still very profitable. I don’t think google/etc are in China for its slave labor.
Certainly, depends on the company/organization. If working in China is a big win then many companies will decide to continue doing so while preparing other options. If it's more marginal, then starting to carefully pull out probably makes sense. Which is more or less what I'm seeing. I'm not sure about ten years but the situation over the last five has started to become pretty clear.
It's not just tech companies, many entertainment companies (e.g. video games, media) do the same, even at the risk of alienating or angering Western customers
> Plenty of large tech companies still play in China
Depending on the segment you are in within the tech industry, the Chinese subsidiary of the foreign company might be a white-labelled Chinese offering (eg. AWS China, Azure China) or entirely unique IP developed in-house by the Chinese subsidiary.
AWS and Azure in China are the literal opposite of white labeling. It's still the global product, branded under the global name, (mostly) operated by the global organizations.
The affiliation with Chinese companies is borderline fine print, white labeling would be if you had "tencent cloud" which happened to behave exactly like Azure.
For many things, it was never a particular net positive. At this point, it makes sense to cleanly and quietly withdraw. This particular release isn't especially quiet but many companies are just stopping doing activities at convenient points.
That's a pretty one-way relationship though. Australia and China are literally thousands of miles away from each other, but China is still important to Australia because all the continental landmasses are far from Australia.
Well unlike most countries Australia and China have a lot of two-way trade (though imports to China are double exports) because China imports lots of raw minerals (as well as food) from them, and noone buys as many raw materials as China does. And thousands of km is still much less than twenty thousand to Europe.
That's not something that's currently planned, and even if it did, it would take years. Besides, it wouldn't apply anyway; Akamai's China business was already routed through a mainland China company, as required by Chinese law.
This can often occur when parts of the application are out of their control. For example, geographical apps are hard to recreate locally on the scale of Google or Apple Maps. It's vastly easier and sometimes cheaper to use their API. But if China has any disputes on how it's represented in the geo data, the software becomes a liability.
> We've had CIFUS for decades. The only change is now we say no to Japan teaching us to make better steel.
I'm curious what you mean by that.
Most of what I know about steel is in relation to knives (and peripherally, swords, axes, and other tools). At least in that (admittedly small) niche, Japanese steel is good, but certainly not superior to US steel (although with the financial trouble that Crucible is having, perhaps things will change a bit).
I used to live in a rust-belt steel town. My knowledge is cursory except for my lived reality of my town being economically depressed by poor management of steel production facilities.
US Steel mostly focuses on commodity steel.
Nippon Steel has a wider portfolio focusing on higher-end specialty steel.
Commodity steel producers like US Steel (they acquired the main steel company in my town) generally play on lower end of the market, and have not invested adequately in technology or modernization. They are not able to make specialty steel because they have not made those kinds of investments. Instead they have chosen to languish with low productivity practices, so they're being outcompeted by foreign companies who are now able to make the same quality of steel but cheaper. They used to laugh at foreign steel companies because of the trash steel they made, but they laughed a bit too long and didn't realize that foreign steel gradually got better year over year. I saw this hubris play out first hand in my town.
My town also has another steel maker, a specialty one. This one was smaller but chose to invest in automation and modern technology, and also on producing new types of high-value steel for specialty applications. They had a profit-sharing scheme and had loyal employees who did not reject automation or modernization, but instead had a growth mindset. They remain competitive today.
This is capitalism's creative destruction playing out, and companies like US Steel that refuse to adapt find themselves outcompeted by more productive players.
The Nippon Steel takeover, even if it had succeeded, might have been too late.
I don't have lived experience with Nucor, but some years ago I read "Plain Talk" by Ken Iverson, who was Nucor's CEO. I don't have context but I remember it being a breezy, folksy management book, and one that had high believability because it came from the horse's mouth -- not some management consultant. It doesn't mean there's no hagiography, but at least you get more practicality out of CEO-written books than management consultant fluff.
Nucor today is the largest steel company in the US (and more profitable than US Steel). I believe they too produce commodity steel, but they were able to stay competitive because of good management, investment in updated technologies like electric arc furnaces, and focus on the quality and service market segments that foreign steelmakers have a hard time competing in.
It was quite obvious to everyone for a long time that commodity steel production will be outcompeted by countries with lower wages, lower energy costs or both. US is not competitive on wages for commodity products, so investing in mass production of regular steel was a bad idea.
Specialty steel has a barrier of entry - not every third world country can make it. Even today China imports most of the bearings for high-speed trains and planes. This grants high margins to the few companies that know how to produce this steel, this is where US can compete.
So what happened there was normal business decisions, not "creative destruction playing out". Economy beats politics for once.
> The only change is now we say no to Japan teaching us to make better steel
Blame Cleveland-Cliffs [0] - Nippon Steel's Brazilian-owed competitor in Ohio.
> Lourenco Goncalves, CEO of steelmaker Cleveland-Cliffs (CLF.N) which made a failed $7 billion bid for U.S. Steel in August 2023, participated in at least nine calls assuring investors that President Joe Biden would scuttle the Nippon Steel merger months before he did so on Friday, according to summaries of investor calls included in a Dec. 17 letter from lawyers for Nippon Steel (5401.T) and U.S. Steel to the Committee on Foreign Investment in the U.S. (CFIUS) and confirmed to Reuters by two participants in the calls.
> “I can’t force U.S. Steel to sell to me, but I can work my magic to make a deal that I don’t agree with not to close," he told investors on a March 13 call hosted by JP Morgan, the letter quoted Goncalves as saying.
> While Goncalves made similar comments about the deal to analysts on three earnings calls this year, his private remarks made throughout 2024 about the deal process show the extent of his effort to cast doubt on Nippon's bid for U.S. Steel. His comments sometimes preceded drops in the U.S. Steel share price, Nippon Steel and U.S. Steel told CFIUS.
China doesn't want to be an opium den for western powers and have their economy underdeveloped by their firms. I think that's okay and it has been paying off. That's why China is able to develop its economy to superpower status while so much of the rest of the world is locked in dependency on the whims of G7.
its really interesting how afghanistan became the world's #1 poppy producer under us occupation and it declined immediately after we left. heroin prices rose, leaving a market opportunity for fent. i still do not believe that china is necessarily the source btw. that needs proof.
Ya, I don’t see that lasting very long since China production is too cheap to compete with. They’ve basically taken over the whole illegal drug business here in the USA (opium has never been a very big drug in the states). I don’t really know what China is trying to prove though, is the PLA involved or is it just a few big Tianjin SOE pharmaceuticals?
It's not lasting it's over. The US pulled out of Afghanistan in 2021 and the Taliban vastly reduced poppy cultivation. That's what caused a sharp rise in heroin prices and a demand for a substitute.
I haven't looked into fent, but given the information is coming mainly from US security agencies who consider china an enemy state, I need to see physical evidence. I won't accept their say-so. This narrative about China smells suspiciously too perfect. China was the subject of 100 years of humiliation, was drugged to death by Americans, British, and others. Now the US is claiming they're doing it to us? The abuser is suddenly claiming we are the symmetric victim? Fishy story.
Great for them. And I guess you are going to tell me China’s fentanyl production and export to the west is in revenge for the opium war that happened in the 19th century?
This has been going on since 2019, fent took over long before the USA left Afghanistan and the Taliban retook control again over poppy production (they were effective at exporting before, so who knows what is going on right now). My guess is that that market doesn’t exist anymore and illicit drugs are controlled by cheap imports of fent from China.
Unlikely. If the precursors are being produced in China, it's just capitalism at work. They don't need to do this kind of stuff, their economic growth is double ours. They just need to mind their house and the world will change in their favor.
> My guess is that that market doesn’t exist anymore and illicit drugs are controlled by cheap imports of fent from China.
That's an interesting take.
EDIT: I am reading some reporting on this.
It seems according to mainstream accounts, most fentanyl is allegedly produced in Mexico from Chinese precursors. It is further added that a significant or all of these precursors are difficult to interdict because they are used for making ordinary consumer items.
"Seizing the precursor chemicals that traffickers need is extremely difficult. Many are used to make basic goods, such as medicines, pesticides and soap."
So one interpretation could be (requiring a more thorough investigation of the precursors list) that China simply produces ordinary bulk chemicals used for many purposes and sends them to people that buy them.
That is because the Jones Act in context to the gross entirety of the USA and PRC economic relations is at best utterly irrelevant. Besides, you might find it interesting to learn that the protectionist parts of the Jones act are in fact common around the world. Many counties prohibit port to port shippments without first visiting an international port. Mind you, many folks here would agree these kind of laws are stupid, and I would personally agree... However, that does not stop this talk about the Jones act from being a Red Herring. I'm glad you found a topic of great interrest to you, and that's great.
What does this have to do with Akamai, and many other American companies, ending services in mainland China? Nothing... What does export tarrifs on Chines exports have to do with Jones act? Very little, asymptoticly little... to the point one might say the topics are orthogonal, or more plainly they might say Red Herring.
Well, I'm glad that we can at least agree that the US has protectionist policies and has had them for a long time.
If you want a bigger protected industry, perhaps checkout any Farm bill from the past several decades. But that's neither here nor there. China is not weird for having protectionist measures - like you said - it's pretty normal - the scope and size is what may be unusual for a country that's not a hegemon.
> What does this have to do with Akamai, and many other American companies, ending services in mainland China?
Glad you asked: China has protectionist policies that are designed to help its domestic tech companies. Consequently, foreign (read American) tech companies like Akamai may find the operating environment much more challenging that they would without those policies. Other non-protectionist policies don't help - such as working with the great firewall, or complying with censorship requests.
TL;DR version: China protects its tech industry like the US protects its farmers and merchant marine. Acknowledging that will help frame whether the respective countries have the political appetite to temp or undo those policies.
People see “framed” facts hammed down their gullet for so long, accept this, be ok with that, this is the new normal even if it’s only “because we say so”, that when an other option is presented, and one isn’t ordered to “accept this because…” it resonates.
I hope you appreciate the irony of your white-hot rage against framing, and the contrast of you saying:
> this is the new normal even if it’s only “because we say so”, that when an other option is presented, and one isn’t ordered to “accept this because…” it resonates.
> Surely you realize this.
Why aren't you just stating the facts and letting parent decide? I don't think you are against providing relevant context (i.e. "framing"), you just have a bee in a bonnet about ... something - journalism as practiced today perhaps, but that has very little to to with the topic at thr root if the thread - until you reframed the discussion to be about journalism.
>China never has and never will offer a level playing field for non-Chinese companies.
Foreign companies in China have to play by Chinese rules, which are more restrictive than the US but so do Chinese companies. When it comes to doing business abroad China imposes very few rules or conditions. Meanwhile the US has been waging a straight up trade war on China with the not just implicit but stated goal of strangling China's technological progress. If anyone should level the playing field it's the US.
This is how this is perceived in the world outside the Anglosphere, it's astonishing how out of touch the American centric commentary on this is. China's open to do business with the world, the US and Western countries are politicizing trade.
Chinese companies don’t have to play by the rules in the same way that foreign companies do. China isn’t a rule of law country, and has explicitly stated that it doesn’t want to be a rule of law country. Instead laws are enforced selectively for whatever outcome the official branch wants at the moment (rule by law). They are perfectly willing to fine American companies when Chinese companies cheat (and not apply anywhere near the same enforcement to the Chinese company that actually cheated).
Foreign companies have to be heavily aware of context just like Chinese companies do, but they will also be treated very differently in how and when laws are applied (not always to their detriment, China will elevate some foreign companies to show they are fair and open to FDI).
>and not apply anywhere near the same enforcement to the Chinese company
About two years ago China imposed regulations on its own tech sector so strict that it killed the entire "ed-tech" sector overnight and wiped out a decent chunk of the country's tech market value. What other country has done this to its own companies?
The US has slapped a 100% tariff on cars, bans on Chinese hardware and software in autonomous vehicles, ripping out 5G equipment, tried to basically destroy Huawei specifically, and the list goes on. Whatever local favoritism may happen in China, you realize that pales to the full blown protectionism that the US engages in. China hasn't even retaliated in kind. You're complaining about being poked with a stick while throwing a sledgehammer around. Am I supposed to believe the "TikTok ban" is anything other than completely absurd security theater? Like do you think the world is so naive they think this is how the "rule of law" works? You realize every time a Western politician utters that term everyone just laughs right?
China has had 20-50% tariffs on car imports since WTO (to protect their car industry, before it was worse!), and it has been successful: they coaxed most foreign car companies into JVs where they could transfer IP. Now that their car industry is secure, to complain about protectionism from the countries they once protected themselves from is just too hypocritical.
Likewise, if you tell me America banning tiktok is unprecedented you’ve never heard of Facebook or YouTube? Again, China has no moral high ground here. If anything, I bet the Chinese government is just wondering what took us so long.
> About two years ago China imposed regulations on its own tech sector so strict that it killed the entire "ed-tech" sector overnight and wiped out a decent chunk of the country's tech market value
Unless you can point to a foreign company that was treated less harshly under these regulations than a Chinese one, this doesn’t disprove the parent’s point.
> The US has slapped a 100% tariff on cars, bans on Chinese hardware and software in autonomous vehicles, ripping out 5G equipment, tried to basically destroy Huawei specifically, and the list goes on.
Besides the car tariffs, these measures were taken to prevent Chinese hacking and intelligence gathering, not to protect domestic industries.
>Unless you can point to a foreign company that was treated less harshly
Less? Foreign companies are supposed to get privileges? My point was they're treated equally by and large. The parent should substantiate their point, how does one disprove an unsubstantiated accusation? I'm genuinely not aware of a company that, as long as they played by domestic rules, was mistreated. Tesla operates in China, they get the same subsidies Chinese EV makers get, as does every German car company.
The other user pointed to tech companies like FB, YT or Google, but they didn't leave because they were treated unfairly, they left when they didn't want to comply with Chinese law. Which is fair enough from a value standpoint or if Americans pressure the companies into it with for example Dragonfly at Google, but that doesn't mean you're treated any worse than domestic companies.
And yet he’s not blocking TikTok. It’s quite astonishing given:
A) they’re a Chinese owned company
B) China asymmetrically has never allowed US social media (amongst practically all major internet properties)
C) We cap ownership of foreign owners of US media. TikTok IS TV for the young generation.
But apparently he thinks it helped him win (would love to know if that was actually true, which is probably impossible to calculate), so who cares about all these other principles!
Genuine question as someone not from the US, wouldn’t that make a significant part of a generation mad at Republicans? Some who might be close to voting age or might even be able to vote already.
Trump isn’t up for re-election, and he’s isn’t interested in the long term performance of the Republican Party (he only became a Republican because that’s the party whose primary he could win).
Both parties are owned by the same oligarchs and security state. The owners don't care who you vote for, the system is a political World Wrestling Federation.
This has nothing to do with democracy, it's about trying to keep a level playing field. The US has been trying to open trade with China since the 70s, and until China stops doing this the US has to adjust trade policy to keep things fair. Free and open trade is a two way street, otherwise it's just charity to the bad faith actor.
That's because China has strongly failed to follow trade agreements under WTO, including the TRIPS agreement (which is an international agreement protecting IP rights), SCM agreement (which prohibits state subsidies of certain trade goods), the National Treatment Principle (which mandates equal treatment of imported and locally produced goods), National Treatment and Market Access commitments (China strongly restricts foreign ownership in certain sectors of their local market, including the tech industry), etc. Mind you most of these agreements were made in the 90s, so China has had plenty of time to comply. It's a bit pointless to keep using the WTO when a country isn't going to comply anyways.
I'm not saying the US is blameless either, but with China, it's nearly shameless. In the tech industry alone, Youtube, Facebook, Instagram, WhatsApp, Twitch, Steam, LinkedIn, Dropbox, ChatGPT and Copilot, Github (partially), etc are all banned in China. Shoot, Google left a decade ago after they kept getting hacked by the Chinese government. When China was still quite poor and developing, this behavior was overlooked, but as they become a rich world power on the global stage, that's not the situation anymore.
The US has a long history of abusive trade relationships. China has seen what the US has done in Latam (and possibly in Africa) and said nope. US is used to have everyone play the US game when trading in Latam and other developing countries. China has been making the US play the China game when trading with China. And now that China is a serious world power, it is possibly a bit late to stop playing.
> whosoever shall smite thee on thy right cheek, turn to him the other also (Jesus)
> I do not like the word tolerance, but could not think of a better one. Tolerance implies a gratuitous assumption of the inferiority of other faiths to one (Gandhi)
I guess the author alludes to the paradox of tolerance first described by Karl Popper 1945 in The Open Society and Its Enemies.
Sometimes deterring punishment is the only thing that keeps people from abusing tolerance or freedom to destroy that tolerance and freedom, but it needs to stay a measure of last resort.
It’s interesting indeed how tolerance and freedom interact with each other’s. To put another perspective, here’s a quote from an unrelated text (0)
> Middle Eastern Muslim culture expert Marvin Zonis notes that Arab societies value the honor and dignity of the individual more than personal liberty.
I 100% about the benefits of (sometime) punishments, but also perhaps a bit less liberty to make space for other values might have benefits. I’m not sure where to place the slider on that scale, as too much honor and dignity can be quite restrictive.
We don’t. But if we keep throwing companies into the grinder like this where all their IP is stolen and they are just replicated in China on an unfair playing field, we end up losing anyways regardless of intent.
Another way to go is not to participate in the hype cycle of pitching investors instead of pitching customers (marketing and sales) and giving away ideas and opportunities that end up being made elsewhere and not supported where the founders currently are.
Doing the market research and opportunity and presenting it on a stage can be very helpful sometimes, other times it can seem a little head scratching.
Does this also mean exploiting their cheap/slave labor for manufacturing and importing materials is also off the table?
Very naive to think the incoming US administration would do anything to help the middle class. He sold voters as “America first”, yet his actions speak for the billionaire class.
In an ideal world the answer to your question is yes, we should not be exploiting slave wages/labor. But most of us are simply not in the position to make that decision.
And the webs of manufacturing middlemen make it difficult to even know at what point in the chain would this practice be happening.
Not to mention the average person would see nothing more than prices going up and complain with 0 knowledge of whats happening downstream
Previous discussion about it: https://news.ycombinator.com/item?id=33678019
reply