Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Amazon Unveils Futuristic Helicopter-Plane Hybrid Drone for Deliveries (bloomberg.com)
152 points by jason_zig on June 5, 2019 | hide | past | favorite | 114 comments



> Its propellers have a slight S-curve designed to lower their noise to make them more acceptable for use in populated neighborhoods, Kimchi said. The craft is far more stable in high winds than more traditional drones, he said.

Noise is my main concern with drone delivery networks. I have been playing with radio controlled aircraft for most of my life. I see drone delivery as inevitable, but living near a noisy drone delivery route could really lower quality of life for the people on the ground.


> Noise is my main concern with drone delivery networks.

What about 20-50 lb flying machines falling onto sidewalks, roofs, and streets?

You don't see amateur RC aviation enthusiasts flying their model airplanes in the middle of the city, and I sure as hell don't want to see a corporation doing this on an industrial scale.

Why not ground-based drone delivery?


This probably isn't for cities. I was at a conference a few years back where Prime Air gave a talk. It was quite a vague talk, but admittedly it was hilarious watching Takeo Kanade grill the speaker on technical details that he refused to elaborate on. Things have no doubt moved on, but they were talking about use cases like delivering to rural areas where people have big gardens to land in. So potentially places where sending a delivery truck miles out of the way would be expensive (e.g. a 20 minute detour to one house). That's where the interesting applications of drone delivery are.

That drone is pretty big, and they're getting extra lift out of the wings which conveniently double as prop guards. Flight time is presumably at least an hour, if it can get something to a customer in half an hour (and has to get back). You'd need quite a few since each drone can therefore only make around 10-20 deliveries a day if they swap out the batteries when it lands. 30 mile round trip range is pretty decent. Also on size - you need somewhere to land it with a few metres wriggle room either side. There's no way that's going into a built up area.

There's no point doing drone delivery in urban environments, the safety case is insane and you're competing with excellent existing delivery channels.


Why are you downvoted? Safety is a thousand times more concerning - the drones will mechanically fail, they will hit bad weather, they will collide with birds, there will be new drone delivery market entrants and will need coordination across systems... If a drone fails at an expected operating altitude, it can kill someone easily on the ground. If it fails at height over a highway, what if it kills a bus driver or causes a tractor trailer to lose control? What if they hit electric lines? What else can go in flight corridors at the same altitudes? Do they container chemicals or voltages that first responders have to be aware of?


There's a reason why it's taking so long for Amazon to get FAA approval for drone delivery and this is exactly why. There are people in important places asking these sorts of questions already.

My guess personally, is that drone delivery will ultimately just be a tool for a self driving truck to get the package from the truck to the door. Drones are not nearly good enough as it stands today to do end to end deliveries at an industrial scale outside of maybe a few rural areas.


The same people in important places that approved the 737 Max?


In the sense that they are human, yes. Even in the sense that they work for the government, which narrows the pool to 20,000,000+

But also, no. Not the same people.


I get your point but the relevant agency division FAA AVS employees 7,200 people [0]. Fewer than that are involved in certifying aircraft safety. Agencies like all other human groups have cultures.

[0] https://www.faa.gov/about/plans_reports/congress/media/fy17_...


Wasn't a critical issue with 737 MAX that incumbent manufacturers are able to self-certify equivalency without any review?

So, to the extent the Boeing 737 MAX team is not reviewing Amazon's applications, no, not the same people.


Yes ... so they have experience in crashing planes, which is relevant. /s


I don't think people fully appreciate the carefulness with which the FAA has approached this, nor the full scope of the scenarios they have considered.

If you look at the first FAA waivers to operate drones industrially -- 5 years ago! -- you will notice that all of the approved drones are military drones. Why military drones? Because the approved drones often had a decade or more of operational use in explicitly hostile environments and very poorly controlled conditions. Being able to reliably operate in those conditions was considered strong evidence they were unlikely to start falling out of the sky when operated under much better controlled environments and less risky conditions.

The challenge for many new drone builders is that their platforms are not built anywhere close to the reliability and resilience standards of military drone systems, which are typical the product of companies with proven aviation/avionics systems engineering experience and bankrolled by US defense spending. Even when they build systems that on paper meet this level of engineering, the lack of operational history and experience puts these drones on the slow path to approval.

Amazon getting the FAA to sign off on industrial use of an exotic drone airframe with no operational history strongly suggests that its safety and resilience characteristics are excellent, at least comparable to drones designed to survive military environments.


The comment is uninteresting and rightly downvoted because (1) it's been discussed to death and has nothing to do with this particular experimental drone model and (2) we already have effective legal apparatuses to deal with liability.


Neither of those are reasons to not discuss this more. We, as a society, prevent all sorts of things from happening without having to just wait for a lawsuit to settle things or whatever.


One has to actually make an argument for the liability system being insufficient. Merely pointing out risks is not enough, especially since no one here has bothered to even do an order-of-magnitude comparison between the risks and the benefits. It's just noise and emoting.


Benefits: You get your stuff quicker from Amazon

Risks: See everything else

Do you really, like I mean _really_ need your stuff from Amazon same day, hours after ordering it? No, you really don't. Nobody does. It's a convenience, a nice-to-have, but is totally unnecessary.

So, I think the burden of proof lays squarely on Amazon and it's proponents of this type of service, to prove beyond any reasonable doubt the risks can be mitigated to the point where the public accepts them as a trade-off.

Having potentially 50lbs of drone fall on your car while you're driving home from work is probably not going to be acceptable... to name just one risk that will have to be proven to be minimized. Or noise levels will be controlled near neighborhoods etc... Other industries have to prove their new thing is safe and acceptable for the public... why should Amazon be any different?


> Do you really, like I mean _really_ need your stuff..hours after ordering it? No, you really don't..It's a convenience

The same could be asked of 2-day home-delivery 20 years back, or 2-day world-travel 200 years back. Yet, here we are.

If necessity is the mother of invention, laziness is the father.


Yes, that's an excellent imitation of the sort of comment of I was complaining about.


Not to mention people's heads! I have faith in agencies like the FAA to keep things safe and reasonable, and where that isn't enough NIMBYs and municipal governments will pick up the slack, no doubt.


> I have faith in agencies like the FAA to keep things safe

Has this faith been shaken by the recent 737 Max issues?


No. Air travel is still incredibly safe, a few high profile accidents notwithstanding.


Interesting point. Maybe instead we should be using self-driving cars with drones (or something ground-based, like RC skateboards) only handling moving packages from the car to the door.


This exists: https://nuro.ai/


It can't be worse than the cars and trucks it will replace, which kill 40,000 people every single year in the USA alone.


How many people do Amazon delivery drivers kill/year? Pretty sure it's not 40,000.


In addition to noise, I'm also concerned about visual clutter, particularly in residential areas.

The last thing I want is to add to the power lines, airplanes, etc. with the visual noise when I'm on my deck trying to enjoy nature.

Further, Amazon is a major player in the advertising space now. I'd be surprised if at some point these weren't leveraged to display ads in some format. I'm a marketer by trade, but personally that's a line I don't want crossed.


Ha, but once again you have no problem with ugly roads, cars and trucks? Ok.


Cars, roads and trucks have been around for a century, and none of us were around to discuss that.

Also now we have a century worth of experience to apply to deciding to add new vehicles to our environment.

And finally for the most part cars, trucks etc are primarily fixed routes - roads. We know where they are, there's planning and committees around design and disruption before they are set in place.

Drones move in a 3 dimensional space, with no set routes, and very few no fly zones.

Basically imagine a moped that can fly over your property or circle where you live looking for an address.

Not to even mention the privacy issues, most cars are still lacking in cameras, or scanning technology. Delivery drones will need that stuff to be standard.


If there is a deck there you probably can’t see nature anyway.


I'm also concerned about visual clutter, particularly in residential areas

Then don't look up in a couple of years when SpaceX has HUNDREDS of satellites ruining the night sky.

https://www.nytimes.com/2019/06/01/science/starlink-spacex-a...


My main concern is the audio and video recording devices that will undoubtedly be attached to these drones (if history is any guide, likely without any disclosure).

I foresee ads for rakes when Amazon notices that there are lots of leaves in my yard.

I don't know why people think this is a good thing right now.


How about just force-delivering a rake with free return?


I admire the way you think. It's horrifying, and disturbing, but I still admire it.


My first thought was "well, maybe you do need a rake". But I guess it probably can't see into your shed/garage/etc to know whether you actually have a rake or not. Ads for things you already have are such a waste.


God forbid I exist for a moment without being bombarded with offers to buy products.

I would gladly trade the inconvenience of not being immediately aware of rake-buying opportunities for privacy and peace of mind.


this is coming, if not already here, from satellites anyway


So I can get my useless bullshit faster and cheaper!


> I have been playing with radio controlled aircraft for most of my life

RC Helis 60 Size Gas back in the old days. You had to manually control everything. You had to built it from scratch. [1] The ones they have now are so simple and easy they are no fun at all. I wonder if you feel the same way?

I am still not getting how this is possible given weather conditions and variability.

[1] Electric? Only thing like that I ever saw was at the hobby shop (remember those) and it had a power cord attached and was plugged into an extension cord and being tested out back.


> You had to build it from scratch

Thats exactly what I'm trying to do here (sort of), I'm like 2 years in on the project and I've only just begun to actually test the thing. https://github.com/castis/currant


You are right that this thing is going to be noisy as hell. I can say this without having heard it because basic physics makes hovering craft with little propellers impossible to make quiet. Welcome to the new world of a flying buzzsaw delivering your neighbor's toothpaste at 7 AM.

No wonder they chose to play classical music over this video.


It'll be quieter than a UPS truck, that's for sure.


If I’m working on a drone in my yard, surrounded by buildings, the noise has a direct line to a few people. If I put that drone at 200 feet, suddenly many more people are directly exposed to that noise. Imagine how loud a UPS truck would be within a 90° cone above your bedroom. Also, consider how many more people would be affected by that noise per delivery.

But to be fair, the frequencies are quite different.


The UPS truck that delivers to my block makes a bit more noise then the average car driving down my street, parks, the driver gets out of it, and spends a couple of minutes making >20 deliveries. A human carrying cardboard boxes does not make much noise. Neither does a powered-off truck... And it's not like UPS does domestic deliveries from the back of 20-wheeler rigs.

This is orders of magnitude less noise than 20 drones, coming and going, throughout the day. (Not to mention the ones passing overhead, to deliver to the next block down...)


Yeah, but if I have enough money/income (which is a big caveat, I know) then I can have some control over the location of the UPS truck by virtue of living someplace not too close to a road.

When was the last time a UPS truck drove over your house / apartment / etc?


Properties owners have rights up to 500 ft from the base of the property (trespass), but this has not been litigated yet.


I did not know that. Noise-wise that's reassuring.

Mind if I ask where the law comes from? I'm assuming that you're talking about the United States - is this a federal/state/local law?


This article has more details (and notes that it's been tested legally at 83 feet): https://www.google.com/amp/s/slate.com/news-and-politics/201...


You don't have an issue with living near a noisy road with trucks driving down it though? I don't understand that.


Not to mention the visual pollution it may cause. I honestly don't think this is going to fly pun intended.


This is a really novel design.. few moving parts, and solves the main issue with multirotor drones.. namely that they are incredibly inefficient in terms of power required to fly.

When the initial hype about Amazon drone delivery came out I remember saying it wouldn't work cause multi-rotor drones are so inefficient.

This will make it work... it'll use a fraction of the electricity if it cruises in "plane mode".

If the max Payload is 5lbs I doubt it's going to weigh 50lbs. Maybe 10-20lbs, but the battery weight is the big thing.


> it'll use a fraction of the electricity if it cruises in "plane mode".

That'll depend how efficient the wing is. Usually there's trade-offs when doing a hybrid design (see V-22 Osprey).

Perhaps the trade-off is still a net benefit in that configuration though...

> If the max Payload is 5lbs I doubt it's going to weigh 50lbs. Maybe 10-20lbs

Not sure, in my RC flying experience, these aircraft can't lift much weight at all. Ounces are a serious concern, let alone pounds. Adding 1/4 extra weight is a big deal for something this small and without large wings to generate lift - the rotors are going to have to do a lot of work to get off the ground.

It'll probably need to be a pretty big aircraft to have enough battery carrying capacity to guarantee a safe round-trip with possible complications with wind, loiter time for safe landing, obstacle avoidance, etc, and big enough to support the 5lbs load without impacting flight characteristics too much.


For sure you are right about it having a lot of challenges and the total weight vs battery weight vs payload weight is a huge one.

But remember the wing design is competing against "no wing" so it is easy to be a huge improvement. Normal drones don't really get to throttle down and they have to do all their flight controls through engine speed adjustments. Normal flight controls use almost no power & when this thing builds up a head of steam it will be able to throttle back to cruise.

It looks like a closed edge biplane when it transitions.. the closed edges help efficiency. It can probably be optimized for a very narrow speed range as well so I bet they can come up with a very nice airfoil design.


> It looks like a closed edge biplane when it transitions.. the closed edges help efficiency

Perhaps. It could also be mostly to stabilize the aircraft in forward flight, since that's more efficient by itself with a "pulling" rotor than a traditional quad-copter setup (using differential thrust from each motor to control movement and stability).

The closed wing doesn't look very large, so we'll have to see just how much lift they actually get out of it. The airfoil design may not generate much or any lift but instead just be used for control, such as a typical horizontal/vertical stabilizer of an aircraft.

Or perhaps it does generate some lift and helps a bit. Just looks rather small for that role, but I could be wrong. RC aircraft often defy principles that apply to their full-sized brethren.

> (FTA): Amazon declined to release some specifics on the device, citing trade secrets

I guess we'll just have to wait for some more specifics.


I'm not sure why they would bother with the effort of this if the wing wasn't generating lift.


I don't know, just speculating it's not providing much lift with it being so small. But, it definitely would help stabilize the aircraft in it's forward flight configuration. You're probably right in that it's doing both.

Was hoping they'd release more details. It's not like RC aircraft have much secret sauce to them... other than Amazon's proprietary package compartment.


It's not really novel. Hybrid quadcopter/plane designs have existed for years. Using the wing as a rotor guard is novel I guess.


> Using the wing as a rotor guard is novel I guess.

Not even that...

http://www.luft46.com/heinkel/helerche.html


I'd like to know more about the support infrastructure they imagine. Even if this doubles flight time compared to current commercial quadcopters it would still have a 40-60min total flight time thus less than half that in range. It's going to require a lot of stations (and battery changes/charging) scattered about as its not like they can launch from warehouses and reach much.


Awesome video of the drone - https://www.youtube.com/watch?v=3HJtmx5f1Fc



Unfortunately no audio of the drone itself so we can judge its noisiness.


Even if you had audio, still can't judge the noise level unless there is a reference sound there too. Such as general highway noise.


Anyone know where that video was filmed? It looks an awful lot like Eastern Washington around the Walla Walla area, but I'm sure there are other parts of the country that look similar.


Could be Washington. https://www.geekwire.com/2015/revealed-how-we-found-amazons-... Meaning, they have tested in WA before.



That piano soundtrack was extra creepy. All I can imagine was Bezos weird face :/


Cool but I can't understand why you'd need such a delivery system? MAYBE you can imagine a scenario where you need to air lift a vial of snake anti-venom very quickly but that's a fantastical, far edge-case. Theres simply no reason that an auto-automobile can't handle 99% of deliveries.

Drones can't go that far. They can only carry tiny payloads. They use a ton of energy just to hover. And lastly- they are much more dangerous than ground-based transport, for the simple reason that _they can fall from the sky_.

Whats to be gained by putting deliveries in the air?


Sorry to be blunt, but you're speaking from ignorance. Take a look at statistics for cars and re-think your position on the relative safety of the two modes of transporting goods.


I believe he's talking about the future when automonomous cars are commonplace.


Zero traffic congestion, close to free to operate, fulfillment in minutes? Under 1 hour order delivery? There’s obviously downsides like sound and then falling out of the sky, but getting anything under 5 pounds two orders of magnitudes faster is a big deal.


One person is happy, thousands of others are negatively affected. This is a selfish mode of transporting goods.


I bet if HN existed in 1890, this would absolutely be a comment on the thread "Introducing the Automobile."

The truth is almost every mode of transport is selfish. It's always selfish, until everyone starts using it, at which point it's revolutionary.


If they tried to introduce cars now, there is no way they would get through. Because they are truly ridiculous and horrible. A million deaths around the world directly, and another million due to pollution? Imagine trying to convince everyone that was a good idea.


Compared to walking dozens of miles everywhere? Remember that the only existing modes of transportation were bicycles and horse-drawn carriages. The car absolutely needed to happen to aid our progress as a civilization.

Everything looks bad in hindsight right? But it is the car that enabled the progression that makes the car impractical.


I agree. This invention does nothing of real value. Perhaps it makes you life a bit more convenient but it's the difference between ordering a pizza in 30 minutes and ordering a pizza in 10 minutes. This can't enhance human well-being- it's simply a toy for rich techies who can afford to pay a premium to get their stuff all the faster.

I wish Mr Bezos would focus on something of real value- asteroid mining and transferal of polluting industries off-world.


Are you sure that the energy cost is lower for a delivery van? Fully loaded (though presumably uncommon for residential deliveries), a commercial van will carry 4 tons and have pretty poor gas mileage. Electric vans may help, but it's still quite a lot of weight to drive around.

The drones have no driver (thus lower margins and/or a richer Amazon, neither of which are necessarily good for inequality, though it does free up more humanity from mundane tasks like delivering parcels), and can be loaded quickly and automatically. The article mentions they can carry 5 lbs, which I suspect is enough to carry a few deliveries at once. As the article mentions, 80% of their sales are less than 5 lbs.

All that said, the noise pollution is going to be horrendous. The visual pollution I don't mind so much, as it _may_ allow cities to lower the number of cars on the road (fewer people driving to run errands), though it's definitely going to feel creepy and dystopian, especially if Amazon are the only ones running these things and there isn't _extremely_ strong privacy oversight.


> And lastly- they are much more dangerous than ground-based transport, for the simple reason that _they can fall from the sky_.

40,000 people are killed every year in the US by ground-based transport.


People want stuff faster. I’d pay more to get stuff faster.

People want it, not need. People pay for things they want as well as need.


Curious how well it will deal with locating the correct place at the correct house to deliver the package? One way to achieve this might be to only use the drone after a human delivery person has delivered a package and tagged in Amazon's map database where packages should be dropped. Or perhaps the house resident uploads a picture of their house with the drop location?


Maybe residents would place beacons where they want the drones to land?


Indeed, the beacon is a landing pad that customers place in their front or backyards (ideally back). See (3) in the following link:

https://www.digitalspy.com/tech/a820748/amazon-prime-air-dro...

As for which house, that's via geocoding (convert postal code to GPS). Reasonably accurate for North America, unclear about other parts of the world.


I think having a human take over via video for the last 30-60 seconds would work pretty well.

So the drone is automated for most of the time, but still uses manual for dropping off on doorstep.


The annular wing is an interesting touch. It should reduce induced drag at the low Reynolds numbers these drones typically fly at, by eliminating the wingtip vortices. I was working on a very similar design for a while, and I placed the props at the wingtips to cancel the vortices instead.

Also it would be interesting to see what their power electronics look like. You get a 20x reduction in thrust demands by going from hover to flight mode, but the problem is that the efficiency of brushless motors also tends to be absolutely dreadful when you run them at 5% throttle.


would it be possible to have 3 of each motor.. the one at low and the one at high.. it seems like in theory you could run all 6 to take off, but let them coast when "flying" versus gaining altitude?


That's absolutely doable, though you would have to prevent the inactive props from spinning and/or feather them to reduce drag. I don't really see variable pitch props on these though, the weight and complexity penalty would probably be excessive.


https://blog.aboutamazon.com/transportation/a-drone-program-...

Lots of talk about vision, no mention of LIDAR. Computer vision is getting better faster than LIDAR is getting cheaper, expect a lidar-less future.


LIDAR provides an additional layer of safety that seems well worth the marginal cost for those producing self-driving cars, where human lives are at stake, even if not applicable to delivery drones.

Do you believe that the future is LIDAR-less, or that the delivery drone future is LIDAR-less?


Did you see the picture of this drone? Do you think it's incapable of killing someone?


Lidar is difficult in drones. That doesn't rule out Lidar as an important sensor in other domains.

Additionally, redundancy is good. Computer vision will not be good enough to make Lidar useless for some time.


This is very clever, but clever usually loses out to practical in the end. Zipline[0] has a delivery drone with even fewer moving parts! It looks an awful lot like an airplane.

[0] https://flyzipline.com/


That doesn't look like a very practical way to drop deliveries for an individual house.


I'd understand the use of this in very spread-out rural areas, but for dense urban areas this is nothing more than a PR exercise to create headlines and improve hiring. Sort of like "self driving cars" and "AI" for Google. You go there to work on exciting stuff and end up getting assigned to some bullshit godforsaken backend nobody wants to work on.


I haven't seen a design quite like this before, but it seems like a natural hybrid of existing Tail-sitter planes and the current popular drone configuration.

https://en.wikipedia.org/wiki/Tail-sitter


Very smart design. No moving parts other than the motors.

Still, hitting one of these in a 152 could be fatal. Pilots are wary of drones for a reason. However, I think they'd be more receptive to the idea of Amazon using them than a random person (assuming Amazon cooperates with the FAA).


Ugh I hate designing around the stupidity of suburbs. Pnumatic tubes in dense cities: so much more efficient.

It's cool but only if you stop thinking how all of this is a hack around the utter failure of planning at any interesting scale in gsi country.


Could someone elaborate on what the use cases would be for something like this? I understand it will carry small items you buy off Amazon prime but that can't be the end of the roadmap for drone based delivery, can it?


walgreens has ~10,000 locations of ~10,000 square feet of products that are overwhelmingly under 5lbs and does ~20B/year in revenue.

whoever solves drone delivery will disrupt the convenience store market. its right there in the name.


How small? Most of my Amazon shipments could be carried by such drone.


Posting about how they're going to take over delivery while being investigated for antitrust stuff is an interesting choice.


Ok Google, "how do I build a faraday net for catching drones and preventing transmission after catch?"


[flagged]


"Eschew flamebait. Don't introduce flamewar topics unless you have something genuinely new to say. Avoid unrelated controversies and generic tangents."

We detached this comment from https://news.ycombinator.com/item?id=20107508 and marked it off-topic.


Are these allowed to fly over my property? And if so, am I allowed to shoot it down? Asking for a friend.


Does your friend also object to planes flying over their property?

(I would point out that the main reason people have objected to drones flying over their property is privacy: up to now, the only use-case for a drone is remote viewing – which some would find creepy/intrusive. I would posit that now drones can theoretically offer tangible real-world benefits --e.g. parcel delivery-- this will gradually fade out as an objection for most people - just as most people would tolerate a postman delivering letters to their house.)


No, because they are 30k feet off the deck and, more importantly obviously, contain people.


Your property goes from ground level to 200 feet; above that it's effectively the domain of the FAA. I would imagine the drones would have to fly between 200 and 500 feet until directly over the delivery target otherwise they would be either violating more restricted FAA airspace or trespassing on private property (under the 200 foot boundary).


If your friend lives anywhere that’s not rural, then no, she probably isn’t allowed to shoot it down, nor should she; bullets and buckshot could end up coming down on someone else.

But, if she uses a satellite TV dish to aim the output of a magnetron at that drone, well, that’s another story.


A much more interesting attack is GPS spoofing. What is Amazon (and everyone else) doing to solve that? GPS spoofing is fairly easy using SDR.

https://hackaday.com/2012/07/01/spoofing-gps-and-getting-you...

https://www.rtl-sdr.com/tag/gps-spoofing/


Maybe they will just prosecute people who do that?


Wait, this guy's family name is seriously Kimchi? The drone is cool but learning Kimchi is a surname is blowing my mind


It’s incredible, right? I don’t want to spoil it for you, but there are folks with last names like Baker, Miller, Smith, etc. that don’t even resemble food dish names! They’re occupations. Mind blown.


Used by people who've adopted those names after being assimilated into the cultures from which those names came about.

Gur Kimchi is evidently Jewish, and his name appears to be a variation of the Sephardic Jewish family name Kimhi. It evolved independently of the Korean name for their side dish, seeing as how they're from different language families


I don't get the reference, would you mind explaining? On a related note, I've always been baffled by Coursera founder & ML fame Andrew Ng's name. How am I supposed to pronounced it?


"Ng" is pronounced somewhat like the "ung" in "sung" but with the vowel shortened.


> How am I supposed to pronounced it?

Push your tongue against the top of your mouth and say 'ung', it's usually close enough (also holds for the common vietnamese name of Nguyen)

Specifically, it's the velar nasal phonome:

https://en.wikipedia.org/wiki/Velar_nasal


A famous Korean side dish?




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: