Hacker News new | past | comments | ask | show | jobs | submit | evanm's comments login

Curious what other languages were evaluated.


From [0] "Based on these criteria, we narrowed our testing languages to C++17, Java, Node.js, and Go. The first three languages were already in use at American Express, whereas Go was not."

[0]: https://americanexpress.io/choosing-go/


I'm........skeptical about all of this. Kid-sized sub built from rocket parts in a matter of hours?


It's mostly just an inflated kevlar bag with some valves.


It's easy to build one of something. Especially when the engineers building on of something usually build several of 1,500,000kg rockets.


Of course an "emergency cave submarine pod" built to specs would have been ideal but emergency situations require emergency measures.


Welding. See: Junkyard Wars.


> You can take an incremental approach and add type annotations as you need them

Most definitely. We've added type safety to new features in our legacy codebase. It's very seamless.


sounds like someone failed a lyft interview


Shrug. I work at Uber and I find this project interesting, even though we're adopting Flow.

Haters gonna hate. If code quality is an issue, contribute, fork or gtfo. Be nice. Etc. Etc. Can we move on now?


[flagged]


The GP was an uncivil swipe, but this is a bannable offense. We'll let you off this time, but not if you do it again.

Instead, please read https://news.ycombinator.com/newsguidelines.html and take the spirit of this site to heart: civil, substantive comments or please don't post.


Wow, this grudge runs deep.

Did a Lyft engineer kill your dog?


jesus, grow up.


A pre-9/11 ticket.


Ah yes, 'twas a simpler time.


ADP is in my axis of evil. Their employee and employer-systems are horrendously out of date and their support is non-existent. Let them fail.


THANKS FOR POSTING THIS.

I was affected and had a repair done. Looks like they're reimbursing customers for past repairs for this issue. Woop!


Outside of most Cirrus aircraft, they're not standard. They work pretty well on light and ultralight (single engine) aircraft.

Think of the extraordinary size and weight a chute or series of chutes would be for a short haul 737-sized aircraft. The cost of lugging those around across a fleet would be absolutely prohibitive. There's also little reason for this type of safety mechanism for such passenger jets. Commercial twin or quad engine planes are extraordinarily safe, and offer many more redundancies than a tiny single engine plane.


I believe you also suffer from square-cube effects. Take a light airplane with a parachute, and scale up all the linear dimensions by a factor of two. You now have something with four times the drag (drag is proportional to cross section) but eight times the weight (weight is proportional to volume). That means your terminal velocity increases by 40% (proportional to the square root of mass/area). In short, if you take a system designed for a SR-22 and just scale it up for a 737, the jet would descend much faster under the canopy, probably to a degree that it would no longer be safe. The parachute would have to be proportionally much larger still.

And as you say, they wouldn't add much safety because there isn't much more to be had. I can only think of two large jet crashes offhand that would have potentially benefitted (the DC-10 that crashed in Sioux City, and that Japanese 747 where the tail fell off) and that's from a period of decades.


Ideally you'd only need chutes big enough for the cabin and that it would detach and parachute down. Would probably still be prohibitively expensive, but a fun thought exercise.


Then you have the remainder of the plane falling, uncontrolled, and potentially towards populated areas. Without a parachute at least the pilot will (hopefully) be trying to aim for an empty patch, and with a parachute at least the rest of the plane is falling rather slowly.


The pilot can aim for an empty patch, then activate the parachute so that the cabin lands softly and the rest of the plane falls on an empty field. Having the chute system installed doesn't remove any options from the pilot.


First, removing the cabin will cause a large shift in the plane's center of gravity; without adjustment of the flight controls, it is likely to stall, dive, tilt to the side, enter a spin, or some combination of the above, and is unlikely to hit anywhere near where it was aimed.

More importantly, though, the parachute is also meant to provide an escape mechanism for unrecoverable spins and stalls, not just engine failures or fuel exhaustion. In a spin or stall 'aiming for an empty patch' is, essentially, not possible, as the control surfaces of the plane are not usable due to reduced airflow. Additionally, in IMC (Instrument Meterological Conditions), it's not possible to see a safe landing spot. This is an ideal use case for a parachute, but if the plane breaks away as you say, the remainder of the plane will land in an unpredictable location.

There's also the increased complexity involved with ensuring all flight control and avionics connectors between the cabin and body of the plane break away reliably and quickly when the parachute is deployed, and never break away otherwise.

In short, adding a breakaway cabin might make the pilot safer (but might not, given the additional complexity for breakaway avionics and flight controls), but it puts more risk on the public at large when deployed, so it's understandable why such a system was never developed (with the possible notable exception of military ejection seats).


More importantly, the force of the parachute(s) braking the plane would put a lot of stress on the plane structure and it's likely to break apart since it's not designed with sagging or hogging in mind (to borrow from naval architecture where those forces are explicitly modelled due to wave action.)


I'm fully aware of the extreme cost of rolling out a fiber infrastructure in new locations, but I still wonder why Google isn't more aggressive with their expansion. While the upfront costs are great, they more or less have guaranteed success in most major cities. Nobody likes the incumbent telcos--they're hated all over the US. I just dream of a world in which we aren't getting raped by Comcast, TWC, and Cox. They are absolutely abysmal and have no incentive to make their customers happy given their market monopoly/oligopoly.


Google does not want to be in the telco business at all, it just has to be to protect the source of its revenue which is people connecting to them via some ISP. They are purposely building out in cheap to build areas. If they started hooking up rural areas and areas with high land values and labor it would be a very expensive exercise. What Google is trying to do is get into enough homes as cheaply as possible to light a fire under other telco companies without actually having to build out the whole country.


> I'm fully aware of the extreme cost of rolling out a fiber infrastructure in new locations,

It's really not that much. I could do my entire town for a little under $9m. Fortunately, Google just announced they'd be coming here (Raleigh-Durham) so hopefully I won't have to.


I'm starting to get interested in wiring my neighborhood (the best current option is 6mb during the day and ~0.5mb in the evenings.) Are there any resources you could send me on what it would take to wire up a small area?


I could talk about this for hours... But really, it's quite a bit of work and really specific to your area. Since you're looking to do a small area, you probably won't have enough capital to bury fiber which is a shame because that's when you start getting the biggest bang for the buck. There's a couple different angles you need to assess.

Politics/Bureaucrats -- what laws are in place to prevent you from deploying in X fashion? Can you bury fiber? Do you have to string across utility poles? If so, who owns the poles? (It'll either be the city or a utility company) What are their rates? What are the requirements to get quotes for their rates? (CenturyLink requires you to be a part of an NDA/Non-Compete Agreement program if you are reselling competitive services)

Business -- how many people are you looking to service? What's the absolute max you could potentially service? (usually subdivisions are zoned with a specific max amount of households) How many would you like to service? What is an acceptable cost? (I found $100 is the sweet spot for gigabit fiber) What kind of customers will you be servicing? (mostly soccor moms who will browse the internet from 11a-3p, young families who embrace video streaming from 8p-11p, retirees who only care about email and facebook from 9a-12p, etc.)

Technical -- how much bandwidth do you really need? Chances are you can actually service 100 or so residential customers with just a single gigabit ethernet connection. How much fiber will you need to run to everyone? The easiest thing to do is run single-mode fiber connected to a media converter in their house. Depending on how you can do "huts" (as google has dubbed them) you will probably do single cisco ToR switch with 48 spf+ ports for 40 customers. You will dedicate 4 ports for uplink connectivity via LACP and have 4 stand-by ports should one be bad. Scale this out as needed.

This will also be specific to your area -- but expect $10k/mile to run conduit to a hut with an additional $2k/mile for different fiber pulls. You would usually run conduit and some kind of fiber at the same time -- probably 96 strand single-mode. But that's up to you.

Not sure what else you need to know. That should be a good starting ground. The bulk of our capital was for fluid insurance bonds. About 80% of our working capital was going to be tied into those while we laid the fiber.


Thanks! (And sorry it took me a while to respond.) In my case, I'm outside of city limits, so that should help with the politics.

I suspect I'm going to have to do copper on poles to make it viable, but how do I find out who owns the poles and what they charge?

Also I don't have any idea how to go about purchasing upstream bandwidth - any advice there?


> I suspect I'm going to have to do copper on poles to make it viable, but how do I find out who owns the poles and what they charge?

That's where a lot of leg work comes in. You'll have to start asking the city who owns them. I started by emailing the first person who sounded like they had the title of "secretary". It took about 5 emails getting batted around before I found out.

> Also I don't have any idea how to go about purchasing upstream bandwidth - any advice there?

Again, this is the leg work on your part. Do all the research you can and figure out who has fiber running nearby. You're going to have to tap into that for service. If it's the provider you go with who owns it then you won't have a seperate dark fiber leasing cost. CenturyLink, AT&T, XO, TW Telecom, Layer3, Qwest (many of these have merged together, but they sometimes use older names when referring to circuits -- so a qwest circuit might be running nearby but you'd go through centurylink to get it) will all be more than happy to take your money, but most of the backbone providers probably won't return your phone calls if you're not talking about 100G connections.


One probable reason is the regulatory hoops they need to jump through to rent/build the infrastructure and roll out the service.


I assume your comment was not meant to come off as hostile, but please don't use the word 'rape' to refer to things other than sexual violence. Doing so creates a hostile environment and minimizes the experiences of rape victims.


It's more a publicity stunt than a business. Google has been per-announcing Google Fiber for five years now, and they've only installed it in parts of three small cities.

They previously tried "free" ("must have Google Account") city-wide WiFi in Mountain View. That lasted from 2009 to 2014 and was shut down on May 3, 2014.


Google has only been expanding to cities that give them extensive tax breaks and other perks that haven't been afforded to most other companies. This is a good thing as it allows Google to expand, but they would never do this unless they got these incentives, plus the guaranteed number of users pre-signed up for the service so that they can guarantee some sort of ROI.


The big telecom companies got massive tax breaks at various times over the last 20 years to deploy high speed broadband, and essentially defrauded customers and tax payers by not doing what they said they would.

One example of many:

https://www.techdirt.com/articles/20131012/02124724852/decad...


Techdirt really has some of the most one sided and consistently terrible articles of any "news" site I've seen.

Here is some actual relevant information related to the Google Fiber concessions, mainly in that they don't have to build out infrastructure to the entire city (unlike Comcast etc), and other breaks not given to previous providers:

http://arstechnica.com/tech-policy/2012/09/how-kansas-city-t...

> Google received stunning regulatory concessions and incentives from local governments, including free access to virtually everything the city owns or controls: rights of way, central office space, power, interconnections with anchor institutions, marketing and direct mail, and office space for Google employees. City officials also expedited the permitting process and assigned staff specifically to help Google. One county even offered to allow Google to hang its wires on parts of utility poles—for free—that are usually off-limits to communications companies.


This is awesome--thanks @mbostock for your work and @erickhill for posting.


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

Search: