The funny thing with the Chicago example is that Chicago actually does have a legacy system for addressing blocks! The examples from River North wouldn't be included (it dates from 1830 and only covers the original 58 blocks centered around the Y of the Chicago River), but you can see the historical map here: http://www.encyclopedia.chicagohistory.org/pages/11175.html
The only popular use of this system in living memory has been the long-in-development Block 37 project in the Loop that finally opened in 2016, about seven years after this post.
this reminds me of issue I had when I heard American on TV or movies use number of blocks to describe a distance, for example: "my house is three blocks away from here". I thought block is a unit, just like meter or kilometer. I've wondered for a long time how many meters in a block there are.
Just for comparison, in Vietnam, we say something like "my house is about 100m away from here" or something like that. If you look at the map of city like Hanoi, Vietnam, you'll see why we cannot use "blocks" to describe distance.
> I thought block is a unit, just like meter or kilometer.
It is a unit, the distance between streets, but it varies widely across boroughs/districts in a single city or other municipality, and they aren’t really a reliable indicator of distance between different cities, though they are vaguely walkable for most places and people.
From Oxford Dictionary:
> (North American English) the area bounded by four streets in a town or suburb
> (North American English) the length of one side of a town block, typically as a measure of distance
Interesting. I didn't know that. Do you do the conversion in your head, like 2 blocks = 160*2 meters, when someone use blocks in a conversation ? Just curious.
also ... LOOK everyone I found an American who uses metric system as first choice in their post!!!
I think it's more navigational? like I know "oh, I'd cross three streets and then be at the house" roughly, more than thinking about distance. It kinda equals effort to walk there and how often you'll stop for traffic more.
This right here. Nobody is converting the local average distance between streets in their heads into ordinary units of length and then converting. In ordinary conversation, to say "so-and-so is three blocks away" means you'd need to cross three streets to get there.
And just like you have an idea of how big a mile or kilometer is without having to get out a tool to measure it, blocks tend to be pretty consistent within a given town or city, so "a block" has a similar rough intuitive distance.
> Interesting. I didn't know that. Do you do the conversion in your head, like 2 blocks = 160*2 meters, when someone use blocks in a conversation ? Just curious.
I do it explicitly some times, but mainly I do 2 blocks -> 5 minutes of walking.
It's also useful to think in blocks because addresses here (and in many places in the US) go up/down by 100 for each block; so if I'm going from 12XX to 7XX (colloquially "The 1200 block to the 700 block") that's 5 blocks.
> also ... LOOK everyone I found an American who uses metric system as first choice in their post!!!
Meters are better than feet in every way even if you use miles for long distances. There are 5280 feet in a mile, but 1600 meters in a mile[1]. One of these numbers is far easier to deal with.
As an aside, I wonder about using centimeters for height; In the US, if you look at a bell-curve for "self reported height of men" you see a drop off just below 6' (~183cm) and a huge spike at 6', distorting the curve[2]. Do you see similar distortions at 180 or 175cm in countries where the metric system is used? Given that far more men are "almost" 180cm or 175cm (in the US, 175 is average) are there more men "rounding up" on their heights in such places?
Athletes, in particular, are incentivized to round-up, as scouts won't even look at you if you are "too short" and scouts like to use even numbers for cutoffs.
1: Technically 1609, but that makes an eighth of a mile estimate off by only about a meter, and the "mile run" event in US high-school track is actually 1600 meters.
2: The center of the bell-curve is also 2-3cm taller than measured heights of men in the US.
>Meters are better than feet in every way even if you use miles for long distances. There are 5280 feet in a mile, but 1600 meters in a mile[1]. One of these numbers is far easier to deal with.
why have you forsaken the furlong, the chain, and the rod?
1. To use all 3 would require remembering more conversion ratios
2. More people (even in the US) are familiar with meters than those units. I was not familiar with the Rod at all, which is the closest to meter in utility for discussing block lengths.
The Chicago grid is actually based on fractions of a mile (1/8 and 1/16 iirc), and when I lived there I would absolutely convert blocks to miles automatically when talking to people.
Usually when I use blocks, in the US, in conversation, it's with a city that both of us are familiar with. Any conversion to other units makes less sense in my head than the visual idea of what I know the city looks like outside.
The block definitely has a New York center of gravity in the American conception. The Manhattan street grid is very regular with “short blocks” and “long blocks” and vast swaths of the rest of the city use a very similar approach.
There’s also certainly an element of it that comes from all the other gridded cities especially Chicago and the older cities of the East and Midwest.
But New York is one of the most written and talked about and filmed places on the planet and I suspect that’s the biggest cause of it being considered standard American English.
I don't think it entered parlance through that usage. It was actually a common feature of the landscape beginning in the 1780s [2] even if we don't count Philadelphia in the 1680s [3] which later became the national capital city.
The Commissioners' Plan which griddified Manhattan was not until 1811 -- but tens of thousands of square miles were already being rapidly surveyed into townships, sections, and ranges, which lend themselves to further square subdividing, after an act of congress back in 1785.[1]
As a fun side note, 19th Century Philadelphia invented the process of adding street numbers to its existing grid using 100-to-a-block addressing, and was also the first American city to add the hundred-block to street signs.
This is cool, because it means that naming matches a sense of closeness. Someone living on the same street can live quite far away from me, but someone living in the same block is close. Christopher Alexander wrote about the advantage of small named units. You could organize a party for a block, but not for a street.
I remember living in Japan around 2001 and having to handle random addresses.
We had several atlas books of regions and neighborhoods. Looking up an address was an exercise in getting the right book for the neighborhood, then using an index which noted which block of the map that house number was in (eg 'c10' using a grid with numbers across the top and letters down the side).
Plotting a course from where you were to this location was also exciting and required patching together multiple maps.
Actually following that course in real life was another challenge altogether and really required visualizing what you saw on the map since, as noted, there aren't many street names.
I'm honestly surprised we were able to succeed at getting to places as often as we did.
I kind of miss that experience, just plugging an address into a phone and following directions is much less exciting.
The first time I went to Japan was also pre-smartphone, and navigating anywhere involved a combination of going to a convenience store with a written address and getting the clerk (hopefully a young girl -- they have better language skills!) to pull out the big white books for the area, look up the address in the index, etc. If you were in the unfortunate situation of needing to navigate in a puzzling neighborhood, you'd have to make a copy of the map or draw a little diagram, and then if you got lost, you'd use the random "neighborhood maps" posted on signs on major streets (these still mostly exist).
Guidebooks would often have little tiny maps drawn in the margins that tried to get you from the nearest train station, but even that could be a nightmare, because the typical downtown Tokyo train station has a dozen or so exits, and just getting to the right exit to use the map was a navigation puzzle in its own right. I got lost for the better part of an hour on my first visit to Tokyo station!
Doing this without any language proficiency meant that getting anywhere could take hours. Fun times. It's sort of mind-blowing how much easier it is for tourists now.
> just plugging an address into a phone and following directions is much less exciting.
Something I realised a year or two back: my mental model of breaking down a larger goal into smaller subgoals is navigation. If you want to get from somewhere in Los Angeles to somewhere in San Francisco, and that step needs to be broken down, it can be analysed into (a) get from LA to Harris Ranch, then (b) get from Harris Ranch to SF.
Obviously people growing up now don't think of things that way any more, because you don't have to be able to think of the equivalent of Harris Ranch on demand; what do you all use instead as a mental model?
First job was a delivery driver pre-Internet and Thomas guide was a lifesaver.
It's hard to explain to younger people my pride in that I was "good" at finding multiple addresses and plotting best course between locations, then knowing where phone booths were to call in on the road, etc. Definitely makes me appreciate GPS and mobile mapping apps.
In Singapore, post/ZIP codes refer to a specific address, building, or block. While letters will still have a full name + address, it seems to mostly be there for error correction and to specify unit numbers.
It should be noted that in spite of the devastation of Tokyo in WW2, many parts of Tokyo retained the pre-war system of neighborhoods and streets, where streets might often not be straight, or start and terminate abruptly within the boundaries of a neighborhood.
Like some other responders, I also lived in Tokyo pre-google maps (in the 90s). The system of location one would use in central Tokyo was:
(city 市 or 都 = such as central Tokyo)->
ward (ku 区 = such as Shibuya-ku, if in a city) ->
area within ward (such as Aoyama) ->
the block / area (chome 丁目 = such as 7 chome, or 7th area)
the house number
For those unfamiliar with an area would first observe the map at the nearest train station to locate the area / chome. As one was walking and potentially got lost, could check street posts for the current chome and an occasional neighborhood map / sign to locate more precisely.
Personally I found navigating this way a lot of fun and an alternative hierarchical way of thinking about location.
(Earth) GPS (Elevation, meters, postfixed) (Unit # or Label)
E.G.
N47°26'39.4" W122°18'03.8" Departures Road Bay 20
A (random) place at the major airport nearest to me (as a not DOX example). Also note now the North / West have been moved to prefix the GPS scalar elements, as the biggest shift in direction.
Contrast:
SEA Seattle International Airport
Departures Road Bay 20
SeaTac WA 98158
For street signs I'd expect MAJOR intersections to get big labels that include the full address, with the most geographically major components (the ones that barely ever change in a city) the smallest.
The algorithm is described here: https://github.com/google/open-location-code/blob/main/Docum... Key notes: Normalize the GPS coords to +/- values (+/-90 and 180), then add 90 or 180 respectively to create an unsigned index along the range. The 10 largest steps are pairs of 0-19 encoding interleaved, while after that it's single step with the remaining fractional parts of Latitude * 2.5e7 (extract mod 5, *4), Long * 8.192e6 (extract mod 4) for each character after the 10th. The specification page does a FAR better job explaining things, but if link rot happens hopefully the summary here is sufficient.
In the United States your address is uniquely identified by an 11 digit number. This is your zip code, the +4 extension, and the last 2 digits of your house number.
If you live on 123 Main Street, Schenectady, NY 12345-6789, then your address can be uniquely represented by 12345678923.
The post office also uses this system for sorting. That way the mail in the truck is in the correct order automatically.
My boil the ocean dream revolves around the idea that addresses are inherently hierarchical but we present them in this screwball mixed format.
the address should be(I will admit your address mixes a few level so my example may be incorrect)
/USA/98158/WA/SeaTac/Departures Road/bay 20/SEA Seattle International Airport
I always store a description with each one.
/country/zip code/city/street/bay/address
Or you could go full ldap
/country=USA/zip=98158/state=WA/city=SeaTac/street=Departures Road/bay=20/address=SEA Seattle International Airport
but I hate the way that looks.
And actually on topic, in many Utah cities the street address are sort of like that 500 south, 3000 west, etc. their reference point is the center of the known universe... the morman tabernacle naturally.
At least for USPS address specification the syntax is roughly:
* The lowest line is the most coarse and sufficient to route to a major sorting facility.
* There's an implicit United States of America line under that, but International mail also uses the lowest line as major routing. (I guess customs I/O comes after that, and then it hits the foreign carrier if handing off.)
* Physical location for a traditional address is the next 'line or two' above that. Usefully for a human holding a thing on a street (after sorting) the numeric address is first on that line set. Then the other bits of the address follow, usually a road designation or some zone thing; this line is little-endien first.
* A building and/or person name are the top, often one line, sometimes two.
This is probably why PlusCodes retains a short human friendly string specification for a major local marker. Like a city or major municipal facility that's unlikely to move over time. Effectively it's the same as that last line or two of USPS addresses. The + mark in the syntax is similar to a numeric decimal point, while also clearly breaking the major and minor flow elements.
It would be useful in some contexts but I wouldn't want to abandon street names and house numbers for it since they do make things easier when giving directions (something like "Take a left at N47°26'39.4 W122°18'03.8" would get out of hand quickly, especially over the phone) and it'd be strange to display something that long on the outside of your house instead of a 3-4 digits.
It may have changed, but last time I was there, most of Ireland outside of Dublin had no addresses, just coordinates. Navigating wasn't too hard though, streets have numbers, and directions were usually a "turn left at road 1310, etc.".
It also helps that most of Ireland is pretty rural.
South Korea was like this too but they thankfully switched recently. It's not uncommon to receive addresses in both the old and new style since most locals grew up on the former.
The new South Korean addressing scheme is kind of a mess and I think a missed opportunity.
i.e. Street names (gil) can be derived from larger roads (ro). e.g. Toegrye-ro is the major road, and Toegye-ro 48-gil is a street that branches from it. Odd number gil are on one side (Toegye-ro 49-gil, 51-gil etc.) and even (48-gil, 50-gil, etc.) on the other.
But then there's not really an addressable road beneath the ro (okay, alleys I guess), so you also end up with multiple Toegye-ro 50-gils, some of which just change names to other multiple numbered gils as you cross intersections, but some don't. So the specific street names are rather localized, and somewhat just replace the function served by navigating by area.
Building numbers are somewhat sane, but don't map to the street frontage, meaning the numbered streets have no alignment to the numbered buildings, yet the building numbers already use an ascending or descending dashed system (e.g. 89-2) that's supposed to help in navigating at the sub-ro level (alleys again).
So the great re-addressing effort didn't really add a lot of navigational power over the previous system. And in practice, even with GPS, many people just find the numbered -gil or use the old address and just ask a local for directions or walk around for a while hoping to find the thing they're looking for.
Example: Take a look at the link below, at the nearby area, and tell me how many Toegye-ro 50-gils there are!
I'm struggling to understand how this would translate, "Go north three miles on High St. Turn left onto 2nd street and go 500m. Then take Park Ave south for 100m."
1) Go to / be pointed the nearest train station, then ask random people for help close by if still can't see the number somewhere or parse the names. People all over are super willing to help even with a language barrier. They'll usually point you up a few streets, at landmarks, or general direction. I'd often end up having to ask like 4 people on the way lol
2) If you're meeting with friends, you meet at the train station then go from there.
I don't know how anyone got around there before mapping software.
I lived in Tokyo before Google Maps, and the answer was: printed maps. Every business card had a map on the back, every ad for an event had a map, every personal invitation had a hand-drawn map. But yeah, if you were going somewhere as a group, you'd meet at a known point (outside train station etc) and then head over together.
If you were roughing it on your own and only had the address, things got more interesting. Train stations always had detailed maps of major landmarks, so finding those was not an issue. If you were looking for something too small to be covered (say, a restaurant), you'd head to the chōme and then start winnowing down. Police boxes (koban) always had detailed neighborhood maps bolted to a wall nearby listing every single business and family by name, albeit usually in handwritten Japanese only, and you could ask the cops for directions too.
The final boss was the non-linear numbering house scheme though. Some friends and I once spent a fruitless hour searching for the HR Giger bar in Tokyo, which we knew was at X-Y-Z, but only managed to find X-Y-(Z+1) and X-Y-(Z-1).
As far as I can tell, in a dense Japanese city, directions start relative to a particular train station, then use distances and landmarks to go from there.
That's pretty much how directions work outside of major cities in the US too, except for a lack of train stations, and the more rural you get the more landmarks are replaced with shibboleths.
"Go about ten minutes that way until you see the old Smith farm. [The Smiths haven't lived there for three generations.] Then take a left and head five minutes down Mill Road. [It's actually called something else, has never had a mill on it in its entire history, but is called that by the locals nonetheless.]" Eventually you start wondering when you tripped and fell into a Discworld novel.
In the UK and Ireland you navigate by pubs, since it is topologically impossible to traverse any connected area without passing at least one pub. The granularity of directions is therefore highest in Dublin (c.f. Joyce, Ulysses, 1922)
Landmarks. It's similar to how directions are given in other countries with large urban areas and shaky addressing systems.
"Turn left at this coffee shop, right at the second church..."
When my wife first moved to the U.S. from Korea I couldn't figure out why she was so insistent on using landmarks to navigate, but eventually learned this was the reason. When we went out to rural areas where I grew up, she would become completely lost because it's not generally as landmark dense and the mental system she had accumulated didn't work as well. That was when she finally had to learn the U.S. system.
For shorter distances you use landmarks, like known chains of convenience stores or restaurants. For longer distances, larger roads do have route numbers (county route 69, national route 3, etc).
Most major roads in Japanese cities have names. Smaller roads do not.
But except for certain cities (Kyoto, Sapporo perhaps) they're just not used for addresses.
The funny thing with the Chicago example is that Chicago actually does have a legacy system for addressing blocks! The examples from River North wouldn't be included (it dates from 1830 and only covers the original 58 blocks centered around the Y of the Chicago River), but you can see the historical map here: http://www.encyclopedia.chicagohistory.org/pages/11175.html
The only popular use of this system in living memory has been the long-in-development Block 37 project in the Loop that finally opened in 2016, about seven years after this post.
Huh, I never really thought about where the name “Block 37” came from, that’s pretty interesting!
Kyoto has this sorted. It's on a grid system, and every horizontal and vertical street has a name.
When giving full addresses it is done like so
京都市中京区堀川通御池西下る西三坊堀川町521番地
Very long! 京都市 (Kyoto city) 中京区 (district/ward) 堀川通 (vertical road) 御池西 (horizontal road) 下る (go south) 西三坊堀川町 (neighbourhood) 521番地 (block number)
In other words you are told the crossing of two roads and which direction to travel from there.
(2009) Some discussion over the years:
57 points, 15 years ago, 25 comments https://news.ycombinator.com/item?id=668197
36 points, 7 years ago, 10 comments https://news.ycombinator.com/item?id=15616782
this reminds me of issue I had when I heard American on TV or movies use number of blocks to describe a distance, for example: "my house is three blocks away from here". I thought block is a unit, just like meter or kilometer. I've wondered for a long time how many meters in a block there are. Just for comparison, in Vietnam, we say something like "my house is about 100m away from here" or something like that. If you look at the map of city like Hanoi, Vietnam, you'll see why we cannot use "blocks" to describe distance.
> I thought block is a unit, just like meter or kilometer.
It is a unit, the distance between streets, but it varies widely across boroughs/districts in a single city or other municipality, and they aren’t really a reliable indicator of distance between different cities, though they are vaguely walkable for most places and people.
From Oxford Dictionary:
> (North American English) the area bounded by four streets in a town or suburb
> (North American English) the length of one side of a town block, typically as a measure of distance
It's only a very vague distance in many places in the US as well, but it's typically between 100 and 200 meters.
In Manhattan (a popular location for US mass media) a block running north-south is 80 meters, but an east-west block is variable and about 250 meters.
Chicago also uses a rectangular block, but with east-west blocks being shorter at 100 meters and north-south blocks being 200 meters.
Where I live, a block is 160 meters downtown (10 blocks to a mile).
Interesting. I didn't know that. Do you do the conversion in your head, like 2 blocks = 160*2 meters, when someone use blocks in a conversation ? Just curious.
also ... LOOK everyone I found an American who uses metric system as first choice in their post!!!
I think it's more navigational? like I know "oh, I'd cross three streets and then be at the house" roughly, more than thinking about distance. It kinda equals effort to walk there and how often you'll stop for traffic more.
This right here. Nobody is converting the local average distance between streets in their heads into ordinary units of length and then converting. In ordinary conversation, to say "so-and-so is three blocks away" means you'd need to cross three streets to get there.
And just like you have an idea of how big a mile or kilometer is without having to get out a tool to measure it, blocks tend to be pretty consistent within a given town or city, so "a block" has a similar rough intuitive distance.
> Interesting. I didn't know that. Do you do the conversion in your head, like 2 blocks = 160*2 meters, when someone use blocks in a conversation ? Just curious.
I do it explicitly some times, but mainly I do 2 blocks -> 5 minutes of walking.
It's also useful to think in blocks because addresses here (and in many places in the US) go up/down by 100 for each block; so if I'm going from 12XX to 7XX (colloquially "The 1200 block to the 700 block") that's 5 blocks.
> also ... LOOK everyone I found an American who uses metric system as first choice in their post!!!
Meters are better than feet in every way even if you use miles for long distances. There are 5280 feet in a mile, but 1600 meters in a mile[1]. One of these numbers is far easier to deal with.
As an aside, I wonder about using centimeters for height; In the US, if you look at a bell-curve for "self reported height of men" you see a drop off just below 6' (~183cm) and a huge spike at 6', distorting the curve[2]. Do you see similar distortions at 180 or 175cm in countries where the metric system is used? Given that far more men are "almost" 180cm or 175cm (in the US, 175 is average) are there more men "rounding up" on their heights in such places?
Athletes, in particular, are incentivized to round-up, as scouts won't even look at you if you are "too short" and scouts like to use even numbers for cutoffs.
1: Technically 1609, but that makes an eighth of a mile estimate off by only about a meter, and the "mile run" event in US high-school track is actually 1600 meters.
2: The center of the bell-curve is also 2-3cm taller than measured heights of men in the US.
>Meters are better than feet in every way even if you use miles for long distances. There are 5280 feet in a mile, but 1600 meters in a mile[1]. One of these numbers is far easier to deal with.
why have you forsaken the furlong, the chain, and the rod?
Two reasons:
1. To use all 3 would require remembering more conversion ratios
2. More people (even in the US) are familiar with meters than those units. I was not familiar with the Rod at all, which is the closest to meter in utility for discussing block lengths.
> Do you see similar distortions at 180 or 175cm in countries where the metric system is used?
No idea, but Vietnamese avg height is 168 cm :D I cannot find any sources published by our gov that has height distribution info.
I'm 69" tall, and I wonder if that would be a spike of its own if we didn't use a combination of inches and feet.
The Chicago grid is actually based on fractions of a mile (1/8 and 1/16 iirc), and when I lived there I would absolutely convert blocks to miles automatically when talking to people.
Usually when I use blocks, in the US, in conversation, it's with a city that both of us are familiar with. Any conversion to other units makes less sense in my head than the visual idea of what I know the city looks like outside.
There is always a bit of a surprise to it, because it can apply equally to really long blocks in the suburbs or really short blocks in a town.
I usually hear it's N avenues away if you need to walk east-west in Manhattan.
The block definitely has a New York center of gravity in the American conception. The Manhattan street grid is very regular with “short blocks” and “long blocks” and vast swaths of the rest of the city use a very similar approach.
There’s also certainly an element of it that comes from all the other gridded cities especially Chicago and the older cities of the East and Midwest.
But New York is one of the most written and talked about and filmed places on the planet and I suspect that’s the biggest cause of it being considered standard American English.
I don't think it entered parlance through that usage. It was actually a common feature of the landscape beginning in the 1780s [2] even if we don't count Philadelphia in the 1680s [3] which later became the national capital city.
The Commissioners' Plan which griddified Manhattan was not until 1811 -- but tens of thousands of square miles were already being rapidly surveyed into townships, sections, and ranges, which lend themselves to further square subdividing, after an act of congress back in 1785.[1]
As a fun side note, 19th Century Philadelphia invented the process of adding street numbers to its existing grid using 100-to-a-block addressing, and was also the first American city to add the hundred-block to street signs.
[1] https://en.m.wikipedia.org/wiki/Seven_Ranges
[2] https://www.bloomberg.com/news/articles/2015-09-16/the-jeffe...
[3] https://dougalleninstitute.org/archives/10148/
[4] https://www.ncesc.com/geographic-pedia/what-is-the-philadelp...
This is cool, because it means that naming matches a sense of closeness. Someone living on the same street can live quite far away from me, but someone living in the same block is close. Christopher Alexander wrote about the advantage of small named units. You could organize a party for a block, but not for a street.
I remember living in Japan around 2001 and having to handle random addresses.
We had several atlas books of regions and neighborhoods. Looking up an address was an exercise in getting the right book for the neighborhood, then using an index which noted which block of the map that house number was in (eg 'c10' using a grid with numbers across the top and letters down the side).
Plotting a course from where you were to this location was also exciting and required patching together multiple maps.
Actually following that course in real life was another challenge altogether and really required visualizing what you saw on the map since, as noted, there aren't many street names.
I'm honestly surprised we were able to succeed at getting to places as often as we did.
I kind of miss that experience, just plugging an address into a phone and following directions is much less exciting.
The first time I went to Japan was also pre-smartphone, and navigating anywhere involved a combination of going to a convenience store with a written address and getting the clerk (hopefully a young girl -- they have better language skills!) to pull out the big white books for the area, look up the address in the index, etc. If you were in the unfortunate situation of needing to navigate in a puzzling neighborhood, you'd have to make a copy of the map or draw a little diagram, and then if you got lost, you'd use the random "neighborhood maps" posted on signs on major streets (these still mostly exist).
Guidebooks would often have little tiny maps drawn in the margins that tried to get you from the nearest train station, but even that could be a nightmare, because the typical downtown Tokyo train station has a dozen or so exits, and just getting to the right exit to use the map was a navigation puzzle in its own right. I got lost for the better part of an hour on my first visit to Tokyo station!
Doing this without any language proficiency meant that getting anywhere could take hours. Fun times. It's sort of mind-blowing how much easier it is for tourists now.
The piecing together of multiple maps doesn't sound much different from the days when we relied upon "The Thomas Guide to Los Angeles & Orange Counties": https://qph.cf2.quoracdn.net/main-qimg-8541cb8ec83526640938d...
> just plugging an address into a phone and following directions is much less exciting.
Something I realised a year or two back: my mental model of breaking down a larger goal into smaller subgoals is navigation. If you want to get from somewhere in Los Angeles to somewhere in San Francisco, and that step needs to be broken down, it can be analysed into (a) get from LA to Harris Ranch, then (b) get from Harris Ranch to SF.
Obviously people growing up now don't think of things that way any more, because you don't have to be able to think of the equivalent of Harris Ranch on demand; what do you all use instead as a mental model?
First job was a delivery driver pre-Internet and Thomas guide was a lifesaver. It's hard to explain to younger people my pride in that I was "good" at finding multiple addresses and plotting best course between locations, then knowing where phone booths were to call in on the road, etc. Definitely makes me appreciate GPS and mobile mapping apps.
In Singapore, post/ZIP codes refer to a specific address, building, or block. While letters will still have a full name + address, it seems to mostly be there for error correction and to specify unit numbers.
It should be noted that in spite of the devastation of Tokyo in WW2, many parts of Tokyo retained the pre-war system of neighborhoods and streets, where streets might often not be straight, or start and terminate abruptly within the boundaries of a neighborhood.
Like some other responders, I also lived in Tokyo pre-google maps (in the 90s). The system of location one would use in central Tokyo was:
(city 市 or 都 = such as central Tokyo)-> ward (ku 区 = such as Shibuya-ku, if in a city) -> area within ward (such as Aoyama) -> the block / area (chome 丁目 = such as 7 chome, or 7th area) the house number
For those unfamiliar with an area would first observe the map at the nearest train station to locate the area / chome. As one was walking and potentially got lost, could check street posts for the current chome and an occasional neighborhood map / sign to locate more precisely.
Personally I found navigating this way a lot of fun and an alternative hierarchical way of thinking about location.
How about a country with neither? Here's a documentary about Costa Rica: https://www.youtube.com/watch?v=wxWfpBQvfLs
I would love for a universal address like...
(Earth) GPS (Elevation, meters, postfixed) (Unit # or Label) E.G.
N47°26'39.4" W122°18'03.8" Departures Road Bay 20
A (random) place at the major airport nearest to me (as a not DOX example). Also note now the North / West have been moved to prefix the GPS scalar elements, as the biggest shift in direction.
Contrast:
For street signs I'd expect MAJOR intersections to get big labels that include the full address, with the most geographically major components (the ones that barely ever change in a city) the smallest. N47°26'39.4" W122°18'03.8" Departures Road Bay 20N47°26'30.0" W122°17'57.5" Departures Road Bay 2
Google's Map also has 'plus codes' CPR2+J7 SeaTac, Washington https://maps.google.com/pluscodes/ and https://github.com/google/open-location-code Which seems like a compelling variation for encoding the GPS value.
The algorithm is described here: https://github.com/google/open-location-code/blob/main/Docum... Key notes: Normalize the GPS coords to +/- values (+/-90 and 180), then add 90 or 180 respectively to create an unsigned index along the range. The 10 largest steps are pairs of 0-19 encoding interleaved, while after that it's single step with the remaining fractional parts of Latitude * 2.5e7 (extract mod 5, *4), Long * 8.192e6 (extract mod 4) for each character after the 10th. The specification page does a FAR better job explaining things, but if link rot happens hopefully the summary here is sufficient.
In the United States your address is uniquely identified by an 11 digit number. This is your zip code, the +4 extension, and the last 2 digits of your house number.
If you live on 123 Main Street, Schenectady, NY 12345-6789, then your address can be uniquely represented by 12345678923.
The post office also uses this system for sorting. That way the mail in the truck is in the correct order automatically.
My boil the ocean dream revolves around the idea that addresses are inherently hierarchical but we present them in this screwball mixed format.
the address should be(I will admit your address mixes a few level so my example may be incorrect)
/USA/98158/WA/SeaTac/Departures Road/bay 20/SEA Seattle International Airport
I always store a description with each one.
/country/zip code/city/street/bay/address
Or you could go full ldap
/country=USA/zip=98158/state=WA/city=SeaTac/street=Departures Road/bay=20/address=SEA Seattle International Airport
but I hate the way that looks.
And actually on topic, in many Utah cities the street address are sort of like that 500 south, 3000 west, etc. their reference point is the center of the known universe... the morman tabernacle naturally.
At least for USPS address specification the syntax is roughly:
* The lowest line is the most coarse and sufficient to route to a major sorting facility.
* There's an implicit United States of America line under that, but International mail also uses the lowest line as major routing. (I guess customs I/O comes after that, and then it hits the foreign carrier if handing off.)
* Physical location for a traditional address is the next 'line or two' above that. Usefully for a human holding a thing on a street (after sorting) the numeric address is first on that line set. Then the other bits of the address follow, usually a road designation or some zone thing; this line is little-endien first.
* A building and/or person name are the top, often one line, sometimes two.
This is probably why PlusCodes retains a short human friendly string specification for a major local marker. Like a city or major municipal facility that's unlikely to move over time. Effectively it's the same as that last line or two of USPS addresses. The + mark in the syntax is similar to a numeric decimal point, while also clearly breaking the major and minor flow elements.
It would be useful in some contexts but I wouldn't want to abandon street names and house numbers for it since they do make things easier when giving directions (something like "Take a left at N47°26'39.4 W122°18'03.8" would get out of hand quickly, especially over the phone) and it'd be strange to display something that long on the outside of your house instead of a 3-4 digits.
It may have changed, but last time I was there, most of Ireland outside of Dublin had no addresses, just coordinates. Navigating wasn't too hard though, streets have numbers, and directions were usually a "turn left at road 1310, etc.".
It also helps that most of Ireland is pretty rural.
For historical interest, you may wish to consult the (cold war era) Usenet practice of giving "icbmto:" coordinates.
And then an earthquake moves the location 4 meters east.
South Korea was like this too but they thankfully switched recently. It's not uncommon to receive addresses in both the old and new style since most locals grew up on the former.
The new South Korean addressing scheme is kind of a mess and I think a missed opportunity.
i.e. Street names (gil) can be derived from larger roads (ro). e.g. Toegrye-ro is the major road, and Toegye-ro 48-gil is a street that branches from it. Odd number gil are on one side (Toegye-ro 49-gil, 51-gil etc.) and even (48-gil, 50-gil, etc.) on the other.
But then there's not really an addressable road beneath the ro (okay, alleys I guess), so you also end up with multiple Toegye-ro 50-gils, some of which just change names to other multiple numbered gils as you cross intersections, but some don't. So the specific street names are rather localized, and somewhat just replace the function served by navigating by area.
Building numbers are somewhat sane, but don't map to the street frontage, meaning the numbered streets have no alignment to the numbered buildings, yet the building numbers already use an ascending or descending dashed system (e.g. 89-2) that's supposed to help in navigating at the sub-ro level (alleys again).
So the great re-addressing effort didn't really add a lot of navigational power over the previous system. And in practice, even with GPS, many people just find the numbered -gil or use the old address and just ask a local for directions or walk around for a while hoping to find the thing they're looking for.
Example: Take a look at the link below, at the nearby area, and tell me how many Toegye-ro 50-gils there are!
https://www.google.com/maps/place/Seoul,+South+Korea/@37.561...
In my experience, hallways in large buildings usually don't have names nor numbers either; only the rooms, and sometimes blocks are numbered too.
Uri, co-founder of Waze, mentioned this on the Lenny's Podcast last week. I wonder if this prompted the post:
https://www.lennyspodcast.com/lessons-from-a-two-time-unicor...
Combine these with the street maps that are not oriented north but the direction you’re facing and you’re gonna have a bad time
> the direction you’re facing
I'm fine with these, but it's like unspoken rule to never indicate the true directions, even if it's gigantic 2m x 2m billboard.
I have gradually come to like those, but I had to unlearn expecting them to be north aligned first
How does giving directions work in Japan?
I'm struggling to understand how this would translate, "Go north three miles on High St. Turn left onto 2nd street and go 500m. Then take Park Ave south for 100m."
When I was in Tokyo, it was basically:
1) Go to / be pointed the nearest train station, then ask random people for help close by if still can't see the number somewhere or parse the names. People all over are super willing to help even with a language barrier. They'll usually point you up a few streets, at landmarks, or general direction. I'd often end up having to ask like 4 people on the way lol
2) If you're meeting with friends, you meet at the train station then go from there.
I don't know how anyone got around there before mapping software.
I lived in Tokyo before Google Maps, and the answer was: printed maps. Every business card had a map on the back, every ad for an event had a map, every personal invitation had a hand-drawn map. But yeah, if you were going somewhere as a group, you'd meet at a known point (outside train station etc) and then head over together.
If you were roughing it on your own and only had the address, things got more interesting. Train stations always had detailed maps of major landmarks, so finding those was not an issue. If you were looking for something too small to be covered (say, a restaurant), you'd head to the chōme and then start winnowing down. Police boxes (koban) always had detailed neighborhood maps bolted to a wall nearby listing every single business and family by name, albeit usually in handwritten Japanese only, and you could ask the cops for directions too.
The final boss was the non-linear numbering house scheme though. Some friends and I once spent a fruitless hour searching for the HR Giger bar in Tokyo, which we knew was at X-Y-Z, but only managed to find X-Y-(Z+1) and X-Y-(Z-1).
As far as I can tell, in a dense Japanese city, directions start relative to a particular train station, then use distances and landmarks to go from there.
That's pretty much how directions work outside of major cities in the US too, except for a lack of train stations, and the more rural you get the more landmarks are replaced with shibboleths.
"Go about ten minutes that way until you see the old Smith farm. [The Smiths haven't lived there for three generations.] Then take a left and head five minutes down Mill Road. [It's actually called something else, has never had a mill on it in its entire history, but is called that by the locals nonetheless.]" Eventually you start wondering when you tripped and fell into a Discworld novel.
As well characterized by Slim Dusty and Ernie Constance in "Just can't miss it, mate" https://archive.org/details/ka-you-just-cant-miss-it-mate
In the UK and Ireland you navigate by pubs, since it is topologically impossible to traverse any connected area without passing at least one pub. The granularity of directions is therefore highest in Dublin (c.f. Joyce, Ulysses, 1922)
I even hear these kind of place names on fire/police radio.
How I navigate in Japan:
1. Get the hotel / company / restaurant's business card with the address in kanji
2. Show that to your taxi driver
3. Taxi driver ignores the address and puts the phone number in their GPS
or
1. Take the train/metro to the approximate location
2. Your destination will be noted in the signage by exit number
Step (2) is especially fun at Shinjuku, at which point you call your friend who says they'll meet you at the South Gate. Which one though?
When I was in Japan in 2008, I found if I asked someone how to get somewhere they would often just walk you there, even if it was several blocks.
Landmarks. It's similar to how directions are given in other countries with large urban areas and shaky addressing systems.
"Turn left at this coffee shop, right at the second church..."
When my wife first moved to the U.S. from Korea I couldn't figure out why she was so insistent on using landmarks to navigate, but eventually learned this was the reason. When we went out to rural areas where I grew up, she would become completely lost because it's not generally as landmark dense and the mental system she had accumulated didn't work as well. That was when she finally had to learn the U.S. system.
For shorter distances you use landmarks, like known chains of convenience stores or restaurants. For longer distances, larger roads do have route numbers (county route 69, national route 3, etc).
Most major roads in Japanese cities have names. Smaller roads do not. But except for certain cities (Kyoto, Sapporo perhaps) they're just not used for addresses.
A really lame article about an actually very interesting system.
Which block is to the left of 40 on that map of Tokyo?
Also part of 40. (In Google Maps, you can click on a building to get its address.) Actually, for me today the ‘40’ renders on the west part.