5 ms·

"It's not as if Python or Ruby have set out to be deliberately difficult. They've grown difficulty because they're used to solve real problems, and real problem

"It's not as if Python or Ruby have set out to be deliberately difficult. They've grown difficulty because they're used to solve real problems, and real problems are hard." The problem of getting from the East coast to the West coast has not changed in 100 years (move body until done), yet when was the last time you had to set the spark and choke and hand-crank a car? Cars solve real problems, and today's cars are sophisticated enough that essentially anyone can use one. To solve real problems. Certainly the problems solved by a car are in a more narrow domain than programming languages. But cars themselves are in a narrower domain than internal combustion engines. Edit: stealth.

mechanical_fish 15y ago
Certainly the problems solved by a car are in a more narrow domain than programming languages. You do realize that this is like saying "certainly two times N is smaller than two to the Nth power"? ;)
jacalata 15y ago
Cars aren't used to cross the continent today, airplanes are. And most people have no idea how to use (pilot) an airplane, they pay someone else to do it for them.
- read_wharf 15y ago
  For now.
  - BerislavLopac 15y ago
    They never will. Before that happens we'll have fully automated planes operated by -- guess what? -- software. Software is not really comparable to cars because it's not a solution -- it's a tool to build solutions. Comparably, we have devised easier ways to drive nails into wood (nailguns instead of hammers), but nails still exist and have to be driven into wood. So I don't think software will ever really be casual, just like building construction will never really be casual. Just too many variables to take into consideration.
drewcrawford 15y ago
> The problem of getting from the East coast to the West coast has not changed in 100 years But computing problems _have changed_. Not necessarily in areas like problem size (although that has happened) but also in areas of usability and convenience. A much more apt analogy would be: the problem of stopping at the grocery store after work is 50 years old, why can't I get an optimal set of directions that accounts for rush hour traffic in 2012? I can give you any number of reasons: we've built more roads, there are more drivers, travelling salesman turned out to be a hard problem, game theory is hard, etc. In some dimensions (cranking the car, proclivity of gas stations, etc.) the situation has improved, but solving some problems re-emphasize the ones we haven't solved.
sophacles 15y ago
I'm sorry, but this analogy is awful. Getting from coast to coast certainly HAS changed in the last 100 years. 100 years ago, you had railroads or mostly dirt roads. A crap ton of infrastructure has been put into it (interstates and the intensely complex logistics of maintaining them). A pile of services exists around this problem (airlines, and the even more complex logistics of keeping them running safely and profitably). Going broader, you are ignoring all types of "get between coasts" that apply to stuff rather than "me" - what is the best way to get a pile of widgets from LA to NYC? Airplane? Train? Some trucks? -- the answer is "well that depends on $logistical_considerations". Going narrower, choosing the right car for you is a pretty complicated process, trading off features, options, gas mileage, capacity and so on - just like any set of programing options, only internalized by the culture". Further, a car is simple to use out of the box, but customization beyond do-dads and bling? Nope, not really easier than it ever was. Different tools? sure. Different base skills? not really. Anyone can do it? Not so much. And before I go into tech equivalencies here, if you think cars are pretty much the same and can be abstracted to "the car" for the purposes of this discussion, go buy a van or pick up truck. Watch your social graph get all buzzy and see the request pour in for help, turns out your form of tradeoff in "car" are very very useful to people, and suddenly a car isn't just a car. Further, I would argue that we have done wonders for the tech equivalent of "car". We have amazing computation devices in our pocket that just do what people never knew they wanted to do in the form of "phone". We have awesome abilities to have web presence unthinkable (outside of internet famous blog machines) 10 years ago, in the form "social graph". These are great and just work for most people, until they need something different... In the infrastructure department: the ability to do semi-custom stuff has grown in ways unimaginable in the car world. Need a website to do your magic? Here are a bunch of platforms like app engine or heroku, all you need to do is get your custom stuff in place, don't worry about the machinery to feed it. Up and down the customization level there are more entry points as well. Going back to the car world, how easy is it really to get a ricer put together? How easy is it to put in crazy hydrolics or a good thumping stereo system? Need to rip off the doors, reshape the body and put in a different engine? Good luck getting a doctor to just do that without serious learning curve. Basically, my point here is that hard stuff is hard. If it isn't your specific field, it is easy to just blow off the stuff that caters to the masses as "they made it easy", and at the same time, it is easy to blow off the stuff in your field which caters to the masses as "not done well". Final thought: a lot of people look at cars and computers the same way, as soon as the littlest thing goes wrong, they freak. or they take it to the mechanic who asks "why haven't you changed the oil in 7 years? (c.f. our IT guys who ask why they don't have anti-virus installed).

mechanical_fish 15y ago

Certainly the problems solved by a car are in a more narrow domain than programming languages. You do realize that this is like saying "certainly two times N is smaller than two to the Nth power"? ;)

jacalata 15y ago

Cars aren't used to cross the continent today, airplanes are. And most people have no idea how to use (pilot) an airplane, they pay someone else to do it for them.

read_wharf 15y ago
For now.
- BerislavLopac 15y ago
  They never will. Before that happens we'll have fully automated planes operated by -- guess what? -- software. Software is not really comparable to cars because it's not a solution -- it's a tool to build solutions. Comparably, we have devised easier ways to drive nails into wood (nailguns instead of hammers), but nails still exist and have to be driven into wood. So I don't think software will ever really be casual, just like building construction will never really be casual. Just too many variables to take into consideration.

drewcrawford 15y ago

> The problem of getting from the East coast to the West coast has not changed in 100 years But computing problems _have changed_. Not necessarily in areas like problem size (although that has happened) but also in areas of usability and convenience. A much more apt analogy would be: the problem of stopping at the grocery store after work is 50 years old, why can't I get an optimal set of directions that accounts for rush hour traffic in 2012? I can give you any number of reasons: we've built more roads, there are more drivers, travelling salesman turned out to be a hard problem, game theory is hard, etc. In some dimensions (cranking the car, proclivity of gas stations, etc.) the situation has improved, but solving some problems re-emphasize the ones we haven't solved.

sophacles 15y ago

I'm sorry, but this analogy is awful. Getting from coast to coast certainly HAS changed in the last 100 years. 100 years ago, you had railroads or mostly dirt roads. A crap ton of infrastructure has been put into it (interstates and the intensely complex logistics of maintaining them). A pile of services exists around this problem (airlines, and the even more complex logistics of keeping them running safely and profitably). Going broader, you are ignoring all types of "get between coasts" that apply to stuff rather than "me" - what is the best way to get a pile of widgets from LA to NYC? Airplane? Train? Some trucks? -- the answer is "well that depends on $logistical_considerations". Going narrower, choosing the right car for you is a pretty complicated process, trading off features, options, gas mileage, capacity and so on - just like any set of programing options, only internalized by the culture". Further, a car is simple to use out of the box, but customization beyond do-dads and bling? Nope, not really easier than it ever was. Different tools? sure. Different base skills? not really. Anyone can do it? Not so much. And before I go into tech equivalencies here, if you think cars are pretty much the same and can be abstracted to "the car" for the purposes of this discussion, go buy a van or pick up truck. Watch your social graph get all buzzy and see the request pour in for help, turns out your form of tradeoff in "car" are very very useful to people, and suddenly a car isn't just a car. Further, I would argue that we have done wonders for the tech equivalent of "car". We have amazing computation devices in our pocket that just do what people never knew they wanted to do in the form of "phone". We have awesome abilities to have web presence unthinkable (outside of internet famous blog machines) 10 years ago, in the form "social graph". These are great and just work for most people, until they need something different... In the infrastructure department: the ability to do semi-custom stuff has grown in ways unimaginable in the car world. Need a website to do your magic? Here are a bunch of platforms like app engine or heroku, all you need to do is get your custom stuff in place, don't worry about the machinery to feed it. Up and down the customization level there are more entry points as well. Going back to the car world, how easy is it really to get a ricer put together? How easy is it to put in crazy hydrolics or a good thumping stereo system? Need to rip off the doors, reshape the body and put in a different engine? Good luck getting a doctor to just do that without serious learning curve. Basically, my point here is that hard stuff is hard. If it isn't your specific field, it is easy to just blow off the stuff that caters to the masses as "they made it easy", and at the same time, it is easy to blow off the stuff in your field which caters to the masses as "not done well". Final thought: a lot of people look at cars and computers the same way, as soon as the littlest thing goes wrong, they freak. or they take it to the mechanic who asks "why haven't you changed the oil in 7 years? (c.f. our IT guys who ask why they don't have anti-virus installed).