Thats just moving things around. At one point in your program, you'll have to pick which class to instantiate, probably from a Person factory of some kind, that will create a new Student, Teacher or Principal. This may be abstracted away with some magic in the ORM that is loading it from your database, it may be a factory you wrote yourself, but that "switch" will exist somewhere. Maybe you're going to use some kind of reflection to do it without a switch, maybe you'll have a lookup table of types, but somewhere that decision has to be made.
Now, if that decision is made somewhere, the same place could configure the behaviors by composition. So you could, at that point, simply have Persons, and have a constructor or setter method that take a move strategy as an argument and saved as a private member. Then the move method would simply invoke it.
This way, a Person can move in any way they need to, without any tight coupling. Here you only have "move", but there could be countless of things people can do, and putting them in distinct buckets falls apart once your system gets any kind of reasonable complexity.
That is when you see insane inheritance hierarchy that get refactored every other day because people forgot that transfer students and transfer teachers have things in common but someone made them inherit from student instead of a class between person and teacher/student.
Composition, mixin, strategies, and other loose coupling patterns that don't associate actions with "is a" relationships are so, so much more flexible for systems you have to maintain for a significant amount of time... Its kind of ironic that out of all languages, Javascript got it right, and did so by accident.
Now that Uber's there, a lot of people choose not to own cars, or people who used to depend on buses and subways go out to certain places that were not convenient before, more often.
So if it goes away, they'll just go back to their old routine, which is fine. Point is, taxi cabs are NOT worth the trouble. If I have the choice between going to restaurant A by subway, or B by Uber, it may just come down to which restaurant I like most. If my choice is A by subway or B by Taxi, its going to be A, no contest.
You are right about the primary reason for this, but its not the only reason. Medias are regulated in a lot of countries, and regulations are different from one to the other. If they didn't go through the legal red time in a given country, they can't necessarily license it. Sometimes the contracts could simply be tied to some marketing campaign, or licensing is delayed to avoid conflicting with something else.
Sometimes...just sometimes...there's even good reasons for it.
Except the reason people are jump hopping has nothing to do with money. It used to be that way during the dotcom recovery, but today? Its been quite a few years since I honestly heard someone hop for money reason. And yes, there's always the occasional exception... Oracle has a low turnover rate in some of their offices. Mainly because its the only big name company that will hire some of these people, so they don't dare leaving.
But in the bay area? Most people hopping are doing it from startup to startup. Not for money reasons, but because they got bored of their previous job, and because honestly, they can. Lately, its almost a given for a lot of people that they'll hop after 1-3 years, regardless of how things are going, just for some change. Loyalty? Its a lot more personal now. Devs are not loyal to their company, they're loyal to the people they work with. These people hop, they hop with one of them.
Now thats not everyone, so of course you'll find exceptions...but even the best places to work for still see the same turnover rate. Most of the ones with low turnover are the places you DONT want to work at.
Devs (like a lot of people in any profession) are pretty damn good at interrupting themselves. People checking their personal email, sport team result, some unrelated chat room, etc. If you watched most devs without them knowing you're looking, without office related interruption, I'd be surprised if half of them went for 30 minutes without interrupting themselves.
Even Google will take you without a degree if you're good enough... I don't have a degree, and honestly no one has even looked in years. I also don't remember when's the last time I heard it come up during an interview review, and I've worked at a lot of places...
I mean sure, if you don't have any experience, people will look for SOMETHING to gauge you to see if its worth their time to phone screen you. But you said "the USA is full of talent". If they're talented, actually talented, they got the skills SOMEWHERE, so they have some kind of experience to show, even if its just a github account.
Then of course there's a lot of people who think they're good but aren't...
hmm? maybe I worded it wrong? All I mean is that if you assume people are always getting better at things they do 40 hours a week, the day they decide to leave your company, is always the day they're the best at what they do (within the context of their employment for you).
I didn't mean people peek after 2 years specifically.
Because it doesn't take very long, and the chef is likely to still be working there in 2 years.
In software engineering, the average time you can reasonably expect someone to stay working for you, regardless of salary or conditions/perks, is about 2~ years. Much less in startup hotbeds like SF.
Now, let say you have some reasonably complex stuff, not too crazy, but not trivial either...maybe you use slightly less common languages (let say Scala over Java...not obscure, but not everyone learns Scala in school), and it takes 4 months for someone to be able to be left alone and do their thing (I'm making numbers up) They still won't be amazingly familiar with your particular business/domain and the ins and outs for a year, and because humans generally always improve with time, they'll be at their peek at the end of the 2 year.
If you hire someone who already knows your technologies, you might be able to reduce that 4 months to 2 months instead. That's 2 more months of peek productivity at the tail end when your engineer is at their best. That could translate in hundreds of thousands, of even millions of dollars depending on the size of your business.
And that is why everyone's hunting down pre-trained people. Of course, then you have to weight that with the cost of not hiring anyone at all, and decide whats best.
As that get widespread, the "big time savings on housing" will be a wash though. Salaries are so high in SF, Boston, NYC, etc, because you have to pay someone that much if you want them to move there, because of housing cost.
And right now, if someone telecommute for one of those companies, salaries stay roughly the same, because that would be a dick move for it not to be.
But companies located in low cost areas don't pay nearly as much.... once its the norm to telecommute, they'll pay you accordingly, since there won't be incentive to do otherwise. You already see it in the couple of telecommute-only companies. Since they have a very large pool of people to pick from, many in low cost areas, they don't need to raise the bid nearly as high to get people.
There's more to it than just that though. My current employer was founded by engineers...they know damn well how it goes, and they've been only hiring people they feel they can trust (break that trust and you're out damn quick...but that pretty much never happens. People are pretty honest).
People work from home (legitimately) left and right, schedules are nearly random (some people come in at 3 pm and work until midnight as their normal schedule, others are in at 5 am and are out right for time for a late lunch...).
We can still get face time when needed and collaborate in real time without friction of shitty tools though (everyone uses Slack or whatever when that's not needed). The inability to have ad-hoc face to face meetings without flying people over is just a deal breaker.
To be able to scale the company, some groups are in separate offices at other locations, but they take care to keep engineering together, sales together, design together, etc. Having engineering and sales videoconference once a month isn't too bad, but having engineering people have to teleconference with each other whenever text isn't good enough would just slow things down.
One day they'll hit the limit of how many they can realistically hire in one location for a given group, then things will change...but you probably want to avoid it if you can.
The vast majority of people won't relocate. People in the tech world are more likely to want to, and people in 3rd world countries are very likely to want to, but in general, people won't.
So if you need to hire 50 great engineers, your best bet is to go where the highest concentration of them are. Even having to compete with hundreds of other companies, its still better (the ones you lose to others can be made up by poaching). If you go and open up shop in the middle of nowhere, you'll never fill up a large team. Now some of the cities you mentionned are ok (ie: Seattle) too. SV isn't the only spot, of course.
Telecommuting only works for a small percentage of top of the top, because phones suck, there's no great videoconference solution out there (No, i know which one you're talking about, it sucks. No, that other one sucks too), and text-only communication makes you lose all the non-verbal, making such communication inefficient for complex matters (it works great as a complement though).
End result: companies need to open up shop in hot spots, and pay the insane amount of $$$ for both real estate and inflated engineer salaries. A few positions can still be filled by remote workers of course, but not the whole thing. Hell, even companies with international offices in the same timezone (ie: a NYC financial with an office in Montreal) have issues with those. It works to some extent, which is why they do it, but its far from ideal.
Yup, thats my problem with movie theaters lately. They stopped enforcing common decency a long time ago, but overall people would be somewhat decent. There was always the ONE dick who wouldn't shut up, but now its the norm more than the exception. And all those people who just can't stop texting continually (if they're not downright talking on the phone). And if you complain, you're the one who "needs to deal with it".
So as everything else in our society, you just have to isolate yourself (because even if you try to just group up with like minded individuals, someone will slip in just to troll you). And then we wonder why there's such big gaps between various groups in the US...
It depends on the field though. The issue with new grads as devs, is that in big tech hubs, the average developer stays at a company around 1 (SF, Seattle) to 2 (Boston, not sure where NYC stands, haven't seen statistics on it) years.
So you get your new grad, you train them....and then they're gone. Offering more benefits or money doesn't help, because they're often leaving for some startup with their friends and aren't even going to get paid because they're "founders" (heck, if you paid them more they'd find they have enough money to quit for a startup sooner).
You have very little time to train someone when you expect them to leave in 12-24 months.
I remember an issue I had a few months ago... we were doing some image processing using HTML canvas element on a web app... Then we wanted a nightly job to use the same code, so we whip out a node.js script. Once it was done, to make sure it worked the same way, we compared the result...
They were different. Spent 2 days trying to debug it (they were using the same code for the most part, wtf?).
At the time, I didn't know about http://en.wikipedia.org/wiki/Canvas_fingerprintingcanvas fingerprinting Most of the time, different computers will generate equivalent, but different at the binary level, images from html canvas.
And there's always the good old floating point operations. ie: 0.2 * 3 = 0.6000000000000001
So its already everywhere, just not everywhere enough that we've been forced to deal with it (those things are usually just afterthought and end up in bugs). Soon, they won't be.
Most of the resources going against piracy are to fight day 1 release (or before) piracy though.
If you had a magical wand, and told the big corps: "100% of people who piracy your stuff before or during the first 2 weeks after release will get caught and get the death penalty. For the rest of 1 year, current copyright laws apply, and we have a magical way to catch everyone who so. After 1 year, the shit is public domain".
They'd probably take that deal. The crazy copyright extensions is just them taking what they can get.
Right, because writing software is just like plugging in to the quadratic equation, and any monkey or "n00b" can do it. I'm extremely curious what field you're in that you have this perspective.
Writing software once the design is figured out IS basically plugging in the equation. If I tell you "make a tool in our image editor that works like photoshop's magic wand to select areas", thats hard.
If I give you a paper with the algorithm to detect an image area as the above, and sit you in front of a codebase that already has the infrastructure to create tools for the editor, it will be downright trivial.
My point is that before you can study it, you have to define what productivity means. The study's conclusion will highly depend on it. If you define productivity as amount of papers written, or amount of line of code written, open layout is -terrible-.
My point is simply that there's a lot more to productivity than: #1 individual productivity and #2 writing stuff as quickly as possible.
Also, even if you take all that in consideration, its still very possible the open office layout will fall short. I'm only saying ONE particular, very narrow argument against it, is wrong. The others may still apply.
How I've seen it done is generally a couple of things.
A) the walls are all using that paint that let you use arbitrary walls as white board. If there's pillars, paint those too. They're not as big, but they're often sufficient for quick sketches, and they have 4 sides.
B) Desks often have short separators. Those are also often whiteboards/magnet.
C) Lots of small huddle rooms good for a quick ad-hoc 2-4 people meeting. Those rooms can't be booked and are so people just hop in and out.
D) Lots and lots of pair programming. Not a whiteboard, but for a lot of usages, it serves the same purpose.
You'd be surprised. Now, don't get me wrong, I worked at a company like that and quit pretty quick because I couldn't stand it. But there's a lot of people out there (a lot!) who are dedicated enough in their work that they'll happily be at 100% from 8 to 7 every day. These people just all work at the same places.
I mean, its not too surprising: there's a lot of jobs where you can't access your phone aside during breaks...and any job where you have to run/stand around all day didn't exactly let you do personal things (again aside during breaks) until cellphones came up.
So obviously people can do it, some good people do. I just prefer working at the countless companies where its not needed:)
We had that discussion at work today over that article. Several people pointed out they were far more productive alone, with the lights off, in a corner, than at their desk and that it proved the open floor plan was bad.
(We're talking software engineers)
My personal take is: almost anyone (who doesn't need babysitting) will be more productive alone in a distraction free area. That is, more productive doing the part of the job that a monkey can do. I can bang out thousands of lines of code very quickly if no one's bothering me, sure.
But here's the catch: that's not the hard part of the job. (almost) anyone could do that. The hard part is the design, architecture, problem solving. Most of the time, those are better done in group. They may seem worse sometimes: arguing feels counterproductive and a waste of time. But no one's perfect and no one knows everything, so being able to bounce off ideas from the person next to you at will can prevent million dollar mistakes. Once the problem is solved, and just typing code as quick as possible is the only thing left to do, sure, work from home if you want, but don't fool yourself that you're doing anything worth a lot.
Then, let's go with the assumption the above is not true: you're a god developer who never makes mistake and figures out everything on their own instantly. There's a lot of people who could use bouncing ideas off of YOU, who could discuss things with you, and may waste time, get blocked, or worse, make mistakes, if they can't get a hold of you in a timely manner. Sure, it will feel like you can't get anything done, but again, once the problem is solved, anyone can implement it: those "n00bs" that are pestering you will be able to do the easy part once they got the info they needed off of you.
And once an office reach a certain size, sending an email or an instant message then waiting 10 minutes so you can be in a good spot to answer adds up to a lot of wasted time. In the end, there's a reason some very successful businesses keep paying a fortune in engineer salaries in SF, Boston, NY or Seattle to keep a critical mass of devs together. There's no substitute and it can often be worth the insane markup.
Now there does come a time when you have to get the easy shit done, and there's a lot of easy shit to do. Library atmosphere sections in an office can take care of that. But if you're always there, or even if you're not but always have the noise cancelling headphones on because you're "OMG SO MUCH MORE PRODUCTIVE", you're honestly part of the problem. You're gonna look good in your yearly review, people may think you're fucking awesome. But as a small part of something bigger, you're just fucking everyone else over.
Even with the same rendering and js engine... some stuff like auto-complete behavior can differ (it was a big plague for AngularJS dev for a while until they changed how they detected textfields updates), security settings can be different (ie: what kind of certificates Chrome consider valid for SSL vs other browsers), a bunch of weirdo edge cases (jsonp over https failing in very specific scenarios on some android versions), and so on and so forth.
Safari's pretty damn popular (relatively speaking. Its too popular to ignore, generally), and has some annoying edge cases that tend to pop up from version to version.
Generally speaking, if you have a reasonably complex website (or usually, web app), and you didn't test a specific version of a specific browser, you have some bugs in it.
The only issue I have with this (and its just a minor hiccup), is that once in the real world, you won't always be able to get things just the right way that suits you. If you're a software engineer, yeah, you probably can hop around until you find just the right thing... Same if you can make your own company, to some extent, and make it successful.
But for most people, they end up having to deal with shit that doesn't suit them. In a way, the one most useful skill you gain from school is just that: learning how to go out of your comfort zone. Even College: honestly what you learn isn't all that useful for the most part, regardless of degree. But you learn to take an arbitrary topic, with a lot of curve balls, in an environment that honestly, kind of sucks, and suck it up for 4 years and get it done.
If everyone gets treated like cute little unique snowflakes, they'll be in for a HELL of a wake up call. Sure, there's some people who deal with it well..but for a lot of people, that will be worse.
Slashdot Mirror
Thats just moving things around. At one point in your program, you'll have to pick which class to instantiate, probably from a Person factory of some kind, that will create a new Student, Teacher or Principal. This may be abstracted away with some magic in the ORM that is loading it from your database, it may be a factory you wrote yourself, but that "switch" will exist somewhere. Maybe you're going to use some kind of reflection to do it without a switch, maybe you'll have a lookup table of types, but somewhere that decision has to be made.
Now, if that decision is made somewhere, the same place could configure the behaviors by composition. So you could, at that point, simply have Persons, and have a constructor or setter method that take a move strategy as an argument and saved as a private member. Then the move method would simply invoke it.
This way, a Person can move in any way they need to, without any tight coupling. Here you only have "move", but there could be countless of things people can do, and putting them in distinct buckets falls apart once your system gets any kind of reasonable complexity.
That is when you see insane inheritance hierarchy that get refactored every other day because people forgot that transfer students and transfer teachers have things in common but someone made them inherit from student instead of a class between person and teacher/student.
Composition, mixin, strategies, and other loose coupling patterns that don't associate actions with "is a" relationships are so, so much more flexible for systems you have to maintain for a significant amount of time... Its kind of ironic that out of all languages, Javascript got it right, and did so by accident.
Now that Uber's there, a lot of people choose not to own cars, or people who used to depend on buses and subways go out to certain places that were not convenient before, more often.
So if it goes away, they'll just go back to their old routine, which is fine. Point is, taxi cabs are NOT worth the trouble. If I have the choice between going to restaurant A by subway, or B by Uber, it may just come down to which restaurant I like most. If my choice is A by subway or B by Taxi, its going to be A, no contest.
You are right about the primary reason for this, but its not the only reason. Medias are regulated in a lot of countries, and regulations are different from one to the other. If they didn't go through the legal red time in a given country, they can't necessarily license it. Sometimes the contracts could simply be tied to some marketing campaign, or licensing is delayed to avoid conflicting with something else.
Sometimes...just sometimes...there's even good reasons for it.
Except the reason people are jump hopping has nothing to do with money. It used to be that way during the dotcom recovery, but today? Its been quite a few years since I honestly heard someone hop for money reason. And yes, there's always the occasional exception... Oracle has a low turnover rate in some of their offices. Mainly because its the only big name company that will hire some of these people, so they don't dare leaving.
But in the bay area? Most people hopping are doing it from startup to startup. Not for money reasons, but because they got bored of their previous job, and because honestly, they can. Lately, its almost a given for a lot of people that they'll hop after 1-3 years, regardless of how things are going, just for some change. Loyalty? Its a lot more personal now. Devs are not loyal to their company, they're loyal to the people they work with. These people hop, they hop with one of them.
Now thats not everyone, so of course you'll find exceptions...but even the best places to work for still see the same turnover rate. Most of the ones with low turnover are the places you DONT want to work at.
Devs (like a lot of people in any profession) are pretty damn good at interrupting themselves. People checking their personal email, sport team result, some unrelated chat room, etc. If you watched most devs without them knowing you're looking, without office related interruption, I'd be surprised if half of them went for 30 minutes without interrupting themselves.
Even Google will take you without a degree if you're good enough... I don't have a degree, and honestly no one has even looked in years. I also don't remember when's the last time I heard it come up during an interview review, and I've worked at a lot of places...
I mean sure, if you don't have any experience, people will look for SOMETHING to gauge you to see if its worth their time to phone screen you. But you said "the USA is full of talent". If they're talented, actually talented, they got the skills SOMEWHERE, so they have some kind of experience to show, even if its just a github account.
Then of course there's a lot of people who think they're good but aren't...
hmm? maybe I worded it wrong? All I mean is that if you assume people are always getting better at things they do 40 hours a week, the day they decide to leave your company, is always the day they're the best at what they do (within the context of their employment for you).
I didn't mean people peek after 2 years specifically.
Because it doesn't take very long, and the chef is likely to still be working there in 2 years.
In software engineering, the average time you can reasonably expect someone to stay working for you, regardless of salary or conditions/perks, is about 2~ years. Much less in startup hotbeds like SF.
Now, let say you have some reasonably complex stuff, not too crazy, but not trivial either...maybe you use slightly less common languages (let say Scala over Java...not obscure, but not everyone learns Scala in school), and it takes 4 months for someone to be able to be left alone and do their thing (I'm making numbers up) They still won't be amazingly familiar with your particular business/domain and the ins and outs for a year, and because humans generally always improve with time, they'll be at their peek at the end of the 2 year.
If you hire someone who already knows your technologies, you might be able to reduce that 4 months to 2 months instead. That's 2 more months of peek productivity at the tail end when your engineer is at their best. That could translate in hundreds of thousands, of even millions of dollars depending on the size of your business.
And that is why everyone's hunting down pre-trained people. Of course, then you have to weight that with the cost of not hiring anyone at all, and decide whats best.
I'm pretty sure they were given an example of an international company...
As that get widespread, the "big time savings on housing" will be a wash though. Salaries are so high in SF, Boston, NYC, etc, because you have to pay someone that much if you want them to move there, because of housing cost.
And right now, if someone telecommute for one of those companies, salaries stay roughly the same, because that would be a dick move for it not to be.
But companies located in low cost areas don't pay nearly as much.... once its the norm to telecommute, they'll pay you accordingly, since there won't be incentive to do otherwise. You already see it in the couple of telecommute-only companies. Since they have a very large pool of people to pick from, many in low cost areas, they don't need to raise the bid nearly as high to get people.
There's more to it than just that though. My current employer was founded by engineers...they know damn well how it goes, and they've been only hiring people they feel they can trust (break that trust and you're out damn quick...but that pretty much never happens. People are pretty honest).
People work from home (legitimately) left and right, schedules are nearly random (some people come in at 3 pm and work until midnight as their normal schedule, others are in at 5 am and are out right for time for a late lunch...).
We can still get face time when needed and collaborate in real time without friction of shitty tools though (everyone uses Slack or whatever when that's not needed). The inability to have ad-hoc face to face meetings without flying people over is just a deal breaker.
To be able to scale the company, some groups are in separate offices at other locations, but they take care to keep engineering together, sales together, design together, etc. Having engineering and sales videoconference once a month isn't too bad, but having engineering people have to teleconference with each other whenever text isn't good enough would just slow things down.
One day they'll hit the limit of how many they can realistically hire in one location for a given group, then things will change...but you probably want to avoid it if you can.
For scaling teams.
The vast majority of people won't relocate. People in the tech world are more likely to want to, and people in 3rd world countries are very likely to want to, but in general, people won't.
So if you need to hire 50 great engineers, your best bet is to go where the highest concentration of them are. Even having to compete with hundreds of other companies, its still better (the ones you lose to others can be made up by poaching). If you go and open up shop in the middle of nowhere, you'll never fill up a large team. Now some of the cities you mentionned are ok (ie: Seattle) too. SV isn't the only spot, of course.
Telecommuting only works for a small percentage of top of the top, because phones suck, there's no great videoconference solution out there (No, i know which one you're talking about, it sucks. No, that other one sucks too), and text-only communication makes you lose all the non-verbal, making such communication inefficient for complex matters (it works great as a complement though).
End result: companies need to open up shop in hot spots, and pay the insane amount of $$$ for both real estate and inflated engineer salaries. A few positions can still be filled by remote workers of course, but not the whole thing. Hell, even companies with international offices in the same timezone (ie: a NYC financial with an office in Montreal) have issues with those. It works to some extent, which is why they do it, but its far from ideal.
Yup, thats my problem with movie theaters lately. They stopped enforcing common decency a long time ago, but overall people would be somewhat decent. There was always the ONE dick who wouldn't shut up, but now its the norm more than the exception. And all those people who just can't stop texting continually (if they're not downright talking on the phone). And if you complain, you're the one who "needs to deal with it".
So as everything else in our society, you just have to isolate yourself (because even if you try to just group up with like minded individuals, someone will slip in just to troll you). And then we wonder why there's such big gaps between various groups in the US...
It depends on the field though. The issue with new grads as devs, is that in big tech hubs, the average developer stays at a company around 1 (SF, Seattle) to 2 (Boston, not sure where NYC stands, haven't seen statistics on it) years.
So you get your new grad, you train them....and then they're gone. Offering more benefits or money doesn't help, because they're often leaving for some startup with their friends and aren't even going to get paid because they're "founders" (heck, if you paid them more they'd find they have enough money to quit for a startup sooner).
You have very little time to train someone when you expect them to leave in 12-24 months.
I remember an issue I had a few months ago... we were doing some image processing using HTML canvas element on a web app... Then we wanted a nightly job to use the same code, so we whip out a node.js script. Once it was done, to make sure it worked the same way, we compared the result...
They were different. Spent 2 days trying to debug it (they were using the same code for the most part, wtf?).
At the time, I didn't know about http://en.wikipedia.org/wiki/Canvas_fingerprintingcanvas fingerprinting Most of the time, different computers will generate equivalent, but different at the binary level, images from html canvas.
And there's always the good old floating point operations. ie: 0.2 * 3 = 0.6000000000000001
So its already everywhere, just not everywhere enough that we've been forced to deal with it (those things are usually just afterthought and end up in bugs). Soon, they won't be.
Most of the resources going against piracy are to fight day 1 release (or before) piracy though.
If you had a magical wand, and told the big corps: "100% of people who piracy your stuff before or during the first 2 weeks after release will get caught and get the death penalty. For the rest of 1 year, current copyright laws apply, and we have a magical way to catch everyone who so. After 1 year, the shit is public domain".
They'd probably take that deal. The crazy copyright extensions is just them taking what they can get.
Right, because writing software is just like plugging in to the quadratic equation, and any monkey or "n00b" can do it. I'm extremely curious what field you're in that you have this perspective.
Writing software once the design is figured out IS basically plugging in the equation. If I tell you "make a tool in our image editor that works like photoshop's magic wand to select areas", thats hard.
If I give you a paper with the algorithm to detect an image area as the above, and sit you in front of a codebase that already has the infrastructure to create tools for the editor, it will be downright trivial.
My point is that before you can study it, you have to define what productivity means. The study's conclusion will highly depend on it. If you define productivity as amount of papers written, or amount of line of code written, open layout is -terrible-.
My point is simply that there's a lot more to productivity than: #1 individual productivity and #2 writing stuff as quickly as possible.
Also, even if you take all that in consideration, its still very possible the open office layout will fall short. I'm only saying ONE particular, very narrow argument against it, is wrong. The others may still apply.
How I've seen it done is generally a couple of things.
A) the walls are all using that paint that let you use arbitrary walls as white board. If there's pillars, paint those too. They're not as big, but they're often sufficient for quick sketches, and they have 4 sides.
B) Desks often have short separators. Those are also often whiteboards/magnet.
C) Lots of small huddle rooms good for a quick ad-hoc 2-4 people meeting. Those rooms can't be booked and are so people just hop in and out.
D) Lots and lots of pair programming. Not a whiteboard, but for a lot of usages, it serves the same purpose.
You'd be surprised. Now, don't get me wrong, I worked at a company like that and quit pretty quick because I couldn't stand it. But there's a lot of people out there (a lot!) who are dedicated enough in their work that they'll happily be at 100% from 8 to 7 every day. These people just all work at the same places.
I mean, its not too surprising: there's a lot of jobs where you can't access your phone aside during breaks...and any job where you have to run/stand around all day didn't exactly let you do personal things (again aside during breaks) until cellphones came up.
So obviously people can do it, some good people do. I just prefer working at the countless companies where its not needed :)
We had that discussion at work today over that article. Several people pointed out they were far more productive alone, with the lights off, in a corner, than at their desk and that it proved the open floor plan was bad.
(We're talking software engineers)
My personal take is: almost anyone (who doesn't need babysitting) will be more productive alone in a distraction free area. That is, more productive doing the part of the job that a monkey can do. I can bang out thousands of lines of code very quickly if no one's bothering me, sure.
But here's the catch: that's not the hard part of the job. (almost) anyone could do that. The hard part is the design, architecture, problem solving. Most of the time, those are better done in group. They may seem worse sometimes: arguing feels counterproductive and a waste of time. But no one's perfect and no one knows everything, so being able to bounce off ideas from the person next to you at will can prevent million dollar mistakes. Once the problem is solved, and just typing code as quick as possible is the only thing left to do, sure, work from home if you want, but don't fool yourself that you're doing anything worth a lot.
Then, let's go with the assumption the above is not true: you're a god developer who never makes mistake and figures out everything on their own instantly. There's a lot of people who could use bouncing ideas off of YOU, who could discuss things with you, and may waste time, get blocked, or worse, make mistakes, if they can't get a hold of you in a timely manner. Sure, it will feel like you can't get anything done, but again, once the problem is solved, anyone can implement it: those "n00bs" that are pestering you will be able to do the easy part once they got the info they needed off of you.
And once an office reach a certain size, sending an email or an instant message then waiting 10 minutes so you can be in a good spot to answer adds up to a lot of wasted time. In the end, there's a reason some very successful businesses keep paying a fortune in engineer salaries in SF, Boston, NY or Seattle to keep a critical mass of devs together. There's no substitute and it can often be worth the insane markup.
Now there does come a time when you have to get the easy shit done, and there's a lot of easy shit to do. Library atmosphere sections in an office can take care of that. But if you're always there, or even if you're not but always have the noise cancelling headphones on because you're "OMG SO MUCH MORE PRODUCTIVE", you're honestly part of the problem. You're gonna look good in your yearly review, people may think you're fucking awesome. But as a small part of something bigger, you're just fucking everyone else over.
Even with the same rendering and js engine... some stuff like auto-complete behavior can differ (it was a big plague for AngularJS dev for a while until they changed how they detected textfields updates), security settings can be different (ie: what kind of certificates Chrome consider valid for SSL vs other browsers), a bunch of weirdo edge cases (jsonp over https failing in very specific scenarios on some android versions), and so on and so forth.
Safari's pretty damn popular (relatively speaking. Its too popular to ignore, generally), and has some annoying edge cases that tend to pop up from version to version.
Generally speaking, if you have a reasonably complex website (or usually, web app), and you didn't test a specific version of a specific browser, you have some bugs in it.
Fucking MacOSX hidden scrollbars.
The only issue I have with this (and its just a minor hiccup), is that once in the real world, you won't always be able to get things just the right way that suits you. If you're a software engineer, yeah, you probably can hop around until you find just the right thing... Same if you can make your own company, to some extent, and make it successful.
But for most people, they end up having to deal with shit that doesn't suit them. In a way, the one most useful skill you gain from school is just that: learning how to go out of your comfort zone. Even College: honestly what you learn isn't all that useful for the most part, regardless of degree. But you learn to take an arbitrary topic, with a lot of curve balls, in an environment that honestly, kind of sucks, and suck it up for 4 years and get it done.
If everyone gets treated like cute little unique snowflakes, they'll be in for a HELL of a wake up call. Sure, there's some people who deal with it well..but for a lot of people, that will be worse.
Didn't he announce he was doing just that last year?
Thats a different discussion. However, what I meant is that you probably shouldn't have your data access and your front end templates interweaving...