I don't understand what your saying. Depends on the doctor/field. In my case oncology so needed tracking different stages of the patient workflow, reports for amount of dose delivered to different types of patients etc. No really good off the shelf solution handles all the parts needed since generally speaking you are using a collection of different vendors CT scanners, EMR, PACS, treatment planning, patent scheduling and radiation treatment machines. Heck even software from the same vendor often does things differently because the Phillips Imaging department doesn't talk to the Phillips Oncology department before designing their software.
Well in my case we are very far from customers. Our software is sold in ~80 countries by distributors and ultimately used by accountants. So addresses might not be the clearest example of this but often we have to make things very generic because one countries tax code differs from another, or they have the same data but need it at different times in their process etc. Often when we ask a distributor they ask a few customers I think and we go with what they say but sometimes no one knows because it is a completely new government requirement/process (ex. recently Australia started requiring their version of a personal retirement savings plan to have yearly audits: not required before so no existing paper based system or whatever to automate, we had to guess what they would need when).
Agreed though, whenever you can actually see a customer working is the best. Even better if you are automating something you do as you know both your current process and what you'd ideally like to have (did that a lot when I worked more on the IT side, or as more of a scientist/technician playing around with accelerators).
Well there is a time issue too. At my current work I make software for accountants, in the past I made it for doctors and scientists either group their time is better spent practicing their expertise not making software. Just because you "can" do something doesn't mean your time is well spent doing it, or that you can do it as well as someone that does it all the time. Sure I could install my new windows, but I'd rather pay a guy to do it because I don't like the work, it would take me at least 4X as long to do so and I'm more likely to screw something up than they are.
Yeah that is scary. You can say because the cost of crap code is lost customers/law suits etc,not to mention moral: knowing you are selling crap but not caring enough to actually try to deliver on your promises. They could still factor in all those costs and say Tata is still cheaper sorry. My degrees are in physics and management so if programming doesn't work out I have other domains to move in. I think that is what we need generally in our profession and probably others: you are just a budget cutback away from your grant being denied, your department outsourced etc. If you are truly smart learn multiple useful professions and hope like hell that they don't all go to crap at the same time.
Part of the problem is the expectation that everyone contributes/carries their wait. I'm sorry but if robots truly replace all menial jobs the idiots out their will have no work that makes sense for them to do, unless you subscribe to the Calvinist work ethic: work itself has value. Drop the Jebus and all that is left is: keep the idiots out of the way while the remaining "useful" people work. It is only our sense of fairness and greed for "more" that makes us think that it is a useful use human life to have someone paid full time to flip a burger when a machine can do it just as well and cheaper.
Amen to that. I both don't think it is possible: not everyone has the temperament to stick with a problem for several days till they figure something out, nor desirable (there are a lot of dumb people out there (or at least dumb when it comes to tech)) for everyone to code. I want my lawyer to focus on kicking ass at being a lawyer, by doctor to focus on kicking ass at making me healthier etc. We ain't the only profession that will matter in the future no matter how much tech is around, there is still going to be the need for good art, good music, good health and other advise, good bjs etc. Everyone has their place.
There is bad code for sure. There is also bad architecture and feature specs. In my experience there are often very fuzzy business requirements, sometimes legitimately: the customer can't tell you what they need they just ask for something "better". Example something I'm currently dealing with: the customer asks to be able to add multiple addresses instead of one to each order. Well is there only one address of a given type allowed (postal, physical, fax, etc) or an arbitrary number of each type? If so how do you know which should be listed first? They aren't sure but they need the feature in a couple months so they can demo it to customers.
So I end up making something as general as possible "just in case" someone somewhere might want to do things that way. Then a few months later I expect someone to complain: we screwed up and added 5 postal addresses to one customer, that make no sense why didn't you warn us?
Anyways, sometimes bad architecture comes about by poorly nailed down design requirements. Sometimes that can be helped by asking the customer more, and sometimes they don't know so are no help All you can do as a dev is tell them it is likely going to be 5X more work to do it in a way that will support all scenarios "just in case" vs doing it the "one true way". If they say fine you need to live with the legacy junk until they feel free to free you up to do a refactor on the code and not spend the whole release trying to pound out new features (fat chance in hell). Sometimes the best you can do is write code that doesn't suck to bad in what the expected typical use case is, and drag it forward as long as you can to meet the current requirements. Once it breaks to the point of no return you have to pay all the tax of the lack of good specs you had all along. This doesn't excuse crappy coding at the method level though: that is a matter of skill and hopefully anyone that considers themselves a professional continues to learn from code reviews, training etc and gets better over time. If you don't you suck and need to go.
It is the exact problem I have with some "free" solutions.There are 1000 people that ever downloaded the thing with maybe 100 current users and 3 contributors. It either does what you want or you need to take the time to understand their code, or you need to keep looking. Script kiddies aren't just caused by amateur coders, they are caused by side projects we make without a proper business need/case. We make it to fill such a niche need that you either have to live with the limitations or move on to another solution. That is what the future will be: a google image search effectively of "hot girl" and countless hours spent finding the one you meant, vs hiring someone with the skill to make yet another "script kiddie' that does exactly what you want, or spending the time to actually learn the market and hiring a professional to make something that is generally applicable and/or innovative.
That is just it. It isn't just a matter of can you understand what an if or for loop is. It IS a skill/psychological problem: you need be the type of person that wants to understand how things work all the way down to the metal, and stubborn enough to keep at it till you can get things to work. If anything the more easy to use computers become the more frustrating it will be to fix that 1% of remaining issues. It takes a special type of idiot to keep bashing their heads against the wall when things are already "good enough".
Depends where you live but at least for teachers there are often school boards or larger organization bodies that insure that even the new ones make better salaries than a lot of programmers. So the distribution is compressed around a high salary. A starting salary in Ontario for example is 51k + a pension, 20 sick days a year that can be banked for early retirement. There pension pays out 60% of their best 5 yrs salary (so don't want to be a principal? Well just do it for 5 years because you'll then make another $10k a year or so when retired). 10yrs in teachers are making about 90k a year more if they are department heads or move into VP or principal positions. They also have full medical and dental with no employee portion of the premiums. So in short at least in my experience in the province, they make an about average to high programmers salary with way more vacation, way better pension, and about a $100 a month savings because not paying anything towards their benefits.
Plumbers, electricians etc all can make really good money too. Some aspects of their job sucks and they might not be able to do it till 65 though, so I guess you're paying towards early retirement for them.
Then you get busy street corners where the only time cars move is on the yellows. I feel for the drivers but it sucks as a pedestrian. Some intersections in Toronto where I work there are always about 50 people on each side trying to cross for each light. The only window the drivers turning get a chance to cross is the small window between when cars start stopping for the yellow and when the green the other way (and the cross light for pedestrians with it) kick in. Cabs seem to be the most aggrressive trying to get through that gap.
Yeah, or deploy spike strips:) That would be awesome actually if the jerk that needs to be unsafe for an hour drive so that they can be 3 car lengths closer to home ends up spending $800 and a few hours getting their car towed and all tires replaced: definitely would change the incentives.
That might be the end result of these automatic "emergency" braking systems: they start to enforce sufficient space between vehicles. After all, the manufacturer doesn't what to get sued when your car fails to stop itself in time to avoid a rearend collision. This might also be one of many things manual drivers will have to give up if they don't opt for a driverless car: regulations will gradually be put into place to make an autopilot take over whenever the driver tries to do something that the government doesn't like: can't speed, can't follow too closely, turn into another lane without enough room etc. It might get to the point where everyone can "drive" a car because you can basically stomp on the gas pedal and the car will steer to avoid everything for you, the only "skill" would be in actually getting to where you want to go though I suppose that could be tied into the system that prevents you from turning into someone too.
True somewhat. If you use all that space between you and the guy in front beause you give him a "love tap" the guy behind you has more space/time to stop. In the majority case you probably are better off slamming your brakes I guess but I can see cases where you end up getting rear ended where you with your slower reaction wouldn't have gotten in an accident at all. Say the guy in front of you slams his brakes then releases, your system slams the brakes and you get rear ended because the guy behind you was closer to you than you were to the guy in front and/or there vehicle needs more space to stop.
they guy with the older car behind you from rear ending you instead? Or are these systems going to optimize between the risk of crashing into the guy in front of you vs the risk it will stop too quickly for the guy behind you to respond (yeah I know we all leave sufficient space between us and the car in front to brake)?
I'm curious, at least the apple.ca site doesn't have the weight or dimensions of the iPad Pro. I wonder if the page was made before the final specs were known or is it a matter of if Apple can't say it is thinner and lighter they want to hide it behind all the new/improved specs?
Yep. The other servers are likely to have source code and such (bug/feature ticketing, Stash repos, TFS etc). Why GMail: not sure. We build collaboration software so our products do need to interact with google, Office 365, SharePoint etc etc not sure if they just figured we all need a business account for G Drive anyways to test our software we might as well get the email service too or what.
In most cases if your internet connection is out work stops anyways. My work we us google for mail, and Skype for IM but otherwise have everything in house. Still if we have a glitch work slows down to a crawl. Every time you try to do something and need to look up a manual page or something, oh crap can't access the site. Huh, okay maybe I can do something else... oh crap that bug has a link to a website, can't work on that... It quickly becomes "lunch time" for everyone in the office for that 30min or whatever where the gateway is having issues.
More geographically concentrated. At least it has been my experience for the jobs I look for in software development about 80-90% of jobs are in major cities. You can find a job pretty much everywhere but the big companies hiring dozens of developers at a time overwhelm the one or two positions smaller companies in smaller communities are hiring. You switch jobs you either find another job in the big city you are in or you move 1hr + away (I'm from Canada: huge country, vastly distributed major cities). I think IT is going to become more like that too. You'll still need the low level support in house but the large number of jobs will be concentrated into areas with huge datacentres. In this case likely more rural locations because of cheap power, colder climate, cheap land etc.
They never say the thing tastes like what it is. Never does wine have a "distinctive note of fermented grapes". I also love to read a couple tasting notes about the same thing. One will say "citris, vanilla and anise" another "watermelon", "bacon", "chocolate". Without fail they seem to have complete different components. Then you get in the room with a wine or whiskey snob and watch them discuss the "peach note" in the drink. Fantastic.
Agreed. You can look at it another way though, a force multiplier effect. If I screw up as an intermediate developer, it is likely we find it in code review or pre-release testing and spend a couple weeks longer than we planned doing something say. If the CEO screws up a negotiation with our UK distributors: there is no money for anyone's bonus, and say 10% of staff laid off. There work isn't necessarily harder, or requiring more skill, but the attention to detail needed, and consequences of screw ups are much higher.
I never got the reality distortion field bit. Even people that prefer windows or FOSS usually at least give Jobs a "brilliant salesman" title. His keynotes were good but that was mainly IMO because the products generally were somewhat interesting (it helps they only release a few of them, a couple times a year and they always go upmarket so they are generally novel, or at least close to cutting edge). Other than that I kind of tuned him out, like I do when an MS exec flogs their latest stuff: while they speak I'm hunting down specs, comparing it to existing stuff etc. I don't listen to everything they say and take it like gospel. Maybe some people did with Steve, never got that though.
Marketing/prioritization is more of what he was good at IMO. Picking the market segments (and sometimes getting projects going that create new ones), focusing on very few things, not trying to satisfy everyone, being content with the most profitable 10% of the market etc. In contrast: Woz did some cool engineering things but he pretty much did them because he found them interesting. If all the stories that are told are to be believed he didn't really care much about making a company out of the computers. In short he kicked ass at building things he wanted to, Jobs kicked ass at getting someone else to build things head wanted and finding people that were willing to pay.
What the other guy does is always easy. That is why as an engineer you get a long list of huge change requests without anyone wanting to nail down a #1 item because "all should be pretty straight forward so should make it into next months release". Or sales guys who's talent is flapping their mouths get no credit by engineers who can barely mumble their way through small talk and often not even that in the common language. Or the guy that dumps his life savings into the company and "picks the winner" big deal he was born with a pile of cash anyone can get richer if they start with 20M. In short everyone's job is easy till you have to do it.
So few shows can have a story arc interesting and complicated enough to justify 16hrs or so a season worth of content. So you get that 2 min flashback scene every episode to drag it out over many seasons. Also, a lot of shows do have a larger story arc but (IMO) lack the ability to develop the characters and so instead get viewers by making every episode a "must see" by killing someone off (Game of Thrones Walking Dead, Sons of Anarchy).
Cop shows: there are so many plot devices that are just insulting too and at least to me actively repulse me: needing to keep a guy on the line to trace a call, infinite zoom on a crappy photo, every police station having that guy that can miraculously hack any computer in seconds etc. I think the variety of real life would be more interesting. The cop bitching about a bum knee might stop him from getting his 20 years in, the secretary blowing the boss. Dirty cops, racist cops, cops that actually try to help people, cops that give a damn but are incompetent etc.
Some shows can be interesting mainly because their main character is interesting even without always having much of a story arc example House, Macgyver, Burn Notice, or occasionally having interesting moral dilemas: Star Trek, some old westerns.
Slashdot Mirror
I don't understand what your saying. Depends on the doctor/field. In my case oncology so needed tracking different stages of the patient workflow, reports for amount of dose delivered to different types of patients etc. No really good off the shelf solution handles all the parts needed since generally speaking you are using a collection of different vendors CT scanners, EMR, PACS, treatment planning, patent scheduling and radiation treatment machines. Heck even software from the same vendor often does things differently because the Phillips Imaging department doesn't talk to the Phillips Oncology department before designing their software.
Well in my case we are very far from customers. Our software is sold in ~80 countries by distributors and ultimately used by accountants. So addresses might not be the clearest example of this but often we have to make things very generic because one countries tax code differs from another, or they have the same data but need it at different times in their process etc. Often when we ask a distributor they ask a few customers I think and we go with what they say but sometimes no one knows because it is a completely new government requirement/process (ex. recently Australia started requiring their version of a personal retirement savings plan to have yearly audits: not required before so no existing paper based system or whatever to automate, we had to guess what they would need when).
Agreed though, whenever you can actually see a customer working is the best. Even better if you are automating something you do as you know both your current process and what you'd ideally like to have (did that a lot when I worked more on the IT side, or as more of a scientist/technician playing around with accelerators).
Well there is a time issue too. At my current work I make software for accountants, in the past I made it for doctors and scientists either group their time is better spent practicing their expertise not making software. Just because you "can" do something doesn't mean your time is well spent doing it, or that you can do it as well as someone that does it all the time. Sure I could install my new windows, but I'd rather pay a guy to do it because I don't like the work, it would take me at least 4X as long to do so and I'm more likely to screw something up than they are.
Yeah that is scary. You can say because the cost of crap code is lost customers/law suits etc,not to mention moral: knowing you are selling crap but not caring enough to actually try to deliver on your promises. They could still factor in all those costs and say Tata is still cheaper sorry. My degrees are in physics and management so if programming doesn't work out I have other domains to move in. I think that is what we need generally in our profession and probably others: you are just a budget cutback away from your grant being denied, your department outsourced etc. If you are truly smart learn multiple useful professions and hope like hell that they don't all go to crap at the same time.
Part of the problem is the expectation that everyone contributes/carries their wait. I'm sorry but if robots truly replace all menial jobs the idiots out their will have no work that makes sense for them to do, unless you subscribe to the Calvinist work ethic: work itself has value. Drop the Jebus and all that is left is: keep the idiots out of the way while the remaining "useful" people work. It is only our sense of fairness and greed for "more" that makes us think that it is a useful use human life to have someone paid full time to flip a burger when a machine can do it just as well and cheaper.
Amen to that. I both don't think it is possible: not everyone has the temperament to stick with a problem for several days till they figure something out, nor desirable (there are a lot of dumb people out there (or at least dumb when it comes to tech)) for everyone to code. I want my lawyer to focus on kicking ass at being a lawyer, by doctor to focus on kicking ass at making me healthier etc. We ain't the only profession that will matter in the future no matter how much tech is around, there is still going to be the need for good art, good music, good health and other advise, good bjs etc. Everyone has their place.
There is bad code for sure. There is also bad architecture and feature specs. In my experience there are often very fuzzy business requirements, sometimes legitimately: the customer can't tell you what they need they just ask for something "better". Example something I'm currently dealing with: the customer asks to be able to add multiple addresses instead of one to each order. Well is there only one address of a given type allowed (postal, physical, fax, etc) or an arbitrary number of each type? If so how do you know which should be listed first? They aren't sure but they need the feature in a couple months so they can demo it to customers.
So I end up making something as general as possible "just in case" someone somewhere might want to do things that way. Then a few months later I expect someone to complain: we screwed up and added 5 postal addresses to one customer, that make no sense why didn't you warn us?
Anyways, sometimes bad architecture comes about by poorly nailed down design requirements. Sometimes that can be helped by asking the customer more, and sometimes they don't know so are no help All you can do as a dev is tell them it is likely going to be 5X more work to do it in a way that will support all scenarios "just in case" vs doing it the "one true way". If they say fine you need to live with the legacy junk until they feel free to free you up to do a refactor on the code and not spend the whole release trying to pound out new features (fat chance in hell). Sometimes the best you can do is write code that doesn't suck to bad in what the expected typical use case is, and drag it forward as long as you can to meet the current requirements. Once it breaks to the point of no return you have to pay all the tax of the lack of good specs you had all along. This doesn't excuse crappy coding at the method level though: that is a matter of skill and hopefully anyone that considers themselves a professional continues to learn from code reviews, training etc and gets better over time. If you don't you suck and need to go.
It is the exact problem I have with some "free" solutions.There are 1000 people that ever downloaded the thing with maybe 100 current users and 3 contributors. It either does what you want or you need to take the time to understand their code, or you need to keep looking. Script kiddies aren't just caused by amateur coders, they are caused by side projects we make without a proper business need/case. We make it to fill such a niche need that you either have to live with the limitations or move on to another solution. That is what the future will be: a google image search effectively of "hot girl" and countless hours spent finding the one you meant, vs hiring someone with the skill to make yet another "script kiddie' that does exactly what you want, or spending the time to actually learn the market and hiring a professional to make something that is generally applicable and/or innovative.
That is just it. It isn't just a matter of can you understand what an if or for loop is. It IS a skill/psychological problem: you need be the type of person that wants to understand how things work all the way down to the metal, and stubborn enough to keep at it till you can get things to work. If anything the more easy to use computers become the more frustrating it will be to fix that 1% of remaining issues. It takes a special type of idiot to keep bashing their heads against the wall when things are already "good enough".
Depends where you live but at least for teachers there are often school boards or larger organization bodies that insure that even the new ones make better salaries than a lot of programmers. So the distribution is compressed around a high salary. A starting salary in Ontario for example is 51k + a pension, 20 sick days a year that can be banked for early retirement. There pension pays out 60% of their best 5 yrs salary (so don't want to be a principal? Well just do it for 5 years because you'll then make another $10k a year or so when retired). 10yrs in teachers are making about 90k a year more if they are department heads or move into VP or principal positions. They also have full medical and dental with no employee portion of the premiums. So in short at least in my experience in the province, they make an about average to high programmers salary with way more vacation, way better pension, and about a $100 a month savings because not paying anything towards their benefits.
Plumbers, electricians etc all can make really good money too. Some aspects of their job sucks and they might not be able to do it till 65 though, so I guess you're paying towards early retirement for them.
Then you get busy street corners where the only time cars move is on the yellows. I feel for the drivers but it sucks as a pedestrian. Some intersections in Toronto where I work there are always about 50 people on each side trying to cross for each light. The only window the drivers turning get a chance to cross is the small window between when cars start stopping for the yellow and when the green the other way (and the cross light for pedestrians with it) kick in. Cabs seem to be the most aggrressive trying to get through that gap.
Yeah, or deploy spike strips :) That would be awesome actually if the jerk that needs to be unsafe for an hour drive so that they can be 3 car lengths closer to home ends up spending $800 and a few hours getting their car towed and all tires replaced: definitely would change the incentives.
That might be the end result of these automatic "emergency" braking systems: they start to enforce sufficient space between vehicles. After all, the manufacturer doesn't what to get sued when your car fails to stop itself in time to avoid a rearend collision. This might also be one of many things manual drivers will have to give up if they don't opt for a driverless car: regulations will gradually be put into place to make an autopilot take over whenever the driver tries to do something that the government doesn't like: can't speed, can't follow too closely, turn into another lane without enough room etc. It might get to the point where everyone can "drive" a car because you can basically stomp on the gas pedal and the car will steer to avoid everything for you, the only "skill" would be in actually getting to where you want to go though I suppose that could be tied into the system that prevents you from turning into someone too.
True somewhat. If you use all that space between you and the guy in front beause you give him a "love tap" the guy behind you has more space/time to stop. In the majority case you probably are better off slamming your brakes I guess but I can see cases where you end up getting rear ended where you with your slower reaction wouldn't have gotten in an accident at all. Say the guy in front of you slams his brakes then releases, your system slams the brakes and you get rear ended because the guy behind you was closer to you than you were to the guy in front and/or there vehicle needs more space to stop.
they guy with the older car behind you from rear ending you instead? Or are these systems going to optimize between the risk of crashing into the guy in front of you vs the risk it will stop too quickly for the guy behind you to respond (yeah I know we all leave sufficient space between us and the car in front to brake)?
I'm curious, at least the apple.ca site doesn't have the weight or dimensions of the iPad Pro. I wonder if the page was made before the final specs were known or is it a matter of if Apple can't say it is thinner and lighter they want to hide it behind all the new/improved specs?
Yep. The other servers are likely to have source code and such (bug/feature ticketing, Stash repos, TFS etc). Why GMail: not sure. We build collaboration software so our products do need to interact with google, Office 365, SharePoint etc etc not sure if they just figured we all need a business account for G Drive anyways to test our software we might as well get the email service too or what.
In most cases if your internet connection is out work stops anyways. My work we us google for mail, and Skype for IM but otherwise have everything in house. Still if we have a glitch work slows down to a crawl. Every time you try to do something and need to look up a manual page or something, oh crap can't access the site. Huh, okay maybe I can do something else ... oh crap that bug has a link to a website, can't work on that ... It quickly becomes "lunch time" for everyone in the office for that 30min or whatever where the gateway is having issues.
More geographically concentrated. At least it has been my experience for the jobs I look for in software development about 80-90% of jobs are in major cities. You can find a job pretty much everywhere but the big companies hiring dozens of developers at a time overwhelm the one or two positions smaller companies in smaller communities are hiring. You switch jobs you either find another job in the big city you are in or you move 1hr + away (I'm from Canada: huge country, vastly distributed major cities). I think IT is going to become more like that too. You'll still need the low level support in house but the large number of jobs will be concentrated into areas with huge datacentres. In this case likely more rural locations because of cheap power, colder climate, cheap land etc.
They never say the thing tastes like what it is. Never does wine have a "distinctive note of fermented grapes". I also love to read a couple tasting notes about the same thing. One will say "citris, vanilla and anise" another "watermelon", "bacon", "chocolate". Without fail they seem to have complete different components. Then you get in the room with a wine or whiskey snob and watch them discuss the "peach note" in the drink. Fantastic.
Agreed. You can look at it another way though, a force multiplier effect. If I screw up as an intermediate developer, it is likely we find it in code review or pre-release testing and spend a couple weeks longer than we planned doing something say. If the CEO screws up a negotiation with our UK distributors: there is no money for anyone's bonus, and say 10% of staff laid off. There work isn't necessarily harder, or requiring more skill, but the attention to detail needed, and consequences of screw ups are much higher.
I never got the reality distortion field bit. Even people that prefer windows or FOSS usually at least give Jobs a "brilliant salesman" title. His keynotes were good but that was mainly IMO because the products generally were somewhat interesting (it helps they only release a few of them, a couple times a year and they always go upmarket so they are generally novel, or at least close to cutting edge). Other than that I kind of tuned him out, like I do when an MS exec flogs their latest stuff: while they speak I'm hunting down specs, comparing it to existing stuff etc. I don't listen to everything they say and take it like gospel. Maybe some people did with Steve, never got that though.
Marketing/prioritization is more of what he was good at IMO. Picking the market segments (and sometimes getting projects going that create new ones), focusing on very few things, not trying to satisfy everyone, being content with the most profitable 10% of the market etc. In contrast: Woz did some cool engineering things but he pretty much did them because he found them interesting. If all the stories that are told are to be believed he didn't really care much about making a company out of the computers. In short he kicked ass at building things he wanted to, Jobs kicked ass at getting someone else to build things head wanted and finding people that were willing to pay.
What the other guy does is always easy. That is why as an engineer you get a long list of huge change requests without anyone wanting to nail down a #1 item because "all should be pretty straight forward so should make it into next months release". Or sales guys who's talent is flapping their mouths get no credit by engineers who can barely mumble their way through small talk and often not even that in the common language. Or the guy that dumps his life savings into the company and "picks the winner" big deal he was born with a pile of cash anyone can get richer if they start with 20M. In short everyone's job is easy till you have to do it.
Yeah that is the one exception to the rule.
So few shows can have a story arc interesting and complicated enough to justify 16hrs or so a season worth of content. So you get that 2 min flashback scene every episode to drag it out over many seasons. Also, a lot of shows do have a larger story arc but (IMO) lack the ability to develop the characters and so instead get viewers by making every episode a "must see" by killing someone off (Game of Thrones Walking Dead, Sons of Anarchy).
Cop shows: there are so many plot devices that are just insulting too and at least to me actively repulse me: needing to keep a guy on the line to trace a call, infinite zoom on a crappy photo, every police station having that guy that can miraculously hack any computer in seconds etc. I think the variety of real life would be more interesting. The cop bitching about a bum knee might stop him from getting his 20 years in, the secretary blowing the boss. Dirty cops, racist cops, cops that actually try to help people, cops that give a damn but are incompetent etc.
Some shows can be interesting mainly because their main character is interesting even without always having much of a story arc example House, Macgyver, Burn Notice, or occasionally having interesting moral dilemas: Star Trek, some old westerns.