This is because bad engineers generate problems in the short term, while bad sales people generate problems in the long term. If you have a bad engineer, you won't deliver a working product on time and you lose money. If you have a bad sales person, you may generate a lot of sales now, but you'll seriously damage your corporate reputation and lose long-term sales. But, by then the sales person will have received their big bonus and moved on...
Totally agree on 1. People care about what your product / service does, not what how it does it. 2 and 3 are more difficult, because often the customer doesn't know. The relevant Henry Ford quote is 'if I'd asked my customers what they wanted, they'd have said a faster horse'. The worst thing you can do when trying to design a system is start with a solution. You need to start with identifying the problem: in Ford's case, you want to go from A to B quickly and efficiently. Then you move on to the existing system: a horse and cart. Then you identify what current technology allows you to do to improve it: replace the horse with an internal combustion engine.
Often, if you ask your customers what they want, they will have a good understanding of their problems, and a bad understanding of potential solutions. They will give you a list of really bad ideas for improving their system. If you actually give them what they ask for, they'll hate you. If you give them what they want, they'll love you. Working out what they want is the difficult bit.
And this post sums up the reasons for the negative attitudes towards sales and marketing: it's not that they're easy, it's that most of the people working in these departments suck at their jobs and, in the case of marketing, have no metrics to be able to judge their quality (metrics are easier in sales).
Identifying potential customers, identifying their needs, and working to fill those needs is very useful and is very difficult. It's much easier just to spam a million people and hope one of them is stupid enough to buy your product.
Given the average standard of computing teachers in the UK, having programming taught as part of the national curriculum is unlikely to make it seem less like magic...
I switched to DuckDuckGo ages ago. Every so often, it comes back with useless results and I go to Google. So far, Google has failed to deliver better results any time I've done that. For any query I've tried, either DDG is 'good enough,' or neither DDG nor Google is.
Instant messaging began with Quantum Link, which eventually became AOL Instant Messenger. ICQ was big, then MSN and Yahoo joined the party. Then there was XMPP, an open, federated, protocol (used, among other things, by Google Talk). IRC and IM don't really fit in the same category - IM is for one-to-one communication, IRC is one-to-many. And I've not seen anything replacing IRC - it's still very actively used.
Email began with BBS email, where each server had its own mailbox system. Eventually some federated. Big online service providers like AOL and Compuserve ran their own proprietary email systems. Internet email completely displaced them all.
You might have a point with usenet, but most things seem to have gone from usenet to mailing lists (often hosted by Google or Yahoo, true), which are still using open protocols. Some people use web interfaces, like gmane, some use their own mail clients.
Web pages, I'm not so sure about. Things like Facebook and MySpace grew quickly, but they still account for a tiny fraction of the total web.
Freedom Box is his latest project. He gave a (quite bad, actually) talk about it at FOSDEM last year and it's been on Slashdot a couple of times. The idea is to produce a cheap plug computer that can run email, chat, and so on services and provides hosting for picture and movie sharing - basically, provide all of the useful features of social networking, but provide them with completely distributed user-controlled implementations in an off-the-shelf package that people with can just buy, take home, plug in, and use.
In the context of social networking, Freedom Box is a very relevant thing to be mentioning...
The issue is not sharing information, it's using a massive single entity as an intermediate for sharing information. If I send on of my friends an email, it goes to my mail server and then to their mail server and then to their computer. I run my own mail server, so the only person who can easily aggregate information about me is me. They often use university or work provided email accounts, so their employers can aggregate information about their employees, but generally don't bother because they're not in the business of doing so. On the other hand, if you send a message via Facebook then a single entity controls both ends of the communication and can easily infer a huge amount of information from correlating various factors (including the IP address used for sending and receiving messages, the user agent, and so on).
The same issue was true with early email systems. If you used Compuserve email, then a single entity could track both endpoints of the communication. If you used Internet email then each ISP could track half of the data. The same holds with XMPP vs proprietary instant messaging protocols.
Having access to the list of correspondents for one person tells you quite a lot, but having access to the list of correspondents for everyone that they correspond with tells you vastly more. You can infer all sorts of things about the social groupings.
It's not that they can't adapt, it's that they won't be able to offer the same old same old college experience for $4000-30000+ per year when you can goto an online school and do it in your spare time for the cost of nothing more than a pc and an internet connection.
All of my university exams had one or two external candidates sitting them. These people aren't enrolled on the course, but they turn up, do the exams, and get a degree at the end of it if they pass them (and any other relevant coursework). The university still charges them, but it charges them quite a bit less than it charges enrolled students. At the end, the degree that they get is indistinguishable from the one I got. Anyone sufficiently motivated could study the syllabus on their own time, take the exams, and get a degree. In spite of this, there were very few external students. I returned briefly last year and taught a module. That year had no external students.
Not sure about where you are, but here the local university will allow anyone, even non-students, to register. They are then allowed to use the library computers to view papers from all of the journals that the university has subscribed to, for free. I know one person who did exactly that while she was writing applications to do a masters (not at the local university, and she was not an alumnus).
A survey a few years ago asked people if they thought they were below average, average, or above average drivers. 80% selected above average. More interestingly, most of the people who said below average were actually above average...
I published one paper after leaving academia. It was no harder than publishing in academia - write the paper, send it off to a journal, wait for the acceptance. Some journals charge for publication, but there is usually at least one open access journal that doesn't in any given field. Decent journals do anonymous review, so the reviewers have no idea of your affiliation.
The main reason I stopped publishing after leaving academia was that I was too lazy to write literature review sections for papers (I actually have several ones on my hard drive that are complete apart from that, but outside academia there is little incentive to rack up the publication count). If I blog about stuff, I get feedback immediately, rather than needing to wait up to a year for journals to get around to publishing a paper.
Does it not where you live? In the UK, the police make the locations of crime reports available online. I checked them when I was looking for a house. One place I looked at had a lot of complaints within a couple of streets of people dealing drugs in the phone boxes and leaving used needles in the street. The place I eventually picked had no complaints of anything more serious than parking illegally for quite a way on all sides.
Or, for those more interested in computers, sudoku has quite a few things in common with a number of error correction techniques and some compression algorithms. This particular result is moderately interesting from an information theoretic perspective and is probably a fairly minor part of a larger project that may well yield practical results.
Out of interest, have you ever tried writing to the senators or representatives for your constituency? On any matter? I most often hear this 'my elected representatives don't represent me' from people who have never bothered letting the politicians that are supposed to represent them know what they think.
In the first meeting of the Welsh Assembly (devolved government), one of the newly elected assembly members referred to another as 'the Honourable Member from...' The speaker interrupted saying 'there are not honourable members here boyo'.
COBOL has things like decimal arithmetic built in. This is very useful for financial applications, where COBOL is still used quite a lot. If you buy a recent IBM machine, you will even find that the CPU has support for decimal floating point types in the FPU and these are used directly by the COBOL compiler.
He said C++. While it is just about possible to write object oriented code in C++, I would be very surprised if more than 1% of C++ programmers managed it. What you are describing is programming with ADTs, which has some overlap with OOP but is as much OOP as defining functions is structured programming.
Slashdot Mirror
Now ask a salesperson how long is a piece of string? They will have a conversation with you and find out what you want the string for
No, that's what a good salesperson will do.
This is because bad engineers generate problems in the short term, while bad sales people generate problems in the long term. If you have a bad engineer, you won't deliver a working product on time and you lose money. If you have a bad sales person, you may generate a lot of sales now, but you'll seriously damage your corporate reputation and lose long-term sales. But, by then the sales person will have received their big bonus and moved on...
Totally agree on 1. People care about what your product / service does, not what how it does it. 2 and 3 are more difficult, because often the customer doesn't know. The relevant Henry Ford quote is 'if I'd asked my customers what they wanted, they'd have said a faster horse'. The worst thing you can do when trying to design a system is start with a solution. You need to start with identifying the problem: in Ford's case, you want to go from A to B quickly and efficiently. Then you move on to the existing system: a horse and cart. Then you identify what current technology allows you to do to improve it: replace the horse with an internal combustion engine.
Often, if you ask your customers what they want, they will have a good understanding of their problems, and a bad understanding of potential solutions. They will give you a list of really bad ideas for improving their system. If you actually give them what they ask for, they'll hate you. If you give them what they want, they'll love you. Working out what they want is the difficult bit.
And this post sums up the reasons for the negative attitudes towards sales and marketing: it's not that they're easy, it's that most of the people working in these departments suck at their jobs and, in the case of marketing, have no metrics to be able to judge their quality (metrics are easier in sales).
Identifying potential customers, identifying their needs, and working to fill those needs is very useful and is very difficult. It's much easier just to spam a million people and hope one of them is stupid enough to buy your product.
Given the average standard of computing teachers in the UK, having programming taught as part of the national curriculum is unlikely to make it seem less like magic...
I switched to DuckDuckGo ages ago. Every so often, it comes back with useless results and I go to Google. So far, Google has failed to deliver better results any time I've done that. For any query I've tried, either DDG is 'good enough,' or neither DDG nor Google is.
Instant messaging began with Quantum Link, which eventually became AOL Instant Messenger. ICQ was big, then MSN and Yahoo joined the party. Then there was XMPP, an open, federated, protocol (used, among other things, by Google Talk). IRC and IM don't really fit in the same category - IM is for one-to-one communication, IRC is one-to-many. And I've not seen anything replacing IRC - it's still very actively used.
Email began with BBS email, where each server had its own mailbox system. Eventually some federated. Big online service providers like AOL and Compuserve ran their own proprietary email systems. Internet email completely displaced them all.
You might have a point with usenet, but most things seem to have gone from usenet to mailing lists (often hosted by Google or Yahoo, true), which are still using open protocols. Some people use web interfaces, like gmane, some use their own mail clients.
Web pages, I'm not so sure about. Things like Facebook and MySpace grew quickly, but they still account for a tiny fraction of the total web.
Freedom Box is his latest project. He gave a (quite bad, actually) talk about it at FOSDEM last year and it's been on Slashdot a couple of times. The idea is to produce a cheap plug computer that can run email, chat, and so on services and provides hosting for picture and movie sharing - basically, provide all of the useful features of social networking, but provide them with completely distributed user-controlled implementations in an off-the-shelf package that people with can just buy, take home, plug in, and use.
In the context of social networking, Freedom Box is a very relevant thing to be mentioning...
Diaspora wasn't designed
This bit, at least, appears to be true.
The issue is not sharing information, it's using a massive single entity as an intermediate for sharing information. If I send on of my friends an email, it goes to my mail server and then to their mail server and then to their computer. I run my own mail server, so the only person who can easily aggregate information about me is me. They often use university or work provided email accounts, so their employers can aggregate information about their employees, but generally don't bother because they're not in the business of doing so. On the other hand, if you send a message via Facebook then a single entity controls both ends of the communication and can easily infer a huge amount of information from correlating various factors (including the IP address used for sending and receiving messages, the user agent, and so on).
The same issue was true with early email systems. If you used Compuserve email, then a single entity could track both endpoints of the communication. If you used Internet email then each ISP could track half of the data. The same holds with XMPP vs proprietary instant messaging protocols.
Having access to the list of correspondents for one person tells you quite a lot, but having access to the list of correspondents for everyone that they correspond with tells you vastly more. You can infer all sorts of things about the social groupings.
If New Labour is what Americans think socialism is, no wonder it's such a dirty word over there.
It's not that they can't adapt, it's that they won't be able to offer the same old same old college experience for $4000-30000+ per year when you can goto an online school and do it in your spare time for the cost of nothing more than a pc and an internet connection.
All of my university exams had one or two external candidates sitting them. These people aren't enrolled on the course, but they turn up, do the exams, and get a degree at the end of it if they pass them (and any other relevant coursework). The university still charges them, but it charges them quite a bit less than it charges enrolled students. At the end, the degree that they get is indistinguishable from the one I got. Anyone sufficiently motivated could study the syllabus on their own time, take the exams, and get a degree. In spite of this, there were very few external students. I returned briefly last year and taught a module. That year had no external students.
Not sure about where you are, but here the local university will allow anyone, even non-students, to register. They are then allowed to use the library computers to view papers from all of the journals that the university has subscribed to, for free. I know one person who did exactly that while she was writing applications to do a masters (not at the local university, and she was not an alumnus).
There's always been privacy on Facebook. Any data that you place on Facebook is made difficult to access for anyone who doesn't pay Facebook.
A survey a few years ago asked people if they thought they were below average, average, or above average drivers. 80% selected above average. More interestingly, most of the people who said below average were actually above average...
I published one paper after leaving academia. It was no harder than publishing in academia - write the paper, send it off to a journal, wait for the acceptance. Some journals charge for publication, but there is usually at least one open access journal that doesn't in any given field. Decent journals do anonymous review, so the reviewers have no idea of your affiliation.
The main reason I stopped publishing after leaving academia was that I was too lazy to write literature review sections for papers (I actually have several ones on my hard drive that are complete apart from that, but outside academia there is little incentive to rack up the publication count). If I blog about stuff, I get feedback immediately, rather than needing to wait up to a year for journals to get around to publishing a paper.
Published, peer-reviewed, papers go in a portfolio for any subject like that...
Last two months? When has anything from infoworld ever been worth reading?
Does it not where you live? In the UK, the police make the locations of crime reports available online. I checked them when I was looking for a house. One place I looked at had a lot of complaints within a couple of streets of people dealing drugs in the phone boxes and leaving used needles in the street. The place I eventually picked had no complaints of anything more serious than parking illegally for quite a way on all sides.
And a self-compiled linker and audited, on a host platform that you compiled yourself...
Or, for those more interested in computers, sudoku has quite a few things in common with a number of error correction techniques and some compression algorithms. This particular result is moderately interesting from an information theoretic perspective and is probably a fairly minor part of a larger project that may well yield practical results.
Out of interest, have you ever tried writing to the senators or representatives for your constituency? On any matter? I most often hear this 'my elected representatives don't represent me' from people who have never bothered letting the politicians that are supposed to represent them know what they think.
In the first meeting of the Welsh Assembly (devolved government), one of the newly elected assembly members referred to another as 'the Honourable Member from...' The speaker interrupted saying 'there are not honourable members here boyo'.
COBOL has things like decimal arithmetic built in. This is very useful for financial applications, where COBOL is still used quite a lot. If you buy a recent IBM machine, you will even find that the CPU has support for decimal floating point types in the FPU and these are used directly by the COBOL compiler.
Object oriented progamming?
He said C++. While it is just about possible to write object oriented code in C++, I would be very surprised if more than 1% of C++ programmers managed it. What you are describing is programming with ADTs, which has some overlap with OOP but is as much OOP as defining functions is structured programming.