Less of a top-ten, and more of a ten-random. What is the domain of this list? It seems like if you can go from Zune to Bluetooth to Biometrics, you should at least touch on something like the Segway HT: the first thing that comes to mind when I hear "tech flop".
...to how television causes social disengagement and bad moods.
does anyone have more info on how television causes social disengagement and bad moods? Are there conclusive studies of this? I'd be interested in reading about that, but TFA doesn't touch that issue. I kind of like how this statement was thrown out there in the story like it's some universal truth that everyone holds to be obvious.
brings up the point that the vast majority of exhibitionists are not out there flaunting their intellectual side come friday night.
I might even say the vast majority of conversation is meaningless banter, and myspace (or livejournal, or wherever...) is just the online incarnation of that.
In what may be the crucial competitive stroke, Microsoft will also allow you to download from its store any song that you've purchased from Apple, unlocking users from iPod's vendor lock-in."
So now when I buy a song from iTunes, it not only supports apple, but bleeds a little from m$? Time to go on a shopping spree!
i'd like to join in the chorus of voices saying basically this:
if you want my interest back in the arcade gaming arena, you need to become more competitive price-wise with other gaming markets, like the consoles and PCs I already own and play for free
I hate to say it, but having been a developer on big databases full of reasonably sensitive information this doesn't surprise me in the least. Operators & developers must have very liberal access to be able to perform their jobs, and they're far too often dangerously undertrained re: basic fundamentals of data security.
"i have to run out, can I just leave you my password to check on this job status in 30 minutes?..."
There's some truth to that, but as a daily professional spreadsheet and database user I dont believe it's in hardly anyone's best interest to increase the (already giant) possible size of the most commonly used spreadsheet. I'd argue that 99.999% of the time the only thing this upgrade is going to do is increase excels capacity for abuse. Excel sheets are an extremely high-overhead storage method for matrixes of data that simply dont need the functions that excel provides for that cost.
I would be really suprised if you couldn't model that data to be scalable with rows instead of columns. That's not to say I haven't been suprised before:)
Yeah I hear that. I've seen some pretty amazing cobbled together stuff in Excel/Access running on some savant's desktop here and there... I guess if it works and it's cheap-as-in-free for essentially the same end result as a costly database, I cant really tear that down.
just remember to look-real-busy when the regulators show up and start asking about quality control
most likely not. I try to put myself a few years back in my head though, and I think if I had encountered a 500 column dataset before I knew anything about data architecture it only would have accelerated my desire to learn.
It's hard to come up with a good analogue to this problem, but if you give people the ability to store their massive datasets in Excel, they'll do it. A person might never ask the question "is there a way to do this already that considers the problems that arise when one does this?"
My lack of knowledge surrounding the xls file format is going to bite me here, but the ability to arbitrarily reference any other "cell" in a spreadsheet would appear to build-in an unnecessary amount of overhead for any kind of large scale dataset storage or processing.
I appreciate that this response made me think about a fundamental difference between spreadsheets and databases.
A spreadsheet is a matrix of cells which can refer to any data or formula in any other cell arbitrarily (ot it can contain static data itself...). A dataset in a flat file or a database is an array of typed (all text if flatfile) columns that can either contain data (such as reference data to other columns, etc...). The overhead required to enable arbitrary cell references from any atom in a dataset is too much for large data sets. This definition of large is different for everyone of course, but in a 1,000,000 row dataset the need for arbitrary cell referencing is 99.999% of the time : overkill.
Wow, generally what kind of dataset has 500 columns? Cant that be denormalized in any way? Do you seriously scan through a list of 500 elements to pick which columns you're interested in at any given point in time? I've worked with a few 500 column oracle tables, but they were the result of poor design without any rational scalability.
I work with both databases and spreadsheets to pay the bills. They serve very different purposes, I dont have much to say other than what a friend once told me in the early days of my developer when all I knew was perl on Linux
"When all you have is a hammer, everything looks like a nail."
If you're dealing with sizeable datasets on a regular basis, it helps to know the tools that were designed to handle them.
I think you just described a new circle of hell, I could hear faint shrieking in my head as I read that.
Do you have any kind of basis for the distributed data assumption? That doesn't seem like an easy feature to sell to consumers, let alone the MS developers who would have to implement it...
no matter what arbitrary limit is set, someone is going to hit the boundry and be upset because they could use "just a little more". I hate recommending the use of something like Microsoft Access, but it's hard for me to imagine Excels interface actually help you accomplish much with tens of thousands of rows of data - that couldn't be accomplished much easier with a query tool. Of course if the tool works for you, then keep on rockin' it. Inspecting big delimited text files is ridiculously easy in a tool like Access if you've never tried it, just fyi.
> WTF? If I've got anyone in IT putting 1,000,000 rows in a spreadsheet, I'm seriously considering demoting them. If you're going to have a million rows, get a database.
It's not the people in IT you have to worry about, it's analysts with 1MM row datasets. At first my thought was "the horror of a 1MM row spreadsheet! blashpemy!...
but now I'm thinking at least they wont be as tempted to us MS Access..
Slashdot Mirror
http://xkcd.com/603/
Peeing in my neighbors gas tank will no longer have the desired effect.
Less of a top-ten, and more of a ten-random. What is the domain of this list? It seems like if you can go from Zune to Bluetooth to Biometrics, you should at least touch on something like the Segway HT: the first thing that comes to mind when I hear "tech flop".
Couldn't agree more, one day I stopped watching TV and my live improved 10 fold. I'd like to read some science that attempts to explain why.
...to how television causes social disengagement and bad moods.
does anyone have more info on how television causes social disengagement and bad moods? Are there conclusive studies of this? I'd be interested in reading about that, but TFA doesn't touch that issue. I kind of like how this statement was thrown out there in the story like it's some universal truth that everyone holds to be obvious.
if we had a news story for every concept a theoretical physicist was speculating on, we'd have /.!
I thought it was a war story right now.
news story? advertisement? what's the difference?
brings up the point that the vast majority of exhibitionists are not out there flaunting their intellectual side come friday night.
I might even say the vast majority of conversation is meaningless banter, and myspace (or livejournal, or wherever...) is just the online incarnation of that.
In what may be the crucial competitive stroke, Microsoft will also allow you to download from its store any song that you've purchased from Apple, unlocking users from iPod's vendor lock-in."
So now when I buy a song from iTunes, it not only supports apple, but bleeds a little from m$? Time to go on a shopping spree!
i'd like to join in the chorus of voices saying basically this:
if you want my interest back in the arcade gaming arena, you need to become more competitive price-wise with other gaming markets, like the consoles and PCs I already own and play for free
isn't this what y2k was supposed to be like?
I hate to say it, but having been a developer on big databases full of reasonably sensitive information this doesn't surprise me in the least. Operators & developers must have very liberal access to be able to perform their jobs, and they're far too often dangerously undertrained re: basic fundamentals of data security. "i have to run out, can I just leave you my password to check on this job status in 30 minutes?..."
There's some truth to that, but as a daily professional spreadsheet and database user I dont believe it's in hardly anyone's best interest to increase the (already giant) possible size of the most commonly used spreadsheet. I'd argue that 99.999% of the time the only thing this upgrade is going to do is increase excels capacity for abuse. Excel sheets are an extremely high-overhead storage method for matrixes of data that simply dont need the functions that excel provides for that cost.
I would be really suprised if you couldn't model that data to be scalable with rows instead of columns. That's not to say I haven't been suprised before :)
Yeah I hear that. I've seen some pretty amazing cobbled together stuff in Excel/Access running on some savant's desktop here and there... I guess if it works and it's cheap-as-in-free for essentially the same end result as a costly database, I cant really tear that down.
just remember to look-real-busy when the regulators show up and start asking about quality control
most likely not. I try to put myself a few years back in my head though, and I think if I had encountered a 500 column dataset before I knew anything about data architecture it only would have accelerated my desire to learn.
It's hard to come up with a good analogue to this problem, but if you give people the ability to store their massive datasets in Excel, they'll do it. A person might never ask the question "is there a way to do this already that considers the problems that arise when one does this?" My lack of knowledge surrounding the xls file format is going to bite me here, but the ability to arbitrarily reference any other "cell" in a spreadsheet would appear to build-in an unnecessary amount of overhead for any kind of large scale dataset storage or processing.
that's a good point, being able to arrange test output in a cell structure much like a spreadsheet would absolutely be useful for smoke testing there.
I appreciate that this response made me think about a fundamental difference between spreadsheets and databases.
A spreadsheet is a matrix of cells which can refer to any data or formula in any other cell arbitrarily (ot it can contain static data itself...). A dataset in a flat file or a database is an array of typed (all text if flatfile) columns that can either contain data (such as reference data to other columns, etc...). The overhead required to enable arbitrary cell references from any atom in a dataset is too much for large data sets. This definition of large is different for everyone of course, but in a 1,000,000 row dataset the need for arbitrary cell referencing is 99.999% of the time : overkill.
Wow, generally what kind of dataset has 500 columns? Cant that be denormalized in any way? Do you seriously scan through a list of 500 elements to pick which columns you're interested in at any given point in time? I've worked with a few 500 column oracle tables, but they were the result of poor design without any rational scalability.
I work with both databases and spreadsheets to pay the bills. They serve very different purposes, I dont have much to say other than what a friend once told me in the early days of my developer when all I knew was perl on Linux
"When all you have is a hammer, everything looks like a nail."
If you're dealing with sizeable datasets on a regular basis, it helps to know the tools that were designed to handle them.
I think you just described a new circle of hell, I could hear faint shrieking in my head as I read that.
Do you have any kind of basis for the distributed data assumption? That doesn't seem like an easy feature to sell to consumers, let alone the MS developers who would have to implement it...
no matter what arbitrary limit is set, someone is going to hit the boundry and be upset because they could use "just a little more". I hate recommending the use of something like Microsoft Access, but it's hard for me to imagine Excels interface actually help you accomplish much with tens of thousands of rows of data - that couldn't be accomplished much easier with a query tool. Of course if the tool works for you, then keep on rockin' it. Inspecting big delimited text files is ridiculously easy in a tool like Access if you've never tried it, just fyi.
> WTF? If I've got anyone in IT putting 1,000,000 rows in a spreadsheet, I'm seriously considering demoting them. If you're going to have a million rows, get a database. It's not the people in IT you have to worry about, it's analysts with 1MM row datasets. At first my thought was "the horror of a 1MM row spreadsheet! blashpemy!... but now I'm thinking at least they wont be as tempted to us MS Access..