Well for one thing what the article is about, the fact that the desktop is extensible with Javascript. We now have something for the GUI that can play a similar roleto shell scripts for the command line.
I agree with you on Gnome 3. When you want to shift user bases you should rename the product. The problem is the developers see "Gnome" as not being the desktop but rather the GNU Object Model which has been updated. The users of the GNU Object Model are the developers not the users of the desktop.
Microsoft wants the Windows 7 community to migrate to Windows 8. On the other hand I think there would be a lot less friction if they had called it "Metro OS the successor to Windows" and had made it clear Windows 7 was the last version of Windows in the traditional sense.
Well we have more of your made up statistics. The comparison for a season company is YOY which is mainly flat. That would be like talking about how clothing stores sales are crushed in the 1st quarter because they aren't doing anything like their christmas volumes.
As for abandoning the desktop: Macbook Pro Retina, New iMac form factor, Mac Pro were all this year.
In all fairness the Gnome 2 userbase was not the desired userbase for Gnome 3. So being "deaf" was part of the design. Gnome wanted to shift its target market.
/. has become ultra conservative when it comes to interface changes. Any substantial change of a piece of software is going to involve some things getting worse in exchange for more things getting better. Which means complex existing workflows likely will have to change. They don't like that even though existing workflows usually stifle innovation.
I hate to tell you your example of Firefox doesn't work. Firefox evolved out of Mozilla product which was Netscape. The Firefox team was separate. Mozilla was then picked up by AOL for 5 years. Firefox was funded. After that Google provided the money. Mozilla foundation has paid the developers and they are scratching whatever itch the big donor wants scratched. There was no "scratch your own itch" almost ever in its development.
Suse is a mainstream distro. TubroLinux is a very mainstream disto (Asian language speaking) etc... There are mainstream distros that use KDE. RedHat is tied to Gnome tightly. Ubuntu is forking Gnome to do their own thing.
Actually the Windows paradigm has not intrenched itself universally. That's part of the problem Windows faces. For people in their mid 20s on up to early 40s they grew up with the evolution of Windows. The paradigms make sense because they evolved gradually. For people older it they never transitioned (in general) and for people younger there was too much culture by the time they got on board. Computer literacy for any fixed age has been going down for a decade.
As for Windows 8, I've just seen the new versions of Dynamix which shows off the new APIs. The direction for business are some very exciting layered applications that will require 8.
First off that's not a citation. Second I think you are confusing the claim being made by the GP. GP was claiming the larger the population the larger the sample sized needed.
A sample of 1 is inaccurate, but it is inaccurate because it is a small sample size. Choosing a sample of 1 tells you very little but it tells you equally little from a population of 3k, 3m, 3b or more.
Similarly sampling bias like your phone example is a problem, there are statistical means for overcoming it. But phone bias issues would apply to a small town as much as to the entire USA. Larger isn't the problem the non-randomness is.
The trouble here is that with our finite sampling process, because we live in the real world we have no way of knowing if our sample is actually random, because there are infinite location we could choose from.
Doesn't actually matter much. If I'm seeing changes across locations with a large sample that drops away as I choose lots (but not an infinite number) of locations. It works just like any other variable. For example I can do probabilities on sets like "all real numbers" and choose randomly from "places" on the number line. I'll miss some properties but I'll only miss properties which happen for a vanishingly small percentage of numbers.
If I were to randomly test 3000 integers I would with probability 0 get all of them odd. Doesn't matter where I test.
Lets make it worse. I were to randomly test 3000 real numbers there is 100% chance (rounded) that all of them will be irrational. I'd then conclude that 100% of the real numbers are irrational. That's a correct statistical statement, for each rational number there are more irrational numbers than the cardinality of the integers. I'd miss the vanishingly small percentage of rationals but for a statistical measure that's fine.
Then after learning the language the new developers are also more productive in Prolog or LISP. The languages are inherently more productive.
Moreover leaving is not a huge deal. You either train or hire for the skill set. These are not impossible find skill sets, rather you get very good quality people because the languages are enjoyable to top notch people.
The markup is not enough to cover the cost of a large inventory almost always. The only thing a large inventory can do is drive sales of small number of items (best sellers for example) and thus offset its costs through marketing. Anyway that's why they don't want to carry what you want to buy, it doesn't offset their costs. And you are absolutely right it would be bad for them if they were dumb enough to try and survive on selling inventory.
. The pols will simply accept your money and go on spending private money.
Money's effectiveness decreases exponentially. But funding challengers you make it very expensive to achieve meaningful changes through fund raising. The minus fund raising in terms of policy compromise might thus have a better chance of overwhelming the positives.
As far as purely issue based pacs. That's fine. That doesn't involve the candidates directly, so they aren't corrupted. A prolife pac supporting a prolife pol isn't a source of corruption.
You can try spending the world into prosperity, but even the U.S. isn't rich enough for that.
We don't have to spend the world into prosperity. A few nice things go a long way in improving our status.
There's nothing you could do for that country they wouldn't see as a threat to their "sovereignty".
We are currently flying drones over Pakistan routinely assassinated their citizens. We are a threat to their sovereignty. They suck but let's be clear they have a legitimate gripe on that one.
Prior to the 2006 there was a strong correlation between socially conservative and supporting neo-conservative foreign policy almost as high as socially conservative issues cross correlated with one another. That's starting to fall apart but the correlation is still very much there. Old correlates strongly with socially conservative.
As an aside I remember the cold war well. My attitude towards the "war on terror" has been "fool me once, shame on you. Fool me twice, shame on me." We all remember being lied to about the Russians so it might cut both ways.
Obama is the man who has to make the humiliating back downs over eg. settlers in the west bank. Contrast this with how the USA regards almost everyone else..
Yes lets contrast it. There are very few if any other countries where the United States tries to make demands about how the country internally organizes domestic housing or distributes population. Israel is uniquely disrespected this not being a purely internal matter.
I don't know Haskell, but in Erlang, which has semantics similar to those you show, maintaining locally accessible mutable state requries fighting the language.
In Haskell state is handled by a monad so it is always isolated. There is no state outside any monadic structure whether parallel or not. Anything not in particular monads can executed in any order by the compiler by default. So state can them be isolated into monads each of which has local but not global state (so it never gets shared) where the monad isn't going to cross threads or state can be in the global single threaded monad which the other threads interact with (this is how I/O works for example). You can also use an actor model where the monads have state and communicate with each other via actors and that gets sequenced via. the I/O monad. This is the default, the easiest models are no different than how a basic print statement works in Haskell.
IIUC that example calculates each small value of the function many times in the course of calculating larger values. Not ideal.
I know I said that. The non-memoized algorithm works wonderfully as an example of parallel code. The memoized version can't really be made parallel but it is efficient:
Spying on more than half the German population is their job? Similar for Brazil?
If they can, yes. There is good reason to object that the US shouldn't be that aggressive but that's been open policy for a long time. Snowden was about a secret policy involving secret laws.
No they were better in Nixons time and pre-Nixon. Do something like a draft which really pisses people off and see how effectively the government can derail the protest movement.
When the government invaded Iraq there was strong public support for the invasion. Support only gradually melted away. As far as opposition to the Wall Street Bailout that was pretty substantial it created an entire political movement and shifted the discourse on economic policies.
Edward Snowden has allowed the big telcos to come clean about the huge number of requests and has forced this issue into public debate. He did tremendous damage to Obama's reputation and may constrict the ability to American telcos to operate abroad. All that is so far, there may be more fallout of the next few years. That ain't bad.
Slashdot Mirror
Well for one thing what the article is about, the fact that the desktop is extensible with Javascript. We now have something for the GUI that can play a similar roleto shell scripts for the command line.
Support for touch
Integrated notification system
I agree with you on Gnome 3. When you want to shift user bases you should rename the product. The problem is the developers see "Gnome" as not being the desktop but rather the GNU Object Model which has been updated. The users of the GNU Object Model are the developers not the users of the desktop.
Microsoft wants the Windows 7 community to migrate to Windows 8. On the other hand I think there would be a lot less friction if they had called it "Metro OS the successor to Windows" and had made it clear Windows 7 was the last version of Windows in the traditional sense.
Well we have more of your made up statistics. The comparison for a season company is YOY which is mainly flat. That would be like talking about how clothing stores sales are crushed in the 1st quarter because they aren't doing anything like their christmas volumes.
As for abandoning the desktop: Macbook Pro Retina, New iMac form factor, Mac Pro were all this year.
In all fairness the Gnome 2 userbase was not the desired userbase for Gnome 3. So being "deaf" was part of the design. Gnome wanted to shift its target market.
/. has become ultra conservative when it comes to interface changes. Any substantial change of a piece of software is going to involve some things getting worse in exchange for more things getting better. Which means complex existing workflows likely will have to change. They don't like that even though existing workflows usually stifle innovation.
I hate to tell you your example of Firefox doesn't work. Firefox evolved out of Mozilla product which was Netscape. The Firefox team was separate. Mozilla was then picked up by AOL for 5 years. Firefox was funded. After that Google provided the money. Mozilla foundation has paid the developers and they are scratching whatever itch the big donor wants scratched. There was no "scratch your own itch" almost ever in its development.
Suse is a mainstream distro.
TubroLinux is a very mainstream disto (Asian language speaking)
etc... There are mainstream distros that use KDE. RedHat is tied to Gnome tightly. Ubuntu is forking Gnome to do their own thing.
Actually the Windows paradigm has not intrenched itself universally. That's part of the problem Windows faces. For people in their mid 20s on up to early 40s they grew up with the evolution of Windows. The paradigms make sense because they evolved gradually. For people older it they never transitioned (in general) and for people younger there was too much culture by the time they got on board. Computer literacy for any fixed age has been going down for a decade.
That's one of the problems Microsoft faces.
OSX productivity hostile? How is that true?
As for Windows 8, I've just seen the new versions of Dynamix which shows off the new APIs. The direction for business are some very exciting layered applications that will require 8.
But it isn't working on independents and democrats.
First off that's not a citation. Second I think you are confusing the claim being made by the GP. GP was claiming the larger the population the larger the sample sized needed.
A sample of 1 is inaccurate, but it is inaccurate because it is a small sample size. Choosing a sample of 1 tells you very little but it tells you equally little from a population of 3k, 3m, 3b or more.
Similarly sampling bias like your phone example is a problem, there are statistical means for overcoming it. But phone bias issues would apply to a small town as much as to the entire USA. Larger isn't the problem the non-randomness is.
Doesn't actually matter much. If I'm seeing changes across locations with a large sample that drops away as I choose lots (but not an infinite number) of locations. It works just like any other variable. For example I can do probabilities on sets like "all real numbers" and choose randomly from "places" on the number line. I'll miss some properties but I'll only miss properties which happen for a vanishingly small percentage of numbers.
If I were to randomly test 3000 integers I would with probability 0 get all of them odd. Doesn't matter where I test.
Lets make it worse. I were to randomly test 3000 real numbers there is 100% chance (rounded) that all of them will be irrational. I'd then conclude that 100% of the real numbers are irrational. That's a correct statistical statement, for each rational number there are more irrational numbers than the cardinality of the integers. I'd miss the vanishingly small percentage of rationals but for a statistical measure that's fine.
Then after learning the language the new developers are also more productive in Prolog or LISP. The languages are inherently more productive.
Moreover leaving is not a huge deal. You either train or hire for the skill set. These are not impossible find skill sets, rather you get very good quality people because the languages are enjoyable to top notch people.
The markup is not enough to cover the cost of a large inventory almost always. The only thing a large inventory can do is drive sales of small number of items (best sellers for example) and thus offset its costs through marketing. Anyway that's why they don't want to carry what you want to buy, it doesn't offset their costs. And you are absolutely right it would be bad for them if they were dumb enough to try and survive on selling inventory.
Money's effectiveness decreases exponentially. But funding challengers you make it very expensive to achieve meaningful changes through fund raising. The minus fund raising in terms of policy compromise might thus have a better chance of overwhelming the positives.
As far as purely issue based pacs. That's fine. That doesn't involve the candidates directly, so they aren't corrupted. A prolife pac supporting a prolife pol isn't a source of corruption.
We don't have to spend the world into prosperity. A few nice things go a long way in improving our status.
We are currently flying drones over Pakistan routinely assassinated their citizens. We are a threat to their sovereignty. They suck but let's be clear they have a legitimate gripe on that one.
Prior to the 2006 there was a strong correlation between socially conservative and supporting neo-conservative foreign policy almost as high as socially conservative issues cross correlated with one another. That's starting to fall apart but the correlation is still very much there. Old correlates strongly with socially conservative.
As an aside I remember the cold war well. My attitude towards the "war on terror" has been "fool me once, shame on you. Fool me twice, shame on me." We all remember being lied to about the Russians so it might cut both ways.
Yes lets contrast it. There are very few if any other countries where the United States tries to make demands about how the country internally organizes domestic housing or distributes population. Israel is uniquely disrespected this not being a purely internal matter.
Find a statistics 101 book that says it.
In Haskell state is handled by a monad so it is always isolated. There is no state outside any monadic structure whether parallel or not. Anything not in particular monads can executed in any order by the compiler by default. So state can them be isolated into monads each of which has local but not global state (so it never gets shared) where the monad isn't going to cross threads or state can be in the global single threaded monad which the other threads interact with (this is how I/O works for example). You can also use an actor model where the monads have state and communicate with each other via actors and that gets sequenced via. the I/O monad. This is the default, the easiest models are no different than how a basic print statement works in Haskell.
I know I said that. The non-memoized algorithm works wonderfully as an example of parallel code. The memoized version can't really be made parallel but it is efficient:
fibs = 0 : 1 : zipWith (+) fibs (tail fibs)
fib n = fibs!!n
Haskell outside of monads does not.
If they can, yes. There is good reason to object that the US shouldn't be that aggressive but that's been open policy for a long time. Snowden was about a secret policy involving secret laws.
9601 for a 95% chance of having an error less than 1%
384 for a 95% chance of having an error less than 5%
Where are you getting that from?
No they were better in Nixons time and pre-Nixon. Do something like a draft which really pisses people off and see how effectively the government can derail the protest movement.
When the government invaded Iraq there was strong public support for the invasion. Support only gradually melted away. As far as opposition to the Wall Street Bailout that was pretty substantial it created an entire political movement and shifted the discourse on economic policies.
Edward Snowden has allowed the big telcos to come clean about the huge number of requests and has forced this issue into public debate. He did tremendous damage to Obama's reputation and may constrict the ability to American telcos to operate abroad. All that is so far, there may be more fallout of the next few years. That ain't bad.
True spying on allies is their job and it is fine. We aren't supposed to be spying on diplomats though. That is a serious breach.