How much would it cost to do it the Right Way from a user's point of view?
Blocking with an error code is the Right Way. That way the sending mail server generates a bounce message and the sender knows that the message didn't get through. The idea of accepting every message so the user can have 50,000 messages in his spambox that will never get looked at for every real message is absurd.
I'm not saying that static binding of event callbacks is the right answer, only that the current model that you describe in such detail isn't the only way to do it.
That's an implementation detail in your GUI toolkit, not a computing problem. Hell, it'd probably be possible to design a programming language that did static binding for event callbacks without changing the behavior - thus reducing that operation to zero runtime computation.
I do know they must be improved to capture more of the market.
My guess is that the number of people who try a Linux distro and give up over one piece of easily-replaceable unsupported hardware is pretty small.
Most people buy a computer knowing what OS it will run. Any significant market share increase for Linux is going to come in the form of people buying Linux boxes (i.e. from Dell or System76). Hell, people randomly installing Linux on their Windows box doesn't even *count* as "market share", because it's not a "sale".
I mean really why do people put up with this? It's almost as if people are too lazy to defend their privacy and too eager to whine about their problems or something.
There are a whole group of people who "defend their privacy" in cases like this simply by avoiding such products and services. These people have no social impact *at all*, because they don't say anything - which means everyone else thinks that "no one cares".
What that means is simple: Yes, you should actively defend your privacy by avoiding intrusive services. But you also need to whine about it on the internet to let others know that someone cares.
Yes. They are. That's what the word "criminal" means: someone who has committed a crime. And there are laws on the books making trivial possession of underage sexual pictures a crime.
I think the question you were looking for is this: Should they be criminals?
You mean to print ballots that are pre-filled out? I could print about one a second. Not that this matters as I could do it at my leisure.
There's a reason I differentiated between an editing attack and a swapping attack. In order to perform a swapping attack on paper ballots, you need to carry great stacks of ballots with you (i.e. cubic meters of them).
Once you get access, how long to upload whatever changes you want to make could take a while.
I'm pretty sure that plugging in the connector would take longer than any data transfer if the attack was properly automated.
Sorry to say it but any retard can stuff a paper ballot box. It takes an experienced hacker to hack an electronic election.
Any idiot with 5 minutes alone with a ballot box can stuff it, assuming that proper fake ballots had been prepared beforehand. Any idiot with 5 seconds alone with a DRE voting machine's input port can stuff it, assuming appropriate pre-preparation. Further, any idiot with 10 seconds alone with a DRE system's central tabulator can "stuff every ballot box" in like 10 seconds.
It's not that there aren't attacks against paper systems, it's that 80 year old poll volunteers know what to watch for and have a lot of chance to see it. Further, a successful attack has a much smaller potential impact. With a clever electronic attack the entire county election may be compromised with nothing apparently sketchy happening at all.
An adder is generally either used by a single user who wants accurate results or by a group of users who all want the same accurate results. Further, adders are generally designed as general-purpose components that will be used in hundreds of different applications - making one that output 3 for 1 + 1 would simply be a poor business decision when it was noticed rather than an effective attack against some specific application.
In contrast, voting machines are specific-purpose devices that are *always* used by large groups of people; and any of those people might want to tamper with the election. It should be obvious that this creates a relatively complex *security* problem rather than a simple electrical engineering / programming problem.
If that level of logging isn't on by default, then the voting machine manufacturers are even more incompetent than I thought. And that's saying a lot, because I doubt they could fasten their own shoes with velcro.
The problem is that paper based elections are no more secure, and if the physical ballots are lost, you're screwed.
They aren't? How many man-seconds alone with the ballots does it take to change the result of a paper election by editing the ballots? How many cubic meters of stuff do you need to carry to swap in forged ballots? Now how about electronically stored ballots?
The general answer to the questions you're asking is this article. Actually read it if you haven't seen it before - it's worth the time.
Ruby is an extention of java, so should I treat it as a fundamental programming language?
Ruby is about as far as you can get from Java while still having classes and mutable variables. There might be an argument for treating Perl, Python, and Ruby as the same language - but even that is stretching it a bit; it'd be like saying that C++ and Java are the same language.
lisp/prolog/haskell - these seem to have a reputation for being developed in academia, for academics. Will they make me the bucks once I leave academia?
Academics aren't just screwing around with pointless stuff for no good reason. Those languages are popular in academia because they allow the users to reason about programming problems in specific interesting ways - coincidentally, being able to reason about programming problems is what "makes you the bucks" as a programmer in industry.
erlang - as many slashdotters have pointed out, erlang's gimmick - being parallel, isn't all that efficient.
Erlang itself isn't wildly efficient because that wasn't its design goal. Erlang was designed for *reliability*, and concurrency was selected as an excellent way to accomplish that (in a well written Erlang program, a server exploding is a non-fatal exception). But... saying that Erlang-style concurrency isn't efficient in general is silly. In fact, it looks like exactly that concurrency model is going to be one of the best ways to effectively program for modern 8+ core CPUs.
By giving away their software for free, they are stealing from the mouths of paid developers!
I'm sure that all of the consultants who make great stacks of money implementing and supporting solutions based on Free Software would disagree with you.
Existing P2P programs would support this sort of thing today if the programmers hadn't been busy trying to work around blocking and traffic shaping.
Seriously... if the ISPs would spend their time *making their network better* instead of *trying to break user applications*, then we could compare P2P programs based on efficient network usage (= download speed) rather than ability to avoid traffic shaping (which = download speed today).
I have never understood this Geek obsession with derivative works.
Geek obsession?
Wanting to hear more about the same characters in the same world is a normal reaction to works of fiction, not limited to geeks at all. Otherwise there wouldn't be so much demand on authors (and movie directors) to release sequels.
Yea. Because interactivity trumps photorealism for every single possible type of game. Oh wait, that's false.
You sound like the people who said that StarCraft was crap because sprites were outdated junk and every good game (like Total Annihilation) had already moved to 3D. Different engineers will make different design choices for different applications, and there is no total order of correctness among a single class of design choice.
The time of a search scales according to the number of records.
False. The time of a search most likely scales according to the log-base-2 of the number of records - so searching two billion records only takes 3% more time than searching one billion records.
Indexing time is going to be proportional to edits - and since people using the site is the whole point, that's not something that should be discouraged.
I'm sorry - if storing or searching text is costing them too much money then they're doing things wrong. If the additional text causes more pageviews, then it must be valuable information - deleting it to reduce the number of pageviews would be idiotic.
So, by that logic, censorship is more evil than malicious libel, yelling "fire" in a public theatre, and even snuff-films, right?
Censorship is when someone in power prevents someone's speech from being heard. So yes... someone in power having enough control over communication to actually prevent those things from being communicated is more evil than a blog post that is maliciously libeling someone or a video file containing footage of someone being killed.
How are you going to censor someone from yelling "fire" in a crowded theater? A gag? Yea, gagging every theater goer is, in fact, more evil than one of them yelling "fire" would be.
Bullshit! The minute they do that, it opens the door for some scumbag politician's power play denying that people possess a right because it's not explicitly enumerated. That's why the bill of rights wasn't written that way in the first fucking place!
Whereas without such a document the politician would deny people *any* rights because there's no reason to think people have rights.
Consider the UN Universal Declaration of Human Rights. Most of the world blatantly ignores it, but it serves an important purpose - it gives us something to point at as a reference point for which rights are basic and universal.
Getting back to the US Bill of Rights, it's not that we would have a right to Privacy if no rights were enumerated, it's that we would have no right to Bear Arms is they weren't.
Fifteen months in, he's still learning, and treating our software product like a research project, as opposed to something we need to ship updates to every few months. It's brutal.
I think you made a mistake, and then misidentified which action was a mistake. Here's a hint: This person would have been a bad hire even if he claimed to be an expert in the technology you were using at hire time.
This suggests a good rule of thumb for replacing a platform:
If a sold-seeming programmer you're about to hire says "You're using platform X? Oh god no, I'm not interested in this job anymore." more than once, it's time to switch platforms.
Slashdot Mirror
Blocking with an error code is the Right Way. That way the sending mail server generates a bounce message and the sender knows that the message didn't get through. The idea of accepting every message so the user can have 50,000 messages in his spambox that will never get looked at for every real message is absurd.
I'm not saying that static binding of event callbacks is the right answer, only that the current model that you describe in such detail isn't the only way to do it.
In fact, some have argued that the problem of interactive GUI applications is innately concurrent: http://video.google.com/videoplay?docid=810232012617965344
That's an implementation detail in your GUI toolkit, not a computing problem. Hell, it'd probably be possible to design a programming language that did static binding for event callbacks without changing the behavior - thus reducing that operation to zero runtime computation.
Yes. There has been a fully functional NTFS driver for a while now.
My guess is that the number of people who try a Linux distro and give up over one piece of easily-replaceable unsupported hardware is pretty small.
Most people buy a computer knowing what OS it will run. Any significant market share increase for Linux is going to come in the form of people buying Linux boxes (i.e. from Dell or System76). Hell, people randomly installing Linux on their Windows box doesn't even *count* as "market share", because it's not a "sale".
There are a whole group of people who "defend their privacy" in cases like this simply by avoiding such products and services. These people have no social impact *at all*, because they don't say anything - which means everyone else thinks that "no one cares".
What that means is simple: Yes, you should actively defend your privacy by avoiding intrusive services. But you also need to whine about it on the internet to let others know that someone cares.
Yes. They are. That's what the word "criminal" means: someone who has committed a crime. And there are laws on the books making trivial possession of underage sexual pictures a crime.
I think the question you were looking for is this: Should they be criminals?
There's a reason I differentiated between an editing attack and a swapping attack. In order to perform a swapping attack on paper ballots, you need to carry great stacks of ballots with you (i.e. cubic meters of them).
I'm pretty sure that plugging in the connector would take longer than any data transfer if the attack was properly automated.
Any idiot with 5 minutes alone with a ballot box can stuff it, assuming that proper fake ballots had been prepared beforehand. Any idiot with 5 seconds alone with a DRE voting machine's input port can stuff it, assuming appropriate pre-preparation. Further, any idiot with 10 seconds alone with a DRE system's central tabulator can "stuff every ballot box" in like 10 seconds.
It's not that there aren't attacks against paper systems, it's that 80 year old poll volunteers know what to watch for and have a lot of chance to see it. Further, a successful attack has a much smaller potential impact. With a clever electronic attack the entire county election may be compromised with nothing apparently sketchy happening at all.
An adder is generally either used by a single user who wants accurate results or by a group of users who all want the same accurate results. Further, adders are generally designed as general-purpose components that will be used in hundreds of different applications - making one that output 3 for 1 + 1 would simply be a poor business decision when it was noticed rather than an effective attack against some specific application.
In contrast, voting machines are specific-purpose devices that are *always* used by large groups of people; and any of those people might want to tamper with the election. It should be obvious that this creates a relatively complex *security* problem rather than a simple electrical engineering / programming problem.
If that level of logging isn't on by default, then the voting machine manufacturers are even more incompetent than I thought. And that's saying a lot, because I doubt they could fasten their own shoes with velcro.
They aren't? How many man-seconds alone with the ballots does it take to change the result of a paper election by editing the ballots? How many cubic meters of stuff do you need to carry to swap in forged ballots? Now how about electronically stored ballots?
The general answer to the questions you're asking is this article. Actually read it if you haven't seen it before - it's worth the time.
Ruby is about as far as you can get from Java while still having classes and mutable variables. There might be an argument for treating Perl, Python, and Ruby as the same language - but even that is stretching it a bit; it'd be like saying that C++ and Java are the same language.
Academics aren't just screwing around with pointless stuff for no good reason. Those languages are popular in academia because they allow the users to reason about programming problems in specific interesting ways - coincidentally, being able to reason about programming problems is what "makes you the bucks" as a programmer in industry.
Erlang itself isn't wildly efficient because that wasn't its design goal. Erlang was designed for *reliability*, and concurrency was selected as an excellent way to accomplish that (in a well written Erlang program, a server exploding is a non-fatal exception). But... saying that Erlang-style concurrency isn't efficient in general is silly. In fact, it looks like exactly that concurrency model is going to be one of the best ways to effectively program for modern 8+ core CPUs.
I'm sure that all of the consultants who make great stacks of money implementing and supporting solutions based on Free Software would disagree with you.
Existing P2P programs would support this sort of thing today if the programmers hadn't been busy trying to work around blocking and traffic shaping.
Seriously... if the ISPs would spend their time *making their network better* instead of *trying to break user applications*, then we could compare P2P programs based on efficient network usage (= download speed) rather than ability to avoid traffic shaping (which = download speed today).
Geek obsession?
Wanting to hear more about the same characters in the same world is a normal reaction to works of fiction, not limited to geeks at all. Otherwise there wouldn't be so much demand on authors (and movie directors) to release sequels.
Yea. Because interactivity trumps photorealism for every single possible type of game. Oh wait, that's false.
You sound like the people who said that StarCraft was crap because sprites were outdated junk and every good game (like Total Annihilation) had already moved to 3D. Different engineers will make different design choices for different applications, and there is no total order of correctness among a single class of design choice.
False. The time of a search most likely scales according to the log-base-2 of the number of records - so searching two billion records only takes 3% more time than searching one billion records.
Indexing time is going to be proportional to edits - and since people using the site is the whole point, that's not something that should be discouraged.
I'm sorry - if storing or searching text is costing them too much money then they're doing things wrong. If the additional text causes more pageviews, then it must be valuable information - deleting it to reduce the number of pageviews would be idiotic.
Why delete when you can tag for improvement?
[This article doesn't provide any references; the information is untrustworthy.]
For any normal use - including accelerated video - the NVidia Linux drivers are solid and have been for years.
There may be edge cases where they have worse performance than other drivers, but not in any area that I've personally seen using the drivers.
Censorship is when someone in power prevents someone's speech from being heard. So yes... someone in power having enough control over communication to actually prevent those things from being communicated is more evil than a blog post that is maliciously libeling someone or a video file containing footage of someone being killed.
How are you going to censor someone from yelling "fire" in a crowded theater? A gag? Yea, gagging every theater goer is, in fact, more evil than one of them yelling "fire" would be.
I'm not seeing how your post is a negative response to my post.
Whereas without such a document the politician would deny people *any* rights because there's no reason to think people have rights.
Consider the UN Universal Declaration of Human Rights. Most of the world blatantly ignores it, but it serves an important purpose - it gives us something to point at as a reference point for which rights are basic and universal.
Getting back to the US Bill of Rights, it's not that we would have a right to Privacy if no rights were enumerated, it's that we would have no right to Bear Arms is they weren't.
I think you made a mistake, and then misidentified which action was a mistake. Here's a hint: This person would have been a bad hire even if he claimed to be an expert in the technology you were using at hire time.
This suggests a good rule of thumb for replacing a platform:
If a sold-seeming programmer you're about to hire says "You're using platform X? Oh god no, I'm not interested in this job anymore." more than once, it's time to switch platforms.