Really? Don't blame Oracle, a huge and well funded IT company that claims to be the best of the best?
The defense of both the healthcare.gov contractor and Oracle keep whining about requirements not being defined until late in the game, but anyone with experience in software development knows that "requirements" evolve over time and iterative development is the only way to do any project of any significant size. Now who do you think is in a better position to know that and manage the project accordingly: A state or federal government official or the head sales guy at an IT consulting firm?
Buyer beware is still a bit true, but in 21st century America, and especially when accepting public money, it should be/seller/ beware. You can't (or shouldn't be allowed to) sell an obviously inadequate product or service and get away with it.
Both Oracle and the healthcare.gov contractor are 100% culpable. They should be in a position to know what they were getting into and should not have gotten into it if it's so poorly defined that they can't deliver.
IBM (another company I hate) actually did this right when they bowed out of that contract for a supercomputer for some university and just paid the associated fines. If you're the IT expert, it's your job to know how these things work and say if and when it can't be done.
Slashdot needs to up the maximum score a comment can get just for this comment. Give it a 10, and give Pearson a -6.02x10^23.
That said, I'm doing pretty well in life because I can figure out WTF the complete idiots at Pearson were thinking slightly more often than not. So possibly preparing kids early for the idiocy that is professional certifications isn't all bad for the kids. But for society, which needs a useful mechanism to determine someone's qualifications, things look pretty bleak.
I think you (and the entirety of the Republican party) misunderstood your microecon textbook. The point of a competitive free market is to reduce the maximum price a company can charge for something to just slightly more than it costs to produce (i.e. a small profit). Any company charging more will loose all their business to the competition who charges less. Free markets actually minimize profits, not maximize them. Monopolies maximize profits, which is to say they allow companies to charge as much as consumers are willing to pay before consumers instead choose to go without the product.
Of course this really only happens for commodities -- goods and services that can both be easily substituted and that the consumer understands well so he can easily decide if the competition is a sufficient substitution. Maybe all heath insurance is mostly the same and easily substituted, but it's definitely too complicated for consumers to understand if and when that's the case.
First off, let's define "hard". You could mean a) absolutely hard: it takes lots of effort to make this work at all b) hard to do well: it takes lots of effort to do this well even though I can do this somewhat acceptably with minimal effort c) time consuming: this takes a lot of f-ing time, and it's unclear that the effort justifies the benefit
a) seems like the most appropriate definition, but judging by the list they seem to mean either b or c.
9. Designing a solution :
b. I can make you some working software based on your off-the-cuff requirements pretty easily. Anticipating what you really meant, what you will ask for next, and writing code that can be easily leveraged to do those things would be 'a'.
8. Writing tests
c. For small projects, automated testing way more time than it's worth. For large projects writing tests is the only way to make it work at all. Of course, all those medium sized projects and those projects that start small but may become large are a challenge. And weather or not the software lends itself and the programming team knows how to use a testing suites make a difference.
7. Writing documentation
c. No one/ever/ reads documentation because we all learn the hard way that it's perpetually out of date. The UI and API/are/ the documentation. If, by "writing documentation" you mean "designing a good UI/API" that makes it obvious to the user what's going on, then this becomes 'a'.
6. Implementing functionality you disagree with
WTF - If you're getting paid, do what you're told. If not, tell 'em to do it themselves. This is only "hard" in any sense if you're a pedantic a-hole. Oh, wait. This is/., so I guess that's all of us.
5. Working with someone else’s code
b - But if they instead had "writing code that isn't a PITA for others to work with", then it's an 'a'.
4. Dealing with other people
That's only because: http://www.dilbert.com/2013-10-10/
But I guess if we spent any time developing our social skills, we wouldn't have had time to learn how to program.
3. Estimating time to complete tasks
Okay, this one really is 'a'. On the other hand, you just shouldn't do this. Instead, you need to get good at getting customers/users on board with iterative development where they wait/pay a bit and get some incremental functionality as you work towards some end goal that neither of you can really predict up front.
2. Explaining what I do (or don’t do)
See #3.
1. Naming things
See #5. Naming things is easy. And my names make perfect sense to me.
Also, queue the penis jokes based on my use of the word "hard" in the subject.
> Look a bit deeper and you'll notice that most of the knocking is done by those who never wrote anything more complicated than a few lines of Perl.
There's something more complicated than a few lines of perl code?
Seriously though, perl is an abomination and, while a revolution when it was first written, Larry Wall should be ashamed of himself for not encouraging people to chose a readable and user friendly scripting language for the 21st century. That said, I write most stuff in perl simply because it's already installed on just about every linux box in existence, packages for just about everything are available via my OS package manager and it actually does work just as well on Windows via ActivePerl when I need it to.
Why can't distros rewrite stuff in Ruby already and NOT install perl by default.
The 3rd party is the government that has to keep the peace. If you have a huge portion of the population up in arms because they work 12+ hours a day and can barley afford to feed themselves, the government has a responsibility to do something about that to protect the private property of the wealthy. There are two options: 1) Pay the police to intimidate, imprison and kill the poor or 2) Provide social wellfare programs to subsidies low wages and pacify the masses.
3rd world countries usually do #1, which is made possibly by buying sophisticated weaponry and technology, mainly from the U.S. 1st world countries usually do #2, but have to charge higher tax rates and/or go into lots debt to do it. The U.S. does both. We have an very high % of our population in prison, medium-high taxes and lots of debt.
We want wages set by the market, but we want that market to naturally set a living wage. The key is figuring out the government policy that can make that happen.
Regulation of pay (minimum wage OR CEO maximum pay) is obviously a horribly idea. However, we do need to provide food and gas subsidies and social wellfare programs (since I think we all agree that #2 is better than #1) to make sure everyone eats, has a place to live and gets some minimal amount of healthcare. That's essentially equivalent to a minimum wage. And we need to tax the hell out of either corporations or rich individuals (the later being preferable) to pay for it, which is essentially a maximum wage.
Subsidies, wellfare and taxes are all bad and equivalent to price and wage controls. They're not the natural fair distribution of wealth via market forces that we want. So what's a government to do?
If wages are low, work on education (get us some good schools rather than just throwing more money at the horrible k-12 and college systems we have today) and technology (fund the basic research that private industry can't because there's no clear path to profit)
If taxes on the rich have to be high to pay for wellfare programs, get people off wellfare by investing. In the long run, scientific research and infrastructure improvement will provide ROI by maintaining the U.S. as the most powerful country (economically and militarily) in the world. In the short run, it will create middle-class jobs. And those middle-class spenders will create demand for less-skilled labor -- reducing the need for wellfare.
When you tax the rich, Mr. Millionaire still buys his $5 coffee at Starbucks and doesn't ask Uncle Sam for a check (though he may bribe his congressman for lower taxes). When you cut government spending, Mr. Middle-class, now unemployed, certainly does drop Starbucks and, in a few years when his savings is exhausted, will be asking Uncle Sam for a check.
Does this remind anyone else of how religious philosophers of days past used to argue over how many angles could dance on the head of a pin? I'm not sure about the angles part, but there are surely some pinheads in this story.
It's probably just some responsible gun owner assuming that since the government has shut down the capitol should be empty and therefore would be the ideal place for a shooting range since there should be no chance of hitting anyone.
Seriously though, $10 says it's a U.S. citizen unhappy with D.C. dysfunction. The terrorists wouldn't waste their bullets. They're home watching CSPAN with a bowl of popcorn and thinking "Mission Accomplished".
It should now be obvious to everyone that we're on a one way train to rampant government censorship enforced at the ISP level with governments exercising legal threats towards ISPs to get their (and by 'their' I mean big corporations, rich religious conservatives and peope who use terrorist fear mongering to keep their cushy jobs.) way, and that western powers, rather than China and the middle east, will be leading the way.
But why is this really a problem? Do I care if they don't let me download pr0n? No. Do I care that they make me actually pay for my entertainment, possibly increasing the price? Not really. Am I scared of the next Hitler coming to power and using his control of the media to exterminate some subset of the population? Seems like a long shot at present. Will censorship prevent a few terrorist attacks by making it harder for them to communicate? Possibly. But all that junk is either unimportant (pr0n and piracy) or unlikely (Hitler and terrorists).
This article demonstrates the real problem with censorship: incompetence. They'll block the wrong stuff and there's nothing I can do about it. There will be a place to report problems, but reports will be ignored, or at least take 6 months to get resolved. The entirety of the Internet will be rendered useless. We may as well all just go back to writing letters and making phone calls (assuming those don't get blocked too).
I need to raise some money to buy a good supply of pens. Anyone want to buy a slightly used keyboard?
They can't do the work with 90 people. But they can outsource parts of it and buy a bunch of off-the-shelf hardware and software to do more with fewer people. This may actually save them money and is how we do it in corporate America. However, foreign governments aren't dieing to get at my datacenters, so I don't really worry about whether or not the firmware on that printer I just installed was hacked in the Chinese factory where it was made. I imagine the NSA is (or should be) a bit different and avoid 3rd party products for just that reason.
You realize you just slashdott'ed/DDOS's an Air Force server, right? That's hacking against the U.S. Government, tantamount to treason. You're goin' to jail, buddy!
That's why I code everything for Silverlight. It's backed by Microsoft, a multibillion dollar enterprise with a long history of excellent support. How could I go wrong?
> Why don't you just compile the damn libraries and 'gcc' for deployment?
Because those savvy enough to have a good legitimate reason to suspect the latest version will solve their problem don't ask me to. They compile it in user space and prove their suspicion correct. Then they ask me to deploy it more widely to others who will use their code, which I do.
> Programmers don't use their own custom versions of libraries because they really really wanted to. Uh, what terribly above-average user base are you lucky enough to support where not a single one of them attempt to procrastinate actually debugging their code by having you install the very latest of anything and everything they can think of just in case it might help? No all of them are that way, but there's always enough of them to make servicing such requests a full-time job if one were to entertain them all.
Sysadmin vs. Programmer Flame War! YES, I'm totally in!!!
As for programmers knowing what they're doing, how about things like: * this code totally worked yesterday, you must have changed something. Oh, wait. I meant it worked on that other system. Where I compiled my own version of libfoo and totally screwed with my LD_LIBRARY_PATH. * I wrote this code in objective-c on my Mac because, but gcc supports objective-c, so you can just recompile on your production Linux system. * I/must/ have the very latest version of libbar because maybe that will fix this strange behavior that I can't explain even though it means compiling from source rather than simply installing packages from the distro and have no idea what changes the latest version even has that might be related to my problem.
Admittedly, many of our "programmers" are scientists who know their science well but code only because computer models drive the science these days, so my degree in CS and experience as a sysadmin (and programmer) probably makes my thoughts on what you need not completely irrelevant. And in my experience, stuff like the above certainly isn't exclusive to scientists playing programmer, but a symptom of the fact that computers and software are overly complicated and it just takes more experience and patience that one would think it should to make them do anything.
But XML is the most abysmal failure of a language ever. SQL is really how we represent most useful data, but if falls down when representing hierarchies. That's where XML comes in. But XML fails horribly at relational data. There are (or were) too many supposedly validating parsers that didn't validate the uniqueness of xml:id attributes. And even if yours does, XML data is ugly and nearly impossible to read and write relational data in manually. And, unlike SQL, which is simple and elegant, xpath almost as bad as ldap in terms of having an ugly and unreadable syntax. So when I need hierarchical-relational data, I'd much rather struggle with SQL's difficulties representing hierarchies than XML's difficulties representing relations.
And even when we're talking just hierarchical data, the APIs for reading XML are horrible. It takes so much code to get anywhere that I can write my own parser for a simple make-it-up-as-I-go config file more easily than I can use the XML APIs in ever so many cases.
And WTF is with attributes? Sure, it sometimes looks cleaner when I'm editing XML by hand, but when coding I have to know not only the name of the thing I want but whether it's an attribute or another tag.
The APIs suck and there's just no obvious and universally agreed upon mapping from an XML file to in-memory object space like ActiveRecord and tons of other ORMs do for SQL.
XML = BIG FAIL! The proof is in the fact that only the Java and M$ idiots with more time than sense jumped on the XML band wagon. The rest of us invented JSON, YAML, or just continued to DOS-style '[block]\n option=value' config files for BOTH configs AND larger data sets.
So no, not every mid-level web programmer should know XML. In fact, programmers of all levels should avoid XML like the plague that it is.
And can someone please invent a language that represents both relational and hierarchical data well, including a text-based data format and schema, query language to grab subsets of a data set, efficient binary protocol to transfer datasets, fast query and storage engine and a simple API to shuffle data from serialized form to objects or arrays in memory. And while I'm at it, can someone do something about the Israelis and Palestinians already.
We run such a shop, but I'm a sysadmin here. I apologize. But you know that guy down the hall. The one who likes to copy.so files around from machine to machine because he thinks he knows what he's doing. The one who installs hundreds of god-only-knows-what pieces of software on his workstation. The one who's computer I can't just reimage because he swears he needs all this stuff that such as process would remove. He's why you're not root. And it turns out there are a lot more of him than you'd think.
> Seriously, web programming is for chumps, Really? The web rules because you don't have to install software on any computer other than your own server and HTTP naturally handles networking and caching for you, allowing your program to run anywhere. True, the W3C is completely impotent and the modern web is a bit of a kluge, but no so much more so than everything else that the no-install, run-anywhere benefits don't dominate. Web is the ONLY way to go.
> Let's talk about having to support multiple version of multiple browser on multiple versions of multiple operating systems on multiple platforms, You probably haven't done it for a while. Since IE8 or so, FF, IE and WebKit are actually reasonably consistent at rendering things. I use reload 8+ browser windows every time I made a change, but now I just use FF and test things across browsers near the end. Works fine.
> all with multiple sized screens. Okay, I'll give you that one, especially with mobile. But if you're not coding a site with more sidebars and ads than actual content, HTML does as good a job as Java's Swing, Awt and layout engines in other GUI languages and is far simpler to code.
> Let's talk about the expectation of being an expert at a horrendous number of technologies like HTML, CSS, Javascript, Ajax, GWT, Java, JSP, EJB, XML, JSF, > Facelets, JPA, JPQL, EL, SQL, PL/SQL, Regex, BASH etc. etc....for the one fucking project! You do need HTML, CSS, Javascript, SQL and one server-side language. Those languages all have different purposes and do them well. But: - Ajax - Jquery wraps this quite nicely - GWT - Just don't - Java - Does anyone actually write applets? And if you picked it for your server-side, shame on you. - JSP, EJB, j* - See above - XML - Why? Haven't we all switched to Json by now? - Facelets - WTF is that? - Regex - If you don't know regex, find another career. And that's not web-specific in any way. - Bash - For the web? I hope you're not sticking bash scripts in cgi-bin anymore?
> Let's talk about the expectation of being an expert at optimising different servers like Apache, Tomcat and JBoss. - WTF are you doing that you need to optimize for the server. Oh ya, you chose Java for your server-side. That was your first mistake.
> Let's talk about the expectation of being an expert at load testing using various load testing suits. I think you're remembering the pre-internet days when you didn't need load-testing because there was no way to connect all these computers to generate a load in the first place. If we just wrote desktop apps that connected to a SQL server, you'd have to load-test that too. Except without HTTP, you'd be writing your own client-side and middleware code to handle all the redirection, load-balancing and caching necessary when your load tests revealed that the expected use would cripple your single database server.
> Let's talk about the dismal state of Flash and Java Applets and HTML5. Ya, I'll give you that one too. But that's why we learned to do most of what we need with HTML/CSS/JS instead.
> I pity the poor web programmer (such as myself), for his or hers is surely a tortured life. Hey, if this was easy, they wouldn't pay you so much to do it.
Despite the fact that there's a huge number of XP users motivated to switch OSs, I predict that 2014 STILL won't be the year of the Linux desktop.
I actually do use it as my desktop at home and at work. But can KDE please make things like adding icons to the desktop and task bar something you don't have to call your sysadmin for. Win8 was a mistake, but KDE (still 2nd only to icwm) is equally bad.
0) Dr. Evil travels back in time and gets the founding fathers to remove the self incrimination part from the 5th amendment. 1) Eve, a criminal, confesses her crime to Paul, the priest. Paul is obviously sworn to secrecy as is customary. 2) The police suspect Paul, of the crime and try him. When he takes the stand swears on the bible that he will tell the truth, the whole truth and nothing but the truth. 3) On the stand, Paul says he's innocent but refuses to answer a more direct question because the answer would information Eve told him in confidence. 4) The jury can't agree to convict Paul of the crime in question, but the prosecutor adds something about refusing to provide evidence to the court and the Jury does agree to convict him on that lesser charge and he is jailed.
In this case, Paul is wrongfully convicted of refusing to self incriminate. The 5th amendment makes it clear that such a thing is not a crime.
Moveover, #3, 'The "benefit" can't be something that benefits all suspects equally, whether they're innocent, guilty of violating a just law, or guilty of violating an unjust law.' ignores the fact that even a small benefit to the innocent is much more valuable than a large benefit to the guilty. I might be able to protect myself from a criminal they couldn't convict, but I've got no chance against a malicious government without checks and balances.
...because hackers have better tools they get for free from the Interwebs. Of course, if it does turn out to be hackers, this Italian firm could always stick a EULA on it and have the BSA enforce. Hackers aren't afraid of the government, but BSA lawyers scare everyone.
It's news that a Google employee is being a dick, since they do have a "do no evil" policy. I hate M$ as much as the next/. reader, but we do have to support windows. We don't put our non-technical friends and family on Linux (still waiting for the year of the Linux desktop). Cut us sysadmins some slack already. @$$.
I have to agree with the majority, DNA is the same as fingerprints and photographs. Of course, them darn coppers shouldn't be allowed to take fingerprints and photographs, especially the latter steals your soul!
So to combat this horribly government overreach, I plan to become a cannibal. With all that DNA from lots of other people constantly in my mouth, they're bound to get a contaminated DNA sample, providing reasonable doubt and ensuring I'm never convicted for my new dietary choices.
Seems better for the 'victim' to leave it up. He can then claim slander and sue whoever posted it AND anyone who reposts it elsewhere. The burden will then be on those douches to have those companies remove the video. Net effect: People will learn to be more careful when posting untrue stuff about others and the Internet will be transformed into a bastion of truth.
> no one seriously expect a tablet to be a PC As a sysadmin that doesn't get to approve IT purchases but has to support them once they're here, that is absolutely NOT true. Most of my users want an iPad. Most need to be told that they ALSO have to have a desktop or laptop to actually do their jobs.
Slashdot Mirror
Really? Don't blame Oracle, a huge and well funded IT company that claims to be the best of the best?
The defense of both the healthcare.gov contractor and Oracle keep whining about requirements not being defined until late in the game, but anyone with experience in software development knows that "requirements" evolve over time and iterative development is the only way to do any project of any significant size. Now who do you think is in a better position to know that and manage the project accordingly: A state or federal government official or the head sales guy at an IT consulting firm?
Buyer beware is still a bit true, but in 21st century America, and especially when accepting public money, it should be /seller/ beware. You can't (or shouldn't be allowed to) sell an obviously inadequate product or service and get away with it.
Both Oracle and the healthcare.gov contractor are 100% culpable. They should be in a position to know what they were getting into and should not have gotten into it if it's so poorly defined that they can't deliver.
IBM (another company I hate) actually did this right when they bowed out of that contract for a supercomputer for some university and just paid the associated fines. If you're the IT expert, it's your job to know how these things work and say if and when it can't be done.
Slashdot needs to up the maximum score a comment can get just for this comment. Give it a 10, and give Pearson a -6.02x10^23.
That said, I'm doing pretty well in life because I can figure out WTF the complete idiots at Pearson were thinking slightly more often than not. So possibly preparing kids early for the idiocy that is professional certifications isn't all bad for the kids. But for society, which needs a useful mechanism to determine someone's qualifications, things look pretty bleak.
maximized and no higher.
Uh... that's a paradox.
I think you (and the entirety of the Republican party) misunderstood your microecon textbook. The point of a competitive free market is to reduce the maximum price a company can charge for something to just slightly more than it costs to produce (i.e. a small profit). Any company charging more will loose all their business to the competition who charges less. Free markets actually minimize profits, not maximize them. Monopolies maximize profits, which is to say they allow companies to charge as much as consumers are willing to pay before consumers instead choose to go without the product.
Of course this really only happens for commodities -- goods and services that can both be easily substituted and that the consumer understands well so he can easily decide if the competition is a sufficient substitution. Maybe all heath insurance is mostly the same and easily substituted, but it's definitely too complicated for consumers to understand if and when that's the case.
First off, let's define "hard". You could mean
a) absolutely hard: it takes lots of effort to make this work at all
b) hard to do well: it takes lots of effort to do this well even though I can do this somewhat acceptably with minimal effort
c) time consuming: this takes a lot of f-ing time, and it's unclear that the effort justifies the benefit
a) seems like the most appropriate definition, but judging by the list they seem to mean either b or c.
9. Designing a solution :
b. I can make you some working software based on your off-the-cuff requirements pretty easily. Anticipating what you really meant, what you will ask for next, and writing code that can be easily leveraged to do those things would be 'a'.
8. Writing tests
c. For small projects, automated testing way more time than it's worth. For large projects writing tests is the only way to make it work at all. Of course, all those medium sized projects and those projects that start small but may become large are a challenge. And weather or not the software lends itself and the programming team knows how to use a testing suites make a difference.
7. Writing documentation /ever/ reads documentation because we all learn the hard way that it's perpetually out of date. The UI and API /are/ the documentation. If, by "writing documentation" you mean "designing a good UI/API" that makes it obvious to the user what's going on, then this becomes 'a'.
c. No one
6. Implementing functionality you disagree with /., so I guess that's all of us.
WTF - If you're getting paid, do what you're told. If not, tell 'em to do it themselves. This is only "hard" in any sense if you're a pedantic a-hole. Oh, wait. This is
5. Working with someone else’s code
b - But if they instead had "writing code that isn't a PITA for others to work with", then it's an 'a'.
4. Dealing with other people
That's only because: http://www.dilbert.com/2013-10-10/
But I guess if we spent any time developing our social skills, we wouldn't have had time to learn how to program.
3. Estimating time to complete tasks
Okay, this one really is 'a'. On the other hand, you just shouldn't do this. Instead, you need to get good at getting customers/users on board with iterative development where they wait/pay a bit and get some incremental functionality as you work towards some end goal that neither of you can really predict up front.
2. Explaining what I do (or don’t do)
See #3.
1. Naming things
See #5. Naming things is easy. And my names make perfect sense to me.
Also, queue the penis jokes based on my use of the word "hard" in the subject.
Apparently not, since I did it multiple times. I blame the US public education system.
> Look a bit deeper and you'll notice that most of the knocking is done by those who never wrote anything more complicated than a few lines of Perl.
There's something more complicated than a few lines of perl code?
Seriously though, perl is an abomination and, while a revolution when it was first written, Larry Wall should be ashamed of himself for not encouraging people to chose a readable and user friendly scripting language for the 21st century. That said, I write most stuff in perl simply because it's already installed on just about every linux box in existence, packages for just about everything are available via my OS package manager and it actually does work just as well on Windows via ActivePerl when I need it to.
Why can't distros rewrite stuff in Ruby already and NOT install perl by default.
The 3rd party is the government that has to keep the peace. If you have a huge portion of the population up in arms because they work 12+ hours a day and can barley afford to feed themselves, the government has a responsibility to do something about that to protect the private property of the wealthy. There are two options: 1) Pay the police to intimidate, imprison and kill the poor or 2) Provide social wellfare programs to subsidies low wages and pacify the masses.
3rd world countries usually do #1, which is made possibly by buying sophisticated weaponry and technology, mainly from the U.S. 1st world countries usually do #2, but have to charge higher tax rates and/or go into lots debt to do it. The U.S. does both. We have an very high % of our population in prison, medium-high taxes and lots of debt.
We want wages set by the market, but we want that market to naturally set a living wage. The key is figuring out the government policy that can make that happen.
Regulation of pay (minimum wage OR CEO maximum pay) is obviously a horribly idea. However, we do need to provide food and gas subsidies and social wellfare programs (since I think we all agree that #2 is better than #1) to make sure everyone eats, has a place to live and gets some minimal amount of healthcare. That's essentially equivalent to a minimum wage. And we need to tax the hell out of either corporations or rich individuals (the later being preferable) to pay for it, which is essentially a maximum wage.
Subsidies, wellfare and taxes are all bad and equivalent to price and wage controls. They're not the natural fair distribution of wealth via market forces that we want. So what's a government to do?
If wages are low, work on education (get us some good schools rather than just throwing more money at the horrible k-12 and college systems we have today) and technology (fund the basic research that private industry can't because there's no clear path to profit)
If taxes on the rich have to be high to pay for wellfare programs, get people off wellfare by investing. In the long run, scientific research and infrastructure improvement will provide ROI by maintaining the U.S. as the most powerful country (economically and militarily) in the world. In the short run, it will create middle-class jobs. And those middle-class spenders will create demand for less-skilled labor -- reducing the need for wellfare.
When you tax the rich, Mr. Millionaire still buys his $5 coffee at Starbucks and doesn't ask Uncle Sam for a check (though he may bribe his congressman for lower taxes). When you cut government spending, Mr. Middle-class, now unemployed, certainly does drop Starbucks and, in a few years when his savings is exhausted, will be asking Uncle Sam for a check.
Does this remind anyone else of how religious philosophers of days past used to argue over how many angles could dance on the head of a pin? I'm not sure about the angles part, but there are surely some pinheads in this story.
It's probably just some responsible gun owner assuming that since the government has shut down the capitol should be empty and therefore would be the ideal place for a shooting range since there should be no chance of hitting anyone.
Seriously though, $10 says it's a U.S. citizen unhappy with D.C. dysfunction. The terrorists wouldn't waste their bullets. They're home watching CSPAN with a bowl of popcorn and thinking "Mission Accomplished".
It should now be obvious to everyone that we're on a one way train to rampant government censorship enforced at the ISP level with governments exercising legal threats towards ISPs to get their (and by 'their' I mean big corporations, rich religious conservatives and peope who use terrorist fear mongering to keep their cushy jobs.) way, and that western powers, rather than China and the middle east, will be leading the way.
But why is this really a problem? Do I care if they don't let me download pr0n? No. Do I care that they make me actually pay for my entertainment, possibly increasing the price? Not really. Am I scared of the next Hitler coming to power and using his control of the media to exterminate some subset of the population? Seems like a long shot at present. Will censorship prevent a few terrorist attacks by making it harder for them to communicate? Possibly.
But all that junk is either unimportant (pr0n and piracy) or unlikely (Hitler and terrorists).
This article demonstrates the real problem with censorship: incompetence. They'll block the wrong stuff and there's nothing I can do about it. There will be a place to report problems, but reports will be ignored, or at least take 6 months to get resolved. The entirety of the Internet will be rendered useless. We may as well all just go back to writing letters and making phone calls (assuming those don't get blocked too).
I need to raise some money to buy a good supply of pens. Anyone want to buy a slightly used keyboard?
They can't do the work with 90 people. But they can outsource parts of it and buy a bunch of off-the-shelf hardware and software to do more with fewer people. This may actually save them money and is how we do it in corporate America. However, foreign governments aren't dieing to get at my datacenters, so I don't really worry about whether or not the firmware on that printer I just installed was hacked in the Chinese factory where it was made. I imagine the NSA is (or should be) a bit different and avoid 3rd party products for just that reason.
You realize you just slashdott'ed/DDOS's an Air Force server, right? That's hacking against the U.S. Government, tantamount to treason. You're goin' to jail, buddy!
That's why I code everything for Silverlight. It's backed by Microsoft, a multibillion dollar enterprise with a long history of excellent support. How could I go wrong?
> Why don't you just compile the damn libraries and 'gcc' for deployment?
Because those savvy enough to have a good legitimate reason to suspect the latest version will solve their problem don't ask me to. They compile it in user space and prove their suspicion correct. Then they ask me to deploy it more widely to others who will use their code, which I do.
> Programmers don't use their own custom versions of libraries because they really really wanted to.
Uh, what terribly above-average user base are you lucky enough to support where not a single one of them attempt to procrastinate actually debugging their code by having you install the very latest of anything and everything they can think of just in case it might help? No all of them are that way, but there's always enough of them to make servicing such requests a full-time job if one were to entertain them all.
Sysadmin vs. Programmer Flame War! YES, I'm totally in!!!
As for programmers knowing what they're doing, how about things like: /must/ have the very latest version of libbar because maybe that will fix this strange behavior that I can't explain even though it means compiling from source rather than simply installing packages from the distro and have no idea what changes the latest version even has that might be related to my problem.
* this code totally worked yesterday, you must have changed something. Oh, wait. I meant it worked on that other system. Where I compiled my own version of libfoo and totally screwed with my LD_LIBRARY_PATH.
* I wrote this code in objective-c on my Mac because, but gcc supports objective-c, so you can just recompile on your production Linux system.
* I
Admittedly, many of our "programmers" are scientists who know their science well but code only because computer models drive the science these days, so my degree in CS and experience as a sysadmin (and programmer) probably makes my thoughts on what you need not completely irrelevant. And in my experience, stuff like the above certainly isn't exclusive to scientists playing programmer, but a symptom of the fact that computers and software are overly complicated and it just takes more experience and patience that one would think it should to make them do anything.
But XML is the most abysmal failure of a language ever. SQL is really how we represent most useful data, but if falls down when representing hierarchies. That's where XML comes in. But XML fails horribly at relational data. There are (or were) too many supposedly validating parsers that didn't validate the uniqueness of xml:id attributes. And even if yours does, XML data is ugly and nearly impossible to read and write relational data in manually. And, unlike SQL, which is simple and elegant, xpath almost as bad as ldap in terms of having an ugly and unreadable syntax. So when I need hierarchical-relational data, I'd much rather struggle with SQL's difficulties representing hierarchies than XML's difficulties representing relations.
And even when we're talking just hierarchical data, the APIs for reading XML are horrible. It takes so much code to get anywhere that I can write my own parser for a simple make-it-up-as-I-go config file more easily than I can use the XML APIs in ever so many cases.
And WTF is with attributes? Sure, it sometimes looks cleaner when I'm editing XML by hand, but when coding I have to know not only the name of the thing I want but whether it's an attribute or another tag.
The APIs suck and there's just no obvious and universally agreed upon mapping from an XML file to in-memory object space like ActiveRecord and tons of other ORMs do for SQL.
XML = BIG FAIL! The proof is in the fact that only the Java and M$ idiots with more time than sense jumped on the XML band wagon. The rest of us invented JSON, YAML, or just continued to DOS-style '[block]\n option=value' config files for BOTH configs AND larger data sets.
So no, not every mid-level web programmer should know XML. In fact, programmers of all levels should avoid XML like the plague that it is.
And can someone please invent a language that represents both relational and hierarchical data well, including a text-based data format and schema, query language to grab subsets of a data set, efficient binary protocol to transfer datasets, fast query and storage engine and a simple API to shuffle data from serialized form to objects or arrays in memory. And while I'm at it, can someone do something about the Israelis and Palestinians already.
We run such a shop, but I'm a sysadmin here. I apologize. But you know that guy down the hall. The one who likes to copy .so files around from machine to machine because he thinks he knows what he's doing. The one who installs hundreds of god-only-knows-what pieces of software on his workstation. The one who's computer I can't just reimage because he swears he needs all this stuff that such as process would remove. He's why you're not root. And it turns out there are a lot more of him than you'd think.
> Seriously, web programming is for chumps,
Really? The web rules because you don't have to install software on any computer other than your own server and HTTP naturally handles networking and caching for you, allowing your program to run anywhere. True, the W3C is completely impotent and the modern web is a bit of a kluge, but no so much more so than everything else that the no-install, run-anywhere benefits don't dominate. Web is the ONLY way to go.
> Let's talk about having to support multiple version of multiple browser on multiple versions of multiple operating systems on multiple platforms,
You probably haven't done it for a while. Since IE8 or so, FF, IE and WebKit are actually reasonably consistent at rendering things. I use reload 8+ browser windows every time I made a change, but now I just use FF and test things across browsers near the end. Works fine.
> all with multiple sized screens.
Okay, I'll give you that one, especially with mobile. But if you're not coding a site with more sidebars and ads than actual content, HTML does as good a job as Java's Swing, Awt and layout engines in other GUI languages and is far simpler to code.
> Let's talk about the expectation of being an expert at a horrendous number of technologies like HTML, CSS, Javascript, Ajax, GWT, Java, JSP, EJB, XML, JSF,
> Facelets, JPA, JPQL, EL, SQL, PL/SQL, Regex, BASH etc. etc....for the one fucking project!
You do need HTML, CSS, Javascript, SQL and one server-side language. Those languages all have different purposes and do them well. But:
- Ajax - Jquery wraps this quite nicely
- GWT - Just don't
- Java - Does anyone actually write applets? And if you picked it for your server-side, shame on you.
- JSP, EJB, j* - See above
- XML - Why? Haven't we all switched to Json by now?
- Facelets - WTF is that?
- Regex - If you don't know regex, find another career. And that's not web-specific in any way.
- Bash - For the web? I hope you're not sticking bash scripts in cgi-bin anymore?
> Let's talk about the expectation of being an expert at optimising different servers like Apache, Tomcat and JBoss.
- WTF are you doing that you need to optimize for the server. Oh ya, you chose Java for your server-side. That was your first mistake.
> Let's talk about the expectation of being an expert at load testing using various load testing suits.
I think you're remembering the pre-internet days when you didn't need load-testing because there was no way to connect all these computers to generate a load in the first place. If we just wrote desktop apps that connected to a SQL server, you'd have to load-test that too. Except without HTTP, you'd be writing your own client-side and middleware code to handle all the redirection, load-balancing and caching necessary when your load tests revealed that the expected use would cripple your single database server.
> Let's talk about the dismal state of Flash and Java Applets and HTML5.
Ya, I'll give you that one too. But that's why we learned to do most of what we need with HTML/CSS/JS instead.
> I pity the poor web programmer (such as myself), for his or hers is surely a tortured life.
Hey, if this was easy, they wouldn't pay you so much to do it.
Despite the fact that there's a huge number of XP users motivated to switch OSs, I predict that 2014 STILL won't be the year of the Linux desktop.
I actually do use it as my desktop at home and at work. But can KDE please make things like adding icons to the desktop and task bar something you don't have to call your sysadmin for. Win8 was a mistake, but KDE (still 2nd only to icwm) is equally bad.
0) Dr. Evil travels back in time and gets the founding fathers to remove the self incrimination part from the 5th amendment.
1) Eve, a criminal, confesses her crime to Paul, the priest. Paul is obviously sworn to secrecy as is customary.
2) The police suspect Paul, of the crime and try him. When he takes the stand swears on the bible that he will tell the truth, the whole truth and nothing but the truth.
3) On the stand, Paul says he's innocent but refuses to answer a more direct question because the answer would information Eve told him in confidence.
4) The jury can't agree to convict Paul of the crime in question, but the prosecutor adds something about refusing to provide evidence to the court and the Jury does agree to convict him on that lesser charge and he is jailed.
In this case, Paul is wrongfully convicted of refusing to self incriminate. The 5th amendment makes it clear that such a thing is not a crime.
Moveover, #3, 'The "benefit" can't be something that benefits all suspects equally, whether they're innocent, guilty of violating a just law, or guilty of violating an unjust law.' ignores the fact that even a small benefit to the innocent is much more valuable than a large benefit to the guilty. I might be able to protect myself from a criminal they couldn't convict, but I've got no chance against a malicious government without checks and balances.
...because hackers have better tools they get for free from the Interwebs. Of course, if it does turn out to be hackers, this Italian firm could always stick a EULA on it and have the BSA enforce. Hackers aren't afraid of the government, but BSA lawyers scare everyone.
It's news that a Google employee is being a dick, since they do have a "do no evil" policy. I hate M$ as much as the next /. reader, but we do have to support windows. We don't put our non-technical friends and family on Linux (still waiting for the year of the Linux desktop). Cut us sysadmins some slack already. @$$.
I have to agree with the majority, DNA is the same as fingerprints and photographs. Of course, them darn coppers shouldn't be allowed to take fingerprints and photographs, especially the latter steals your soul!
So to combat this horribly government overreach, I plan to become a cannibal. With all that DNA from lots of other people constantly in my mouth, they're bound to get a contaminated DNA sample, providing reasonable doubt and ensuring I'm never convicted for my new dietary choices.
Seems better for the 'victim' to leave it up. He can then claim slander and sue whoever posted it AND anyone who reposts it elsewhere. The burden will then be on those douches to have those companies remove the video. Net effect: People will learn to be more careful when posting untrue stuff about others and the Internet will be transformed into a bastion of truth.
Possibly I'm a bit over-optimistic.
> no one seriously expect a tablet to be a PC
As a sysadmin that doesn't get to approve IT purchases but has to support them once they're here, that is absolutely NOT true. Most of my users want an iPad. Most need to be told that they ALSO have to have a desktop or laptop to actually do their jobs.