I'll make a bet with you. You set up the daytime server on your desktop machine. Give me the name of your machine (not the IP). I will try to resolve your machines name and connect to the daytime server once a minute for the next 90 days. If I can hit your daytime server every single time, I will give you $100 cash. If I can not, you will give me $100 cash.
If you can't do something as simple as keep a machine available for 90 days, is it because your OS sucks? Because I have an old Pentium 266 machine sitting under my desk that I've been able to hit, every single time, for over a year.
I'm assuming that sometime in the next week one of the slashdot editors will be trolled with an article like:
I'm rather new with programming and stumbled across the article
Go To Statement Considered Harmful from 1968. I tried some of the examples in this article, and noted that some of those issue reported no longer seem valid whereas quite a few still very much are around. What has been improved since the article was written? Will the new 64-bit architectures finally fix all the problems with Go To Statements, or is this something that the hardware designers still need to work on?
I agree with you that if he's serious about running a server he should just go buy a used server.
But, unless he buys a used eMachine from a guy in a back of a van, I doubt the hardware is going to be the significant factor in his machine's availability. Some of the problems are going to be hardware related -- a UPS fails, or the server room air conditioning fails, or a network switch fails, or a land-line gets cut by a backhoe, or the janitor hits the machine hard with the floor buffer.
But common problems are often going to be soft failures -- your ISP changes your IP block and forgets to tell you, or the company providing you with DNS goes through its end-of-life death rattles, or your server room is scheduled to be rewired, or your leased line stops working and none of the three companies involved admits responsibility.
Of course, the good old-fashioned wet failures are probably going to be the most common on a club server admin'd by a rotating staff of part-time inexperienced admins. Someone discovers, after a kernel upgrade, that they forgot to compile in support for the off-the-wall filesystem some past admin stuck on a filesystem. Or, someone hoses the apache configuration, and can't fix the problem for two weeks. Or, some guy who had the root password for two weeks back three semesters ago installed an unpatched PHPnuke website, and now more scriptkiddies have accounts than regular users.
Servers are a pain in the ass. Dual hot-swappable power supplies are the final step in setting up a high-availabilty machine. They're also the cheapest part of the whole process.
I don't know what kind of club you have, or what kind of users you have. But, I assume that the things you have to worry about, in order from most important to least important, are
data reliability
administration policy
availability, and last,
performance
.
The biggest chunk of your budget (which is time and cash) should probably go to your backup solution and your security audit policy. Remember, RAID is NOT a backup. Nothing will torque your users more than losing all of their files when your RAID array is corrupted when you kick out the power cord at 3 am while doing routine maintenance. Having good backups is a must. However, your users probably will be nearly as torqued when some luser's PHP website goes bad and all the database passwords are sprayed across the web, or when one of the several monthly security patches doesn't get applied, and a l33t dude decides to take down the box.
The second biggest chunk of your time should probably go to your administration policies -- who gives out accounts, who terminates accounts, who helps with account problems, who deals with the results of the security audits, who is on call for server problems, who is given the administrator's cell phone number, etc.
When a user mistakenly does an rm -rf * on his entire web directory, who does he call? When a user wants to get back an email that he recieved sometime in June of 2001, who does he call? When a user wants to get his database backed up before he starts making big changes to it, who does he call? When a user needs a Perl module installed for his website, who does he call? When a user wants to add an entry to the DNS server, who does he call?
These are the things that your users will actually care about. They're also the biggest pains in the ass you can possibly imagine. This is why people pay for server administrators.
Next, think about availability. This includes simple things like how often stuff will break, and how quickly you will be to get the cash you'll need to replace the broken stuff. It also includes stuff like the DNS servers you'll be using, and the network line you'll be using, and the power supply to the building, and even the quality of the air conditioning in the room you'll have. Also, if you do have a secure location, who has the keys you'll need to get in there at 3 in the morning when you have to hit the reset button?
The LAST thing to consider is performance. It sounds like your entire server will fit on an old Pentium 66 with 128 mb of RAM. And, I imagine you'll be using the school's network, so I doubt you have to worry about paying the recurring network line lease costs.
You're looking at all of the sexy stuff with the server. Unfortunately, servers are not sexy. They're a pain in the ass. Having a Dual Pentium Xeon 2.8 GHz machine with 8 Gb of RAM is fun. Having 250 pissed off users calling you when a power outage corrupts your RAID array during finals week is not fun.
What, aside from decoding some jpegs, is fp used for these days? It's faster to do the calculations as integers and then divide into a floating point value when completed.
I was not aware of this fact. Do you have a link?
Integer calculations certainly take much less silicon. But I was not aware that they were faster in modern desktop processors. I would have assumed that for just about all desktop machines the exact opposite was true -- floating point operations would be more heavily optomized, simply because they're more heavily used, and silicon is so cheap. I guess I'm wrong. I learn something new every day.
Re:None is so blind as those who will not see
on
Jaguar is Over
·
· Score: 1
Your math is wrong. The "initial cost" of Mac OS X is free. It costs $0. Second, you're purposely misrepresenting Microsoft's release rate. If you are talking about "home versions" of their operating system, the release cycle has been:
1995: Win95a
1997: Win98b (this is not a typo, and b was not a downloadable upgrade)
1998: Win98
2000: WinME
2002: WinXP
There has never been even two years in a row, much less three, where you got all the updates from Microsoft "for free", despite your misleading claim to the contrary.
If we assume a the "Home" verison of each operating system, purchased in the year 2000, the math goes something like this:
Microsoft, home versions:
2000: Windows ME, full version: $200
2002: Windows XP Home, upgrade: $100
Total: $300
Microsoft, full featured versions:
2000: Windows 2000 Pro, full version: $400
2002: Windows XP Pro, upgrade: $200
Total: $600
Apple, home and business versions:
2000: Mac OS X 10.1, full version: $0
2001: Mac OS X 10.2, upgrade: $130
2002: Mac OS X 10.3, upgrade: $130
Total: $260.
I do not see any scenario where the Apple price is higher. Please point out where my math is wrong.
Microsoft WinXP Pro, full version: $400. Microsoft WinXP Home, full version, $300. Mac OS X 10.3, full version: $0.
Microsoft WinXP Pro, upgrade: $200. Microsoft WinXP Home, upgrade: $100. Mac OS X 10.3, upgrade: $130.
Yes, Mac OS X 1.3 upgrade is $30 cheaper than WinXP Home upgrade. If you're clalim is "Microsoft sells something -- anything -- for less than $130" then you've won your arguement. Congratulations. I was claiming that Mac OS X was not cheaper than Win XP. I'm pretty sure I've won my arguement too, hands down.
I guess we were just arguing two different things. By the way, Microsoft also sells mice. They have more than one button. They also cost even less than $100.
Next time, perhaps you could argue that Mac OS X is more expensive than a Microsoft Mouse. You'd win that one too.
I'll just keep to my "Mac OS X is not more expensive than other operating systems" arguement, since that seems to be the valid comparison. I've seen nothing to contradict such a claim.
You would prefer that ANYONE can watch your public actions
You've perfectly defined "public actions." Was this intentional?
You're perfectly happy letting some $11 an hour security guards build their own person "america's raciest street videos" for their own wanking pleasure?
You would have no problem with someone following you around every moment you're in public, watching you closely, and keeping notes, as long as that person has a shiney badge on. However, you want to be absolutely certain that no-one without shiney badges be allowed to follow around the people with the badges.
You have far more trust in shiney badges than I do.
You're right -- the $60 is not for the RedHat OS. It's the "basic" level of the RedHat network. The prices for the RedHat network are:
$60 per computer per year -- Basic RedHat Network.
$96 per computer per year -- Enterprise RedHat Network.
I don't know where the $39 or even the $149 price you cite comes from, though -- I don't see any prices that low on RedHat's site. Their current OS price list is something like
$2,499 per computer per year -- Enterprise AS Operating System (Premium)
$1,499 per computer per year --- Enterprise AS Operating System (Standard)
$799 per computer per year -- Enterprise ES Operating System (Standard)
$349 per computer per year -- Enterprise ES Operating System (Basic)
$299 per computer per year -- Enterprise WS Operating System (Standard), and
$179 per computer per year -- Enterprise WS Operating System (Basic)
On any comparison between Apple's prices and RedHat's prices, Apple wins. Hand down.
It's not enough to ask, "is this location being watched by a public agency." The question that must be answered is, "how can I get a copy of the recording."
If these are public cameras, being paid for by public funds, with the justification that they are recording public space, then only one conclusion is possible. Every person must be allowed complete and uncensored access to these cameras. There can be no argument that anything recorded by these cameras should not be available to the public. Any argument to that effect would imply immediately that these cameras are not recording public information, but are recording something else entirely.
If these cameras are not, in fact, public cameras recording public actions in public places, freely available to any and all members of the public, then they simply should not exist.
I was replying to a person who made the absolutely nonsensical claim that WinXP Pro is somehow a "server" version of the Microsoft OS, and made the grossly innapropriate comparison between the $199 WinXP Pro upgrade price and the $499 Mac OS X 10-client server price..
I'll assume that you're *not* the person I was originally replying to, and agree with me that his line of reasoning was totally broken. You've certainly made no attempt to support his arguement, so I imagine this is a safe assumption.
You're bringing up an entirely different, and totally unrelated, question -- is it more appropriate to compare the $129 MacOS X Panther upgrade with the $99 WinXP Home upgrade, or the $199 WinXP Pro upgrade. I have absolutely no desire to answer such a question, and am really only posting this reply to make it clear that I have never attempted to answer such a question. To claim that I somehow did answer that question, and that I answered it in a way unsatisfactory to you, is total nonsense.
The "server" version of RedHat costs several thousands of dollars. The $60 version of redhat is the desktop version. Do you have any idea what software costs?
What are you talking about? Windows 2000 Pro is a desktop OS. Windows XP Pro is a desktop OS. Windows Longhorn will be, I assume, a desktop OS.
Windows 2000 Server with a 10-CAL license is far more than $499. Windows 2003 Server with a 10-CAL license is far more than $499. The server version of Longhorn with a 10-CAL license will be, I assume, far more than $499.
Please, explain your comment to me. Do you honestly believe that the $199 upgrade versions Windows 2000 Pro or Windows XP Pro are the "server" version of those operating systems? Do you have any idea what software costs?
Re:Let me be the first to say...
on
Jaguar is Over
·
· Score: 1
Apple has decide there is a sane upgrade price for users of Mac OS X 10.x. The sane upgrade price they've decided on is $129.
I imagine they could charge people with Mac OS 9 even more, but I understand why they're reluctant to do that. Doing that would be difficult, expensive to administer, and just cause resentment.
Windows 2000 Pro came out in 2000. The upgrade price was about $200. Windows XP Pro came out in '02. The upgrade price was about $200. Longhorn is going to be out in another year. The upgrade is going to be about $200.
So... in 4 years, you're spending $600 on the OS.
Pleae tell me how is this different than Apple?
Re:Let me be the first to say...
on
Jaguar is Over
·
· Score: 1
The upgrade price for Panther is $129. The non-upgrade price is... get ready... $0. It comes for free with new machines.
Seriously, what are you asking? Do you have an Apple computer somewhere that doesn't already have an operating system license?
The Dell Latitudes are not marketed to people who want high performance machines. They're marketed to people who want to be certain that if they order multiple machines with identical specifications they'll get multiple identical machines. Not only will you get several identical machines if you order them all at once (which is much more unusual than you may realize), but you'll also be able get identical machines if you order them again three months later (which is a much more common requirement than you may realize). And, you'll still have identical machines after the various warranty replacements you'll inevitably have over the next three years.
If you don't get any value from having multiple identical machines, then you don't want to buy Dell Latitudes. They don't offer anything of value to you. But don't act as if that means they don't offer anything of value to anyone.
I agree too. The world would be a much worse place if we couldn't all see your mother in the shower.
You're correct. Only law enforcement agents should be able to see the cameras. Because, we all know that law enforcement agents never use the information they're given for ill purposes.
Those tatoos look like shit. If those poor tatoo "artists" didn't have skin to doodle on, they'd have to use black velvet or the cover of their third-hour study hall notebooks.
I'm sure there are good looking tatoos, on good looking people, appreciated by people with artistic sensibility. You've managed to prove that this simply isn't always the case.
Find the person in management who made the business decision to get this project done on the client's deadline. This person is probably not your boss. Schedule a one-on-one meeting with that person. Ask this person, "what are the incentives for getting this project done on the client's deadline?"
If you can't get a one-on-one meeting with the person making the money decisions, it's time to seriously examine your career. You're as valuable to your company as a fry cook is to McDonald's. You'll have to ask some serious soul searching questions, like "am I a shitty employee, or am I working for a shitty company, or both?" and "is this a problem that I'm willing to expend the effort on to fix?"
Also, make sure you're actually negotiating. Know what you're offering to your company, and know what you're willing to accept in return. Know these things before you walk into that meeting. Last, note that it would have been much better to do this before your company agreed to the clients demands. Your company is as stuck as you are right now -- it doesn't sound like they have very much wiggle room left in the contract, and they may not have much wiggle room left in the developer budget, either.
Think very carefully before you negotiate a resolution like "you will give me a bonus next time, or I will not be working for you anymore."
And whatever you do, never, ever, ever bitch about work at work. All good managers appreciate you negotiating for what you want and what you deserve. No manager appreciates and employee who bitches and moans about what he has.
Actually, it arguably doesn't matter and arguably helps American business since our products are cheaper abroad making them more attractive ovreseas.
It's too bad that businesses don't vote. All economists agree that the falling dollar hurts the american public. And the american public votes.
No, wait. We don't vote. Nevermind.
Re:windows vs *nix - un-informed is un-informed
on
Yet Another Windows Worm
·
· Score: 2, Insightful
This reflects one of the design philosophies of *nix: only give users the privileges they need, and have a huge, well defined wall between them and the system.
You're smoking a huge crack pipe, my friend. In unix, I need suid to change my password, 'fer christ's sake.
I mean, it's painfully obvious that you have no unix experience whatsoever. It's just sad that you got modded up on a site like slashdot, which used to be moderated by geeks.
varargs? Why? It is an object-oriented language - take use of that polymorphism!
How are varargs and polymorphism related? Or, rather, for a language using single dispatch, like Java, how are varargs and polymorphism related, in any way?
I don't know if you've noticed, but Java doesn't do method dispatch based on the runtime type of a methods arguements. In other words, the methods's arguements have nothing to do with the polymorphism of a method. So, how could something like varargs, which is a change that effects the arguements to a method, be replaced by single-dispatch polymorphism?
90 day uptime is actually pretty good.
I'll make a bet with you. You set up the daytime server on your desktop machine. Give me the name of your machine (not the IP). I will try to resolve your machines name and connect to the daytime server once a minute for the next 90 days. If I can hit your daytime server every single time, I will give you $100 cash. If I can not, you will give me $100 cash.
If you can't do something as simple as keep a machine available for 90 days, is it because your OS sucks? Because I have an old Pentium 266 machine sitting under my desk that I've been able to hit, every single time, for over a year.
What type of downtime is acceptable?
I agree with you that if he's serious about running a server he should just go buy a used server.
But, unless he buys a used eMachine from a guy in a back of a van, I doubt the hardware is going to be the significant factor in his machine's availability. Some of the problems are going to be hardware related -- a UPS fails, or the server room air conditioning fails, or a network switch fails, or a land-line gets cut by a backhoe, or the janitor hits the machine hard with the floor buffer.
But common problems are often going to be soft failures -- your ISP changes your IP block and forgets to tell you, or the company providing you with DNS goes through its end-of-life death rattles, or your server room is scheduled to be rewired, or your leased line stops working and none of the three companies involved admits responsibility.
Of course, the good old-fashioned wet failures are probably going to be the most common on a club server admin'd by a rotating staff of part-time inexperienced admins. Someone discovers, after a kernel upgrade, that they forgot to compile in support for the off-the-wall filesystem some past admin stuck on a filesystem. Or, someone hoses the apache configuration, and can't fix the problem for two weeks. Or, some guy who had the root password for two weeks back three semesters ago installed an unpatched PHPnuke website, and now more scriptkiddies have accounts than regular users.
Servers are a pain in the ass. Dual hot-swappable power supplies are the final step in setting up a high-availabilty machine. They're also the cheapest part of the whole process.
$4,000 per month? or $4,000 per year?
I don't know what kind of club you have, or what kind of users you have. But, I assume that the things you have to worry about, in order from most important to least important, are
- data reliability
- administration policy
- availability, and last,
- performance
.The biggest chunk of your budget (which is time and cash) should probably go to your backup solution and your security audit policy. Remember, RAID is NOT a backup. Nothing will torque your users more than losing all of their files when your RAID array is corrupted when you kick out the power cord at 3 am while doing routine maintenance. Having good backups is a must. However, your users probably will be nearly as torqued when some luser's PHP website goes bad and all the database passwords are sprayed across the web, or when one of the several monthly security patches doesn't get applied, and a l33t dude decides to take down the box.
The second biggest chunk of your time should probably go to your administration policies -- who gives out accounts, who terminates accounts, who helps with account problems, who deals with the results of the security audits, who is on call for server problems, who is given the administrator's cell phone number, etc.
When a user mistakenly does an rm -rf * on his entire web directory, who does he call? When a user wants to get back an email that he recieved sometime in June of 2001, who does he call? When a user wants to get his database backed up before he starts making big changes to it, who does he call? When a user needs a Perl module installed for his website, who does he call? When a user wants to add an entry to the DNS server, who does he call?
These are the things that your users will actually care about. They're also the biggest pains in the ass you can possibly imagine. This is why people pay for server administrators.
Next, think about availability. This includes simple things like how often stuff will break, and how quickly you will be to get the cash you'll need to replace the broken stuff. It also includes stuff like the DNS servers you'll be using, and the network line you'll be using, and the power supply to the building, and even the quality of the air conditioning in the room you'll have. Also, if you do have a secure location, who has the keys you'll need to get in there at 3 in the morning when you have to hit the reset button?
The LAST thing to consider is performance. It sounds like your entire server will fit on an old Pentium 66 with 128 mb of RAM. And, I imagine you'll be using the school's network, so I doubt you have to worry about paying the recurring network line lease costs.
You're looking at all of the sexy stuff with the server. Unfortunately, servers are not sexy. They're a pain in the ass. Having a Dual Pentium Xeon 2.8 GHz machine with 8 Gb of RAM is fun. Having 250 pissed off users calling you when a power outage corrupts your RAID array during finals week is not fun.
What, aside from decoding some jpegs, is fp used for these days? It's faster to do the calculations as integers and then divide into a floating point value when completed.
I was not aware of this fact. Do you have a link?
Integer calculations certainly take much less silicon. But I was not aware that they were faster in modern desktop processors. I would have assumed that for just about all desktop machines the exact opposite was true -- floating point operations would be more heavily optomized, simply because they're more heavily used, and silicon is so cheap. I guess I'm wrong. I learn something new every day.
There has never been even two years in a row, much less three, where you got all the updates from Microsoft "for free", despite your misleading claim to the contrary.
If we assume a the "Home" verison of each operating system, purchased in the year 2000, the math goes something like this:
Microsoft, home versions:
Microsoft, full featured versions:
Apple, home and business versions:
I do not see any scenario where the Apple price is higher. Please point out where my math is wrong.
Actually, here are some numbers:
Microsoft WinXP Pro, full version: $400.
Microsoft WinXP Home, full version, $300.
Mac OS X 10.3, full version: $0.
Microsoft WinXP Pro, upgrade: $200.
Microsoft WinXP Home, upgrade: $100.
Mac OS X 10.3, upgrade: $130.
Yes, Mac OS X 1.3 upgrade is $30 cheaper than WinXP Home upgrade. If you're clalim is "Microsoft sells something -- anything -- for less than $130" then you've won your arguement. Congratulations. I was claiming that Mac OS X was not cheaper than Win XP. I'm pretty sure I've won my arguement too, hands down.
I guess we were just arguing two different things. By the way, Microsoft also sells mice. They have more than one button. They also cost even less than $100.
Next time, perhaps you could argue that Mac OS X is more expensive than a Microsoft Mouse. You'd win that one too.
I'll just keep to my "Mac OS X is not more expensive than other operating systems" arguement, since that seems to be the valid comparison. I've seen nothing to contradict such a claim.
WinXP is not cheaper than Mac OS X. Where did you get such an idea?
You would prefer that ANYONE can watch your public actions
You've perfectly defined "public actions." Was this intentional?
You're perfectly happy letting some $11 an hour security guards build their own person "america's raciest street videos" for their own wanking pleasure?
You would have no problem with someone following you around every moment you're in public, watching you closely, and keeping notes, as long as that person has a shiney badge on. However, you want to be absolutely certain that no-one without shiney badges be allowed to follow around the people with the badges.
You have far more trust in shiney badges than I do.
- $60 per computer per year -- Basic RedHat Network.
- $96 per computer per year -- Enterprise RedHat Network.
I don't know where the $39 or even the $149 price you cite comes from, though -- I don't see any prices that low on RedHat's site. Their current OS price list is something likeOn any comparison between Apple's prices and RedHat's prices, Apple wins. Hand down.
It's not enough to ask, "is this location being watched by a public agency." The question that must be answered is, "how can I get a copy of the recording."
If these are public cameras, being paid for by public funds, with the justification that they are recording public space, then only one conclusion is possible. Every person must be allowed complete and uncensored access to these cameras. There can be no argument that anything recorded by these cameras should not be available to the public. Any argument to that effect would imply immediately that these cameras are not recording public information, but are recording something else entirely.
If these cameras are not, in fact, public cameras recording public actions in public places, freely available to any and all members of the public, then they simply should not exist.
I was replying to a person who made the absolutely nonsensical claim that WinXP Pro is somehow a "server" version of the Microsoft OS, and made the grossly innapropriate comparison between the $199 WinXP Pro upgrade price and the $499 Mac OS X 10-client server price..
I'll assume that you're *not* the person I was originally replying to, and agree with me that his line of reasoning was totally broken. You've certainly made no attempt to support his arguement, so I imagine this is a safe assumption.
You're bringing up an entirely different, and totally unrelated, question -- is it more appropriate to compare the $129 MacOS X Panther upgrade with the $99 WinXP Home upgrade, or the $199 WinXP Pro upgrade. I have absolutely no desire to answer such a question, and am really only posting this reply to make it clear that I have never attempted to answer such a question. To claim that I somehow did answer that question, and that I answered it in a way unsatisfactory to you, is total nonsense.
The "server" version of RedHat costs several thousands of dollars. The $60 version of redhat is the desktop version. Do you have any idea what software costs?
What are you talking about? Windows 2000 Pro is a desktop OS. Windows XP Pro is a desktop OS. Windows Longhorn will be, I assume, a desktop OS.
Windows 2000 Server with a 10-CAL license is far more than $499. Windows 2003 Server with a 10-CAL license is far more than $499. The server version of Longhorn with a 10-CAL license will be, I assume, far more than $499.
Please, explain your comment to me. Do you honestly believe that the $199 upgrade versions Windows 2000 Pro or Windows XP Pro are the "server" version of those operating systems? Do you have any idea what software costs?
Apple has decide there is a sane upgrade price for users of Mac OS X 10.x. The sane upgrade price they've decided on is $129.
I imagine they could charge people with Mac OS 9 even more, but I understand why they're reluctant to do that. Doing that would be difficult, expensive to administer, and just cause resentment.
Windows 2000 Pro came out in 2000. The upgrade price was about $200. Windows XP Pro came out in '02. The upgrade price was about $200. Longhorn is going to be out in another year. The upgrade is going to be about $200.
So... in 4 years, you're spending $600 on the OS.
Pleae tell me how is this different than Apple?
The upgrade price for Panther is $129. The non-upgrade price is... get ready... $0. It comes for free with new machines.
Seriously, what are you asking? Do you have an Apple computer somewhere that doesn't already have an operating system license?
The Dell Latitudes are not marketed to people who want high performance machines. They're marketed to people who want to be certain that if they order multiple machines with identical specifications they'll get multiple identical machines. Not only will you get several identical machines if you order them all at once (which is much more unusual than you may realize), but you'll also be able get identical machines if you order them again three months later (which is a much more common requirement than you may realize). And, you'll still have identical machines after the various warranty replacements you'll inevitably have over the next three years.
If you don't get any value from having multiple identical machines, then you don't want to buy Dell Latitudes. They don't offer anything of value to you. But don't act as if that means they don't offer anything of value to anyone.
I agree too. The world would be a much worse place if we couldn't all see your mother in the shower.
You're correct. Only law enforcement agents should be able to see the cameras. Because, we all know that law enforcement agents never use the information they're given for ill purposes.
Cops are good, mm'kay?
Those tatoos look like shit. If those poor tatoo "artists" didn't have skin to doodle on, they'd have to use black velvet or the cover of their third-hour study hall notebooks.
I'm sure there are good looking tatoos, on good looking people, appreciated by people with artistic sensibility. You've managed to prove that this simply isn't always the case.
Find the person in management who made the business decision to get this project done on the client's deadline. This person is probably not your boss. Schedule a one-on-one meeting with that person. Ask this person, "what are the incentives for getting this project done on the client's deadline?"
If you can't get a one-on-one meeting with the person making the money decisions, it's time to seriously examine your career. You're as valuable to your company as a fry cook is to McDonald's. You'll have to ask some serious soul searching questions, like "am I a shitty employee, or am I working for a shitty company, or both?" and "is this a problem that I'm willing to expend the effort on to fix?"
Also, make sure you're actually negotiating. Know what you're offering to your company, and know what you're willing to accept in return. Know these things before you walk into that meeting. Last, note that it would have been much better to do this before your company agreed to the clients demands. Your company is as stuck as you are right now -- it doesn't sound like they have very much wiggle room left in the contract, and they may not have much wiggle room left in the developer budget, either.
Think very carefully before you negotiate a resolution like "you will give me a bonus next time, or I will not be working for you anymore."
And whatever you do, never, ever, ever bitch about work at work. All good managers appreciate you negotiating for what you want and what you deserve. No manager appreciates and employee who bitches and moans about what he has.
You can buy 1 thing off the dollar menu, and put the remaining $12 into health insurance.
You better get health insurance, 'cuz I don't want to absorb the costs of treating a disgusting fat diabetic while his toes rot off.
Actually, it arguably doesn't matter and arguably helps American business since our products are cheaper abroad making them more attractive ovreseas.
It's too bad that businesses don't vote. All economists agree that the falling dollar hurts the american public. And the american public votes.
No, wait. We don't vote. Nevermind.
This reflects one of the design philosophies of *nix: only give users the privileges they need, and have a huge, well defined wall between them and the system.
You're smoking a huge crack pipe, my friend. In unix, I need suid to change my password, 'fer christ's sake.
I mean, it's painfully obvious that you have no unix experience whatsoever. It's just sad that you got modded up on a site like slashdot, which used to be moderated by geeks.
varargs? Why? It is an object-oriented language - take use of that polymorphism!
How are varargs and polymorphism related? Or, rather, for a language using single dispatch, like Java, how are varargs and polymorphism related, in any way?
I don't know if you've noticed, but Java doesn't do method dispatch based on the runtime type of a methods arguements. In other words, the methods's arguements have nothing to do with the polymorphism of a method. So, how could something like varargs, which is a change that effects the arguements to a method, be replaced by single-dispatch polymorphism?