Why waste 2 hours digging for an obscure Windows bug when all the Windows troubleshooting guidelines say "Have you rebooted the system"? It's far too often the fastest fix to get the system back online, especially for Exchange servers or Windows machines running Oracle.
No, I was missing it on the download page. Thank you for the correction. The conditions under which they publish it are still pretty difficult to work with: getting that "Download Center" account involves giving your home address, phone number, real name, and email account, all all required. But as near as I can tell, you can't easily bundle it into a DEB or RPM with those licensing restrictuions without some serious risk of violating Sun's licensing, for reasons that are pretty obvious if you look at the rest of the Debian history.
Whether one can compile what Sun actually publishes from that source is something I'll have XÀDðest.
Similar things happened with Pine, which has similarly restrictive licensing. But Sun doesn't even publish source, which makes it even harder to deal with legally.
Those inanities are discussed in good depth at http://www.asty.org/articles/20010702pine.html. I highly recommend it as a good discussion of how bad licensing of "free" software can actually prevent it from being "open" and deliberately hinder people who want to work with it. Sun's licensing is similarly restrictive, with that caveat that Sun doesn't develop source code for Java. Washington University at least publishes their source code, even if you're not allowed to publish modifications of it under any circumstances.
You are a target user of the software authors, such as Sun in this case. Debian is, correctly, noticing that the licensing of Sun does not permit them to distribute software they way they normally do it, with intact and compilable source and automatic compilation as needed.
These are not small matters: supporting such packages is a security and software management problem for Debian and other distributions, who often haven't included the Java packages for exactly this reason. Fedora Core 5, for example, seems to have discarded the jdk and jre (or whatever they're called this week) that they included in older OS releases. And Sun has been doing weirdness like slapping their installable RPM inside an executable binary, then renaming the basic packages, then insisiting on jumping through licensing agreement hoops simply to download the software. It's not appropriate in Linux or other open source systems, and is a compelling reason for extensive development of genuinely open source Java software, such as the gcc/Java projects.
Sun is duplicating, badly, what Adobe attempted to do with Postscript in creating a "public" but heavily licensed standard that only they can control the standards for. It's understandable, but is helping actually sideline Sun's particular Java version in favor of more robust ones such as IBM's, and fostering the same sort of open source replacement that ghostscript became for Postscript, and with gaining popularity of the open source version to the point where Sun will want to come up with a new software to license, much as Adobe is dumping Postscript and proceeding to PDF wherever feasible.
You're clearly not familiar with the annual Victoria's Secret catalog show. It causes noticeable bandwidth problems every year: they're manageable, but noticeable on really core routers around the world.
It used to be possible, but with Akamai in place, it's easy to plan ahead and pay them some money to do their caching or streaming services to handle the load better. They've been doing it successfully for years, for things like the Emmy's, so they can certainly handle the traffic, and I understand they can set up their services within one hour for a new exciting customer like this.
I don't expect a big problem, except in workplaces where 30 people turn the game on at the same time and they only have a DSL uplink.
On Linux, all the PDF printers are built on top of ghostscript: that's the core of the PDFCreator tool for printing PDF in Windows, ImageMagick as used in Linux and UNIX and CygWin, and a lot of other open source tools. Ghostscript is what Adobe wishes their commercial versions of Postscript and PDF tools would be, an easily plugged-in utility for document creation and conversion, but ghostscript adheres to the published standards better and is much more portable by avoiding "business-plan" driven feature additions.
Note that there are two versions of ghostscript: the GNU version and the AFPL version. The GNU version is GPL licensed, the AFPL is not and has more interesting licensing, but the AFPL works better for recent Postscript and PDF features. Of course, watching Microsoft break themselves trying to integrate ghostscript licensing into MS Office and throw out Adobe's weirder licensing would be an amazing vindication of open source tools: I'd pay the price of an Adobe license just to sit at ringside and watch the fight.
If they only need to print or save as PDF, use PDFCreator, available at http://sourceforge.net/projects/pdfcreator/. It's free, faster, more reliable, and its generated PDF is both smaller and more likely to print well. Of course, it's baed on Ghostscript instead of Adobe's commercial products, but that's life.
Leave Adobe Acrobat for people who need a $500 license for software to create fancy forms.
I strongly suggest you get over your concerns. This has obvious uses in drug testing and labwork to learn how nerve tissue works, and for learning how to do good nerve electrodes for spinal tap electrodes, prosthetic limb control, or sensory replacemtn techniques such as artificial vision and hearing.
Creating a mind by chopping up a brain this way is so far away in the future, and so unlikely to even be feasible, that it's not worth wasting your time. The brain works by having both local chemistry and 3-D connections among the neurons, which change and shift to adapt to new memory, new reflexes, and chemical stimuli. Replacing that with brain slices is like replacing your computer by chopping up a chain saw and expecting it to work.
And Marathon: I used to borrow time on company Mac's to play that. Unfortunately, the creators of it switched to the XBox and created Halo (which has numerous Marathon references subtlely embedded).
Don't forget the corporate loopholes for purchasing licenses to spam, such as occurred with Microsoft's SenderID keys which are almost always a sign of being a spammer, and the trivial break-ins to "authorized" mail sending servers to do your spam through a permitted server.
Oddly, some tools such as SPF provide some good techniques for reducing the forged email a great deal, which has been a traditional problem with spam and phishing. But it's been submarined by Microsoft trying to "embrace and extend" it with the fundamentally flawed SenderID keys.
( ) Sorry dude, but I don't think it would work.
( ) This is a stupid idea, and you're a stupid company for suggesting it.
( ) Nice try, assh0le! I'm going to find out where you live and burn your house down!
(X) This is the real world, not a high school freshman "Intro to Computing" class.
Then included a few of the similar examples of it and why it has failed, such as Microsoft's SenderID keys.
You guess wrong. While such systems are rarely directly net accessible, they're often accessible to machines in a company's internal network or DMZ, precisely for remote administration or debugging. And once the virus or attack is inside, they can use the "feel-good" use of an external firewall to enter every improperly secured system.
Ask any administrator who's had to deal with spamware showing up inside their company's network from some VP's traveling laptop, which they refuse to update lest it break or until the company buys them a new one.
Laptop users and Microsoft Outlook users need to use some local directory for their material, and a lot of Windows software has the use of "My Documents" to store information hardcoded, and it's almost impossible to fix it in all the software that does it by default. If you set your profile to use the network share as your home directory, you'll just encrypt that.
I'm sorry, but Hitchhiker's Guide was a frustrating nightmare of hours of wasted time to get even a single joke to work, when it wasn't crashing randomly or wedgning pointlessly. For big fans of Douglas Adams it might have been fun, but I found it completely unplayable due to that pitiful parser of commands and the frequent bugs that left it wedged.
Neverland, however, was funny in plots, puzzles, and animation, with animation that worked on a very modest 200 MHz machine with a 15" screen and 640x480 animation. And it still works well, for children 4 and above to play and 10 and above to actually play through, and for anyone in the room watching it to laugh at.
This does not differ from most corporate setups: actually securing SMB is difficult enough, and actually securing NFS is historically a very bad joke. It's called "No F***ing Security" for a reason.
A cheap box like this can actually help, by providing good guides on how to properly authenticate the server, and not forcing users to home-brew their SSH or Kerberos or NFSv4 or Samba or Netatalk setups from scratch, on OS's that do some of them well but not others.
I agree with some of what you say. Steel cases instead of cheap fragile plastic are definitely worth it. But some of the strangest things wind up being really important, such as whether the case forces you to cut your hand trying to get the power supply mounted, or inevitably drop screws in the case inextricably into weird slots while mounting things, or break badly designed feet. And it's also important to look at whether fans can be cleaned, because let's face it: many gamers are slobs and have their rooms filled with dust. Worse, many games are inclined to rest a Big Gulp on their desk where they will leave it for 3 days, half-filled, until the cheap paper cup dissolves and spills onto things, such as your suggested blowhole fan.
Cases with side fans are inevitably put in "computer desks" that have closely fitted walls that block the side fans. Front to back is the way to go for cooling, baby, with big blower fans in unused drive bays. Noise is fairly irrelevant if you're wearing good headsets with earpieces that actually cover your ears, instead of these foolish wienie headsets that try to stuff things inside your ear canal. Blocking out even hideously large case fans, and cubicle naighbors, and screaming kids nearby, with good "over-the-entire-ear" $100 headsets gives you a vastly better sound experience than even $1000 stereo systemas and sound cards.
A case that is well designed against spills on top of it is priceless. One that is well designed to allow you to reach components, and not grab a network cable inextricably into the case is also worth quite a bit. I've had the devil of a time extracting CAT5 cables from some otherwise cheap cases, and seen too many people actually set things down on top of or pour them into fans or ventholes on top of cases. And Cases with funky feet that tilt, or prevent you from setting things down safely on top of the case are merely foolish, because people will inevitabl pile them up, drop them, and wonder why their CD that's been sitting on the dirty, dusty, scratchy floor for a week with the dong walking on them don't work right.
And those stupid cases with "Exciting! Graphical! Clocks!" that have incredibly cheap built-in chips that drift a minute a day are worse than useless, they're actually worse than the "blinking 12:000" clocks that can't be set by anyone over the age of 6.
Amnesty International, like the ACLU, has been consistent about protecting speech and other freedoms, even where they detest the message or the people they're protecting. Accusing them of selective political enforcement is unfair and unfounded. They've been big believers in free speech being allowed to expose truth and debate ideas, rather than censoring idiots who write revisionist histories.
How about we make Oracle a trade? If they throw out the end-user licenses and become liable for the flaws and damage caused by the flaws in their software, we'll protect them from security flaws published without telling Oracle at least 3 months before publication?
Yes, and your gun is safe left loaded and unlocked because you don't have any kids in your house. Until you have a houseguest, or kids visiting with their folks, or plumbers or cleaners in to work on your systems, or you bring it in for repair because the hard drive is crashing. And you forget you kept the gun lying around: now you don't have a safe place to put it, in a rush.
Except that bridges get potholes, and cars hit and scrape the anti-rust paint, and roads get renamed so they need new signs, or other roads get changed so the traffic doubles, etc. Bridges are hardly static: Oracle has been inexcusably static with their software, precisely because it is a lumbering behemoth and nearly impossible except for a highly trained person to manage with a huge investment in time.
They can't tell which pothole, when fixed, will cause all cars to go faster and skid off the place they didn't bother to put a safety rail, or which security hole fixed will trigger failure of some unrelated part of the operating system that happens to use that as a feature. It's too big, too monolithic, and too static to be used for anything but those very few products that demand such absolutely high levels of number crunching and justify the money. Some customers crave that level of "stability".
But handling new demands means change, and the money for Oracle is usually better invested in programmer time breaking down your problems and running them on cheaper hardware with cheaper tools: I've seen this in several cases now, for people switching to both MySQL and PostgreSQL.
Oh, it's worse. It's like Simson Garfinkel's "UNIX Security" book: he waves his hands at some complex vulnerability without making it clear, and ignores the real risks. For SSH, the real risk is idiots who refuse to use ssh-agent and keep copies of their private key, unencrypted, on insecure machines. Worse, people keep unencrypted private keys on poorly managed NFS shares so that any shmuck who plugs into a local network with a laptop can look in "/home" and get all the SSH keys.
Even more fun, many sites running NIS or gateway servers still fail to enable "shadow passwords", which means that any random user can look at the encrypted password list and run Alec Moffett's old "crack" program against them, and get roughly one-in-ten of all passwords decrypted. (Yes, MD5 passwords help, but a lot of people still don't use that.)
Slashdot Mirror
Why waste 2 hours digging for an obscure Windows bug when all the Windows troubleshooting guidelines say "Have you rebooted the system"? It's far too often the fastest fix to get the system back online, especially for Exchange servers or Windows machines running Oracle.
No, I was missing it on the download page. Thank you for the correction. The conditions under which they publish it are still pretty difficult to work with: getting that "Download Center" account involves giving your home address, phone number, real name, and email account, all all required. But as near as I can tell, you can't easily bundle it into a DEB or RPM with those licensing restrictuions without some serious risk of violating Sun's licensing, for reasons that are pretty obvious if you look at the rest of the Debian history.
Whether one can compile what Sun actually publishes from that source is something I'll have XÀDðest.
Similar things happened with Pine, which has similarly restrictive licensing. But Sun doesn't even publish source, which makes it even harder to deal with legally.
Those inanities are discussed in good depth at http://www.asty.org/articles/20010702pine.html. I highly recommend it as a good discussion of how bad licensing of "free" software can actually prevent it from being "open" and deliberately hinder people who want to work with it. Sun's licensing is similarly restrictive, with that caveat that Sun doesn't develop source code for Java. Washington University at least publishes their source code, even if you're not allowed to publish modifications of it under any circumstances.
You are a target user of the software authors, such as Sun in this case. Debian is, correctly, noticing that the licensing of Sun does not permit them to distribute software they way they normally do it, with intact and compilable source and automatic compilation as needed.
These are not small matters: supporting such packages is a security and software management problem for Debian and other distributions, who often haven't included the Java packages for exactly this reason. Fedora Core 5, for example, seems to have discarded the jdk and jre (or whatever they're called this week) that they included in older OS releases. And Sun has been doing weirdness like slapping their installable RPM inside an executable binary, then renaming the basic packages, then insisiting on jumping through licensing agreement hoops simply to download the software. It's not appropriate in Linux or other open source systems, and is a compelling reason for extensive development of genuinely open source Java software, such as the gcc/Java projects.
Sun is duplicating, badly, what Adobe attempted to do with Postscript in creating a "public" but heavily licensed standard that only they can control the standards for. It's understandable, but is helping actually sideline Sun's particular Java version in favor of more robust ones such as IBM's, and fostering the same sort of open source replacement that ghostscript became for Postscript, and with gaining popularity of the open source version to the point where Sun will want to come up with a new software to license, much as Adobe is dumping Postscript and proceeding to PDF wherever feasible.
Akamai has a fairly successful setup for precisely this: It's described at http://www.akamai.com/en/html/services/streaming.h tml.
You're clearly not familiar with the annual Victoria's Secret catalog show. It causes noticeable bandwidth problems every year: they're manageable, but noticeable on really core routers around the world.
It used to be possible, but with Akamai in place, it's easy to plan ahead and pay them some money to do their caching or streaming services to handle the load better. They've been doing it successfully for years, for things like the Emmy's, so they can certainly handle the traffic, and I understand they can set up their services within one hour for a new exciting customer like this.
I don't expect a big problem, except in workplaces where 30 people turn the game on at the same time and they only have a DSL uplink.
On Linux, all the PDF printers are built on top of ghostscript: that's the core of the PDFCreator tool for printing PDF in Windows, ImageMagick as used in Linux and UNIX and CygWin, and a lot of other open source tools. Ghostscript is what Adobe wishes their commercial versions of Postscript and PDF tools would be, an easily plugged-in utility for document creation and conversion, but ghostscript adheres to the published standards better and is much more portable by avoiding "business-plan" driven feature additions.
Note that there are two versions of ghostscript: the GNU version and the AFPL version. The GNU version is GPL licensed, the AFPL is not and has more interesting licensing, but the AFPL works better for recent Postscript and PDF features. Of course, watching Microsoft break themselves trying to integrate ghostscript licensing into MS Office and throw out Adobe's weirder licensing would be an amazing vindication of open source tools: I'd pay the price of an Adobe license just to sit at ringside and watch the fight.
If they only need to print or save as PDF, use PDFCreator, available at http://sourceforge.net/projects/pdfcreator/. It's free, faster, more reliable, and its generated PDF is both smaller and more likely to print well. Of course, it's baed on Ghostscript instead of Adobe's commercial products, but that's life.
Leave Adobe Acrobat for people who need a $500 license for software to create fancy forms.
I strongly suggest you get over your concerns. This has obvious uses in drug testing and labwork to learn how nerve tissue works, and for learning how to do good nerve electrodes for spinal tap electrodes, prosthetic limb control, or sensory replacemtn techniques such as artificial vision and hearing.
Creating a mind by chopping up a brain this way is so far away in the future, and so unlikely to even be feasible, that it's not worth wasting your time. The brain works by having both local chemistry and 3-D connections among the neurons, which change and shift to adapt to new memory, new reflexes, and chemical stimuli. Replacing that with brain slices is like replacing your computer by chopping up a chain saw and expecting it to work.
And Marathon: I used to borrow time on company Mac's to play that. Unfortunately, the creators of it switched to the XBox and created Halo (which has numerous Marathon references subtlely embedded).
Don't forget the corporate loopholes for purchasing licenses to spam, such as occurred with Microsoft's SenderID keys which are almost always a sign of being a spammer, and the trivial break-ins to "authorized" mail sending servers to do your spam through a permitted server.
Oddly, some tools such as SPF provide some good techniques for reducing the forged email a great deal, which has been a traditional problem with spam and phishing. But it's been submarined by Microsoft trying to "embrace and extend" it with the fundamentally flawed SenderID keys.
You're too gentle, sir. I'd have selected:
Furthermore, this is what I think about you:
( ) Sorry dude, but I don't think it would work.
( ) This is a stupid idea, and you're a stupid company for suggesting it.
( ) Nice try, assh0le! I'm going to find out where you live and burn your house down!
(X) This is the real world, not a high school freshman "Intro to Computing" class.
Then included a few of the similar examples of it and why it has failed, such as Microsoft's SenderID keys.
You guess wrong. While such systems are rarely directly net accessible, they're often accessible to machines in a company's internal network or DMZ, precisely for remote administration or debugging. And once the virus or attack is inside, they can use the "feel-good" use of an external firewall to enter every improperly secured system.
Ask any administrator who's had to deal with spamware showing up inside their company's network from some VP's traveling laptop, which they refuse to update lest it break or until the company buys them a new one.
Laptop users and Microsoft Outlook users need to use some local directory for their material, and a lot of Windows software has the use of "My Documents" to store information hardcoded, and it's almost impossible to fix it in all the software that does it by default. If you set your profile to use the network share as your home directory, you'll just encrypt that.
I'm sorry, but Hitchhiker's Guide was a frustrating nightmare of hours of wasted time to get even a single joke to work, when it wasn't crashing randomly or wedgning pointlessly. For big fans of Douglas Adams it might have been fun, but I found it completely unplayable due to that pitiful parser of commands and the frequent bugs that left it wedged.
Neverland, however, was funny in plots, puzzles, and animation, with animation that worked on a very modest 200 MHz machine with a 15" screen and 640x480 animation. And it still works well, for children 4 and above to play and 10 and above to actually play through, and for anyone in the room watching it to laugh at.
This does not differ from most corporate setups: actually securing SMB is difficult enough, and actually securing NFS is historically a very bad joke. It's called "No F***ing Security" for a reason.
A cheap box like this can actually help, by providing good guides on how to properly authenticate the server, and not forcing users to home-brew their SSH or Kerberos or NFSv4 or Samba or Netatalk setups from scratch, on OS's that do some of them well but not others.
Your points are well-made: I've sometimes heard "squirrel cage" fans referred to as "blowers", and shouldn't have used that name for them.
Too bad they can't write English. What H1B gifted idiot wrote their web page?
I agree with some of what you say. Steel cases instead of cheap fragile plastic are definitely worth it. But some of the strangest things wind up being really important, such as whether the case forces you to cut your hand trying to get the power supply mounted, or inevitably drop screws in the case inextricably into weird slots while mounting things, or break badly designed feet. And it's also important to look at whether fans can be cleaned, because let's face it: many gamers are slobs and have their rooms filled with dust. Worse, many games are inclined to rest a Big Gulp on their desk where they will leave it for 3 days, half-filled, until the cheap paper cup dissolves and spills onto things, such as your suggested blowhole fan.
Cases with side fans are inevitably put in "computer desks" that have closely fitted walls that block the side fans. Front to back is the way to go for cooling, baby, with big blower fans in unused drive bays. Noise is fairly irrelevant if you're wearing good headsets with earpieces that actually cover your ears, instead of these foolish wienie headsets that try to stuff things inside your ear canal. Blocking out even hideously large case fans, and cubicle naighbors, and screaming kids nearby, with good "over-the-entire-ear" $100 headsets gives you a vastly better sound experience than even $1000 stereo systemas and sound cards.
A case that is well designed against spills on top of it is priceless. One that is well designed to allow you to reach components, and not grab a network cable inextricably into the case is also worth quite a bit. I've had the devil of a time extracting CAT5 cables from some otherwise cheap cases, and seen too many people actually set things down on top of or pour them into fans or ventholes on top of cases. And Cases with funky feet that tilt, or prevent you from setting things down safely on top of the case are merely foolish, because people will inevitabl pile them up, drop them, and wonder why their CD that's been sitting on the dirty, dusty, scratchy floor for a week with the dong walking on them don't work right.
And those stupid cases with "Exciting! Graphical! Clocks!" that have incredibly cheap built-in chips that drift a minute a day are worse than useless, they're actually worse than the "blinking 12:000" clocks that can't be set by anyone over the age of 6.
Amnesty International, like the ACLU, has been consistent about protecting speech and other freedoms, even where they detest the message or the people they're protecting. Accusing them of selective political enforcement is unfair and unfounded. They've been big believers in free speech being allowed to expose truth and debate ideas, rather than censoring idiots who write revisionist histories.
How about we make Oracle a trade? If they throw out the end-user licenses and become liable for the flaws and damage caused by the flaws in their software, we'll protect them from security flaws published without telling Oracle at least 3 months before publication?
Yes, and your gun is safe left loaded and unlocked because you don't have any kids in your house. Until you have a houseguest, or kids visiting with their folks, or plumbers or cleaners in to work on your systems, or you bring it in for repair because the hard drive is crashing. And you forget you kept the gun lying around: now you don't have a safe place to put it, in a rush.
Except that bridges get potholes, and cars hit and scrape the anti-rust paint, and roads get renamed so they need new signs, or other roads get changed so the traffic doubles, etc. Bridges are hardly static: Oracle has been inexcusably static with their software, precisely because it is a lumbering behemoth and nearly impossible except for a highly trained person to manage with a huge investment in time.
They can't tell which pothole, when fixed, will cause all cars to go faster and skid off the place they didn't bother to put a safety rail, or which security hole fixed will trigger failure of some unrelated part of the operating system that happens to use that as a feature. It's too big, too monolithic, and too static to be used for anything but those very few products that demand such absolutely high levels of number crunching and justify the money. Some customers crave that level of "stability".
But handling new demands means change, and the money for Oracle is usually better invested in programmer time breaking down your problems and running them on cheaper hardware with cheaper tools: I've seen this in several cases now, for people switching to both MySQL and PostgreSQL.
Oh, it's worse. It's like Simson Garfinkel's "UNIX Security" book: he waves his hands at some complex vulnerability without making it clear, and ignores the real risks. For SSH, the real risk is idiots who refuse to use ssh-agent and keep copies of their private key, unencrypted, on insecure machines. Worse, people keep unencrypted private keys on poorly managed NFS shares so that any shmuck who plugs into a local network with a laptop can look in "/home" and get all the SSH keys.
Even more fun, many sites running NIS or gateway servers still fail to enable "shadow passwords", which means that any random user can look at the encrypted password list and run Alec Moffett's old "crack" program against them, and get roughly one-in-ten of all passwords decrypted. (Yes, MD5 passwords help, but a lot of people still don't use that.)