a) Just because your PC has a 500W PSU, does not mean that it draws 500W. Odds are that your PC is only drawing 80-120W depending on what you are doing (SSDs are 1-3W, traditional drives are 6-12W when active, CPUs are 45W up to 125W maximum with the ability to clock down to only a handful of watts, etc.). Stick a meter on the system and see.
b) Sleep mode is admittedly hit/miss, tends to work fine on the better laptops, but is frequently broken due to poor device drivers or cheap devices (i.e. you get what you pay for, most of the time).
c) You're spot-on when it comes to PC performance. In the 90s and right up until 2003-2005, power doubled every 1.5-2 years. So a 3-4 year old PC might be 4x-6x slower then a brand new PC. Then they hit the wall in the early to mid 2000s and performance only increased 5-10% per year. A system purchased in 2013 might only have single-core performance that is 30-40% higher then one from 3-4 years ago. It's sole advantage is then that it has more cores, which may or may not be a benefit to the software that you are running.
Personal experience is that once multi-core CPUs dropped below $200, we moved into the period where PCs only need to be replaced every 5-8 years instead of every 3-5. A multi-core unit from 2007, given a SSD and enough RAM and maybe a faster video card, is still a viable system for general desktop work. The newer quad/hex/octo core units are just going to make that cycle even longer (7-12 years) because the multiple cores means that the machine stays responsive for much longer.
Users may not notice brute speed, but they definitely notice responsiveness. Adding SSDs to the system improves how fast the hard drive can serve up random file XYZ, and adding cores means that the operating system UI almost always has a place to park itself and meet the user's needs quickly.
Our current desktops were all installed in the 2006-2009 era. They're currently being upgraded with SSDs, more RAM, and Win7. We don't plan on replacing the systems until 2014-2017, which is a 7-10 year lifespan. I worry more about capacitor failure in these systems now instead of worrying that they are too slow to meet the needs. Fortunately, it's all whitebox, and replacement innards are commodity items that can just be dropped in with minimal fuss.
They should have let the DOJ split them up into multiple companies way back when. It would have been a net benefit over the long-term.
But they're too focused on keeping the customers locked in. So I really wonder if they'll be around in 10 years. The landscape has drastically changed in the last 5 years - interoperability with open standards is now a lot more important than it used to be. The customers now have many more options, most of which are compatible with each other.
IBM's SPSS runs on Linux, and that's a big bucks package. And that's just one small example.
There's only one thing that ties us to Windows at the office, and that's MS-Access. Everything else, we could switch to Linux desktops (or Macs for the creatives).
Expect your laptop battery to be effectively useless within 3 years of purchase, particularly if it's kept plugged in at all times.
IBM/Lenovo Thinkpads offer a way in their battery gauge / power management software to prevent recharging until the battery falls below X%. I usually set my users at 85% or 90% as the threshold before charging begins. At which point, it charges from X% up to 100%.
Usually only happens about once every 1-4 weeks, depending on how much they leave it hooked up to the wall outlet vs carrying it around.
Other brands might offer similar features, but we only use Thinkpads.
I'm glad to see a high bar set for the certification of LED bulbs. CFL lights rarely hit their expected life span, among other problems
My CFLs last anywhere from 3-10 years. Top killers are heat and dirty switches (with a lot of "bounce"). Are LED bulbs better? Probably, but there are bad actors and substandard designs in both types.
We were originally told by our vendor that Win7 would stop being sold last October. So we stocked up on licenses for future builds (we whitebox the PCs).
That being said... you can still buy Win7 licenses from the various vendors. And there's always the Win8 Pro - downgrade option.
(I suspect that Microsoft is muddying the waters for the past year. Doing their usual FUD in order to drive people towards Win8 by saying that Win7 is not available.)
All of those vendors sell laptops with Win7 Professional. My guess is that your IT department is trying to use consumer-quality laptops instead of buying the business-class laptops.
Lenovo T530 offer anything from Win7 Home Premium, to Win7 Professional to Win8 or Win8 Pro. Dell Latitude E5530 come with Win7 Professional. Granted the Dell is only an i3, but the Lenovo comes with an i7 as an option.
End user computing experience moved away from java, and onto HTML5/Javascript
Which has a lot more to do with centralizing the code, instead of having to do upgrades of software on thousands of individual machines. Now you can point just about any old browser (current within the last 12-18 months) at your web based application without worrying about what version of module XYZ is installed on machine ABC.
Give it a few years, the centralized applications will end up ossified and users will start demanding things that the central IT department can't give them. Which will shift the balance of power away from the back offices / mainframes again, back to the desktops / department level.
Well, just like XP brought standardized WiFi settings (I despised the hell that was Win2k WiFi, where every vendor had their own proprietary UI), Win7 brings things like better search (especially for programs in the start menu), SSD TRIM support, better security features, and 64bit. The window preview (the thumbnails as you mouseover items in the program list) is also very nice.
Took me a week or two to get used to Win7. There's still a few stupid decisions, but overall it stays out of my way and lets me get work done.
We're upgrading all our XP desktops to Win7 this year and hoping that we won't have to upgrade the O/S again until 2016-2019. That is, assuming that the existing hardware (dual-core CPUs, with 4-8GB RAM and SSDs) isn't overly slow by then. Maybe by that point, MS will have released another "good" operating system - or they'll have cratered and release MS Office for Linux.
From what I recall, IBM's problem with the PS/2 brand was:
1) They tried to shift everyone to MCA instead of the more open ISA/EISA, mostly because they were trying to stuff the genie back in the bottle and retake control of the industry.
2) The lower end of the PS/2 line was garbage, which tarnished the upper-end.
We had a few PS/2 server towers to play with. They were rather over-engineered and expensive, and the Intel / Compaq / AT&T commodity systems were faster and less expensive.
Outlook is garbage unless you're talking to an Exchange server. It barely supports IMAP.
Not to say that TBird can't be improved (especially task/calendar stuff), but for large mailboxes over IMAP it's far better then Outlook. Plus better support for multiple email address scenarios where you need to either have multiple aliases for a mailbox or have multiple mailboxes.
Works fine as long as the controller / RAID software syncs the remaining good disk to the newly installed disk and not the other way around.
If you plan on doing something like that regularly, you should consider doing a 3-way mirror (3 disks, RAID-1) so that you're left with two good disks after pulling one. Easily done in Linux software RAID or the better (real hardware) RAID controllers.
(My feeling is that if you are going to dedicate a spare disk to a RAID-1 array, you may as well make use of it by making it the 3rd mirror. Depending on the controller, you lose no write speed but you gain more security and possibly slightly better read performance.)
Run IMAP with a local cached copy of everything. Thunderbird searches will search locally unless you tell it to search on the IMAP server.
And emails older then a year should be split out into annual archive folders. Which is a fast and dirty way of organizing. If you remember that you talked about something in 2011, it's probably easier to find with good subject lines / searching then trying to keep track of a complex hierarchy of folders.
Worst case, you have to search both the 2010 and 2012 folders to find something.
Thunderbird also has a "fast filter" system where you can quickly filter a folder based on keywords in either the subject, sender, receiver or body (or all of the above).
XP ran like garbage in 256MB of RAM, unless you never installed anything other then the basic operating system and never opened more then one program at a time. By the time that SP3 rolled around you really needed a minimum of 512MB.
And multi-taskers needed at least 1GB, with 2GB being a common minimum spec since 2006-ish.
As others have said, no database ports should ever be exposed to the world at large. You should have a firewall in place that only allows traffic to/from an extremely limited IP address range. Which mitigates a whole lot of issues, even if the database software is vulnerable.
Sure, I'll need to update my pgsql instances, but because they're firewalled off from the outside world, I don't have to lose sleep over it until the fix comes out.
Ours was 192.0.x.y. Took me about 5 years to finally get us swapped over to the 172.16.x.y - 172.31.x.y range. Seems like a lot of companies didn't grasp that only 192.168.x.y was valid for private use. The main reason we finally switched was that the old 254 address space was too small for our growing needs so we upgraded to a 2046 size address space.
Arch regularly makes changes that will leave your system thoroughly hosed if you update without watching the news feed.
That sounds a lot like Gentoo's methods. Little or no quality assurance in the main feeds which can leave you hosed if you don't treat updates like unexplored minefields. Until they get serious about that, they're a hobbyist distro.
We started with Gentoo way back when. It *sounded* like a good idea at the time because money was tight, hardware was under-spec, and being able to tweak the kernel and other things to run as fast as possible seemed attractive. Unfortunately, the reality is that hardware is relatively cheap, and my time is increasingly rare and expensive. Plus, having all the drivers in the kernel is a good thing because it lets you move entire systems from one set of hardware to another in a pinch.
We moved to CentOS / SciLinux / RHEL because it:
- Almost always works
- Less variation in setups
- Has a public company backing it
- Far easier to get support for RHEL vs Gentoo
- Has the mind share and market share
Mint w/ Mate is probably the best of the breed out of that last group. Debian and a nice desktop, without all the nonsense of Ubuntu's "chase the next rainbow" flavor. I used to recommend Ubuntu, but they jumped off the deep end a few years back so now I recommend Mint.
we're going to have a whole lot of trouble anyway since we're going to have to readjust to a world where energy isn't cheap*, at least in the medium term.
* For various definitions of cheap
I said it 10 years ago, I said it 5 years ago, and I'll say it again. If the price of oil rises *gradually* (over a period of years), you will not see doom-and-gloom happen. There will be market shifts, there will be many things that become more expensive or no longer cost-effective.
But if the rise in prices is spread across multiple years or decades, then people have time to adjust habits, purchase more fuel-efficient vehicles / houses / appliances, develop new energy technologies, build new alternative energy facilities, go after the more expensive oil fields, etc.
Frankly, I worry more about access to fresh water in the next 20 years then I do about the price of oil.
Tape has a vital place in the IT administration world.
Tape is expensive, fragile and requires special hardware. Removable or external magnetic hard drives, OTOH, are cheap, sturdy and will work on any system that you can scrounge up.
Given the costs of tape drives and tape media, it's not surprising that a lot of small / medium businesses just use hard drives for backups. External 2.5" 1TB drives are dirt cheap and you could do weekly off-site backups using them with 13 generations for less then $2000. You can't even buy a large capacity tape drive for $2000. Much less the tapes needed to run a proper backup cycle.
Unless there are legal reasons to keep 5-10 years of backups, or you are dealing in more then 3-5 TB of storage to be backed up, or taking things off-site daily via courier tape is just too expensive.
If it's a password that they use every single day, any user with a brain larger then a goldfish can remember it after a week or two. Those who can't - should probably be working the checkout line at the local grocery store and not handling sensitive data.
However, this means you should not be requiring them to change the password without good cause. Weekly/Monthly/Quarterly resets are not a good enough reason to force a password reset.
We give our users the instructions to put the password on a folded slip of paper in their wallet/purse with no other identifying information on it. People are generally pretty good about keeping track of their money and not letting unauthorized people access to it. So you may as well take advantage of that during the period that it takes the user to remember the password.
The bigger problem is passwords which are used for sensitive systems, where the user accesses them on an infrequent basis. Storage of those passwords is a big problem because users have a hard time remembering them.
Slashdot Mirror
Yes, you need to look at the XML functions. There are options ranging from outputting a single field as XML up to an entire table/query as XML.
Instead of salting and hashing, use a key derivation function (e.g., bcrypt, scrypt).
Which is just "hashing" by another name. You're exchanging one "one-way" function for a different "one-way" function that just takes longer.
You still need to be randomly salting each user's password so that the attackers cannot use a single pre-computed rainbow table against you.
a) Just because your PC has a 500W PSU, does not mean that it draws 500W. Odds are that your PC is only drawing 80-120W depending on what you are doing (SSDs are 1-3W, traditional drives are 6-12W when active, CPUs are 45W up to 125W maximum with the ability to clock down to only a handful of watts, etc.). Stick a meter on the system and see.
b) Sleep mode is admittedly hit/miss, tends to work fine on the better laptops, but is frequently broken due to poor device drivers or cheap devices (i.e. you get what you pay for, most of the time).
c) You're spot-on when it comes to PC performance. In the 90s and right up until 2003-2005, power doubled every 1.5-2 years. So a 3-4 year old PC might be 4x-6x slower then a brand new PC. Then they hit the wall in the early to mid 2000s and performance only increased 5-10% per year. A system purchased in 2013 might only have single-core performance that is 30-40% higher then one from 3-4 years ago. It's sole advantage is then that it has more cores, which may or may not be a benefit to the software that you are running.
Personal experience is that once multi-core CPUs dropped below $200, we moved into the period where PCs only need to be replaced every 5-8 years instead of every 3-5. A multi-core unit from 2007, given a SSD and enough RAM and maybe a faster video card, is still a viable system for general desktop work. The newer quad/hex/octo core units are just going to make that cycle even longer (7-12 years) because the multiple cores means that the machine stays responsive for much longer.
Users may not notice brute speed, but they definitely notice responsiveness. Adding SSDs to the system improves how fast the hard drive can serve up random file XYZ, and adding cores means that the operating system UI almost always has a place to park itself and meet the user's needs quickly.
Our current desktops were all installed in the 2006-2009 era. They're currently being upgraded with SSDs, more RAM, and Win7. We don't plan on replacing the systems until 2014-2017, which is a 7-10 year lifespan. I worry more about capacitor failure in these systems now instead of worrying that they are too slow to meet the needs. Fortunately, it's all whitebox, and replacement innards are commodity items that can just be dropped in with minimal fuss.
They should have let the DOJ split them up into multiple companies way back when. It would have been a net benefit over the long-term.
But they're too focused on keeping the customers locked in. So I really wonder if they'll be around in 10 years. The landscape has drastically changed in the last 5 years - interoperability with open standards is now a lot more important than it used to be. The customers now have many more options, most of which are compatible with each other.
IBM's SPSS runs on Linux, and that's a big bucks package. And that's just one small example.
There's only one thing that ties us to Windows at the office, and that's MS-Access. Everything else, we could switch to Linux desktops (or Macs for the creatives).
Expect your laptop battery to be effectively useless within 3 years of purchase, particularly if it's kept plugged in at all times.
IBM/Lenovo Thinkpads offer a way in their battery gauge / power management software to prevent recharging until the battery falls below X%. I usually set my users at 85% or 90% as the threshold before charging begins. At which point, it charges from X% up to 100%.
Usually only happens about once every 1-4 weeks, depending on how much they leave it hooked up to the wall outlet vs carrying it around.
Other brands might offer similar features, but we only use Thinkpads.
I'm glad to see a high bar set for the certification of LED bulbs. CFL lights rarely hit their expected life span, among other problems
My CFLs last anywhere from 3-10 years. Top killers are heat and dirty switches (with a lot of "bounce"). Are LED bulbs better? Probably, but there are bad actors and substandard designs in both types.
W76 is about 164kg (100kt warhead), W87 is 200-270kg (300-475kt), W88 is 360kg (up to 475kt).
The WWII era bombs: Little Boy was 4400 kg and 28" across, Fat Man was 4,633 kg and 5' across.
Minimum size seems to be about 150kg. They're not quite there yet.
We were originally told by our vendor that Win7 would stop being sold last October. So we stocked up on licenses for future builds (we whitebox the PCs).
That being said... you can still buy Win7 licenses from the various vendors. And there's always the Win8 Pro - downgrade option.
(I suspect that Microsoft is muddying the waters for the past year. Doing their usual FUD in order to drive people towards Win8 by saying that Win7 is not available.)
You're being sold a bill of goods.
All of those vendors sell laptops with Win7 Professional. My guess is that your IT department is trying to use consumer-quality laptops instead of buying the business-class laptops.
Lenovo T530 offer anything from Win7 Home Premium, to Win7 Professional to Win8 or Win8 Pro. Dell Latitude E5530 come with Win7 Professional. Granted the Dell is only an i3, but the Lenovo comes with an i7 as an option.
End user computing experience moved away from java, and onto HTML5/Javascript
Which has a lot more to do with centralizing the code, instead of having to do upgrades of software on thousands of individual machines. Now you can point just about any old browser (current within the last 12-18 months) at your web based application without worrying about what version of module XYZ is installed on machine ABC.
Give it a few years, the centralized applications will end up ossified and users will start demanding things that the central IT department can't give them. Which will shift the balance of power away from the back offices / mainframes again, back to the desktops / department level.
Well, just like XP brought standardized WiFi settings (I despised the hell that was Win2k WiFi, where every vendor had their own proprietary UI), Win7 brings things like better search (especially for programs in the start menu), SSD TRIM support, better security features, and 64bit. The window preview (the thumbnails as you mouseover items in the program list) is also very nice.
Took me a week or two to get used to Win7. There's still a few stupid decisions, but overall it stays out of my way and lets me get work done.
We're upgrading all our XP desktops to Win7 this year and hoping that we won't have to upgrade the O/S again until 2016-2019. That is, assuming that the existing hardware (dual-core CPUs, with 4-8GB RAM and SSDs) isn't overly slow by then. Maybe by that point, MS will have released another "good" operating system - or they'll have cratered and release MS Office for Linux.
From what I recall, IBM's problem with the PS/2 brand was:
1) They tried to shift everyone to MCA instead of the more open ISA/EISA, mostly because they were trying to stuff the genie back in the bottle and retake control of the industry.
2) The lower end of the PS/2 line was garbage, which tarnished the upper-end.
We had a few PS/2 server towers to play with. They were rather over-engineered and expensive, and the Intel / Compaq / AT&T commodity systems were faster and less expensive.
Outlook is garbage unless you're talking to an Exchange server. It barely supports IMAP.
Not to say that TBird can't be improved (especially task/calendar stuff), but for large mailboxes over IMAP it's far better then Outlook. Plus better support for multiple email address scenarios where you need to either have multiple aliases for a mailbox or have multiple mailboxes.
Works fine as long as the controller / RAID software syncs the remaining good disk to the newly installed disk and not the other way around.
If you plan on doing something like that regularly, you should consider doing a 3-way mirror (3 disks, RAID-1) so that you're left with two good disks after pulling one. Easily done in Linux software RAID or the better (real hardware) RAID controllers.
(My feeling is that if you are going to dedicate a spare disk to a RAID-1 array, you may as well make use of it by making it the 3rd mirror. Depending on the controller, you lose no write speed but you gain more security and possibly slightly better read performance.)
Run IMAP with a local cached copy of everything. Thunderbird searches will search locally unless you tell it to search on the IMAP server.
And emails older then a year should be split out into annual archive folders. Which is a fast and dirty way of organizing. If you remember that you talked about something in 2011, it's probably easier to find with good subject lines / searching then trying to keep track of a complex hierarchy of folders.
Worst case, you have to search both the 2010 and 2012 folders to find something.
Thunderbird also has a "fast filter" system where you can quickly filter a folder based on keywords in either the subject, sender, receiver or body (or all of the above).
XP ran like garbage in 256MB of RAM, unless you never installed anything other then the basic operating system and never opened more then one program at a time. By the time that SP3 rolled around you really needed a minimum of 512MB.
And multi-taskers needed at least 1GB, with 2GB being a common minimum spec since 2006-ish.
As others have said, no database ports should ever be exposed to the world at large. You should have a firewall in place that only allows traffic to/from an extremely limited IP address range. Which mitigates a whole lot of issues, even if the database software is vulnerable.
Sure, I'll need to update my pgsql instances, but because they're firewalled off from the outside world, I don't have to lose sleep over it until the fix comes out.
Ours was 192.0.x.y. Took me about 5 years to finally get us swapped over to the 172.16.x.y - 172.31.x.y range. Seems like a lot of companies didn't grasp that only 192.168.x.y was valid for private use. The main reason we finally switched was that the old 254 address space was too small for our growing needs so we upgraded to a 2046 size address space.
Arch regularly makes changes that will leave your system thoroughly hosed if you update without watching the news feed.
That sounds a lot like Gentoo's methods. Little or no quality assurance in the main feeds which can leave you hosed if you don't treat updates like unexplored minefields. Until they get serious about that, they're a hobbyist distro.
We started with Gentoo way back when. It *sounded* like a good idea at the time because money was tight, hardware was under-spec, and being able to tweak the kernel and other things to run as fast as possible seemed attractive. Unfortunately, the reality is that hardware is relatively cheap, and my time is increasingly rare and expensive. Plus, having all the drivers in the kernel is a good thing because it lets you move entire systems from one set of hardware to another in a pinch.
We moved to CentOS / SciLinux / RHEL because it:
- Almost always works
- Less variation in setups
- Has a public company backing it
- Far easier to get support for RHEL vs Gentoo
- Has the mind share and market share
Mint w/ Mate is probably the best of the breed out of that last group. Debian and a nice desktop, without all the nonsense of Ubuntu's "chase the next rainbow" flavor. I used to recommend Ubuntu, but they jumped off the deep end a few years back so now I recommend Mint.
SuSE has still the best hardware detection and fool-proof installation system of all distros - yes, even better than Ubuntu and Ubuntu derivatives.
We debated SuSE, and it was a contender right up until Novell bought it and then setup that deal-with-the-devil over patents with Microsoft.
Now it's horribly tainted and dead to us.
we're going to have a whole lot of trouble anyway since we're going to have to readjust to a world where energy isn't cheap*, at least in the medium term.
* For various definitions of cheap
I said it 10 years ago, I said it 5 years ago, and I'll say it again. If the price of oil rises *gradually* (over a period of years), you will not see doom-and-gloom happen. There will be market shifts, there will be many things that become more expensive or no longer cost-effective.
But if the rise in prices is spread across multiple years or decades, then people have time to adjust habits, purchase more fuel-efficient vehicles / houses / appliances, develop new energy technologies, build new alternative energy facilities, go after the more expensive oil fields, etc.
Frankly, I worry more about access to fresh water in the next 20 years then I do about the price of oil.
Tape has a vital place in the IT administration world.
Tape is expensive, fragile and requires special hardware. Removable or external magnetic hard drives, OTOH, are cheap, sturdy and will work on any system that you can scrounge up.
Given the costs of tape drives and tape media, it's not surprising that a lot of small / medium businesses just use hard drives for backups. External 2.5" 1TB drives are dirt cheap and you could do weekly off-site backups using them with 13 generations for less then $2000. You can't even buy a large capacity tape drive for $2000. Much less the tapes needed to run a proper backup cycle.
Unless there are legal reasons to keep 5-10 years of backups, or you are dealing in more then 3-5 TB of storage to be backed up, or taking things off-site daily via courier tape is just too expensive.
If it's a password that they use every single day, any user with a brain larger then a goldfish can remember it after a week or two. Those who can't - should probably be working the checkout line at the local grocery store and not handling sensitive data.
However, this means you should not be requiring them to change the password without good cause. Weekly/Monthly/Quarterly resets are not a good enough reason to force a password reset.
We give our users the instructions to put the password on a folded slip of paper in their wallet/purse with no other identifying information on it. People are generally pretty good about keeping track of their money and not letting unauthorized people access to it. So you may as well take advantage of that during the period that it takes the user to remember the password.
The bigger problem is passwords which are used for sensitive systems, where the user accesses them on an infrequent basis. Storage of those passwords is a big problem because users have a hard time remembering them.