Single-mode is the long haul stuff, not multi-mode.
You need: 1. Power at both ends of the line 2. a ditch with conduit 3. a spool of single-mode 4. a professional with the tools to terminate the ends of the fiber 5. two single mode to ethernet media converters, http://www.cdw.com/shop/products/default.aspx?EDC=884092
The music industry is one of the few industries that get away with this.
If you went to the grocer and wanted to buy a couple of oranges only to find out you need to buy the tree with all it's oranges and discover there are only two edible fruits on it, you'd be outraged. Sorry eggs only come in sterling silver 12 packs with 10 cracked.
Sure you occasionally have to buy multi-packs of things, but they're generally of equal quality.
Digital delivery lets you buy what you want. It will eventually win.
So when the major routes are too congested, it'll start telling all the people to take a back road, thus immediately congesting that "less traveled" back route.
I dunna think this is gonna work.
Traffic speed is less a factor of number of cars on a stretch of road, more of number of cars trying to make a decision. (change lanes, exit, enter)
Think automated traffic cameras at merge areas, ticket people who get out of line, race up and cut back in line. Automated systems fire directed sound when you're less than 2 carlengths from the next guy.
Meh, by that standard, nothing in the airport should be labeled.
If 50% of the people in a given place would benefit from the sign, put it up, make it big and clear.
BWI has one, but it's rather innocuous. When I get on the plane, they try to train me 7 ways to Sunday on how to crash. When you're hopping from airport to airport it's hard to tell where the damn lines start and end.
In MIA you carry your check bags to a tsa screener and then get in line, I've been through San Antonio a couple of times and they can't decide whether to take my shoes off or not.
They could use the cellphone data network for the uplink. the cell companies are mostly worried about broadband throughput, if they could provide 128 up, the system would be passible
Or weight sensors under the tank, or optically via shooting visible light through the tank top to bottom and reading the loss, or acoustic range finding from top down or PSI sensors at the bottom of the tank.
There are a million ways to do this. Some tech firm just came up with an idea to combine laser transponders with solar cells and is trying to find something to do with it.
I saw once on TV that they add agents to the fuel to make it even less flammable and a red dye of some sort but it was years ago and that might be outdated information shouldn't really matter for your plan you could use light diffraction.
Put a window at the top of the tank, shoot a narrow beam of light at 45 degrees, place light sensing material on the bottom and side of the tank, as the level recedes, the beam will split more toward the true 45 degree slant it's initial path was set to.
Hospital costs and spending are paramount to DOD costs and spending. A $50,000 prosthetic leg is made up of $1000 in raw materials and $49,000 in labor and design? How much medical imaging equipment does $1,000,000 buy you? 1 MRI... It's a circular racket. The insurance companies pay the hospitals, The hospital charges and outrageous price, the vendors charge the hospital a lot for their equipment, The people working for the hospital get a cut of all that. You really don't expect the boss of 6 digit income workers to make 30k.
Doctors aren't evil, they're just making hard decisions and getting their cut from the corrupt finance model.
Negative, It might not have been gsm in your area but I had a Nokia 6620 on AT&T long before Cingular. What ticked me off was getting the third rate treatment after Cingular took over. I had picked up a sony vaio with an integrated edge data modem. Cingular took over ATT and was sending me two bills. I asked for the accounts to be merged, they told me i'd have to scrap my old phone number. I told them I'd like it ported they said sorry it's the same company. I unlocked the phone myself and took it to T-mobile and bought a Sprint USB EVDO device.
The OP really doesn't give enough detail to give decent advice, the bug could be huge or tiny. The system could be unnecessarily complex or just gigantic.
4-5 weeks for a maintenance patch is fine, but if this bug is stopping your client from doing their work, you should be getting single fixes out the door much faster than that. Patches need to be prioritized. Go back to your last rev, fix only the problem in question and release that small subset as a hotfix.
IMO unless you're writing an operating system, I'd say that taking 4-5 weeks to release a hotfix would mean there's something wrong in your architecture. Perhaps the system was designed for a much smaller implementation and people have just been growing jungles in the existing framework to meet new requirements?
The project should be broken down in to manageably small subsystems. Each subsystem should be autonomous enough to work nearly independently of the other systems. That way you should be limiting the mainstay of your regression testing to the area of the bug itself. Each subsystem should have internal check code and hooks for testing. (a debug mode) Even the best systems still have Achilles heels, but they should be kept few and designed with expertise. Any well crafted system should have limited the damage by design.
This attention to detail in design isn't always possible, deadlines are funny like that. If companies and clients only understood that spending 2x the time up front reduces the backside time by 10x, our jobs would all be a little easier.
Many vendors (MS included) develop the patch and allow the users who call in with the problem to install at their own risk before exhaustive testing has been completed.
Off Topic Well we have very different needs. I have 70 servers in 26 sites comprising about an 8TB footprint. Synthetic fulls and high throughput are saving my bacon. I upgraded from backup exec and a 132T with 2 LTO2's to a 6 drive 128 tape LTO3 solution. We spent a few bucks.
Dells Commvault (CV) support is lacking. It's been hinted that I can go through Dell to renegotiate direct CV support but I don't know how that works with the lite versions. I'm not even sure the interface is the same.
I also took the class which was fantastic. I'd still be wandering around the app blindly.
A little hint... The one thing they drilled in to our heads at the class... I thought it was propaganda at the time: If you're having a problem, it's most likely going to be in your hardware or your network. Reverse DNS HAS to be PERFECT. Connections have to be solid. Tape drives need to be the latest firmware revision, drivers have to be up to date. When in doubt, use Microsoft Backup and try to back up target, if that works at decent speed, it's CV or a CV setting.
The most impressive thing I've seen so far, right after install we were doing our first full. It was about 80% done and in the middle of a multi-terabyte set. We were dorking with the commserve at a very low level and destroyed it. Backups all locked up. We ended up reloading from scratch and restoring the DR. After the DR finished and rebooted the backups picked back up, right at the 64KB block they left off. It was surreal, like nothing happened at all.
The only two things I have bad to say are: 1. It could be a little more forgiving on network issues. and 2. The default mode for a failed backup is no data protection. By default if the backup completes with failure, you cannot restore from it without a special license and a data explorer. I'm used to BE failing all the time but being able to get whatever i can back out of it as if nothing went wrong. You need to stay on top of your backups with CV.
If you want to chew the fat about it, I'm doormouse where that g mail happens.
Meh, my vendors will not "support" it, but if you've paid for support from them, they'll "try" to get it working. Which of course almost always works identically to RHEL.
and by my vendors I mean Dell, Commvault, EMC. and by metaphorically, I mean get your coat.
Slashdot Mirror
It's electric, just where do you expect them to put the batteries... wait, I don't want to know.....
Single-mode is the long haul stuff, not multi-mode.
You need:
1. Power at both ends of the line
2. a ditch with conduit
3. a spool of single-mode
4. a professional with the tools to terminate the ends of the fiber
5. two single mode to ethernet media converters, http://www.cdw.com/shop/products/default.aspx?EDC=884092
The music industry is one of the few industries that get away with this.
If you went to the grocer and wanted to buy a couple of oranges only to find out you need to buy the tree with all it's oranges and discover there are only two edible fruits on it, you'd be outraged. Sorry eggs only come in sterling silver 12 packs with 10 cracked.
Sure you occasionally have to buy multi-packs of things, but they're generally of equal quality.
Digital delivery lets you buy what you want. It will eventually win.
So Microsoft will do what they always do, buy Sony...
Would it be better, the readership of Slashdot be reduced to raving mad lunatics having grasped the presence of the all great sleeper?
Ohh wait, that might already have happened.
meh, I bet they were working on 3.5 ton (3175kg) beef cattle.
What happens when the advertiser decided to edit pages their ads show up on for the purpose of endorsing their own service as a reported consumer.
I'm fairly certain you could leave the ram in the machine and read it while running if you were careful.
LCD on one side, camera on the other.....
So when the major routes are too congested, it'll start telling all the people to take a back road, thus immediately congesting that "less traveled" back route.
I dunna think this is gonna work.
Traffic speed is less a factor of number of cars on a stretch of road, more of number of cars trying to make a decision. (change lanes, exit, enter)
Think automated traffic cameras at merge areas, ticket people who get out of line, race up and cut back in line. Automated systems fire directed sound when you're less than 2 carlengths from the next guy.
Meh, by that standard, nothing in the airport should be labeled.
If 50% of the people in a given place would benefit from the sign, put it up, make it big and clear.
BWI has one, but it's rather innocuous. When I get on the plane, they try to train me 7 ways to Sunday on how to crash. When you're hopping from airport to airport it's hard to tell where the damn lines start and end.
In MIA you carry your check bags to a tsa screener and then get in line, I've been through San Antonio a couple of times and they can't decide whether to take my shoes off or not.
*stocks up on lantern batteries and large nails*
They could use the cellphone data network for the uplink.
the cell companies are mostly worried about broadband throughput, if they could provide 128 up, the system would be passible
04 08 15 16 23 42
PS3 + TVersity (tversity.com) makes a fantastic media server.
Or weight sensors under the tank, or optically via shooting visible light through the tank top to bottom and reading the loss, or acoustic range finding from top down or PSI sensors at the bottom of the tank.
There are a million ways to do this. Some tech firm just came up with an idea to combine laser transponders with solar cells and is trying to find something to do with it.
I saw once on TV that they add agents to the fuel to make it even less flammable and a red dye of some sort but it was years ago and that might be outdated information shouldn't really matter for your plan you could use light diffraction.
Put a window at the top of the tank, shoot a narrow beam of light at 45 degrees, place light sensing material on the bottom and side of the tank, as the level recedes, the beam will split more toward the true 45 degree slant it's initial path was set to.
Hospital costs and spending are paramount to DOD costs and spending. A $50,000 prosthetic leg is made up of $1000 in raw materials and $49,000 in labor and design? How much medical imaging equipment does $1,000,000 buy you? 1 MRI... It's a circular racket. The insurance companies pay the hospitals, The hospital charges and outrageous price, the vendors charge the hospital a lot for their equipment, The people working for the hospital get a cut of all that. You really don't expect the boss of 6 digit income workers to make 30k.
Doctors aren't evil, they're just making hard decisions and getting their cut from the corrupt finance model.
http://www.cyberdriveillinois.com/publications/rules_of_the_road/rr_chap08.html
When you're not sitting at tracks:
Flashing Red means stop and yeald the right of way.
Flashing Yellow means proceed with caution.
Flashing red for trains is usually accompanied by a bell and often a pike.
Trolleys in SF are the odd man out with their own unique system.
Here's a pretty good set of Right/Left turn on red rules for North America
http://en.wikipedia.org/wiki/Right_turn_on_red
Negative, It might not have been gsm in your area but I had a Nokia 6620 on AT&T long before Cingular. What ticked me off was getting the third rate treatment after Cingular took over. I had picked up a sony vaio with an integrated edge data modem. Cingular took over ATT and was sending me two bills. I asked for the accounts to be merged, they told me i'd have to scrap my old phone number. I told them I'd like it ported they said sorry it's the same company. I unlocked the phone myself and took it to T-mobile and bought a Sprint USB EVDO device.
You don't have 5V or 12V AC service in your house either.
Voltage Drop for AC or DC is pretty much the same, it's about how much current you're trying to pass.
http://nooutage.com/vdrop.htm
120V AC or DC could be reduced at the end point.
I LOLed,
IP is safe for work
The OP really doesn't give enough detail to give decent advice, the bug could be huge or tiny. The system could be unnecessarily complex or just gigantic.
4-5 weeks for a maintenance patch is fine, but if this bug is stopping your client from doing their work, you should be getting single fixes out the door much faster than that. Patches need to be prioritized. Go back to your last rev, fix only the problem in question and release that small subset as a hotfix.
IMO unless you're writing an operating system, I'd say that taking 4-5 weeks to release a hotfix would mean there's something wrong in your architecture. Perhaps the system was designed for a much smaller implementation and people have just been growing jungles in the existing framework to meet new requirements?
The project should be broken down in to manageably small subsystems. Each subsystem should be autonomous enough to work nearly independently of the other systems. That way you should be limiting the mainstay of your regression testing to the area of the bug itself. Each subsystem should have internal check code and hooks for testing. (a debug mode) Even the best systems still have Achilles heels, but they should be kept few and designed with expertise. Any well crafted system should have limited the damage by design.
This attention to detail in design isn't always possible, deadlines are funny like that. If companies and clients only understood that spending 2x the time up front reduces the backside time by 10x, our jobs would all be a little easier.
Many vendors (MS included) develop the patch and allow the users who call in with the problem to install at their own risk before exhaustive testing has been completed.
meh, off my soapbox, need to go to work myself.
But they saved on lawyers fees because they didn't try to chase down housewives and drag them through lengthy legal battles.
Not to mention probably the best PR money can buy.
Let's hope this raises the bar.
Off Topic
Well we have very different needs. I have 70 servers in 26 sites comprising about an 8TB footprint.
Synthetic fulls and high throughput are saving my bacon. I upgraded from backup exec and a 132T with
2 LTO2's to a 6 drive 128 tape LTO3 solution. We spent a few bucks.
Dells Commvault (CV) support is lacking. It's been hinted that I can go through Dell to renegotiate direct CV support
but I don't know how that works with the lite versions. I'm not even sure the interface is the same.
I also took the class which was fantastic. I'd still be wandering around the app blindly.
A little hint... The one thing they drilled in to our heads at the class... I thought it was propaganda at the time:
If you're having a problem, it's most likely going to be in your hardware or your network. Reverse DNS HAS to be PERFECT. Connections have to be solid. Tape drives need to be the latest firmware revision, drivers have to be up to date. When in doubt, use Microsoft Backup and try to back up target, if that works at decent speed, it's CV or a CV setting.
The most impressive thing I've seen so far, right after install we were doing our first full. It was about 80% done and in the middle of a multi-terabyte set. We were dorking with the commserve at a very low level and destroyed it. Backups all locked up. We ended up reloading from scratch and restoring the DR. After the DR finished and rebooted the backups picked back up, right at the 64KB block they left off. It was surreal, like nothing happened at all.
The only two things I have bad to say are: 1. It could be a little more forgiving on network issues. and 2. The default mode for a failed backup is no data protection. By default if the backup completes with failure, you cannot restore from it without a special license and a data explorer. I'm used to BE failing all the time but being able to get whatever i can back out of it as if nothing went wrong. You need to stay on top of your backups with CV.
If you want to chew the fat about it, I'm doormouse where that g mail happens.
Meh, my vendors will not "support" it, but if you've paid for support from them, they'll "try" to get it working. Which of course almost always works identically to RHEL.
and by my vendors I mean Dell, Commvault, EMC.
and by metaphorically, I mean get your coat.