You can change energy from one form to another but can't create or destroy it.
If you MUST have 4 computers running - they will eat something like 500-800+ watts - so you need to use or get rid of that amount of energy. (note that CRTs use similar amounts - use LCDs!)
Most refrigerators use something less than 500-800 watts to cool their interiors - but many (most) use motors to compress coolant and dump the heat out to cooling coils on the back (heating the surrounding room). Why not instead use a 'fridge that uses an absorption cycle (like the ones you find in a camper that use either electricity or gas to provide heat but don't use a motor) and run it off the heat of the computers?
Note that you'll have to provide some dump for the heat - so you'll probably want to put the computers (and some part of the fridge) near an outside wall so you can dump heat outside (choose the side away from the sun) - use long monitor cables etc. to bring the video/keyboard/mouse connections to where you want them.
In Summer - open the windows near the computers to let the heat out
In Winter - put a fan there and blow the excess heat into the rest of the house.
Of course you might want to run some copper tubing from the water supply near these heat sources to pre-heat the water prior to running it into the normal water heater too. You can also plumb them into solar panels on the roof in Summer (and even in Winter in some areas) to preheat water. A little bit of electronics might be necessary to ensure you don't overheat the water - I've almost burned my hand off the panels I have on the roof for my pool:)
Lots of ways of (re)using the energy you take into the house before letting it out - and taking advantage of the energy that is freely available from the sun when it shines.
I have a number of Unix/Linux users who use their systems as desktop workstations and don't use root (at all - I set them up and do all maintenance remotely)
Their systems do daily backups of home directories to a protected area that is read-only by their IDs. Whether or not the overall systems are less virus/worm prone is not really the issue, the fact is that only an attack that can get root access can actually do (locally) irretrievable damage.
The better thing IMHO about Linux/Unix is that there is transparency about what actually needs to be backed up in most cases (some require a bit of sleuthing but even they can be made transparent) - the home directory and maybe a major application data directory (MySQL for example)
Only these need to be dealt with - the rest of the machine's resources can be replicated/restored/reinstalled and add the data and go on your happy way.
You're the CTO - got the data and the key - problem?
TFA is talking about the potential for loss given loss of a USB drive (thumb or whatever) that might or might not have been authorized to copy the original info. If authorized then chances are that DRM might have saved the info from exposure. If not (and today this is the most likely scenario) then whatever is on the USB drive is now public info.
DRM is not yet widely used or even liked. And none of my customers use it for anything (and some are publishers)
I wonder how much communications had to do with our perception of the number of geniuses at that time. Maybe Joe Smoe of the 9th century developed comparable work to that of Albert Einstein but just didn't get published.
Many have posted posits about a problem with the ground and maybe bonding the neutral to ground. There are instances where even though the electrician has done his job correctly, there can still be a voltage difference between the grounds at various outlets to each other.
There are two ways this can happen:
current flow in the neutral on circuits with more than one plug (many offices put 2 or 4 plugs on a single circuit) causes a voltage drop across the wire due to the resistance in the length of the cable. If the plugs are close together this is next to nothing, but if they are far apart (and many electricians spread them by hop-scotching the circuits around the room so two plugs close together are on separate circuits but the same circuit shows up 20' away again (and maybe 40' away again)
Note that this can be induced in otherwise properly bonded circuits by the use of daisy-chained power strips/bars. It is the act of plugging in a (typically high-power) item on the end of the line (so the draw through the whole line is high) and connecting it via a signal cable (Ethernet, serial, USB, etc.) to something plugged closer to the junction box (electrically) that then ends up routing some of the ground current through the low-voltage signal line.
imbalance between the 3 legs of the 3-phase causing current flow in the common neutral which causes a voltage difference between neutral and building ground
In general, the cure is to return to the bonded-circuit of yesterday designated by the orange plugs where they've been installed. These consist of a single plug per circuit (never more than one!) where ground is bonded to neutral at the connection box so there is no possibility that there may be a voltage drop due to current between individual recepticals.
We seem to have gotten away from the specification and use of these (more expensive to install) power recepticals. I for one continue to specify them for most commercial installations and have yet to see any of the kinds of things mentioned here when they have been properly installed and used.
The worst case I saw of this was in an office that was long and skinny in a brand new building (they were the top floor) with retail below. The office went in first, and the retail later proved to include a dry-cleaner that used quite a bit of power off two phases of the 3.
A Unix computer in the center of the long office fed dumb terminals the length of both directions. The reception area was the farthest out - about 100' by wire - RS232 shield.
The terminal at reception kept doing wierd things: hanging, mystery characters, and in fact died - 3 times! Lights on but nobody home!
It turned out the serial interface was dieing - due to about 5 volts between ground and neutral which was pulling current through the cable's cladding and buring out the chip. The electrician poh-poh'd it saying "5 volts on a 120volt line was nothing to worry about" but in fact it was 5 volts on a 12 volt circuit and carried current in what should have been a voltage (high impedance) system. No wonder the interfaces were flaky and buring out.
Replacing the power plugs with "home-run" single bonded ones fixed the problem.
Sounds like Encyclopedia Galactica (Asimov)
on
Forecasting Doomsday
·
· Score: 2, Interesting
Foundation series: Civilization is falling - accumulate all knowledge in a set of books and make copies to send to the far reaches of the known universe.
Or do you subscribe to Heinlein and his survivor stories like Farnham's Freehold?
With the various governments' movements to ban guns and such I'm beginning to smell a conspiracy theory here somewhere:)
Me? I'll probably be one of the first ones to die when I can't get the drugs that keep me alive - of course Darwin is at work there too. "Whatever doesn't kill you makes you stronger" and "go lemmings!" are my two favourite catch-phrases.
and you have a camera (or microphone, or radio, or video camera) - and nobody can stop you from putting any software you like into it.
The CPUs that come in today's toys and appliances are able to be adapted to do almost anything. Brute force and minimal hardware can be made to do things that yesterday took dedicated processors and million man-hours of programming to do.
Look at today's software radio that can tune literally any channel and/or use any type of encoding scheme. Put that up against a radio station that is sending out DRM tagged audio.
Same thing with video. I have one of the old Timex (Microsoft) data watches - a 1 bit video (bar code) reader on it. The first pictures from Mars were done with similar hardware - 1 pixel camera with rotating mirror to build up a picture over time.
If someone sells me access to "the Internet" and blocks ports defined in RFCs then it isn't "the Internet" it is something else.
Back when AOL and Compuserve were BBSs (networks unto themselves with minimal/no connection to other services) their customers demanded access to Internet E-mail and got it; eventually bundled in as opposed to for extra charge.
The ISPs will have to realize that there are ways to circumvent their blockages and all it takes is one person to come up with it and the whole world knows.
How about "port knocking" as a data transport? I hesitate to list some of the other methods our group of gurus has discussed over the past few years, but you can be assured that there are lots, and the black hats have been using them for some time now.
How about someone providing a service that tunnels other traffic via an unblocked port? Unencrypted there would be not much extra overhead - encrypted it would be proof against almost any blocking since the tunnel service provider can use any port they want and the ISP can't block them all or what's the use of calling it a network. Port 80 sounds like a good choice.
And if the ISP blocks the service's address block, how about something that does a shared-bandwidth service such as bittorrent does now?
Pretty soon the ISPs will get it through their thick skulls that blocking ports isn't the way - providing lower latency for similar service (to that provided by someone farther away by net) or making partnerships (franchises, etc.) with the data/service/application providers is really the only way to differentiate.
Using the routers is easy - but it will not prevail.
OK - 5-50KW for 1/10 second isn't much - but it would light a ton of LED signs for a looonnnng time.
These guys really need to give their collective heads a shake - ~&25,000 will purchase a hell of a lot of LED lights, a battery/capacitor bank and Solar Array (OK - Britain doesn't get as much sun as some places - but its possible, OK?)
Put this one up there with the ones who think there is a perpetual motion machine.
1 on a Windoze box with a TV card in it running Win2000 on a V95 Optiquest 19" at 1280x1024
2 NEC Accusync LCD71VM at 1280x1024 on a nVidia dual-head and
another V95 Optiquest at 1600x1200 on a 3D Rage (for editing graphics sometimes but see below for normal use)
Ignoring the Windows box as a necessary evil for times when I have to debug a customer who hasn't switched yet, we move on to the Linux box (still running RH9 but in the process of moving to FC4)
I run 3 separate desktops instead of a unified one - mostly because I have not found a way to have enough virtual desktops to satisfy me any other way. I run 2x15 (30) in each screen for a total of 90 virtual desktops - and have seen times when I have something in all of them - but not always.
I'm generally in one of 3 mental modes - and use the screens differently for each:
1 - general system monitoring and maintenance - read e-mail, watch logs, read documentation, etc.
2 - major system updates/maintenance - configuring and copying software, etc.
3 - working with graphics and web site creation and updates
In 1 I have system monitoring visuals on the right (1600x1200) monitor of various flavors - xload, xosview, eximon - started from a single script (showall - see below), working screens (xterms, web browsers, etc.) with things like slashdot and groklaw plus other documentation and web sites I'm working with in the middle (and sometimes on the left) and e-mail on the left. I'll preempt the e-mail for more screens if necessary but tend to have it back most times.
In 2 I'll have screens on the left on a new system, on the right on the old and documentation in the middle
In 3 I'll do editing on the V95 because it has reasonable color fidelity compared to the LCDs, while I select and categorize (gThumb) on the central screen and do command line stuff on the left LCD (ImageMagick etc.)
I start programs in the various desktops in a consistent manner all the time. Each of the 3 monitors has a slightly different mix with most consitency at each end and the central ones more "ad hoc"
In general, the right screens on all desktops are used for access to the major servers I monitor. Most active on the right, second most active to its left, and so on toward the left (currently 8 machines are showing)
On the left screen's desktops I start at the left with e-mail and local root and personal items, then web screens and on the right, xterms to systems
On the center screen's desktops I start on the left with local (to my workstation) xterms, then web screens and on the right xterms to systems
On the right screen's desktops I start with music (xmms and volume control) upper left with the "showall" screen next right. Bottom left is vmWare with another Win2k instance in it, used for "dangerous" stuff because I can restore it with a copy command from the command line. Other vmWare sessions to it's right as necessary. In the center desktops I'll have more web screens and on the right are xterms
In the center portion of each screen I start a browser as necessary. I'll typically have one each of Mozilla, Firefox and something else up on the 3 screens, with Firefox in the center one with as many as 8-10 sessions (press ^N 7 times and use the mouse to distribute the screens to desktops) on each of the center 8-10 virtual desktops.
Left-most top web screen (5th from left actual desktop) has Slashdot and below that is Groklaw. Next right to each of them are Nagios and MRTG on my main monitoring machines and to the Nagios' right is a session into the main e-mail configuration machine. The rest home on Google or my main web menu (which times out to refresh to google after 2 minutes) and may have anything that strikes my fancy on
Two are LCD and one is a "normal" CRT since the LCDs don't yet have the broad contrast range that CRTs do and I sometimes edit pictures.
Having VmWare is a must if you need to test on anything but your normal OS and version.
And at least 1 Meg RAM per MHz of CPU no matter what people tell you otherwise! If you are not actually using the RAM for programs, it will speed up your disk access and give you response that fits with your requirements. This of course means that dual-core CPUs should have double the RAM and a pair of them 4 times - dual Xeon 3GHz means something like 12 Gigs of RAM.
And last but not least - almost no SWAP! But keep a swap-formatted file around just in case you need it. The fact that you have almost no swap (I have 10 Megs) means your system will pretty much never swap out anything you are actually using - like VmWare if you only use it infrequently like I do. you should also likely set the paging file for things like Windows under VmWare to almost nothing, but ramp up the amount of RAM the virtual machine has available as the latest version of VmWare will allow you to oversubscribe RAM if you need to.
In addition, get at least a base set of fast/wide SCSI drives (15k RPM 320FW) and use them for any compiles. Two or three on a controller max if you are striping for speed. Use a separate set of UDMA or SATA drives for archive and low(er)-speed storage. Of course if you can spend the money, make sure they're all fast SCSI. SATA is coming up on the performance curve but doesn't yet seem to be there. Use them if you can't afford the SCSIs.
I've long felt that putting a good programmer in front of the fastest hardware is a win situation. If you pay someone the big bucks, spending a few extra on a fast system is worth it. I've seen companies put a top flight programmer in front of a machine I wouldn't give my ailing mother (and I love my mother); what a waste!
While it is not cp or copy, using rsync will achieve this for backups. Do a cp -al to an archive area which simply links the files (and thereby saves the space but eats up lots of inodes) and when rsync next looks at the original and finds a difference it copies the original and unlinks it from the copy - then applies the changes to the new file leaving the old one alone.
so the functionality is somewhere buried in rsync - maybe a plugin can be made.
I've managed lots of programming staff and observe that every time a compile or other "large" process takes more than a couple of minutes with the CPU or some other critical system maxed out (i.e. can't do something else 'cause my system's too slow) the programmer will get up and go get a coffee or water or a stretch or whatever - and not get back until some time after the process has completed.
My partner Stuart, for instance, is addicted to capucino (has a $10,000 machine of his own at home where he does most of his work) and when he does a major kernel compile it takes his maxed-out system 5-10 minutes to complete. This is just long enough for him to go get a cup and mostly finish it off. If the compile took less than 2 minutes, he'd probably wait in his chair instead.
His recently purchased 15,000 RPM ultra 320 SCSI drives can mostly keep his current dual CPU system choked at the CPU, but if we put in one of the faster CPU quads or other system it will languish waiting for the disk.
The use of solid-state drives, even as small as 4 Gigs (although I'd probably go for 4-8 of them to increase the size and throughput by RAID0) would keep him in his chair more, and more than pay for the cost in increased productivity.
Of course if you look at this from the point of view of a server instead of a workstation, the economic reasoning may be easier.
I use almost exactly the above noted rsync_snapshots technique to deal with several data stores that are in excess of 500Gigs each on customer machines. The nightly/weekly backups of most take a few minutes to an hour or so of transfer time - then the "cp -al" takes a while on the backup machine (some of the stores are literally millions of files)
One system has a very large database - but the fact that rsync will pick the changes out of the center of a file makes the process fast even though some of the files are Gigs in size.
This (rsync + cp -al to allow for incrementals) is by far the best way I've found for backing up either large amounts of files or large files that have changes somewhere in them but not necessarily everywhere in them.
The backup servers are RAID-5 with hot spares and mdmonitor turned on plus NAGIOS watching too.
Note that the first copy can take lots of time if the connection is slow (I have one that is still running after a week at about 50Bytes/sec to allow their VOIP to continue to work) but once it is done, the only things that are copied are changes.
#!/bin/sh # daily-backup
BW=1000
for dir in bin boot data dev etc home lib mnt net root sbin usr tmp var opt do
do this for each of the similar systems then with a script do a cp -al to a numbered directory. Delete the highest number and move the lower ones up. This assumes you don't have the ability to mount locally via smbmount or nfsmount.
I'm guessing that what you say really doesn't matter so much as that you say the same thing each time and the system recognizes the resonances and attributes of the sound that are shaped by your various body cavities and your vocal chords.
In other words, this is really just another form of biometrics.
Of course the fact that you have a gallon of phlegm in your lungs, a hoarse voice from barfing and fever induced hiccups today because you've just caught avian flu may make paying for your medical visit a bit problematic.
Slashdot Mirror
If you MUST have 4 computers running - they will eat something like 500-800+ watts - so you need to use or get rid of that amount of energy. (note that CRTs use similar amounts - use LCDs!)
Most refrigerators use something less than 500-800 watts to cool their interiors - but many (most) use motors to compress coolant and dump the heat out to cooling coils on the back (heating the surrounding room). Why not instead use a 'fridge that uses an absorption cycle (like the ones you find in a camper that use either electricity or gas to provide heat but don't use a motor) and run it off the heat of the computers?
Note that you'll have to provide some dump for the heat - so you'll probably want to put the computers (and some part of the fridge) near an outside wall so you can dump heat outside (choose the side away from the sun) - use long monitor cables etc. to bring the video/keyboard/mouse connections to where you want them.
In Summer - open the windows near the computers to let the heat out
In Winter - put a fan there and blow the excess heat into the rest of the house.
Of course you might want to run some copper tubing from the water supply near these heat sources to pre-heat the water prior to running it into the normal water heater too. You can also plumb them into solar panels on the roof in Summer (and even in Winter in some areas) to preheat water. A little bit of electronics might be necessary to ensure you don't overheat the water - I've almost burned my hand off the panels I have on the roof for my pool :)
Lots of ways of (re)using the energy you take into the house before letting it out - and taking advantage of the energy that is freely available from the sun when it shines.
I have a number of Unix/Linux users who use their systems as desktop workstations and don't use root (at all - I set them up and do all maintenance remotely)
Their systems do daily backups of home directories to a protected area that is read-only by their IDs. Whether or not the overall systems are less virus/worm prone is not really the issue, the fact is that only an attack that can get root access can actually do (locally) irretrievable damage.
The better thing IMHO about Linux/Unix is that there is transparency about what actually needs to be backed up in most cases (some require a bit of sleuthing but even they can be made transparent) - the home directory and maybe a major application data directory (MySQL for example)
Only these need to be dealt with - the rest of the machine's resources can be replicated/restored/reinstalled and add the data and go on your happy way.
During off-peak time, pump water uphill to a holding reservoir - a big swimming pool on the roof might do.
:)
Heat the water with the waste heat from the cooling units.
Sell access for swimming - nice warm water (well, here in Canada we like it warm
During peak hours, drain the pool back via generators to make electricity. (make sure you tell people first)
Use warm water to cool more - generate steam.
Run steam through turbines to generate electricity.
Use electricity to pump more water to pool on roof
continue as needed
Somewhere along the way I chose things electronic (and computational) and here I am...
What does the education system expose your kids to today?
TFA is talking about the potential for loss given loss of a USB drive (thumb or whatever) that might or might not have been authorized to copy the original info. If authorized then chances are that DRM might have saved the info from exposure. If not (and today this is the most likely scenario) then whatever is on the USB drive is now public info.
DRM is not yet widely used or even liked. And none of my customers use it for anything (and some are publishers)
Personally I'd rather know that it's been used and let the experts go looking for it :)
got your rubber goves handy?
My personal hacker (12 years old, immune from prosecution) just duplicated your key-fob's ID. What are you going to do about it?
Check - and Mate!
The log files don't lie!
Of course if you can't find them, then it doesn't matter, does it? Does WinXX create a log file of USB insertion - damned if I know!
I wonder how much communications had to do with our perception of the number of geniuses at that time. Maybe Joe Smoe of the 9th century developed comparable work to that of Albert Einstein but just didn't get published.
There are two ways this can happen:
Note that this can be induced in otherwise properly bonded circuits by the use of daisy-chained power strips/bars. It is the act of plugging in a (typically high-power) item on the end of the line (so the draw through the whole line is high) and connecting it via a signal cable (Ethernet, serial, USB, etc.) to something plugged closer to the junction box (electrically) that then ends up routing some of the ground current through the low-voltage signal line.
In general, the cure is to return to the bonded-circuit of yesterday designated by the orange plugs where they've been installed. These consist of a single plug per circuit (never more than one!) where ground is bonded to neutral at the connection box so there is no possibility that there may be a voltage drop due to current between individual recepticals.
We seem to have gotten away from the specification and use of these (more expensive to install) power recepticals. I for one continue to specify them for most commercial installations and have yet to see any of the kinds of things mentioned here when they have been properly installed and used.
The worst case I saw of this was in an office that was long and skinny in a brand new building (they were the top floor) with retail below. The office went in first, and the retail later proved to include a dry-cleaner that used quite a bit of power off two phases of the 3.
A Unix computer in the center of the long office fed dumb terminals the length of both directions. The reception area was the farthest out - about 100' by wire - RS232 shield.
The terminal at reception kept doing wierd things: hanging, mystery characters, and in fact died - 3 times! Lights on but nobody home!
It turned out the serial interface was dieing - due to about 5 volts between ground and neutral which was pulling current through the cable's cladding and buring out the chip. The electrician poh-poh'd it saying "5 volts on a 120volt line was nothing to worry about" but in fact it was 5 volts on a 12 volt circuit and carried current in what should have been a voltage (high impedance) system. No wonder the interfaces were flaky and buring out.
Replacing the power plugs with "home-run" single bonded ones fixed the problem.
Or do you subscribe to Heinlein and his survivor stories like Farnham's Freehold?
With the various governments' movements to ban guns and such I'm beginning to smell a conspiracy theory here somewhere :)
Me? I'll probably be one of the first ones to die when I can't get the drugs that keep me alive - of course Darwin is at work there too. "Whatever doesn't kill you makes you stronger" and "go lemmings!" are my two favourite catch-phrases.
The CPUs that come in today's toys and appliances are able to be adapted to do almost anything. Brute force and minimal hardware can be made to do things that yesterday took dedicated processors and million man-hours of programming to do.
Look at today's software radio that can tune literally any channel and/or use any type of encoding scheme. Put that up against a radio station that is sending out DRM tagged audio.
Same thing with video. I have one of the old Timex (Microsoft) data watches - a 1 bit video (bar code) reader on it. The first pictures from Mars were done with similar hardware - 1 pixel camera with rotating mirror to build up a picture over time.
Today's fast chips (and you don't have to purchase a hobbled one from Intel or AMD - you can build your own with GPL VHDL code) can be used to create any camera you want.
All it takes is one person to do it and the rest of the world will know how.
"gee Mr. government man - I was just experimenting" :)
If someone sells me access to "the Internet" and blocks ports defined in RFCs then it isn't "the Internet" it is something else.
Back when AOL and Compuserve were BBSs (networks unto themselves with minimal/no connection to other services) their customers demanded access to Internet E-mail and got it; eventually bundled in as opposed to for extra charge.
The ISPs will have to realize that there are ways to circumvent their blockages and all it takes is one person to come up with it and the whole world knows.
How about "port knocking" as a data transport? I hesitate to list some of the other methods our group of gurus has discussed over the past few years, but you can be assured that there are lots, and the black hats have been using them for some time now.
How about someone providing a service that tunnels other traffic via an unblocked port? Unencrypted there would be not much extra overhead - encrypted it would be proof against almost any blocking since the tunnel service provider can use any port they want and the ISP can't block them all or what's the use of calling it a network. Port 80 sounds like a good choice.
And if the ISP blocks the service's address block, how about something that does a shared-bandwidth service such as bittorrent does now?
Pretty soon the ISPs will get it through their thick skulls that blocking ports isn't the way - providing lower latency for similar service (to that provided by someone farther away by net) or making partnerships (franchises, etc.) with the data/service/application providers is really the only way to differentiate.
Using the routers is easy - but it will not prevail.
OK - 5-50KW for 1/10 second isn't much - but it would light a ton of LED signs for a looonnnng time.
These guys really need to give their collective heads a shake - ~&25,000 will purchase a hell of a lot of LED lights, a battery/capacitor bank and Solar Array (OK - Britain doesn't get as much sun as some places - but its possible, OK?)
Put this one up there with the ones who think there is a perpetual motion machine.
Reboot, Reinstall, Upgrade
OTOH - I might want one of those butt plugs - I have a bit of travelling to do...
Here's what I have on my desktop as I write this:
4 monitors on 2 systems:
1 on a Windoze box with a TV card in it running Win2000 on a V95 Optiquest 19" at 1280x1024
2 NEC Accusync LCD71VM at 1280x1024 on a nVidia dual-head and another V95 Optiquest at 1600x1200 on a 3D Rage (for editing graphics sometimes but see below for normal use)
Ignoring the Windows box as a necessary evil for times when I have to debug a customer who hasn't switched yet, we move on to the Linux box (still running RH9 but in the process of moving to FC4)
I run 3 separate desktops instead of a unified one - mostly because I have not found a way to have enough virtual desktops to satisfy me any other way. I run 2x15 (30) in each screen for a total of 90 virtual desktops - and have seen times when I have something in all of them - but not always.
I'm generally in one of 3 mental modes - and use the screens differently for each:
1 - general system monitoring and maintenance - read e-mail, watch logs, read documentation, etc.
2 - major system updates/maintenance - configuring and copying software, etc.
3 - working with graphics and web site creation and updates
In 1 I have system monitoring visuals on the right (1600x1200) monitor of various flavors - xload, xosview, eximon - started from a single script (showall - see below), working screens (xterms, web browsers, etc.) with things like slashdot and groklaw plus other documentation and web sites I'm working with in the middle (and sometimes on the left) and e-mail on the left. I'll preempt the e-mail for more screens if necessary but tend to have it back most times.
In 2 I'll have screens on the left on a new system, on the right on the old and documentation in the middle
In 3 I'll do editing on the V95 because it has reasonable color fidelity compared to the LCDs, while I select and categorize (gThumb) on the central screen and do command line stuff on the left LCD (ImageMagick etc.)
I start programs in the various desktops in a consistent manner all the time. Each of the 3 monitors has a slightly different mix with most consitency at each end and the central ones more "ad hoc"
In general, the right screens on all desktops are used for access to the major servers I monitor. Most active on the right, second most active to its left, and so on toward the left (currently 8 machines are showing)
On the left screen's desktops I start at the left with e-mail and local root and personal items, then web screens and on the right, xterms to systems
On the center screen's desktops I start on the left with local (to my workstation) xterms, then web screens and on the right xterms to systems
On the right screen's desktops I start with music (xmms and volume control) upper left with the "showall" screen next right. Bottom left is vmWare with another Win2k instance in it, used for "dangerous" stuff because I can restore it with a copy command from the command line. Other vmWare sessions to it's right as necessary. In the center desktops I'll have more web screens and on the right are xterms
In the center portion of each screen I start a browser as necessary. I'll typically have one each of Mozilla, Firefox and something else up on the 3 screens, with Firefox in the center one with as many as 8-10 sessions (press ^N 7 times and use the mouse to distribute the screens to desktops) on each of the center 8-10 virtual desktops.
Left-most top web screen (5th from left actual desktop) has Slashdot and below that is Groklaw. Next right to each of them are Nagios and MRTG on my main monitoring machines and to the Nagios' right is a session into the main e-mail configuration machine. The rest home on Google or my main web menu (which times out to refresh to google after 2 minutes) and may have anything that strikes my fancy on
oh... wait... hmmm...
too late
Two are LCD and one is a "normal" CRT since the LCDs don't yet have the broad contrast range that CRTs do and I sometimes edit pictures.
Having VmWare is a must if you need to test on anything but your normal OS and version.
And at least 1 Meg RAM per MHz of CPU no matter what people tell you otherwise! If you are not actually using the RAM for programs, it will speed up your disk access and give you response that fits with your requirements. This of course means that dual-core CPUs should have double the RAM and a pair of them 4 times - dual Xeon 3GHz means something like 12 Gigs of RAM.
And last but not least - almost no SWAP! But keep a swap-formatted file around just in case you need it. The fact that you have almost no swap (I have 10 Megs) means your system will pretty much never swap out anything you are actually using - like VmWare if you only use it infrequently like I do. you should also likely set the paging file for things like Windows under VmWare to almost nothing, but ramp up the amount of RAM the virtual machine has available as the latest version of VmWare will allow you to oversubscribe RAM if you need to. In addition, get at least a base set of fast/wide SCSI drives (15k RPM 320FW) and use them for any compiles. Two or three on a controller max if you are striping for speed. Use a separate set of UDMA or SATA drives for archive and low(er)-speed storage. Of course if you can spend the money, make sure they're all fast SCSI. SATA is coming up on the performance curve but doesn't yet seem to be there. Use them if you can't afford the SCSIs.
I've long felt that putting a good programmer in front of the fastest hardware is a win situation. If you pay someone the big bucks, spending a few extra on a fast system is worth it. I've seen companies put a top flight programmer in front of a machine I wouldn't give my ailing mother (and I love my mother); what a waste!
so the functionality is somewhere buried in rsync - maybe a plugin can be made.
My partner Stuart, for instance, is addicted to capucino (has a $10,000 machine of his own at home where he does most of his work) and when he does a major kernel compile it takes his maxed-out system 5-10 minutes to complete. This is just long enough for him to go get a cup and mostly finish it off. If the compile took less than 2 minutes, he'd probably wait in his chair instead.
His recently purchased 15,000 RPM ultra 320 SCSI drives can mostly keep his current dual CPU system choked at the CPU, but if we put in one of the faster CPU quads or other system it will languish waiting for the disk.
The use of solid-state drives, even as small as 4 Gigs (although I'd probably go for 4-8 of them to increase the size and throughput by RAID0) would keep him in his chair more, and more than pay for the cost in increased productivity.
Of course if you look at this from the point of view of a server instead of a workstation, the economic reasoning may be easier.
One system has a very large database - but the fact that rsync will pick the changes out of the center of a file makes the process fast even though some of the files are Gigs in size.
This (rsync + cp -al to allow for incrementals) is by far the best way I've found for backing up either large amounts of files or large files that have changes somewhere in them but not necessarily everywhere in them.
The backup servers are RAID-5 with hot spares and mdmonitor turned on plus NAGIOS watching too.
Note that the first copy can take lots of time if the connection is slow (I have one that is still running after a week at about 50Bytes/sec to allow their VOIP to continue to work) but once it is done, the only things that are copied are changes.
do this for each of the similar systems then with a script do a cp -al to a numbered directory. Delete the highest number and move the lower ones up. This assumes you don't have the ability to mount locally via smbmount or nfsmount.In other words, this is really just another form of biometrics.
Of course the fact that you have a gallon of phlegm in your lungs, a hoarse voice from barfing and fever induced hiccups today because you've just caught avian flu may make paying for your medical visit a bit problematic.
university gives students recording device
university either has something in place that allows them to have lectures recorded, or allows them to allow others to record.
No problem. Either that or the university's lawyers should be flogged.