The best advice I can give you is to get someone who knows this stuff. A lot of the information out there (including a lot of the other responses to this post) contain just plain bad information.
The most important piece of the puzzle was left off of this question. At what speed is the connectivity needed? Another question would be what level of reliablity.
First of all, let me qualify the following information with the fact that water changes things. A lot. Especially at frequencies up in the ghz, where water is essentially a mirror. That said, running a link across this much water isn't impossible, it just takes some engineering work.
The first piece of information I want to provide is that you can do a 2.4ghz 802.11b link at 11mhz over this distance with the right antenna system without even breaking a sweat. There is sufficient link margin for 99.99+% uptime. I have links running over 20 miles without any problems at all. You should also have almost no interference problems in the environment I envision. Again water will be an issue, but you should be able to work around that. This would provide a 11mb/s radio rate (About 5-8mb/s usable) link. This would be cheap. About $2000 an end when you get done putting together all the pieces. Towers not included.
Using different equipment, I can go up to 15 miles (24 km) at 93Mb/s. Assuming this is 10km or less, I can also go up to 7 miles (11km) at 420Mb/s (Full Duplex, 840mb/s aggregate). Of course these are much much much more expensive.
All of the above options are license-exempt, which basically means that at least in the US you don't need a license. I think that Canada is in the same boat, but I would have to do some more research. There are also licensed options.
To hopefully drive home the point above. You need to find someone qualified to do the engineering on this. It is NOT as easy as it looks. I find that most "experts" on this either are in the category of "I've done a couple of these and they work" or "I couldn't make them work". You need to find someone who understands the inner workings of why these work or don't. If they don't know about path loss calculations, freznel zones, etc. etc. etc. then they're probably not going to be any good to you.
I normally don't try to market my services when I post. However, if you can't find someone you feel is qualified or if you'd like me to look over someone elses design, drop me an email.
I hand-wired an aftermarket keyboard to my Timex/Sinclair 1000 (Commerically assembled ZX81). I soon learned that there was another reason for the membrane keyboard - I could type faster than the computer could scan the keyboard.
To try to add something relevant to the original discussion: If I was the original poster, I would go to a used computer store and try out everything they had till I found one I liked.
STRONG BUY ALERT FOR PHARMECUTICAL COMPANIES
on
Mir Deathwatch
·
· Score: 1
We have issued a strong buy alert for all Pharmecutical Companies around the world, and especially those with experience with producing pharmaceuticals which have a strong effect on fungus. We predict that over the next few months, a fungus from space will cause much grief for everyone around the world. Thus, the stock for these companies are bound to rise in value.
We also issue a strong buy alert for that company which makes Lysol, and also for Clorox, as these have proven themselves useful in curing other types of fungal problems.
Any resemblance of this message to an actual piece of spam, which either exists or has been deleted is strictly intended.
If you manage to get through the whole thing without meeting someone technical, that's a bad sign!
Not Necessarily. Maybe the right job for a person is to be the sole tech guy in a smaller office (10 or so people). There is something to be said for being able to run the IT show anyway you want to. There are quite a few of these jobs out there where they realize they need a full time person because their consultants can't give them the help they need.
I'm trying to envision how you could do this in a reasonably fast manner on a very large database. From what I know about regular expressions, they can be VERY cpu intensive even on a small file. For instance, how would you match:
somewhere.*over.*there
Across an entire internet sized search engine? I guess you could pre-select documents containing somewhere and over and there and then proceed with screening them through a "standard" regexp search.
However, doing the prelimiary match using the regexp would definately be resource-prohibitive. In the above example, you would have to read the text of each file in to do the regexp. Not to mention the cost of keeping the text around.
That said, I can see how you could implement a regexp-like front end to a search tool if you had some restrictions as to what you could do with the regular expressions. However, I suspect the idea was more to be able to do advanced conditionals and other funky stuff within the regular expressions, and limiting this would probably limit the usefullness of the product.
So, maybe to summarize my rambling, the initial hurdle would be to re-invent the way normal regexps work in order to be efficient in a multi-giabyte database.
It used to be common folklore that the (cough) more interesting (cough cough) alt newsgroup names were a result of news server administrators having one too many on friday nights.
I'm not sure what you meant by cheap, but I've used something like this mixer which is available from Radio Shack. There are always few listed at eBay.
I would recommend finding one with a real slider for each audio channel. The radio shack one has four line level inputs, each with a slider. Some mixers might have like 6 inputs but they share them between 2 sliders. This won't work all that well for what you want.
Most of these have a headphone level output jack, so you can use headphones instead of speakers, or both. I've been thinking about getting one of the plantronics dual-use headsets which let me use the same headset for phone and computer use.
One word of warning - most computer speakers will require a headphone-level output to work. They won't work on the standard line-level output on the back of these mixers. Actually that's not 100% accurate - they will work but might not be able to get all that loud - which might be exactly what you want.
I used to successfully destroy a mouse about every 2-3 months. It was kinda like this....
1) Go buy the cheapest mouse you could find.
2) Accidentally drop and/or just simply use mouse for 2-3 months.
3) Determine mouse is in fact dead, usually at the worst possible time.
4) Repeat #1
Finally I got pissed and "splurged" and bought the $20 logitech. This was at least 8 years ago and that mouse is still going strong. It has been through at least 6 moves (being thrown in boxes with other computer crap) and has outlived numerous MOUSE PADS (they just get too caked with that wonderful grungy dirt)....
The one on my desk (a different mouse) gets it's "tail" yanked fairly regularly as when I move the computer around, it often gets tangled with the other cables. I just grab the mouse and pull - although fairly gently - to free additional cable.
I have at various times on other machines ended up with other varieties, including expensive microsoft mice, and they have all died within a year or so. The microsoft ones tend to die when the grunge builds up on the rollers and they're almost impossible to get clean once that occurs.
I can't however vouch for any of the optical varieties. I suspect they would solve some of the physical roller problems. I would be buying a logitech one though if I bought one.
Oh, did I mention that I really Like Logitech mice?
There is one other point which I don't think anyone else hit on.
I've been on the "employer" side of several things similar to this. The real problem is trust. In my environment, letting people do real work requires in most cases letting the people have root level access to several unix servers.
I'm NOT going to let just anyone off the street have those. In fact, usually I don't give them to an employee until they have been there some time and then only if it is really necessary.
As a result, the first few months are usually what at least we consider grunt work - or more specifically, tasks which don't involve fiddling with the systems at a root-level access.
I suspect in an engineering environment this is similar. You don't want people you have there only a short period of time to be fiddling with pieces of your business which is critical to it's functioning.
Imagine a surgeon in a hospital sending their interns into the operating room to do surgery on patients. Not going to happen.
The only option to let these people do meaningful work is to let them do the work, with a supervisor breathing down their necks. This is going to take longer than it would have if the supervisor would have just done it themselves, and is going to be more costly than if you didn't have the intern. Especially if you are paying the intern anything.
If you were hiring the employee long-term this would be considered training expense, because after a while, the supervisor can let the employee go on on their own.
About the only thing I could think of which might be useful company-wise to use an intern for in the context of the original question is to team up each intern with an engineer, and while you probably can't let the intern do the actual engineering, the intern can do the engineer's "grunt work" such as copying, etc, and the intern will most likely learn a LOT from the engineer they are teamed with.
From what I read from the article itself is that what they are trying to create is a forum that the People who are responsible for fixing the bugs can openly discuss the exploits and fix them before it becomes public knowledge. I doubt they would say that it was wise to move to a "closed source" model.
In fact, Paul said that they were going to continue to distribute code in the same way they always had - which I take to mean including source that anyone can look at. I doubt there will be much change in the way 99.9% of the community interreacts with Bind and ISC.
Today, what happens is that someone finds a bug, ISC gets told about it, and then they FIX IT IN PUBLIC, while thousands of script kiddies run around and try the exploit on machines that don't even have a patch available yet. Even Microsoft is usally given the opportunity to fix major security bugs prior to them being posted publically on Bugtraq. Why should ISC be any different?
From the bugtraq faq:
A sensible protocol to follow while reporting a security vulnerability is as follows:
Contact the product's vendor or maintainer and give them a one week period to respond. If they don't respond post to the list.
If you do hear from the vendor give them what you consider appropriate time to fix the vulnerability. This will depend on the vulnerability and the product. It's up to you to make and estimate. If they don't respond in time post to the list.
If the contact you asking for more time consider extending the deadline in good faith. If they continually fail to meet the deadline post to the list.
Now, assuming that people DO this, where today can ISC discuss and fix the bug in private? I think having a list with only selected individuals on - specifically the ones Vixie mentioned, including the people who maintain bind for the major Open Source Distributions (FreeBSD, Linux, etc.) - where they can prepare fixes prior to CERT releasing the information and (hopefully) prior to it being seen on bugtraq.
I'm not sure the reason behind the fees. Perhaps it is to help pay for the additional coordination required for doing this. Perhaps it's just to keep the script kiddies out.
I'm also not sure about the private access to the CVS database. This is the only part of this that concerns me. I can think of several reasons why this might be a good idea. Does anyone know if you can get a hold of the CVS database today?
BTW forget the ID checks at airports. If my sister was able to get a fake ID at 16 to go drinking; I think that a terrorist could get one.
I recently attended a traning session in Ohio with an acquantance. His drivers license had expired and so the day before he went to the DMV and got it renewed and basically ended up with a piece of paper which showed he had renewed.
His ID was checked at least 4 times. When picking up the tickets both ways - and when renting a car. PLUS, we visited NASA Glenn and the security guard there looked at it also.
Not one of the people who looked at the ID saw that it was expired. I can understand the airport people and perhaps NASA, as they probably didn't really care if he had driving privledges but THE RENTAL CAR AGENCY?
In any case, to try to salvage some on-topicness from this post, I fully agree that the security systems in place are trivial to "bypass". For instance, I carried onto the plane a laptop case containing the laptop, 3 pcmcia cards, an entire ethernet hub, a whole truckload of cables, and a whole bunch of other stuff I figured the people looking at the X-Ray machine would have freaked over, but they didn't even make me turn it on.
Then there's the other extreme. The same acquantance as above walked into a Federal Building to deliver a replacement monitor. For "security" reasons there are no parking spaces near the building and so he basically illegally parked, took the monitor in, talked to the security guard, made sure it was ok to leave the monitor box by the front door and then went and legally parked his vehicle.
By the time he got back, the guards had changed and the new guard on duty was trying to stuff the box through the xray machine - tearing the box to try to make it fit. As they needed the box to ship the old monitor back, he asked them what they were doing. When they figured out it was his box, they pushed him up against the wall and basically interrogated him. Of course, he wasn't a very happy camper. The excuse they gave was that "ever since the Oklahoma? bombing they are a little touchy".
Then, when I was in San Francisco for a contract, I see a ryder truck parked next to the federal building for about 2-3 hours.
Both FreeBSD and Linux (and others) use the HLT instruction when idle to reduce CPU power consumption.
When the HLT instruction is executed, the CPU sleeps until it gets an interrupt - usually a timer tick or an i/o of some sort.
I know that anything which increases the load on a FreeBSD system or changes the duty cycle of HLT causes the CPU temperature to change. It is amazing how a slight load change on these systems can overload marginally-adequate cooling systems and cause CPU problems.
So to tie this back to the original question - some servers actually go into a "micro" sleep mode - hundreds of times a minute.
I guess I should have thought twice before posting. If you find a pm2er (note the r) it can also act as a T1 router which would eliminate the cisco, if you're really being cheap.
The other piece I left out is that you'll need a T1 CSU/DSU. This will be about $500 or so used.
I've run pops without the cisco before and they work, but I usually like to have a cisco somewhere to be able to have knobs to turn to fix weird problems. But again that's me.
Why don't you start a real ISP and then you could provide the service below cost?:)
Depending on how small you are, you could really just do this with a Cisco 2509 or 2511 - 8 or 16 asyncronous ports for modems and a couple sync ports for upstream and an ethernet port. Plus external modems (Recommend Courier V.Evertyhing), of course.
PErsonally, I'd probably be looking for a Cisco 2501 plus a Livingston portmaster instead of the 2511.
It looks like the 2511 (16 ports) are going for between $1000-$1500. 2501's are about $700, Portmasters are under $300.
You can authenticate using the 2511 or Portmaster without using any authentication server. You can probably get away with using your upstream's DNS server, and for email you can probably get away with either outsourcing or telling the users to get a free web-based solution.
If you'd like to a FreeBSD (or I guess I better also say Linux) box to the mix and you can then run your own DNS, authenticate using Radius, run your own mail and web servers, among others.. For this few users, you don't need much more than a mid-range pentium or low-end pentium II with an IDE disk.
Generally, getting into the V.90 game requires a significantly larger investment. You are probably talking about buying equipment which will handle 24 simultaneous users at a pop and you'll need to buy a T carrier circuit from the telco. And, if you're in a rural area, V.90 a lot of times performs worse than a good V.everything modem at v.34 speeds.
You should also verify your upstream cost. The circuit alone could cost you thousands a month.
Guess/var/spool/news must really "abuse the filesystem" then. Odd that in nearly 20 years noone has come up with an alternative.
Not so....From the INN (Internet News Faq):
CNFS is the Cyclic News File System, written by Scott Fritchie. It is a high-performance method of storing news articles, designed to avoid the high overhead involved in interacting with the file system when storing articles in individual files. From INN's INSTALL file:
CNFS stores articles sequentially in pre-configured buffer files. When the end of the buffer is reached, new articles are stored from the beginning of the buffer, overwriting older articles.
INN can either store it the "traditional" way or as described above, in a rotating spool file. Typhoon, Cyclone, and Breeze (from bCandid, formerly Highwind) all use a similar format to this. This WORKS for news. This WOULDN'T WORK for mail, unless you want a "most recent x bytes of mail" type of storage. Diablo does something similar, although I am not a diablo expert by any means.
The url is www.wavelan.com.
There may be another one also since there are actually two parts of lucent doing the 802.11 stuff.
My understanding is that the airport is actually one of the Wavelan RG-1000's that has been relabeled, but I don't know that for sure.
If this is the lucent product, then my understanding about the way lucent handles under-warranty repairs is consistent with the apple policy - They don't repair them, they replace them. I suspect it's cheaper to just replace the hardware (since it is obviously not very expensive) than to pay someone to figure out what in the (#*@$ blew up.
There was an article in the November QST about using Meteor scatter for Routine Communications. From the sounds of the article, that with proper operations that you can pretty much use meteor scatter year round.
"Hi I'm Forrest with Montana Internet Corporation. I was wondering who I needed to talk to about a routing problem which is affecting one of my customers."
I think this immediately clarifies that they are talking to someone other than the end user and are much more likely to call someone if they can't figure it out.
This reminds me of a story that the owner of a house we have a wireless repeater on told us.
When he was building his house USWest charged him an outrageous figure for the install for his phone line since he was quite a ways from the nearest telephone line. I think it was nearly if not over 5 figures.
So, one day he gets contacted by USWest. It seems that they want to drag a Fiber across his property. He basically said to them, "Well, back when I put a phone line into my house you charged me x for the install. That's what it's going to cost you to drag that across my property." Evidentally they paid up.
By the way, we are swapping out service for the use of their house as a repeater location. And seeing as we have a 25mb/s pipe to them to service the repeater, they are getting really GOOD service too... Fiber is a different can of worms though, as there isn't exactly an ethernet port handy, and singlemode fiber is a real pain to splice and terminate.
being able to suspend bits of metal in the air over some stove burner looking thingy
The things they were suspending were actually magnets. If you drop a magnet on top of a superconductor, it won't get any closer than a certain distance due to the fact that if you move a conductor through a magnetic field (or move a magnetic field "through" a conductor) it generates electricity. Since there is no resistance to electricity - and thus, no way for the superconductor to eliminate the potential energy, the magnet will essentially float above the superconductor.
The stove burner was really some sort of cooler with a superconducting wafer on top.
(almost?) Every MRI machine in the country uses superconductors to produce the magnetic field. Basically, they charge up the superconducting wire and then connect the ends together.
One of my acquantances up here services the things. It's really fun when he tries to sell a zero volt 1000 amp power supply to someone else and they say that it is impossible to have current without voltage. But since you can have current in a conductor with no resistance, you can also have lots of current and no voltage.
There are some other real applications for this stuff. The revolution was being able to do this at a much higher temperature than say the 1 degree above absolute zero discussed above.
Slashdot Mirror
The most important piece of the puzzle was left off of this question. At what speed is the connectivity needed? Another question would be what level of reliablity.
First of all, let me qualify the following information with the fact that water changes things. A lot. Especially at frequencies up in the ghz, where water is essentially a mirror. That said, running a link across this much water isn't impossible, it just takes some engineering work.
The first piece of information I want to provide is that you can do a 2.4ghz 802.11b link at 11mhz over this distance with the right antenna system without even breaking a sweat. There is sufficient link margin for 99.99+% uptime. I have links running over 20 miles without any problems at all. You should also have almost no interference problems in the environment I envision. Again water will be an issue, but you should be able to work around that. This would provide a 11mb/s radio rate (About 5-8mb/s usable) link. This would be cheap. About $2000 an end when you get done putting together all the pieces. Towers not included.
Using different equipment, I can go up to 15 miles (24 km) at 93Mb/s. Assuming this is 10km or less, I can also go up to 7 miles (11km) at 420Mb/s (Full Duplex, 840mb/s aggregate). Of course these are much much much more expensive.
All of the above options are license-exempt, which basically means that at least in the US you don't need a license. I think that Canada is in the same boat, but I would have to do some more research. There are also licensed options.
To hopefully drive home the point above. You need to find someone qualified to do the engineering on this. It is NOT as easy as it looks. I find that most "experts" on this either are in the category of "I've done a couple of these and they work" or "I couldn't make them work". You need to find someone who understands the inner workings of why these work or don't. If they don't know about path loss calculations, freznel zones, etc. etc. etc. then they're probably not going to be any good to you.
I normally don't try to market my services when I post. However, if you can't find someone you feel is qualified or if you'd like me to look over someone elses design, drop me an email.
To try to add something relevant to the original discussion: If I was the original poster, I would go to a used computer store and try out everything they had till I found one I liked.
We also issue a strong buy alert for that company which makes Lysol, and also for Clorox, as these have proven themselves useful in curing other types of fungal problems.
Any resemblance of this message to an actual piece of spam, which either exists or has been deleted is strictly intended.
Not Necessarily. Maybe the right job for a person is to be the sole tech guy in a smaller office (10 or so people). There is something to be said for being able to run the IT show anyway you want to. There are quite a few of these jobs out there where they realize they need a full time person because their consultants can't give them the help they need.
[JjFf][ae][nb]\s*[0-9]{1,2},{0,1}\s*[0-9]{2,4}
This gets really messy really fast and exactly *HOW* are you going to do a query on a keyed database using this?
No!!!! Please, anywhere but Montana, we have enough loonies up here already (myself included.).
somewhere.*over.*there
Across an entire internet sized search engine? I guess you could pre-select documents containing somewhere and over and there and then proceed with screening them through a "standard" regexp search.
However, doing the prelimiary match using the regexp would definately be resource-prohibitive. In the above example, you would have to read the text of each file in to do the regexp. Not to mention the cost of keeping the text around.
That said, I can see how you could implement a regexp-like front end to a search tool if you had some restrictions as to what you could do with the regular expressions. However, I suspect the idea was more to be able to do advanced conditionals and other funky stuff within the regular expressions, and limiting this would probably limit the usefullness of the product.
So, maybe to summarize my rambling, the initial hurdle would be to re-invent the way normal regexps work in order to be efficient in a multi-giabyte database.
It used to be common folklore that the (cough) more interesting (cough cough) alt newsgroup names were a result of news server administrators having one too many on friday nights.
Seems strangley fitting for some reason...
If you don't get it you obviously haven't seen the boxes their computers come in....
I would recommend finding one with a real slider for each audio channel. The radio shack one has four line level inputs, each with a slider. Some mixers might have like 6 inputs but they share them between 2 sliders. This won't work all that well for what you want.
Most of these have a headphone level output jack, so you can use headphones instead of speakers, or both. I've been thinking about getting one of the plantronics dual-use headsets which let me use the same headset for phone and computer use.
One word of warning - most computer speakers will require a headphone-level output to work. They won't work on the standard line-level output on the back of these mixers. Actually that's not 100% accurate - they will work but might not be able to get all that loud - which might be exactly what you want.
I will NEVER buy a non-logitech mouse ever again.
I used to successfully destroy a mouse about every 2-3 months. It was kinda like this....
1) Go buy the cheapest mouse you could find.
2) Accidentally drop and/or just simply use mouse for 2-3 months.
3) Determine mouse is in fact dead, usually at the worst possible time.
4) Repeat #1
Finally I got pissed and "splurged" and bought the $20 logitech. This was at least 8 years ago and that mouse is still going strong. It has been through at least 6 moves (being thrown in boxes with other computer crap) and has outlived numerous MOUSE PADS (they just get too caked with that wonderful grungy dirt)....
The one on my desk (a different mouse) gets it's "tail" yanked fairly regularly as when I move the computer around, it often gets tangled with the other cables. I just grab the mouse and pull - although fairly gently - to free additional cable.
I have at various times on other machines ended up with other varieties, including expensive microsoft mice, and they have all died within a year or so. The microsoft ones tend to die when the grunge builds up on the rollers and they're almost impossible to get clean once that occurs.
I can't however vouch for any of the optical varieties. I suspect they would solve some of the physical roller problems. I would be buying a logitech one though if I bought one.
Oh, did I mention that I really Like Logitech mice?
There is one other point which I don't think anyone else hit on.
I've been on the "employer" side of several things similar to this. The real problem is trust. In my environment, letting people do real work requires in most cases letting the people have root level access to several unix servers.I'm NOT going to let just anyone off the street have those. In fact, usually I don't give them to an employee until they have been there some time and then only if it is really necessary.
As a result, the first few months are usually what at least we consider grunt work - or more specifically, tasks which don't involve fiddling with the systems at a root-level access.
I suspect in an engineering environment this is similar. You don't want people you have there only a short period of time to be fiddling with pieces of your business which is critical to it's functioning.
Imagine a surgeon in a hospital sending their interns into the operating room to do surgery on patients. Not going to happen.
The only option to let these people do meaningful work is to let them do the work, with a supervisor breathing down their necks. This is going to take longer than it would have if the supervisor would have just done it themselves, and is going to be more costly than if you didn't have the intern. Especially if you are paying the intern anything.
If you were hiring the employee long-term this would be considered training expense, because after a while, the supervisor can let the employee go on on their own.
About the only thing I could think of which might be useful company-wise to use an intern for in the context of the original question is to team up each intern with an engineer, and while you probably can't let the intern do the actual engineering, the intern can do the engineer's "grunt work" such as copying, etc, and the intern will most likely learn a LOT from the engineer they are teamed with.
In fact, Paul said that they were going to continue to distribute code in the same way they always had - which I take to mean including source that anyone can look at. I doubt there will be much change in the way 99.9% of the community interreacts with Bind and ISC.
Today, what happens is that someone finds a bug, ISC gets told about it, and then they FIX IT IN PUBLIC, while thousands of script kiddies run around and try the exploit on machines that don't even have a patch available yet. Even Microsoft is usally given the opportunity to fix major security bugs prior to them being posted publically on Bugtraq. Why should ISC be any different?
From the bugtraq faq:
Now, assuming that people DO this, where today can ISC discuss and fix the bug in private? I think having a list with only selected individuals on - specifically the ones Vixie mentioned, including the people who maintain bind for the major Open Source Distributions (FreeBSD, Linux, etc.) - where they can prepare fixes prior to CERT releasing the information and (hopefully) prior to it being seen on bugtraq.I'm not sure the reason behind the fees. Perhaps it is to help pay for the additional coordination required for doing this. Perhaps it's just to keep the script kiddies out.
I'm also not sure about the private access to the CVS database. This is the only part of this that concerns me. I can think of several reasons why this might be a good idea. Does anyone know if you can get a hold of the CVS database today?
I recently attended a traning session in Ohio with an acquantance. His drivers license had expired and so the day before he went to the DMV and got it renewed and basically ended up with a piece of paper which showed he had renewed.
His ID was checked at least 4 times. When picking up the tickets both ways - and when renting a car. PLUS, we visited NASA Glenn and the security guard there looked at it also.
Not one of the people who looked at the ID saw that it was expired. I can understand the airport people and perhaps NASA, as they probably didn't really care if he had driving privledges but THE RENTAL CAR AGENCY?
In any case, to try to salvage some on-topicness from this post, I fully agree that the security systems in place are trivial to "bypass". For instance, I carried onto the plane a laptop case containing the laptop, 3 pcmcia cards, an entire ethernet hub, a whole truckload of cables, and a whole bunch of other stuff I figured the people looking at the X-Ray machine would have freaked over, but they didn't even make me turn it on.
Then there's the other extreme. The same acquantance as above walked into a Federal Building to deliver a replacement monitor. For "security" reasons there are no parking spaces near the building and so he basically illegally parked, took the monitor in, talked to the security guard, made sure it was ok to leave the monitor box by the front door and then went and legally parked his vehicle.
By the time he got back, the guards had changed and the new guard on duty was trying to stuff the box through the xray machine - tearing the box to try to make it fit. As they needed the box to ship the old monitor back, he asked them what they were doing. When they figured out it was his box, they pushed him up against the wall and basically interrogated him. Of course, he wasn't a very happy camper. The excuse they gave was that "ever since the Oklahoma? bombing they are a little touchy".
Then, when I was in San Francisco for a contract, I see a ryder truck parked next to the federal building for about 2-3 hours.
Go figure.
When the HLT instruction is executed, the CPU sleeps until it gets an interrupt - usually a timer tick or an i/o of some sort.
I know that anything which increases the load on a FreeBSD system or changes the duty cycle of HLT causes the CPU temperature to change. It is amazing how a slight load change on these systems can overload marginally-adequate cooling systems and cause CPU problems.
So to tie this back to the original question - some servers actually go into a "micro" sleep mode - hundreds of times a minute.
The other piece I left out is that you'll need a T1 CSU/DSU. This will be about $500 or so used.
I've run pops without the cisco before and they work, but I usually like to have a cisco somewhere to be able to have knobs to turn to fix weird problems. But again that's me.
Depending on how small you are, you could really just do this with a Cisco 2509 or 2511 - 8 or 16 asyncronous ports for modems and a couple sync ports for upstream and an ethernet port. Plus external modems (Recommend Courier V.Evertyhing), of course.
PErsonally, I'd probably be looking for a Cisco 2501 plus a Livingston portmaster instead of the 2511.
It looks like the 2511 (16 ports) are going for between $1000-$1500. 2501's are about $700, Portmasters are under $300.
You can authenticate using the 2511 or Portmaster without using any authentication server. You can probably get away with using your upstream's DNS server, and for email you can probably get away with either outsourcing or telling the users to get a free web-based solution.
If you'd like to a FreeBSD (or I guess I better also say Linux) box to the mix and you can then run your own DNS, authenticate using Radius, run your own mail and web servers, among others.. For this few users, you don't need much more than a mid-range pentium or low-end pentium II with an IDE disk.
Generally, getting into the V.90 game requires a significantly larger investment. You are probably talking about buying equipment which will handle 24 simultaneous users at a pop and you'll need to buy a T carrier circuit from the telco. And, if you're in a rural area, V.90 a lot of times performs worse than a good V.everything modem at v.34 speeds.
You should also verify your upstream cost. The circuit alone could cost you thousands a month.
Not so....From the INN (Internet News Faq):
CNFS is the Cyclic News File System, written by Scott Fritchie. It is a high-performance method of storing news articles, designed to avoid the high overhead involved in interacting with the file system when storing articles in individual files. From INN's INSTALL file:
CNFS stores articles sequentially in pre-configured buffer files. When the end of the buffer is reached, new articles are stored from the beginning of the buffer, overwriting older articles.
INN can either store it the "traditional" way or as described above, in a rotating spool file. Typhoon, Cyclone, and Breeze (from bCandid, formerly Highwind) all use a similar format to this. This WORKS for news. This WOULDN'T WORK for mail, unless you want a "most recent x bytes of mail" type of storage. Diablo does something similar, although I am not a diablo expert by any means.
My understanding is that the airport is actually one of the Wavelan RG-1000's that has been relabeled, but I don't know that for sure.
If this is the lucent product, then my understanding about the way lucent handles under-warranty repairs is consistent with the apple policy - They don't repair them, they replace them. I suspect it's cheaper to just replace the hardware (since it is obviously not very expensive) than to pay someone to figure out what in the (#*@$ blew up.
There was an article in the November QST about using Meteor scatter for Routine Communications. From the sounds of the article, that with proper operations that you can pretty much use meteor scatter year round.
"Hi I'm Forrest with Montana Internet Corporation. I was wondering who I needed to talk to about a routing problem which is affecting one of my customers."
I think this immediately clarifies that they are talking to someone other than the end user and are much more likely to call someone if they can't figure it out.
SunOS 5.x (Solaris 2...) is SysV derived. It's a real pain to have to remember 'ps -ef' on the one remaining solaris box I administer.
This reminds me of a story that the owner of a house we have a wireless repeater on told us.
When he was building his house USWest charged him an outrageous figure for the install for his phone line since he was quite a ways from the nearest telephone line. I think it was nearly if not over 5 figures.
So, one day he gets contacted by USWest. It seems that they want to drag a Fiber across his property. He basically said to them, "Well, back when I put a phone line into my house you charged me x for the install. That's what it's going to cost you to drag that across my property." Evidentally they paid up.
By the way, we are swapping out service for the use of their house as a repeater location. And seeing as we have a 25mb/s pipe to them to service the repeater, they are getting really GOOD service too... Fiber is a different can of worms though, as there isn't exactly an ethernet port handy, and singlemode fiber is a real pain to splice and terminate.
being able to suspend bits of metal in the air over some stove burner looking thingy
The things they were suspending were actually magnets. If you drop a magnet on top of a superconductor, it won't get any closer than a certain distance due to the fact that if you move a conductor through a magnetic field (or move a magnetic field "through" a conductor) it generates electricity. Since there is no resistance to electricity - and thus, no way for the superconductor to eliminate the potential energy, the magnet will essentially float above the superconductor.
The stove burner was really some sort of cooler with a superconducting wafer on top.
One of my acquantances up here services the things. It's really fun when he tries to sell a zero volt 1000 amp power supply to someone else and they say that it is impossible to have current without voltage. But since you can have current in a conductor with no resistance, you can also have lots of current and no voltage.
There are some other real applications for this stuff. The revolution was being able to do this at a much higher temperature than say the 1 degree above absolute zero discussed above.