My girlfriend is a type 1 diabetic. Instead of regular injections, she uses an insulin pump. This pump is an external device, about the size of a pager, that feeds insulin into her body via a short tube.
Several months ago she upgraded to a new pump. This new model (a Medtronic MiniMed) wirelessly communicates with a number of devices. It receives blood glucose data from a continuous glucose monitor. It also receives her regular readings from her standard "prick your finger" blood sugar tests via her test kit. And, it has a wireless key fob that allows her to adjust the pumps settings without having to dig through pockets and clothes to get at the unit.
My first comment to her was "With all of this wireless control, how easy is it for someone to use this wireless interface to put you into a diabetic coma, or worse, kill you?" She thinks it's a fairly ridiculous concept, citing encryption, receiver range, and "Why would anyone want to kill me?", among other reasons.
Well, I say that anything that has any type of wireless interface is hackable. There are, of course, no published documents that I can find detailing what steps have been taken to secure these devices. I'm seriously concerned as to whether or not the companies that make insulin pumps, pace makers, implants, etc, may not be taking these concerns seriously.
One night after having dinner with a few friends, one of the group noted that she needed to stop by a Verizon store to get a new phone. Hers had been having troubles and she was eligible for a free upgrade (with two year contract extension, of course...)
We arrived at the Verizon store, and the representative pulled out a few different models from Motorola, LG, and Samsung for her to look at. I casually pulled my iPhone out of my pocket and sat it in the mix. The Verizon rep exclaimed "OH, is that an iPhone?!", grabbed it, and spent a few minutes playing with it. His breach of salesperson protocol must have snuck into his head, as his face changed to one of concern. He put it down, and started talking about how it doesn't have 3G like Verizon's smart phones, and how the lack of replaceable battery made every other phone on the table a much better choice.
I walked away with a grin on my face, knowing that even the Verizon rep was excited by a phone only available through another carrier.
I remember thinking "when will I ever need this" through algebra and trig during high school... More often than I thought.
Here are two real-world questions I've had to solve with math over the past few years. I'd start by learning to answer them, and then identify other problems to solve! Yes, there are a lot of people who agree with me, but I didn't see too many examples.
1) In a previous life I co-founded a wireless ISP. I'd often need to calculate how tall a tower was, how high up on the tower an antenna was, and where the signal from said antenna would reach the ground. So, find a cellular or some other tower and figure out how tall it is. Then, find an antenna array on the tower (sometimes at the top, sometimes not). If you can't tell, just pick something obvious and assume it's an antenna.:) Calculate its height.
Now, assume that the antenna has a 30 degree vertical beamwidth. At what distance would the "beam" reach the ground. This calculates what I call the "umbrella effect", or in other words, the area where you are actually too CLOSE to a tower to be within it's coverage area.
(Most carriers will angle their sectors downwards as to not waste beamwidth going up into the atmosphere and maximize coverage, but for the purposes of calculation assume that the center of the "beam" is perfectly horizontal.)
2) Find a swimming pool and figure out it's depth. Then, figure out how much water is in it. Calculate how much water would need to be added to increase the water level by one foot.
I know the above doesn't really tell you how to learn, but hopefully having some interesting real life questions to answer might help. Of course, you might not find the above problems remotely interesting. If that's the case, I apologize for wasting time and bits!
Actually, if you run Apache in a chroot, this just means that the attacker will install their phishing sites somewhere within the chroot, and Apache will still serve them out.
Same with suexec. Regardless of which user the exploit code is owned by, it can still be served out happily by Apache. Suexec certainly does make it easy to see which website the attackers came in via though, instead of everything being owned by the apache or nobody user.
Now, if you would have said "tighten up php.ini", I might have agreed.
The company I work for performs emergency Linux support services. We get a lot of calls from peoples boxes who are attacked. I've seen at least two eBay/PayPal phishing sites recently. In both cases, it had nothing at all to do with Linux itself.
Case #1: Customer running a web server had vulnerable PHP applications (I believe it was an outdated WordPress). Someone was able to use this vulnerability to wget a few php scripts and bury them under some subfolders.
Case #2: Customer had a non-root account with a weak password. This account was in the "root" group, giving it write access to a number of system files. Cracker was able to brute force the password quite easily, make a directory called eBay under/var/www/html, and stick some php code in there.
In both cases, the php scripts were logging username and password guesses into a text file. The text file was within the same web root, allowing the cracker to easily grab the latest passwords over http instead of needing to re-crack. Also, in both cases, there were at least a dozen usernames and passwords in the text files.
The lesson: Keep your web apps up to date, use strong passwords, and don't add anyone to the root group.
I use Firefox on Mac (intel) and Windows, with the latest versions on both. I can have Firefox open for a full week on Windows without any problems, however on either Mac I have to restart Firefox about once every day or two, otherwise browsing slows to a crawl. At extremes the whole machine will start to bog down until I "force quit" (kill -9) Firefox. I'll also experience oddness where images will just stop loading.
Running "bare bones" on all Firefox installs, no plugins other than whatever may have been included with the base distribution.
Does anyone else notice this? I've switched back to Safari on the Mac in the meantime.
As a lot of people have noted here, how a sysadmin's job is usually very transparent because they are doing their job when things are not broken. Sometimes there are servers to install, updates to apply, etc, and those are easy enough to track. However I think uptime is the #1 way to show that people are doing their job.
In my case, I prepare regular reports for the PHBs showing helpdesk requests, helpdesk response time, backup results, SMTP traffic, and most importantly, a very large month-by-month uptime graph on the first page.
In my case, at salary review time, I can plot our uptime percentage from a few months prior to me "taking over" to present, draw an arrow showing where I took over, and show that our uptime percentage has gone up significantly since then. The data can be just as valuable to you as it is for the PHB.
Toshiba has also been pushing some deals and promotions on their entry-level HD-DVD player. As an example, when I recently purchased an LCD HDTV, I had the option of buying an HD-A2 for $0.01. Of course I accepted.
And no, this wasn't a "build the player cost into the TV" trick. The price I paid was $500 less than MSRP and about on-par with what the major online sites were selling the set for.
I personally think the format war is really hurting both camps. Just look at some past format wars...
SACD vs DVD-A? Neither, both are dead.
DVD+r/-r? Neither, both are equally supported after many years.
56Flex vs X2? Neither, we now have v.90 and of course broadband.
Having an open source CA is one thing. Having the root certificate included in major browsers is an expensive endeavor. The www.cacert.org site has an FAQ entry about this:
Summary: Lots of open source browsers already have the cert; Mozilla/Firefox will have it soon. Internet Explorer (and apparently Apple's Safari) won't have it unless they come up with a way to pay for the $75,000+ plus $10,000 a year for a AICPA WebTrust audit.
I have a thing for Sun Sparc 20's -- they are VERY upgradeable and extremely reliable. Two of mine have quad 150mhz Ross processors making them snappy ehough to serve out some Apache/PHP/MySQL, host a little e-mail for a few domains, and do some secondary DNS. They're small, don't use a TON of power, and just plain cool.
Oh, and they'll run Linux or a few of the BSD's just fine..
Here is the uptime of one of my production Sparc 20's hosting a bit of email and DNS:
Is it possible that GoDaddy made this change to "Unknown" in January to give Microsoft an idea of what kind of a NetCraft jump to expect if that "Unknown" became "IIS"? I mean, it did happen in suspiciously close proximity to this new GoDaddy/MS deal.
if imperfections make the ride, cedar point has plenty to offer you. might i recommend the magnum, which was their record-setting steel coaster 10-ish years ago. its age has roughened it up, and it is one fun ride.
if you have a chiropractor on speed dial, i'll recommend the mean streak - their at-the-time world setting wooden coaster. that thing is painful, but soooo much fun.
cedar point really is the coaster capital, they have something for everyone.
the millenium force, their record setter that came out before TTD, lasts more than 26 seconds, and literally approach blackout on the downhill every time.
and having ridden the "TTD" many times, i can' say it's not remotely boring.
i don't work at cp (although i worked at a fry stand there when i was 16 -- the year raptor opened). i'm just pointing out that a) 120+mph in 3 seconds up a 400+ foot almost-completely-vertical tower is hardly boring, and b) if it's not your bag, they have something for everyone......and damn if i don't love living 15 minutes away from CP:)
Have you ever had a flight where the person next to you decided to tell you their life story? You know, putting on headphones and closing your eyes doesn't seem to give them the hint?
I speak from recent experience -- perhaps without the cellphone ban, she could have annoyed someone else.:)
I was one of those geocachers!!!!!! (mattopia on geocaching.com)
We used a deep cycle battery and a power inverter. The cam was an Axis 2120 network camera. We stripped the insulation and one of the pairs from the ethernet cable, used the remaining pairs for data and a hacked power over ethernet solution for the cam.
Set the axis up to FTP an image to an iBook at the other end every 5 seconds. We ended up with over 300 images to dig through and found about 20 decent ones.
The balloons necessary to lift it were expensive. We had over $60 in helium and balloons. The Axis we had laying around, (I also run eriecam.com and tend to have extras. The Axis makes a fairly high quality camera (Sony CCD), so the images were pretty decent quality -- not like a $20 walmart webcam. Resolution was limited to 640x480 though.
It's also worth noting that we used the ethernet cable as the tether/string for the balloons too, to save the weight of having a separate string.
I should send the images to Axis to see if it qualifies as a unique application.
More pictures can be found here, including a few more of the ground setup.
The one I worked for was a bit different. Small Internet Providers throughout the country contracted us to handle their technical support for them. Since many of these were "mom and pop" operations with just a few hundred customers in one city, they relied heavily on maintaining that local image. As a result, we were NEVER, EVER allowed to give any indication that we were not located in the area the person was calling from. I remember talking to customers of a Florida ISP about how nice the weather is, when in fact I'm sitting in Toledo, Ohio (hint to the company's identity?) in a snowstorm. If we were asked for our location, we had to respond that we were not permitted to give out details on our location due to security concerns. I had to give that line a few times a day.
We also had to be crafty. Although some "premium" customers had dedicated phone numbers so that we could find out which ISP they were calling for, many of the individual ISP's calls were routed to a common toll free number, so we'd have no idea as to which of the hundreds of ISP's we do support for the caller is from. We answer the call generically ("Tech Support, how may I assist you?") and usually asked for the customer's e-mail address for an indication of which ISP they were with. The domain name would give away the ISP. Unfortunately, people often did not give the domain name, or had offsite e-mail accounts. Since we couldn't give away that we were not with "their ISP", I couldn't flat out ask. I'd have to narrow it down by area code, and then search between ISP's in that area to find out who they were with -- often taking 10-15 minutes.
I remember one time management signed a deal and gave the call center side a chance to prepare. It was a HUGE customer - larger than all of our other ISP's combined. One night, on my shift, they simply forwarded the tech support number over to us. We went from an average 3 minute call queue time to well over an hour. We did not have the staff to handle the calls, and had no information at all about the specifics of the ISP -- dialup numbers, e-mail servers, etc. It was days before we even had the correct info to give customers. In the meantime, we just had to go with it.
And finally, we had no training program at all, so the company tried to hire people from an outsourcer in the area who had already been through their hideous training program. We paid a dollar an hour more, so it was usually pretty easy to do. Unfortunately, we supported dialup customers, and the company we stole people from supported cable modems, so new hirees usually knew nothing of dialup.
I lasted about six months there surprisingly. When I started it was a small operation with only a dozen or so techs. By the time I left, they had on average 30-40 people per shift. We grew so fast that they had to temporarily build a room in the warehouse and put up folding tables to make room for the new call center people. I'm sure they are much bigger by now, but probably still working out of the warehouse.
Yes, you are correct. It's been awhile on both movies for me as well. In Ferris Bueller, he changes the attendance while Rooney is on the phone with his mom.
And along the same times, how about Matthew Broderick? Not only did he hack into WOPR for a game of global thermonuclear war in "War Games", but he also changed his grades in "Ferris Bueller's Day Off." Now that takes some talent.
We have an 8 or 9 story condo complex we do this in. On the roof, we have two radios - one to bring the connection in from our wireless WAN, and one going into a roof-mounted antenna. That antenna is specifically designed for this type of application. It's a cross polarized 180 degree panel antenna. We have a 250mw amplifier sitting between that and a Cisco 340 series AP. About 1/2 way down in one of the customers condos we have a repeater to help the guys in the lower apartments. It works very well. For your application, if you use a low power amplifer (higher power ones may be illegal based on the FCC's ISM-band regulation) and the right antenna, you can probably do this for under $1,000.
If it is easy to run cable between floors and you don't mind some significant labor, a cheap AP on each floor with a good antenna will do very well too. Make sure to pay close attention to your channel plan when installing multiple AP's. Also, never use the built in omni antennas. You can get 6dbi - 8dbi patch antennas for very cheap. Only get enough spread (120-180 degrees I imagine) to cover the area you need to. With 4 floors, I don't see you needing more than one or two access points. A lot of this depends on the building itself. If its a newer building, with the materials used, signal is likely to travel farther.
www.hyperlinktech.com has a very good selection of antennas. We get a number of the ones we use from here. Their tech staff can probably help you with layout and design as well. I don't work for them, I've just had good experience in the past with them.
Another thing to consider is the client cards. Most off the shelf cards have cheap internal antennas and are low power. The Cisco cards we use are 100mw cards.
Having a powerful card for the clients will help quite a bit too. The Cisco 350 series cards are 100mw cards, which is double or more most cheap off-the-shelf cards. You will pay a premium, but you'll have a lot less phone calls from your users about signal dropping out.
Another solution we haven't tried is to actually locate the access point and antenna outside the building, like where the dumpsters are, and focus the signal in from the side through the windows. A WISP friend of ours has done this in a few areas, and has had very good luck covering buildings much larger than yours with just two AP's and sector antennas.
Some people have suggested going with the "G" standard. Considering you are not doing anything that has super high bandwidth use, I would recommend against it. The lower the bandwidth, the better the range. Most all of the access points in our wireless system (well over 50) are running at 5.5mb, or even 2mb. You will have to test and see what works best in your system.
And finally, as many have also said, are you sure you want to do this? Three years ago, my father and I started a wireless ISP as a hobby. We never anticipated getting as big as we have. Trust me when I say it will end up being less of a hobby and more of a job. Even a small network like that will take maintenance, and you will end up doing tech support for the users in the building. If you do move forward, don't spend all of that money. Have the other users and perhaps even the landlord subsidize some of it. You are providing a marketable service that would make your building stand out over others. Don't give that additional marketing power to the landlord for free! Have EVERYTHING down on paper before starting the install. If the equipment is mounted on the landlord's building, there's no other way to prove that it is yours.
I salute you! I remember retrofitting one of those boards in a 128k mac, which not only added SCSI, but added additional memory (512k or 1mb, can't remember which). Combined it with an external 40mb SCSI drive, and I had a killer system! Even back then people were modding old stuff to have the functionality of newer systems.
I do remember it was a real pain to get installed, and we damn near ruined the system a few times. Bent the pins of the ROM all to hell.
We are a DSL provider. I know what Verizon charges us for a 768/128 DSL. If we charged $30/month for a 768k DSL circuit, we would be loosing about $8/month. This does not even take into account support costs, our link to Verizon, or our bandwidth from our upstream providers.
For a broadband ISP to make money by selling DSL, they need to either own the network themselves (ie Verizon, SBC, CLEC's, etc) or have major quantities of customers to get any type of discount from the ILEC/CLEC.
Luckily our broadband wireless lets us provide a decent broadband product at a decent price and actually make a bit of money off of it.
And when you work with high band stuff, up in the Ghz range, and you need the really BIG (1 5/8" Andrew) cable, offer to tear it out of an old cell site or other transmission tower for free.
And, if you're crafty (like my father tends to be), you can make a nice side business of selling used cable to other ham operators.
The install cost sounds about right if a tower needs to be erected.
You'd be surprised how far bandwidth can go. Most of our customers get 384k packages, many shared with quite a few machines. Our largest customer has over 130 machines sharing a 512k wireless Internet connection with no issues.
Now if you have 3 power users that chug at the p2p networks 24/7, then 256k will not go very far, especially on your outbound.
I appreciate everyone's input. However it seems most of the discussion so far has been hardware related. That end of things is fine - I have plenty of x86 boxen running Debian or FreeBSD. As the article mentions, I'm mainly looking for software solutions to provide our customers with a hosting control panel (Webmin is really not designed for this in a shared server environment) that will run on one of our existing platforms - Debian/x86 or FreeBSD/x86.
I co-own a primarily wireless ISP. One of the main reason most WISP's mainly cater to business is the equipment cost. A "good" client install generally costs several hundred dollars, and most 'residential' customers are used to paying less than 100, or anything, for the modem and install, as is usually the case with cable and dsl.
If you are behind a lot of thick trees, you will most definitely need your antenna to be above them. We have gotten by with strapping 20 foot of mast pipe to a chimney, or using existing towers. Sometimes you can get signal through the trees if they are fairly sparse, but your reliability will greatly diminish.
If you're within several miles of their tower and you can get a clear line of sight to them, and if they know what they are doing, you should have no problems whatsoever.
One thing to watch out for -- a lot of ISP's who offer wireless really don't know much about RH (the 'wireless' end of things). We have an RF engineer on staff, and without him we would be nowhere. I would ask some questions about how THEIR towers are setup. There are many important things that network guys with no RF experience do not take into account -- proper waterproofing, cable loss, amplification and preamplification, etc. Some WISP's will install high power amplifiers with lots of receive gain to extend their coverage area, but that extra receive gain can be a real pain by amplifying noise and reflections, so proper care and consideration needs to be factored in.
Even if you don't know what the answers should be, you can judge how knowledgeable they are about the RF design of their network and whether or not it is properly done by their answers just by how 'certain' they sound when they answer.
If you'd like more info, shoot me an e-mail. I'm fairly certain you are not in our coverage area, so this isn't a sales pitch.:)
Slashdot Mirror
My girlfriend is a type 1 diabetic. Instead of regular injections, she uses an insulin pump. This pump is an external device, about the size of a pager, that feeds insulin into her body via a short tube.
Several months ago she upgraded to a new pump. This new model (a Medtronic MiniMed) wirelessly communicates with a number of devices. It receives blood glucose data from a continuous glucose monitor. It also receives her regular readings from her standard "prick your finger" blood sugar tests via her test kit. And, it has a wireless key fob that allows her to adjust the pumps settings without having to dig through pockets and clothes to get at the unit.
My first comment to her was "With all of this wireless control, how easy is it for someone to use this wireless interface to put you into a diabetic coma, or worse, kill you?" She thinks it's a fairly ridiculous concept, citing encryption, receiver range, and "Why would anyone want to kill me?", among other reasons.
Well, I say that anything that has any type of wireless interface is hackable. There are, of course, no published documents that I can find detailing what steps have been taken to secure these devices. I'm seriously concerned as to whether or not the companies that make insulin pumps, pace makers, implants, etc, may not be taking these concerns seriously.
Interesting story...
One night after having dinner with a few friends, one of the group noted that she needed to stop by a Verizon store to get a new phone. Hers had been having troubles and she was eligible for a free upgrade (with two year contract extension, of course...)
We arrived at the Verizon store, and the representative pulled out a few different models from Motorola, LG, and Samsung for her to look at. I casually pulled my iPhone out of my pocket and sat it in the mix. The Verizon rep exclaimed "OH, is that an iPhone?!", grabbed it, and spent a few minutes playing with it. His breach of salesperson protocol must have snuck into his head, as his face changed to one of concern. He put it down, and started talking about how it doesn't have 3G like Verizon's smart phones, and how the lack of replaceable battery made every other phone on the table a much better choice.
I walked away with a grin on my face, knowing that even the Verizon rep was excited by a phone only available through another carrier.
I remember thinking "when will I ever need this" through algebra and trig during high school... More often than I thought.
:) Calculate its height.
Here are two real-world questions I've had to solve with math over the past few years. I'd start by learning to answer them, and then identify other problems to solve! Yes, there are a lot of people who agree with me, but I didn't see too many examples.
1) In a previous life I co-founded a wireless ISP. I'd often need to calculate how tall a tower was, how high up on the tower an antenna was, and where the signal from said antenna would reach the ground. So, find a cellular or some other tower and figure out how tall it is. Then, find an antenna array on the tower (sometimes at the top, sometimes not). If you can't tell, just pick something obvious and assume it's an antenna.
Now, assume that the antenna has a 30 degree vertical beamwidth. At what distance would the "beam" reach the ground. This calculates what I call the "umbrella effect", or in other words, the area where you are actually too CLOSE to a tower to be within it's coverage area.
(Most carriers will angle their sectors downwards as to not waste beamwidth going up into the atmosphere and maximize coverage, but for the purposes of calculation assume that the center of the "beam" is perfectly horizontal.)
2) Find a swimming pool and figure out it's depth. Then, figure out how much water is in it. Calculate how much water would need to be added to increase the water level by one foot.
I know the above doesn't really tell you how to learn, but hopefully having some interesting real life questions to answer might help. Of course, you might not find the above problems remotely interesting. If that's the case, I apologize for wasting time and bits!
Actually, if you run Apache in a chroot, this just means that the attacker will install their phishing sites somewhere within the chroot, and Apache will still serve them out.
Same with suexec. Regardless of which user the exploit code is owned by, it can still be served out happily by Apache. Suexec certainly does make it easy to see which website the attackers came in via though, instead of everything being owned by the apache or nobody user.
Now, if you would have said "tighten up php.ini", I might have agreed.
The company I work for performs emergency Linux support services. We get a lot of calls from peoples boxes who are attacked. I've seen at least two eBay/PayPal phishing sites recently. In both cases, it had nothing at all to do with Linux itself.
/var/www/html, and stick some php code in there.
Case #1: Customer running a web server had vulnerable PHP applications (I believe it was an outdated WordPress). Someone was able to use this vulnerability to wget a few php scripts and bury them under some subfolders.
Case #2: Customer had a non-root account with a weak password. This account was in the "root" group, giving it write access to a number of system files. Cracker was able to brute force the password quite easily, make a directory called eBay under
In both cases, the php scripts were logging username and password guesses into a text file. The text file was within the same web root, allowing the cracker to easily grab the latest passwords over http instead of needing to re-crack. Also, in both cases, there were at least a dozen usernames and passwords in the text files.
The lesson: Keep your web apps up to date, use strong passwords, and don't add anyone to the root group.
I use Firefox on Mac (intel) and Windows, with the latest versions on both. I can have Firefox open for a full week on Windows without any problems, however on either Mac I have to restart Firefox about once every day or two, otherwise browsing slows to a crawl. At extremes the whole machine will start to bog down until I "force quit" (kill -9) Firefox. I'll also experience oddness where images will just stop loading.
Running "bare bones" on all Firefox installs, no plugins other than whatever may have been included with the base distribution.
Does anyone else notice this? I've switched back to Safari on the Mac in the meantime.
As a lot of people have noted here, how a sysadmin's job is usually very transparent because they are doing their job when things are not broken. Sometimes there are servers to install, updates to apply, etc, and those are easy enough to track. However I think uptime is the #1 way to show that people are doing their job.
In my case, I prepare regular reports for the PHBs showing helpdesk requests, helpdesk response time, backup results, SMTP traffic, and most importantly, a very large month-by-month uptime graph on the first page.
In my case, at salary review time, I can plot our uptime percentage from a few months prior to me "taking over" to present, draw an arrow showing where I took over, and show that our uptime percentage has gone up significantly since then. The data can be just as valuable to you as it is for the PHB.
(and no, I'm not tweaking the numbers!)
Toshiba has also been pushing some deals and promotions on their entry-level HD-DVD player. As an example, when I recently purchased an LCD HDTV, I had the option of buying an HD-A2 for $0.01. Of course I accepted.
And no, this wasn't a "build the player cost into the TV" trick. The price I paid was $500 less than MSRP and about on-par with what the major online sites were selling the set for.
I personally think the format war is really hurting both camps. Just look at some past format wars...
SACD vs DVD-A? Neither, both are dead.
DVD+r/-r? Neither, both are equally supported after many years.
56Flex vs X2? Neither, we now have v.90 and of course broadband.
Having an open source CA is one thing. Having the root certificate included in major browsers is an expensive endeavor. The www.cacert.org site has an FAQ entry about this:
http://wiki.cacert.org/wiki/InclusionStatus
Summary: Lots of open source browsers already have the cert; Mozilla/Firefox will have it soon. Internet Explorer (and apparently Apple's Safari) won't have it unless they come up with a way to pay for the $75,000+ plus $10,000 a year for a AICPA WebTrust audit.
I have a thing for Sun Sparc 20's -- they are VERY upgradeable and extremely reliable. Two of mine have quad 150mhz Ross processors making them snappy ehough to serve out some Apache/PHP/MySQL, host a little e-mail for a few domains, and do some secondary DNS. They're small, don't use a TON of power, and just plain cool.
Oh, and they'll run Linux or a few of the BSD's just fine..
Here is the uptime of one of my production Sparc 20's hosting a bit of email and DNS:
[matt@darkside]$ uptime
9:43AM up 953 days, 16:03, 1 user, load averages: 0.11, 0.11, 0.08
It would be well over 1,000 if a UPS hadn't needed replaced 953 days ago.
Quick speculation here...
Is it possible that GoDaddy made this change to "Unknown" in January to give Microsoft an idea of what kind of a NetCraft jump to expect if that "Unknown" became "IIS"? I mean, it did happen in suspiciously close proximity to this new GoDaddy/MS deal.
I'll go back to my conspiracy cave now...
if imperfections make the ride, cedar point has plenty to offer you. might i recommend the magnum, which was their record-setting steel coaster 10-ish years ago. its age has roughened it up, and it is one fun ride.
...and damn if i don't love living 15 minutes away from CP :)
if you have a chiropractor on speed dial, i'll recommend the mean streak - their at-the-time world setting wooden coaster. that thing is painful, but soooo much fun.
cedar point really is the coaster capital, they have something for everyone.
the millenium force, their record setter that came out before TTD, lasts more than 26 seconds, and literally approach blackout on the downhill every time.
and having ridden the "TTD" many times, i can' say it's not remotely boring.
i don't work at cp (although i worked at a fry stand there when i was 16 -- the year raptor opened). i'm just pointing out that a) 120+mph in 3 seconds up a 400+ foot almost-completely-vertical tower is hardly boring, and b) if it's not your bag, they have something for everyone...
Have you ever had a flight where the person next to you decided to tell you their life story? You know, putting on headphones and closing your eyes doesn't seem to give them the hint?
:)
I speak from recent experience -- perhaps without the cellphone ban, she could have annoyed someone else.
I was one of those geocachers!!!!!! (mattopia on geocaching.com)
We used a deep cycle battery and a power inverter. The cam was an Axis 2120 network camera. We stripped the insulation and one of the pairs from the ethernet cable, used the remaining pairs for data and a hacked power over ethernet solution for the cam.
Set the axis up to FTP an image to an iBook at the other end every 5 seconds. We ended up with over 300 images to dig through and found about 20 decent ones.
The balloons necessary to lift it were expensive. We had over $60 in helium and balloons. The Axis we had laying around, (I also run eriecam.com and tend to have extras. The Axis makes a fairly high quality camera (Sony CCD), so the images were pretty decent quality -- not like a $20 walmart webcam. Resolution was limited to 640x480 though.
It's also worth noting that we used the ethernet cable as the tether/string for the balloons too, to save the weight of having a separate string.
I should send the images to Axis to see if it qualifies as a unique application.
More pictures can be found here, including a few more of the ground setup.
The one I worked for was a bit different. Small Internet Providers throughout the country contracted us to handle their technical support for them. Since many of these were "mom and pop" operations with just a few hundred customers in one city, they relied heavily on maintaining that local image. As a result, we were NEVER, EVER allowed to give any indication that we were not located in the area the person was calling from. I remember talking to customers of a Florida ISP about how nice the weather is, when in fact I'm sitting in Toledo, Ohio (hint to the company's identity?) in a snowstorm. If we were asked for our location, we had to respond that we were not permitted to give out details on our location due to security concerns. I had to give that line a few times a day.
We also had to be crafty. Although some "premium" customers had dedicated phone numbers so that we could find out which ISP they were calling for, many of the individual ISP's calls were routed to a common toll free number, so we'd have no idea as to which of the hundreds of ISP's we do support for the caller is from. We answer the call generically ("Tech Support, how may I assist you?") and usually asked for the customer's e-mail address for an indication of which ISP they were with. The domain name would give away the ISP. Unfortunately, people often did not give the domain name, or had offsite e-mail accounts. Since we couldn't give away that we were not with "their ISP", I couldn't flat out ask. I'd have to narrow it down by area code, and then search between ISP's in that area to find out who they were with -- often taking 10-15 minutes.
I remember one time management signed a deal and gave the call center side a chance to prepare. It was a HUGE customer - larger than all of our other ISP's combined. One night, on my shift, they simply forwarded the tech support number over to us. We went from an average 3 minute call queue time to well over an hour. We did not have the staff to handle the calls, and had no information at all about the specifics of the ISP -- dialup numbers, e-mail servers, etc. It was days before we even had the correct info to give customers. In the meantime, we just had to go with it.
And finally, we had no training program at all, so the company tried to hire people from an outsourcer in the area who had already been through their hideous training program. We paid a dollar an hour more, so it was usually pretty easy to do. Unfortunately, we supported dialup customers, and the company we stole people from supported cable modems, so new hirees usually knew nothing of dialup.
I lasted about six months there surprisingly. When I started it was a small operation with only a dozen or so techs. By the time I left, they had on average 30-40 people per shift. We grew so fast that they had to temporarily build a room in the warehouse and put up folding tables to make room for the new call center people. I'm sure they are much bigger by now, but probably still working out of the warehouse.
Yes, you are correct. It's been awhile on both movies for me as well. In Ferris Bueller, he changes the attendance while Rooney is on the phone with his mom.
And along the same times, how about Matthew Broderick? Not only did he hack into WOPR for a game of global thermonuclear war in "War Games", but he also changed his grades in "Ferris Bueller's Day Off." Now that takes some talent.
We have an 8 or 9 story condo complex we do this in. On the roof, we have two radios - one to bring the connection in from our wireless WAN, and one going into a roof-mounted antenna. That antenna is specifically designed for this type of application. It's a cross polarized 180 degree panel antenna. We have a 250mw amplifier sitting between that and a Cisco 340 series AP. About 1/2 way down in one of the customers condos we have a repeater to help the guys in the lower apartments. It works very well. For your application, if you use a low power amplifer (higher power ones may be illegal based on the FCC's ISM-band regulation) and the right antenna, you can probably do this for under $1,000.
If it is easy to run cable between floors and you don't mind some significant labor, a cheap AP on each floor with a good antenna will do very well too. Make sure to pay close attention to your channel plan when installing multiple AP's. Also, never use the built in omni antennas. You can get 6dbi - 8dbi patch antennas for very cheap. Only get enough spread (120-180 degrees I imagine) to cover the area you need to. With 4 floors, I don't see you needing more than one or two access points. A lot of this depends on the building itself. If its a newer building, with the materials used, signal is likely to travel farther.
www.hyperlinktech.com has a very good selection of antennas. We get a number of the ones we use from here. Their tech staff can probably help you with layout and design as well. I don't work for them, I've just had good experience in the past with them.
Another thing to consider is the client cards. Most off the shelf cards have cheap internal antennas and are low power. The Cisco cards we use are 100mw cards.
Having a powerful card for the clients will help quite a bit too. The Cisco 350 series cards are 100mw cards, which is double or more most cheap off-the-shelf cards. You will pay a premium, but you'll have a lot less phone calls from your users about signal dropping out.
Another solution we haven't tried is to actually locate the access point and antenna outside the building, like where the dumpsters are, and focus the signal in from the side through the windows. A WISP friend of ours has done this in a few areas, and has had very good luck covering buildings much larger than yours with just two AP's and sector antennas.
Some people have suggested going with the "G" standard. Considering you are not doing anything that has super high bandwidth use, I would recommend against it. The lower the bandwidth, the better the range. Most all of the access points in our wireless system (well over 50) are running at 5.5mb, or even 2mb. You will have to test and see what works best in your system.
And finally, as many have also said, are you sure you want to do this? Three years ago, my father and I started a wireless ISP as a hobby. We never anticipated getting as big as we have. Trust me when I say it will end up being less of a hobby and more of a job. Even a small network like that will take maintenance, and you will end up doing tech support for the users in the building. If you do move forward, don't spend all of that money. Have the other users and perhaps even the landlord subsidize some of it. You are providing a marketable service that would make your building stand out over others. Don't give that additional marketing power to the landlord for free! Have EVERYTHING down on paper before starting the install. If the equipment is mounted on the landlord's building, there's no other way to prove that it is yours.
Matt
I salute you! I remember retrofitting one of those boards in a 128k mac, which not only added SCSI, but added additional memory (512k or 1mb, can't remember which). Combined it with an external 40mb SCSI drive, and I had a killer system! Even back then people were modding old stuff to have the functionality of newer systems.
I do remember it was a real pain to get installed, and we damn near ruined the system a few times. Bent the pins of the ROM all to hell.
We are a DSL provider. I know what Verizon charges us for a 768/128 DSL. If we charged $30/month for a 768k DSL circuit, we would be loosing about $8/month. This does not even take into account support costs, our link to Verizon, or our bandwidth from our upstream providers.
For a broadband ISP to make money by selling DSL, they need to either own the network themselves (ie Verizon, SBC, CLEC's, etc) or have major quantities of customers to get any type of discount from the ILEC/CLEC.
Luckily our broadband wireless lets us provide a decent broadband product at a decent price and actually make a bit of money off of it.
And when you work with high band stuff, up in the Ghz range, and you need the really BIG (1 5/8" Andrew) cable, offer to tear it out of an old cell site or other transmission tower for free.
And, if you're crafty (like my father tends to be), you can make a nice side business of selling used cable to other ham operators.
The install cost sounds about right if a tower needs to be erected.
You'd be surprised how far bandwidth can go. Most of our customers get 384k packages, many shared with quite a few machines. Our largest customer has over 130 machines sharing a 512k wireless Internet connection with no issues.
Now if you have 3 power users that chug at the p2p networks 24/7, then 256k will not go very far, especially on your outbound.
I appreciate everyone's input. However it seems most of the discussion so far has been hardware related. That end of things is fine - I have plenty of x86 boxen running Debian or FreeBSD. As the article mentions, I'm mainly looking for software solutions to provide our customers with a hosting control panel (Webmin is really not designed for this in a shared server environment) that will run on one of our existing platforms - Debian/x86 or FreeBSD/x86.
I co-own a primarily wireless ISP. One of the main reason most WISP's mainly cater to business is the equipment cost. A "good" client install generally costs several hundred dollars, and most 'residential' customers are used to paying less than 100, or anything, for the modem and install, as is usually the case with cable and dsl.
:)
If you are behind a lot of thick trees, you will most definitely need your antenna to be above them. We have gotten by with strapping 20 foot of mast pipe to a chimney, or using existing towers. Sometimes you can get signal through the trees if they are fairly sparse, but your reliability will greatly diminish.
If you're within several miles of their tower and you can get a clear line of sight to them, and if they know what they are doing, you should have no problems whatsoever.
One thing to watch out for -- a lot of ISP's who offer wireless really don't know much about RH (the 'wireless' end of things). We have an RF engineer on staff, and without him we would be nowhere. I would ask some questions about how THEIR towers are setup. There are many important things that network guys with no RF experience do not take into account -- proper waterproofing, cable loss, amplification and preamplification, etc. Some WISP's will install high power amplifiers with lots of receive gain to extend their coverage area, but that extra receive gain can be a real pain by amplifying noise and reflections, so proper care and consideration needs to be factored in.
Even if you don't know what the answers should be, you can judge how knowledgeable they are about the RF design of their network and whether or not it is properly done by their answers just by how 'certain' they sound when they answer.
If you'd like more info, shoot me an e-mail. I'm fairly certain you are not in our coverage area, so this isn't a sales pitch.
-dot- net
As an ISP, I will not be billing my customers for any such thing, and will gladly move away from any upstream provider that decides to bill us.