Any projector that works well enough to be used in even a moderately lit room will weigh at least as much as an ultrabook and has the added disadvantage of needing AC power to be of any use. Just buy an ultrabook or an iPad. I also find it kind of odd that a few pounds (the difference between an ultrabook and a laptop) would make much of a difference. My business has folks flying all over the world for 25-30 weeks out of the year and they all manage fine with full weight Dell laptops. Never had one complaint related to weight restrictions. Some do complain that it's a pain to carry a laptop, but that's a different issue.
I have never worked for a company which didn't clearly state in the employee handbook that company-owned technology assets are for work purposes only, may be monitored at any time with no notice, for any reason, or for no reason. It only makes sense for them to put that in there because it allows them to do whatever they want without worrying about you, and that's the way all contracts are written by default (to favor the party writing it). If you don't agree with this, then you should tell them and see if they will change the policy (most will not, you are not that important to them).
There are many excellent reasons for companies to proxy https traffic. Just off the top of my head...
It can help troubleshooting network / application issues
It helps them monitor what you're doing online and make sure you're not sending private data places it shouldn't go (gmail, and many social networks use https, so they want to watch these sites)
Anything with high system requirements is not a great candidate for virtualization. One of the major selling points is taking ten servers that average 5% CPU and combining them onto a single host to realize cost savings over having to replace those 10 physical servers at some point.
If you have a quad cpu db server then you're likely going to have to dedicate a host to run that single VM, so you're still paying the same hardware costs, adding in VM software costs, and slightly reducing the performance by introducing the overhead of a hypervisor.
Other problems are systems that have specialized hardware, i.e. something that you have to connect to a serial port. That doesn't work well in a virtualized environment.
My preference is always to keep the system running backups on a separate physical server. Reason for this is 1) the backup server should not be so critical as to impact your production apps if there is a problem with it, 2) restoring from a VM catastrophe is a lot easier if you don't need the VM infrastructure to work to perform a restore, 3) it's a lot easier to throw a server and a tape drive in your trunk if you need to run to escape some sort of disaster than it is to throw multiple VM hosts, SAN, etc in there, and 4) I prefer not to share bus bandwidths with running VMs when running backups.
Since most others are telling you why this is a bad idea, etc...I'll try to actually provide a useful answer.
In a situation like you describe where the condo likely owns it's own wire plant, DSL is probably a pretty good solution. WiFi is going to be problematic and expensive because you have to worry about interference, installing, securing and maintaining APs, etc. Ethernet is also going to be crazy expensive to install.
Assuming you have one or two centralized wiring closets where all of the telco wiring comes in to the building, you could install your own DSLAM there and sell or lease DSL modems to your owners/tenants. The DSL signal would go from the DSLAM in the wiring closet, over the existing phone wiring to the individual units where the subscribers would connect their modems. While DSL is normally pretty slow compared to cable, this is usually because you have to deal with distance limitations, but in a situation like this you shouldn't have any runs over a 400-500 feet which would put them well within the max speed distance. 80 units over 20 stories means 4 units per story, so there shouldn't be too much horizontal wiring, and figuring about 10-12' of height per story we come up with a max vertical distance of around 250 feet.
Versatek is one company that makes equipment designed for this purpose that I found with a quick Google search. They have DSLAMs that they claim can deliver 24 Mbps per subscriber, which isn't the fastest service, but it's well above the national average.
Doing some quick math...
80 units times average normal broadband speed of 10Mbps = 800 Mbps
Figure maybe 10% actual average utilization (assuming you don't have anyone that is BT crazy) = 80 Mbps
My guess would be that you would probably have plenty of bandwidth at 80 Mbps, but you could start there and scale up/down as needed.
The real issues I see are:
1) Who is going to support this when a subscriber has a problem at 9PM on a Friday night?
2) You'll need some technical expertise to install and provision the service, which increases costs and reduces the savings associated with your economy of scale
3) You'll have to pay for commercial service to be able to legally resell it
4) You may also incur additional costs to ensure you have enough IP addresses (don't be a dick and put everyone behind NAT, if you have any techies this will make them mad)
This sounds like a pretty cool project. You can DIY and save some money and pass those savings on to the subscribers, or you can just hire a company to do it all for you, but then you sort of negate the cost savings.
As the submitter notes, the defendant knows that suing them over a few thousand dollars probably doesn't make financial sense, so they are free to screw you with no fear of recourse. So why, exactly, would they be interested in participating in this type of service if offers them nothing and potentially opens them up to costs and liability?
Not sure how this turned into a 2A discussion, but I think perhaps I replied to the wrong person. I was replying to the person that asks if gunshots were a crime, and presuming that he was meaning in a developed area, since that's where the devices in the article were being used, so that's the context my reply makes sense in.
Perhaps 2A doesn't deal with defense of individual liberties and personal safety, it is there to make sure that the government cannot become tyrannical by removing the population's ability to deal with an unjust government.
But you are right, it is unnecessary to write self defense into a law because it has already been dealt with by stating that no person can be unjustly deprived of life, liberty, etc, and it is also a basic human right which supercedes the law of any government.
In just about every municipality in the USA, yes. It is illegal to discharge a firearm with a few exceptions, i.e. acting in self defense, at a designated shooting range, etc. Otherwise, yes, discharging a firearm is quite illegal and definitely something I would want law enforcement to be notified of. Would you really feel safe in a community where folks could just randomly go outside and start shooting stuff?
Seems like Microsoft has created a product to fit the exact scenario you are in with Office 365. They provide hosted Exchange, Sharepoint (and Lync if you like it enough to use it). As I recall, it's about $100/user/year If you can live with web based everything or about $250/user/year if you want the full Office Professional desktop suite as well.
As soon as they find out they're being monitored, they'll just turn off WiFi and use 3G instead. Good luck monitoring that.
But seriously...
1) if you're trying to build this from scratch and are asking slashdot - it's not going to work, and
2) what ever happened to simply stating the rules and expecting everyone to follow them? It sounds like this guy is a serious control freak.
Your comment makes no sense, and I'll explain why in really simple terms.
Let's say you have a network in your house. You and two roommates each have a PC. On each PC there are various files such as music, videos, etc that are shared on the network. Each person can access the files on all of the computers across the network and it's all free because when you build the house you ran your own network lines and installed a switch. But one day one of the roommates gets bored with the music and videos because he's already heard/seen them all and decides he would like to connect to other computers, outside the house. Now you have to buy an Internet connection because the content he wants to access is not located on your network.
So now you've got cost associated with getting the content you want. As the type of content being accessed went from mp3 to avi to mkv files, the filesizes also rose and you needed to add more bandwidth, so your costs went up. So I think this clearly indicates how internal content is cheaper - you don't have to pay anyone for access to other networks to get it.
Everyone gives comcast and other ISPs a ration of shit for exempting their own services from bandwidth caps, but there really is a solid justification behind doing so - ASSUMING that all internal content would be treated the same way. Since there isn't an easy way for Comcast to measure your data usage at the edge of their network instead of at your CPE I kind of understand them not exempting all internal content but it's still kinda shitty that they don't.
You have to take into account the inherent protocol overhead. For TCP/IP over ethernet I have seen different numbers but I usually figure about 10-15%. If you have DSL that means you likely also have PPPoE to deal with, which introduces further overhead. If we assume 10% overhead, that puts you right at 2.7 Mbps, so it seems that your line is performing exactly as it should when it gets 2.7 Mbps.
Furthermore, your ISP probably states that you will receive speeds "up to" 3 Mbps, which you are.
Also, they probably also state that they guarantee the "up to" 3 Mbps speed to the DSLAM which is the only part of your connection which does not share bandwidth with other users.
So for all of those reasons, I'd say you have a minimal chance of your ISP caring about the trouble your're experiencing. If their network is overloaded to the point that you get sub 1-Mbps speeds during peak demand then they ought to upgrade their capacity but they're not obligated to do so.
I suppose it is, but in this day and age, a cell phone or tablet will just as easily get you to Picasa, Flickr or where ever else you happen to have those photos stored online:)
Agree with others that have suggested "go big or go home". If you're the type of person that thinks about installing a server rack in your house, a 6U or 12U rack probably isn't going to be big enough for long enough. Plus, as you've noted, the smaller racks are really designed for network equipment, not servers. If you're going to put a server in there, you may as well go for at least a half rack. And I'd really recommend a four post setup with sliding rails for whatever servers you end up putting in there - it just makes things so much easier. You'll inevitably end up needing to upgrade RAM or something, and it's a huge PITA if you have to pull the whole server out of the rack. Plus, four post racks are readily available second hand on craigslist.
As to your comment about two cat5 jacks per room and four in the living room - as someone who did exactly that four years ago when I build my house I can tell you that you're boxing yourself in. If you're going to have a home office, you probably want to put at least four jacks in there. For your living room, just think about all the stuff you may someday have that will want cat5. I have an Apple TV, Samsung TV, Xbox, Bluray, and a PC. I ended up putting a 5 port switch in my entertainment center, and it may well be more cost effective to only run one network jack to a small switch at locations where you need higher port density. Most home stuff doesn't have really high throughput requirements, so losing the single highspeed backplane of a centralized switch isn't a huge issue. The only issue here would be (as in my case) if you need PoE on some of those ports.
Here are a few other things I'd do differently if I was building again and had a budget for this sort of thing:
1) Run a min 20A (30A is even better) dedicated circuit with a twist lock connector to the rack's location. If you want to get a rackmount UPS in the 2000-3000VA range, it will probably require this.
2) Install sound deadening around the rack - network equipment is typically pretty noisy
3) Plan for cooling - if you can run an A/C duct to the rack's location that's good, but plan for how you're going to keep stuff cool when the rest of the house has reached your target temperature and the central A/C turns off for 30 minutes. You may want to look at one of the smaller portable A/C units that you can duct either into the crawlspace (check local codes about this, you may run into problems with mold if you duct moist air up there), or outside.
4) Run CONDUIT - this is probably the biggest tip I can give you. If you're able to install wall boxes and such before drywall goes up, spend $100 on a roll of blue flex conduit and run that from your wall boxes up into the attic/crawlspace. Make sure to stub the conduit up high enough so that any blown-in insulation doesn't cover the top of it. You'll be so glad you did this in a few years when you want to upgrade or add more wiring.
5) Cable management - don't overlook it. Patch panels and wire management to and at the rack make life so much nicer. You can get by without it but if you do a really nice job you'll find yourself wanting to show it off to your friends to impress them (unfortunately, it usually just draws a blank stare).
6) K.I.S.S. - since I'm a network engineer, I built my home network with a cisco router, AP and switch, created VLANs and public/private WiFi networks, then realized that most consumer level tech isn't designed to be compatible with that in the least. Take Apple TV for instance - it relies on network broadcasts. If you have an iPad or iPhone with airplay and want to send a stream to your Apple TV, the two devices have to be on the same network. There's no way to manually enter device IP addresses to get two devices on separate subnets to talk to each other. Naturally, if your PCs are on a private network with the Apple TV, that's well and good, but what if a friend comes over and connects to your public WiFi and you want him to be able to use airplay from his iPhone? I've had similar problems with Slingbox, etc. That kind of stuff just isn't designed to work on a network any more complicated than what you get with your standard Linksys router.
Assuming an average cell phone plan runs them $75/month (taking into account estimated discounted pricing for an account w/ 175 phones), that's $630k for 4 years of service. If they just offered some type of app people could download to their phones which would allow them to be remotely monitored in exchange for $20/mo, I'm sure they'd get more than 175 people that would be willing to be monitored in exchange for the $20. Still, the same 175 people would only cost $168k over four years.
Better still, market the app to parents of teens and include functionality that would allow the parents to monitor their teens activity online for free. I'm sure you would get thousands upon thousands of parents that would consent to have their teens phone service monitored in exchange for being able to monitor it themselves. That would totally eliminate the cost of providing cell service and make the only costs be the development of the app and the hosting of the servers, so call the app $50k and call the servers $15k/year and then your four year cost is only $110k and you'd have data from thousands if not tens of thousands of teens instead of only 175.
I remember back in the nineties when you'd order PC parts from Computer Shopper or the like to build a new PC and you had a whole room full of PC parts, there was a very distinct smell. You don't get that so much anymore since it seems like only hardcore folks build PCs these days instead of buying them. I wish someone made an air freshener that had that smell so I could hang it in my car.
The issue is that 90% of Americans will not buy it because it's initial cost is too high. Most people only look at the cost of the bulb and disregard the cost of replacement bulbs and the electricity required to light it.
If you look at a 5 or 10 year TCO I'm sure the bulb makes a lot of sense, but when most people in America are living paycheck to paycheck, saving $200 over the next 10 years isn't worth spending $100 today.
Unless their subscribers also happen to use Netflix (or Hulu, or iTunes, or anything in the AWS cloud, or...) and have issues accessing it, then they dump Comcast and get FIOS or something else.
And more realistically, if Comcast doesn't peer with L3, Comcast pays out the ass for transit to them because a much larger portion of the Internet is on-net for L3 than for Comcast. L3 is connected to just about everybody because they are truly a tier 1 provider while Comcast is not.
Without disputing what you're claiming about Comcast and Level3, I will say that it doesn't really make sense to me. Usually that arrangement is the other way around (i.e. the person sending more data is the one that gets paid). Think about Joe Blow ISP that has 100 subscribers, if they go to L3, Comcast, Sprint, etc and say hey I want to peer with you, the other ISP will say 99% of the traffic will be from us to you, so you are not peering, you are essentially using us as an ISP so you have to pay to peer with us. Also, ISPs want to be well connected to CDNs so their subscribers have fast access to content, so there are incentives for ISPs to peer with CDNs.
At any rate, traffic within the Comcast network is going to be simpler for them to manage. If you use a company analogy, thing about a company that has an office and inside that office they have a video server. Now when the guys from finance want to watch videos, their traffic is contained on the Internal company network where they have fast 1Gbps switches and don't have to worry about bandwidth, but when they go to Youtube the traffic has to traverse the Internet connection which is very likely much slower than the internal LAN. So the IT group charges a portion of the cost of the Internet connection back to the finance group because they're using a portion of the bandwidth, but the internal bandwidth is free because the infrastructure that is using is just a cost of doing business.
The difference is that the Xfinity app doesn't have to leave the Comcast network to get its' content. Netflix, etc, do.
Just think of them metering your bandwidth as it leave the Comcast network instead of when it leaves your house. As a network operator, I don't really see this as being evil. ISPs have to pay for bandwidth that leaves their network, while content inside their network is free. Naturally ISPs want their subscribers to pay for content they access outside the ISP's network since the ISP itself incurs costs for that.
With that having been said, I'm sure Comcast is large enough that it is probably peering with just about everyone, so they don't actually have to pay for that bandwidth, but still, the peering points are usually the choke points in a network, so it makes sense for them to institute caps that would be metered at those points versus at your modem.
Slashdot Mirror
Any projector that works well enough to be used in even a moderately lit room will weigh at least as much as an ultrabook and has the added disadvantage of needing AC power to be of any use. Just buy an ultrabook or an iPad. I also find it kind of odd that a few pounds (the difference between an ultrabook and a laptop) would make much of a difference. My business has folks flying all over the world for 25-30 weeks out of the year and they all manage fine with full weight Dell laptops. Never had one complaint related to weight restrictions. Some do complain that it's a pain to carry a laptop, but that's a different issue.
Seriously, how do these questions even get to Slashdot? Call a moving company or PODS and tell them to pack it up. Problem solved.
Like if you live in an area with no cell coverage and you want to listen to music while mowing the grass...which happened to me yesterday.
It DOES violate the YouTube TOS. What did they expect was going to happen?
I have never worked for a company which didn't clearly state in the employee handbook that company-owned technology assets are for work purposes only, may be monitored at any time with no notice, for any reason, or for no reason. It only makes sense for them to put that in there because it allows them to do whatever they want without worrying about you, and that's the way all contracts are written by default (to favor the party writing it). If you don't agree with this, then you should tell them and see if they will change the policy (most will not, you are not that important to them).
There are many excellent reasons for companies to proxy https traffic. Just off the top of my head...
It can help troubleshooting network / application issues
It helps them monitor what you're doing online and make sure you're not sending private data places it shouldn't go (gmail, and many social networks use https, so they want to watch these sites)
It can help in terms of caching https content
Anything with high system requirements is not a great candidate for virtualization. One of the major selling points is taking ten servers that average 5% CPU and combining them onto a single host to realize cost savings over having to replace those 10 physical servers at some point.
If you have a quad cpu db server then you're likely going to have to dedicate a host to run that single VM, so you're still paying the same hardware costs, adding in VM software costs, and slightly reducing the performance by introducing the overhead of a hypervisor.
Other problems are systems that have specialized hardware, i.e. something that you have to connect to a serial port. That doesn't work well in a virtualized environment.
My preference is always to keep the system running backups on a separate physical server. Reason for this is 1) the backup server should not be so critical as to impact your production apps if there is a problem with it, 2) restoring from a VM catastrophe is a lot easier if you don't need the VM infrastructure to work to perform a restore, 3) it's a lot easier to throw a server and a tape drive in your trunk if you need to run to escape some sort of disaster than it is to throw multiple VM hosts, SAN, etc in there, and 4) I prefer not to share bus bandwidths with running VMs when running backups.
Since most others are telling you why this is a bad idea, etc...I'll try to actually provide a useful answer.
In a situation like you describe where the condo likely owns it's own wire plant, DSL is probably a pretty good solution. WiFi is going to be problematic and expensive because you have to worry about interference, installing, securing and maintaining APs, etc. Ethernet is also going to be crazy expensive to install.
Assuming you have one or two centralized wiring closets where all of the telco wiring comes in to the building, you could install your own DSLAM there and sell or lease DSL modems to your owners/tenants. The DSL signal would go from the DSLAM in the wiring closet, over the existing phone wiring to the individual units where the subscribers would connect their modems. While DSL is normally pretty slow compared to cable, this is usually because you have to deal with distance limitations, but in a situation like this you shouldn't have any runs over a 400-500 feet which would put them well within the max speed distance. 80 units over 20 stories means 4 units per story, so there shouldn't be too much horizontal wiring, and figuring about 10-12' of height per story we come up with a max vertical distance of around 250 feet.
Versatek is one company that makes equipment designed for this purpose that I found with a quick Google search. They have DSLAMs that they claim can deliver 24 Mbps per subscriber, which isn't the fastest service, but it's well above the national average.
Doing some quick math...
80 units times average normal broadband speed of 10Mbps = 800 Mbps
Figure maybe 10% actual average utilization (assuming you don't have anyone that is BT crazy) = 80 Mbps
My guess would be that you would probably have plenty of bandwidth at 80 Mbps, but you could start there and scale up/down as needed.
The real issues I see are:
1) Who is going to support this when a subscriber has a problem at 9PM on a Friday night?
2) You'll need some technical expertise to install and provision the service, which increases costs and reduces the savings associated with your economy of scale 3) You'll have to pay for commercial service to be able to legally resell it 4) You may also incur additional costs to ensure you have enough IP addresses (don't be a dick and put everyone behind NAT, if you have any techies this will make them mad)
This sounds like a pretty cool project. You can DIY and save some money and pass those savings on to the subscribers, or you can just hire a company to do it all for you, but then you sort of negate the cost savings.
As the submitter notes, the defendant knows that suing them over a few thousand dollars probably doesn't make financial sense, so they are free to screw you with no fear of recourse. So why, exactly, would they be interested in participating in this type of service if offers them nothing and potentially opens them up to costs and liability?
Not sure how this turned into a 2A discussion, but I think perhaps I replied to the wrong person. I was replying to the person that asks if gunshots were a crime, and presuming that he was meaning in a developed area, since that's where the devices in the article were being used, so that's the context my reply makes sense in.
Perhaps 2A doesn't deal with defense of individual liberties and personal safety, it is there to make sure that the government cannot become tyrannical by removing the population's ability to deal with an unjust government.
But you are right, it is unnecessary to write self defense into a law because it has already been dealt with by stating that no person can be unjustly deprived of life, liberty, etc, and it is also a basic human right which supercedes the law of any government.
In just about every municipality in the USA, yes. It is illegal to discharge a firearm with a few exceptions, i.e. acting in self defense, at a designated shooting range, etc. Otherwise, yes, discharging a firearm is quite illegal and definitely something I would want law enforcement to be notified of. Would you really feel safe in a community where folks could just randomly go outside and start shooting stuff?
Microsoft have stores?
Seems like Microsoft has created a product to fit the exact scenario you are in with Office 365. They provide hosted Exchange, Sharepoint (and Lync if you like it enough to use it). As I recall, it's about $100/user/year If you can live with web based everything or about $250/user/year if you want the full Office Professional desktop suite as well.
As soon as they find out they're being monitored, they'll just turn off WiFi and use 3G instead. Good luck monitoring that.
But seriously...
1) if you're trying to build this from scratch and are asking slashdot - it's not going to work, and
2) what ever happened to simply stating the rules and expecting everyone to follow them? It sounds like this guy is a serious control freak.
Your comment makes no sense, and I'll explain why in really simple terms.
Let's say you have a network in your house. You and two roommates each have a PC. On each PC there are various files such as music, videos, etc that are shared on the network. Each person can access the files on all of the computers across the network and it's all free because when you build the house you ran your own network lines and installed a switch. But one day one of the roommates gets bored with the music and videos because he's already heard/seen them all and decides he would like to connect to other computers, outside the house. Now you have to buy an Internet connection because the content he wants to access is not located on your network.
So now you've got cost associated with getting the content you want. As the type of content being accessed went from mp3 to avi to mkv files, the filesizes also rose and you needed to add more bandwidth, so your costs went up. So I think this clearly indicates how internal content is cheaper - you don't have to pay anyone for access to other networks to get it.
Everyone gives comcast and other ISPs a ration of shit for exempting their own services from bandwidth caps, but there really is a solid justification behind doing so - ASSUMING that all internal content would be treated the same way. Since there isn't an easy way for Comcast to measure your data usage at the edge of their network instead of at your CPE I kind of understand them not exempting all internal content but it's still kinda shitty that they don't.
You have to take into account the inherent protocol overhead. For TCP/IP over ethernet I have seen different numbers but I usually figure about 10-15%. If you have DSL that means you likely also have PPPoE to deal with, which introduces further overhead. If we assume 10% overhead, that puts you right at 2.7 Mbps, so it seems that your line is performing exactly as it should when it gets 2.7 Mbps.
Furthermore, your ISP probably states that you will receive speeds "up to" 3 Mbps, which you are.
Also, they probably also state that they guarantee the "up to" 3 Mbps speed to the DSLAM which is the only part of your connection which does not share bandwidth with other users.
So for all of those reasons, I'd say you have a minimal chance of your ISP caring about the trouble your're experiencing. If their network is overloaded to the point that you get sub 1-Mbps speeds during peak demand then they ought to upgrade their capacity but they're not obligated to do so.
I suppose it is, but in this day and age, a cell phone or tablet will just as easily get you to Picasa, Flickr or where ever else you happen to have those photos stored online :)
Out of curiosity why do you recommend using film? I took about 200 pictures of my home before drywall went up but I used a digital camera.
Agree with others that have suggested "go big or go home". If you're the type of person that thinks about installing a server rack in your house, a 6U or 12U rack probably isn't going to be big enough for long enough. Plus, as you've noted, the smaller racks are really designed for network equipment, not servers. If you're going to put a server in there, you may as well go for at least a half rack. And I'd really recommend a four post setup with sliding rails for whatever servers you end up putting in there - it just makes things so much easier. You'll inevitably end up needing to upgrade RAM or something, and it's a huge PITA if you have to pull the whole server out of the rack. Plus, four post racks are readily available second hand on craigslist.
As to your comment about two cat5 jacks per room and four in the living room - as someone who did exactly that four years ago when I build my house I can tell you that you're boxing yourself in. If you're going to have a home office, you probably want to put at least four jacks in there. For your living room, just think about all the stuff you may someday have that will want cat5. I have an Apple TV, Samsung TV, Xbox, Bluray, and a PC. I ended up putting a 5 port switch in my entertainment center, and it may well be more cost effective to only run one network jack to a small switch at locations where you need higher port density. Most home stuff doesn't have really high throughput requirements, so losing the single highspeed backplane of a centralized switch isn't a huge issue. The only issue here would be (as in my case) if you need PoE on some of those ports.
Here are a few other things I'd do differently if I was building again and had a budget for this sort of thing:
1) Run a min 20A (30A is even better) dedicated circuit with a twist lock connector to the rack's location. If you want to get a rackmount UPS in the 2000-3000VA range, it will probably require this.
2) Install sound deadening around the rack - network equipment is typically pretty noisy
3) Plan for cooling - if you can run an A/C duct to the rack's location that's good, but plan for how you're going to keep stuff cool when the rest of the house has reached your target temperature and the central A/C turns off for 30 minutes. You may want to look at one of the smaller portable A/C units that you can duct either into the crawlspace (check local codes about this, you may run into problems with mold if you duct moist air up there), or outside.
4) Run CONDUIT - this is probably the biggest tip I can give you. If you're able to install wall boxes and such before drywall goes up, spend $100 on a roll of blue flex conduit and run that from your wall boxes up into the attic/crawlspace. Make sure to stub the conduit up high enough so that any blown-in insulation doesn't cover the top of it. You'll be so glad you did this in a few years when you want to upgrade or add more wiring.
5) Cable management - don't overlook it. Patch panels and wire management to and at the rack make life so much nicer. You can get by without it but if you do a really nice job you'll find yourself wanting to show it off to your friends to impress them (unfortunately, it usually just draws a blank stare). 6) K.I.S.S. - since I'm a network engineer, I built my home network with a cisco router, AP and switch, created VLANs and public/private WiFi networks, then realized that most consumer level tech isn't designed to be compatible with that in the least. Take Apple TV for instance - it relies on network broadcasts. If you have an iPad or iPhone with airplay and want to send a stream to your Apple TV, the two devices have to be on the same network. There's no way to manually enter device IP addresses to get two devices on separate subnets to talk to each other. Naturally, if your PCs are on a private network with the Apple TV, that's well and good, but what if a friend comes over and connects to your public WiFi and you want him to be able to use airplay from his iPhone? I've had similar problems with Slingbox, etc. That kind of stuff just isn't designed to work on a network any more complicated than what you get with your standard Linksys router.
Hopefully some of this is helpful to you.
Remember this started four years ago. The iPhone had barely been available for a year and Blackberry was still a fairly "good" phone.
Assuming an average cell phone plan runs them $75/month (taking into account estimated discounted pricing for an account w/ 175 phones), that's $630k for 4 years of service. If they just offered some type of app people could download to their phones which would allow them to be remotely monitored in exchange for $20/mo, I'm sure they'd get more than 175 people that would be willing to be monitored in exchange for the $20. Still, the same 175 people would only cost $168k over four years.
Better still, market the app to parents of teens and include functionality that would allow the parents to monitor their teens activity online for free. I'm sure you would get thousands upon thousands of parents that would consent to have their teens phone service monitored in exchange for being able to monitor it themselves. That would totally eliminate the cost of providing cell service and make the only costs be the development of the app and the hosting of the servers, so call the app $50k and call the servers $15k/year and then your four year cost is only $110k and you'd have data from thousands if not tens of thousands of teens instead of only 175.
I remember back in the nineties when you'd order PC parts from Computer Shopper or the like to build a new PC and you had a whole room full of PC parts, there was a very distinct smell. You don't get that so much anymore since it seems like only hardcore folks build PCs these days instead of buying them. I wish someone made an air freshener that had that smell so I could hang it in my car.
The issue is that 90% of Americans will not buy it because it's initial cost is too high. Most people only look at the cost of the bulb and disregard the cost of replacement bulbs and the electricity required to light it.
If you look at a 5 or 10 year TCO I'm sure the bulb makes a lot of sense, but when most people in America are living paycheck to paycheck, saving $200 over the next 10 years isn't worth spending $100 today.
Unless their subscribers also happen to use Netflix (or Hulu, or iTunes, or anything in the AWS cloud, or ...) and have issues accessing it, then they dump Comcast and get FIOS or something else.
And more realistically, if Comcast doesn't peer with L3, Comcast pays out the ass for transit to them because a much larger portion of the Internet is on-net for L3 than for Comcast. L3 is connected to just about everybody because they are truly a tier 1 provider while Comcast is not.
Without disputing what you're claiming about Comcast and Level3, I will say that it doesn't really make sense to me. Usually that arrangement is the other way around (i.e. the person sending more data is the one that gets paid). Think about Joe Blow ISP that has 100 subscribers, if they go to L3, Comcast, Sprint, etc and say hey I want to peer with you, the other ISP will say 99% of the traffic will be from us to you, so you are not peering, you are essentially using us as an ISP so you have to pay to peer with us. Also, ISPs want to be well connected to CDNs so their subscribers have fast access to content, so there are incentives for ISPs to peer with CDNs.
At any rate, traffic within the Comcast network is going to be simpler for them to manage. If you use a company analogy, thing about a company that has an office and inside that office they have a video server. Now when the guys from finance want to watch videos, their traffic is contained on the Internal company network where they have fast 1Gbps switches and don't have to worry about bandwidth, but when they go to Youtube the traffic has to traverse the Internet connection which is very likely much slower than the internal LAN. So the IT group charges a portion of the cost of the Internet connection back to the finance group because they're using a portion of the bandwidth, but the internal bandwidth is free because the infrastructure that is using is just a cost of doing business.
The difference is that the Xfinity app doesn't have to leave the Comcast network to get its' content. Netflix, etc, do.
Just think of them metering your bandwidth as it leave the Comcast network instead of when it leaves your house. As a network operator, I don't really see this as being evil. ISPs have to pay for bandwidth that leaves their network, while content inside their network is free. Naturally ISPs want their subscribers to pay for content they access outside the ISP's network since the ISP itself incurs costs for that.
With that having been said, I'm sure Comcast is large enough that it is probably peering with just about everyone, so they don't actually have to pay for that bandwidth, but still, the peering points are usually the choke points in a network, so it makes sense for them to institute caps that would be metered at those points versus at your modem.