I was in the middle of watching a video on Hulu just a few moments ago when it suddenly stopped buffering. I peeked over at my cable modem, and sure enough, it had lost the signal. 99.9% advertised uptime does imply 0.1% downtime, after all. Luckily, I had the foresight to have previously installed the tethering carrier patch on my iPhone! I turned on tethering, connected my iPhone through USB, and I was off browsing the web again in no time (I opted to suspend my hulu video until later). I ended up tethering for just over 20 minutes until my cable service returned. I continued to browse the web normally (although I put my hulu video on hold), including browsing these forums, viewing the 3GS unboxing video, and looking up information about the Windows Mobile Store. AT&T's mobile data plan allows for 5 GB at a bandwidth up to 3.2 Mbps (no minimum specified) for $30 per month. Even for this small convenience - 20 minutes in 30 days - AT&T wants to charge me an extra $30 per month! They want to charge users double merely to access the same amount of data and bandwidth via tethering, claiming the adverse affects of tethering to their network. AT&T's service agreement states that devices can not be used to tether to a computer, but are they justified? I decided to keep track of the network traffic just to see how much data and bandwidth I used. Here's the data: Stats: Total test time: 24 mins Total data accessed: 17.6 MB Average connection speed: 937 Kbps Analysis: Average network load: 100.1 Kbps Time to reach 5 GB cap: 4 days, 20 hours, and 22 minutes (continuous browsing) Although network data is accessed in spurts at rates equivalent to the average connection speed (937 Kbps), the average network load is calculated as the total amount of data used over the total test time. This is more representative of actual network load as users will not all be requesting spurts of data simultaneously; in other words, data requests for multiple users will average out into a constant rate of data. The average network load for this type of browsing is 10.7% of the average bandwidth provided by AT&T during the test (937 Kbps). It would require that 24.2% of one's total available time be spent in continuous browsing over a tethered 3G connection to reach the 5 GB data cap per month. This value assumes 16 waking hours per day in a 30-day month. Conclusions: The average load on the network for web browsing on a tethered connection represents only 10.7% of the bandwidth provided per user. Additionally, the browsing habits that would be required to reach the monthly data cap with this type of web browsing is unrealistic. This does not indicate that web browsing with internet tethering has a significantly adverse to the wireless network. AT&T's policy of no tethering is not justified for three reasons. 1) There is no provision for occasional (limited use) tethering for no additional cost which would benefit users in my situation and result in no adverse impact on the wireless network. 2) Justification that tethering is prohibited due to high-bandwidth applications is redundant as there are already prohibitions on using such applications in the service agreement. This includes services such as server hosting, web broadcasting, P2P file sharing, and streaming television signals. 3) Justification that tethering is prohibited due to high-bandwidth applications or extreme data use is irrelevant as AT&T has control of both the bandwidth available to users and the amount of data they may access per month. They have the power to limit bandwidth to users such that their use will not adversely affect their network, as well as to deny wireless data service for users that exceeded the data cap within any given month. If you agree with these conclusions, contact the FCC and voice your opinion that it is time for carriers to admend their policies to consumers' needs despite apologists citing the service agreements that users must accept.