The Tablet Debate: 3G Or Wi-Fi? 395
Barence writes "We regularly review tablets and mention the fact that X tablet has Wi-Fi, Y tablet has 3G, but how many people are interested in each? Do most people view 3G as a must-have extra, or is Wi-Fi plenty for a device most commonly used in the home? We asked our readers for their opinions and the responses were fairly evenly split between those on both sides, with a healthy proportion also saying they may not choose it but like to have 3G as an option. What do Slashdotters think? Is 3G a must-have for tablets or will a tethered smartphone do the job?"
They should not be separate devices (Score:5, Interesting)
My laptop, my phone, and my tablet should all just be viewports and ways of interacting with one homogenous device. They should all be integrated parts of a whole.
To that end, I do not care which thing has which feature. I just care that I can seamlessly access the Internet no matter where I am.
The MiFi option works best for me... (Score:3, Interesting)
It depends on your environment (Score:5, Interesting)
In my area, Cablevision, Time Warner, and Comcast have lined the major highways with Wifi access points. Even better, I can register my mobile device MAC addresses on the Cablevision web site, so I don't even have to go through a sign-in (it looks like an open access point to a registered MAC), and there's no extra charge for WiFi connectivity (it's included in the home internet cost).
My wife has been driving while I Googled - in a moving vehicle - with no problem. The hand-offs seem to work fine. The more ubiquitous WiFi is, the less important 3G is. Why pay a monthly fee for something you don't need?
I've received a survey from my cable company asking a series of questions about cellular use, and inquiring whether I'd be interested in a completely unlimited, uncapped voice/text/data Android smartphone for $29/month. I suspect that the cable companies will be offering WiFi based service in the not-too-distant future over their internet/WiFi infrastructure.
Re:tethered via adhoc wifi will do the job (Score:5, Interesting)
GSM OTOH uses time-domain multiplexing (basically the phones take turns talking to the tower). That's ok for voice, but horribly wasteful of bandwidth for data. Consequently, GSM providers had to develop an entirely new system and protocol for data, which is why it took them 1-2 years to catch up to CDMA's 3G data. When they finally did roll it out, it needed an entirely separate radio (you had to upgrade phones), which added to the complexity, cost, and power consumption of the phone. But a second radio carries with it the advantage of doing voice and data simultaneously. In terms of use, there are very few times when you actually need to use them simultaneously. The CDMA carriers didn't package a second dedicated 3G data radio in their phones simply because there wasn't enough demand for it. It only became an issue when Verizon and AT&T got into an advertising war.
Fast forward to today. The 4G on CDMA carriers uses a different technology for data and, just like GSM phones, a different radio than that used for voice. If your CDMA phone has a 4G connection, it can do simultaneous voice and data just fine. It has nothing to do with CDMA or GSM, everything to do with how many radios are on your phone. (The same is why phones can do simultaneous voice and WiFi. 802.11b/g/n uses yet another different radio.)
Re:tethered via adhoc wifi will do the job (Score:4, Interesting)
It's amusing how a CDMA superiority has become spun as a GSM advantage
What is quite a lot more amusing is to watch someone claim a feature that is significantly less useful for the user is in any way a mark of "superiority". Just because it has a nicer technical design does not make it superior if the end result to the user is inferior.
That's the thing that technical minded people on Slashdot continue to find elusive, this notion that technology exists to serve people and not the other way round.
What I would like to see, personally: (Score:5, Interesting)
What I would like to see is a non-profit org get created for the sole purpose of acquiring licensed spectrum for long-range packet radio, and to create an open-licensed wireless protocol to use it-- Then release it publicly FOSS style.
Because it would be long distance (1 to 2 miles would be the ideal coverage area for access points I would think.. could be wrong though. That's just a guess.) it would need to prevent abuse as part of the protocol itself, and so I personally would like to see encryption be a requirement for devices to connect-- Not some gutless password based encryption either, I mean PKI-style encryption with issued certificates kind.
One of the neat things about public/private key encryption is that you can theoretically have multiple public keys to a single private key, which could then be independently enforced.
The idea is to replace 802.11x (A, B, G, N) straight up, and to make consumer boxes that serve as access points just like wifi routers. For businesses offering complimentary internet, (who wouldnt want to be a free ISP to everyone within 1-2 miles-- and only offer to paying customers) a simple near-feild communication plate built into the counter at the store could supply a time-leased certificate to the device (think really short range bluetooth), giving it permission to access the AP, which would then get revoked after the time elapses. The ability to have multiple public keys per private key would let this work. The business's AP would keep a "pool" of public certs, and would track their use against unique hardware IDs from the connecting devices. (The AP would check that the cert is valid, then check to see who it was issued to-- If the unique ID does not match, no connection.) This would keep people from being repeat customers at a specific place, and eventually having every cert in the pool pushed to their device over time, and then no longer needing to make a purchase to gain access. It would also prevent people from using what is assumed to be a unique public key at the same time somebody else is, and causing problems. (There would be 2 levels of uniqueness-- Unique public key, and unique device MAC. The AP would check both, and decline connection if either is invalid.)
Certificate checking would be strictly enfoced, but 1:1 correlations between certs and devices would not be, based on how the AP is configured. For people wanting to run "Open" connections, (Equivilent of unprotected wifi), a default certificate set of 1 private key (burned into the AP as part of specification compliance), and 1 public key would be ubiquitous to all devices, and would fascilitate that configuration. The over-air data would still be encrypted, just with this defacto key set. (Useless from a security standpoint because everyone has the keys, but useful from a protocol design standpoint because you can always populate the encryption type feild of the datagram header, even in "open" mode.)
The ideal situation would allow deployment of user-generated key sets right out of the box, built on strong encryption bit depths---
It would be the beginning of the open-mesh network everyone seems to want so badly, myself included.
Given that whitespaces have been enabled for public use, maybe I should order an FPGA kit and cobble together a proof of concept some time...
I really would like to see a non-profit org created to administer a project like this though. Could even use whitespace spectrum instead of licensing dedicated.