Slashdot Mirror
Bluetooth Wireless Devices Delayed
Sean Fitzpatrick writes: "From this article on www.arizonarepublic.com:
The future has been postponed, at least slightly. Bluetooth, the wireless technology to link computers, home entertainment appliances and other gadgets will not debut this summer as some manufacturers had hoped.
Story at http://www.arizon arepublic.com/news/articles/0721bluetoothon.html (free registration required)." The registration isn't too onerous, but be warned: it's a short article. The upshot is that Bluetooth won't arrive to Jes' Folks until at least the (actual) turn of the century, according to Joyce Putscher, director of a an Arizona research firm called Cahners In-Stat Group, and "several [unnamed] manufacturers." A shame, but when was the last time cool technology arrived conspicously early? If wishes were horses, then we'd already have Bluetooth.
YuNicks sez:
Nonsense: there's 802.11b, already running and on the ground. Coupled with IPv6 and the QoS guarantees it has, tell me what I can do with Bluetooth that I can't do with wireless Ethernet?
Bluetooth is NOT a wireless networking protocol. It's a wireless cabling protocol. Sure, it can do some of the same things, but it isn't intended for the massive amounts of data that pass through an Ethernet connection, but instead for the piddling bits and bytes that pass from keyboards to PCs and such. Bluetooth is low-power and will be adaptable to PDAs and cell phones; 802.11 last I heard required at the very least a PCMCIA card.
They're intended for entirely different purposes and situations; don't even think of them as competitors. 802.11 offers, as you note, the kinds of QOS -- throughput, robustness, security -- that corporations require to connect up laptops. It will make office networking much easier. Bluetooth isn't about that; it's about getting all the piddly little electronic toys you use to interoperate, exchange address books, and such.
----
lake effect weblog
{Network engineer in Chicago--looking for work!}
Also, try playing with IR. I've got a TRGpro and a Kodak DC265 that talk to my PC using IR. Not quite as magic as RF, but hopefully you can get to the end of the year without shelling out $2k for a developer's kit.
Just take it one day at a time and you'll make it...
You're right, 802.11 does use the same frequencies as Bluetooth. However, lots of folks have realized this and 802.15 is supposed to be a new standard, based on Bluetooth that's (more) compatible with 802.11. If you want to read the nitty gritty about what effect a bunch of Bluetooth tranceivers have on a 801.11 network, take a look at this: http://grouper.ieee.org/groups/802/15/pub/1999/Sep 99/Misc/BT-80211-Coexist.ppt The presentation only talks about the effect that BT has on 802.11, not the other way around. If I read this correctly, even with a bunch of BT piconets, 802.11 can keep chugging away. However, I wonder if the lower power BT signals will get blasted out of the sky by 802.11 packets.
Can't wait for Bluetooth? Then look into HomeRF. It's a mid-level wireless protocol that isn't going to be an IEEE standard and will probably eventually disappear, positioned as it is somewhere between the device-level connectivity of Bluetooth and the robust networking of 802.11b.
...
Mainly available to consumers in the form of Intel Anypoint and running at 1.6Mpbs, it's a decent solution for SOHO and hobbyist applications, and most important, it's in stores now, where a one-PC kit (no cards, USB pluggable) retails for around $60.
The minimal 802.11b "wireless ethernet" (or "Wi-Fi", ecch) configuration right now requires an expensive ($200-1500) base unit that's basically an Ethernet hub with robust IR ports, as well as IR Ethernet cards (~$100) for each PC to be connected. This will be fine for professional environments, who will demand the 11Mpbs throughput, but a bit steep for most consumers. (FYI, 802.11 is the protocol that Apple Airport and its PC sibling, Lucent's Orinoco, as well as future products like Cisco Aironet and 3Com AirConnect, run. For reasons probably related to "Steve Jobs", Airport base stations plus card can be had for as little as $400 total, while Orinoco and other companies' PC-compatible offerings run $1000 and up for essentially the same equipment. If all 802.11 products were available at the Airport price level, there would be no market for HomeRF at all.)
OK, if you are one of those "only the highest possible bandwidth will do" users, you will want to jump straight to the wireless ethernet offering, damn the pricetag. But for most people the most taxing thing they'll run is web-browsing, gameplaying, or Napstering, and since 1.6Mpbs is superior to almost any home internet connection, it won't bottleneck you. HomeRF gives you freedom of movement in your apartment or even backyard if you like for a reasonable cost.
I do expect Bluetooth to ultimately take over this market, since it will be available on many more devices than just computers, and may finally make things like remotely programmable air conditioners an affordable reality.
Bottom line: there are three wireless standards out there, and you're advised to read labels carefully the next couple of years. Me, I'll be trawling the 10%-off-returns shelves at Micro Center
[Disclaimer: I can't report yet on how well it works, I've just looked closely at the product so far.]
----
lake effect weblog
{Network engineer in Chicago--looking for work!}
I'm a QoS specialist at a vendor of policy-based QoS management software - trust me, nobody is going to migrate to IPv6 for QoS purposes.
IPv6 has precisely one extra feature for QoS beyond IPv4 - the flow label, which helps reduce classification load in core routers in RSVP networks. This is a rather unlikely scenario, for details see the RSVP over DiffServ drafts at http://www.ietf.org, in the issll group.
Apart from this one feature, all standard QOS features such as DiffServ, RSVP, COPS, COPS-PR, etc, will work identically in IPv4 and IPv6. The sole speed advantage of IPv6 is its more regular header structure; its disadvantage is that more memory references will be needed for each forwarding and classification decision, so there'll need to be good ASIC or network processor support (as in Intel's IXA architecture) before IPv6 can really take off. However, I expect IPv6 to happen within 2-5 years.
IMO people will migrate to IPv6 to support mobile IP efficiently (its routing header lets you efficiently route to mobile nodes without risking source address spoofing), huge user populations (think millions of mobile phones, tablets, cable modem/ADSL users, etc), and for ease of re-addressing (think companies that are merging or divesting business units).
802.11b is a completely different technology to Bluetooth, with much greater range, about 10 times the bandwidth, and greater power consumption. Not at all comparable to Bluetooth.
No, really. We've spent the last ten years plowing out tech at such an incredible rate that first-or-close-second-to-market has become a priority over other considerations such as quality, reliability etc. We're starting to see a backlash against that mentality in the non-OSS secotor (OSS has more of a to-market-when-I-get-around-to-it mentality) at long last. First OS X shows a slipping release date, despite DP4 demonstrating that it clearly works and could probably be pushed to market soonish. Now Bluetooth, which also works "well enough" to meet the current standard of quality, is rolled back. More time = better product. So I'm happy with the delay.
2 1337 4 u!