Slashdot Mirror
You Don't Know Jack about VoIP
gManZboy writes "Phil Sherburne and Cary Fitzgerald, two senior technologists over at Cisco, have written an in-depth overview of VoIP for developers and the like (not for everyone who's ever used a phone). Like Queue's earlier You Don't Know Jack about Disks, this article covers the history, the basic technologies, how they work, and where they're headed. If you found the blog post yesterday lacking, check this one out."
Unless you read /.
Waaaa, I can't find the link!!!!!
here you go
Trolling is a art,
I think Cisco is just ahead of their time, as usual. We're just catching up with where they could have served us back in 1999, when I introduced the NYC insurance industry to VoIP with a project using their new Celsius gear.
--
make install -not war
voip will take over. Voice can be transmitted at such a low bandwidth, and all the cost to make a connection anywhere in the world is the cost of your ISP. I think they have them, but you need to have some sort of program always listening on a port from your IP, and transfer incoming calls to a usb connected phone that rings. Then you'd have all sorts of spam bots calling everyone's IP, so you'd have a list of approved incoming IP's or a numerical code that allows your call.
I know I'm going to be modded up on this
Before you waste time trying to get VoIP (or paying for VoIP from a provider) going it is worth testing your connection to see if it can support VoIP calls at a reasonable quality. You might want to test your line at various times during the day... I get crappier calls in the evening.
Anyway, http://testyourvoip.com/ provides a decent free testing serice just using a web browser.
-ben
This is a great statement to read while eating some jumbo shrimp.
I don't think it's a fad, this year I am getting VIOP from Primus Canada. Finally I can say "screw Telus" because I am getting cable Internet from Shaw and phone through Primus. It's a good feeling.
--- to swing on the spiral...
What, indeed, do we know about Jack?
--
make install -not war
You do not need cisco for voip.
All you need is asterisk.
You will be suprised how many actual deployed asterisk solution exist.
It is here and now.
If you're in the market for a VOIP service, Geekbooks did a pretty decent comparison of different services. Does anyone have any other links?
So you're getting VoIP just to screw the telco? Are you actually saving money? How are you measuring ROI? What will your ROI be?
I'm just curious. I've had several places look into it but have never found any way to justify it.
Insert offensive troll-style sig here. Please mod or respond appropriately.
This is almost akin to a retraction of yesterday's blog post, which was indeed lacking. bravo.
Moo.
Last time I talked to Cisco they wanted about 75k (in USD) for a solution which had the main selling points of user login/call routing. ooooo!
Still no phone2phone encryption!
The unslient majority wants:
- video phones
- encryption
- Cell+Wifi in one device with auto switch over
- Server software that runs on Linux for those of us that like a standard back office.
Yeah, Primus VIOP is only $20 /m when whith Telus it's at leas $30 /m for basic service. Plus you can have a extra line for $4 extra that you can place anywhere in Canada. So even though I live in Edmonton I could have a local number in Toronto that I could make local calls to there from. Also people could call me on that line locally and it would ring here in E-town.
--- to swing on the spiral...
...is here!
and all the cost to make a connection anywhere in the world is the cost of your ISP
Unless you're one of the unlucky who has to use a DSL provider which requires you to pay for a landline to get said DSL service. Then you're stuck in a bit of a pickle. Hopefully that will change, I seem to remember hearing about laws regarding that problem.
Not just because of what Cringely said, but my phone works in a power outage, and still sounds way better than cell, for example. http://www.pbs.org/cringely/pulpit/pulpit20040624. html
No way, my company pays over $10,000/ month in long distance for our 1-800 nubmers and all the long distance we use. The bandwidth you could get for that price is pretty good.
I knew Jack.
He's actually kinda friendly.
/* We dance to the sounds of sirens and we watch genocide to relax*/
VoIP makes a lot of sense when your company already has frame relay to its international offices. Why pay for international telephone calls when you can do it over your existing network?
Internationally, though, voice is still a cash cow. That may last a while longer.
Voice over IP is more of an advantage for companies with elaborate internal telecommunications infrastructures. The VoIP gear is cheaper.
How well does 911, or your local countries' equivalent, work with it? When you dial 911 from a voip phone, does it report the location of the phone, or the billing location of the phone?
Best Slashdot Co
Sounds like your system is not set up correctly. You should be able to transfer a caller to (extension)#2 and send the caller to (extension)'s voicemail. There is another config to allow prepending a digit or * to the (extension) to send to voicemail.
As long as the phones and a voice gateway have power, the survivability feature should keep some voice services active in the event of power failure.
911 works well, as long as there is a gateway with a POTS line at each site. Otherwise, you've got to do the E911 stuff, and maintain the data.
Don't pick up the pho*(@)$*@&@!@ NO CARRIER
We tried VOIP awhile back but I do have an issue with reliablity. It maybe cheaper to make calls on VOIP but I had many drop calls because of network issues. Another issue was quality of Also during power outages unless your network and equipment is on UPS and generator you can't call anyone since there will be no power to your equipment. Unless you have a PBX and maintain the infrastucture for both and the PBX is at its end of life then I would consider VOIP.
There are many other questions and issues that needs to be raised before jumping to VOIP but these are just a few basic ones that I ran into.
Wow. As much as that tells you something about VoIP it is mighty far from in depth. Entire books are devoted to nothing but QoS and H.323. You don't know anything if you don't know these.
Sincerely,
Jellyvision, Inc.
Mine isn't. Well, it is for general usage, but my DSL is down for one reason or another a few hours every month. I've never had a cable modem but if it's anything like my horrible spotty Adelphia digital cable (which seems to be out a few hours a week, and has constant lags and glitches) I would expect the same.
Neither of these problems is so bad, but if a DSL glitch meant I couldn't use the phone either I would really be up shit creek. (I suppose most home VOIP users would also have a cell phone, but what about, say, businesses who rely on incoming calls?)
Land lines may be archaic but they are very dependable. Even when the power goes out, they're there. Since VOIP relies on both power and your broadband service-- both of which are prone to occasional glitches, especially if you live in a less-than-urban area-- I would never trust one to be my sole phone line.
Of course if I lived in an area where land lines were horribly expensive-- like the Caribbean, or areas of Europe and Asia-- VOIP would be a Godsend.
my password is private, but unchanged.
ENUM. <snip> ENUM calls for telephone numbers to be written DNS-style, rooted at the domain e164.arpa. So, 1.212.543.6789 becomes 9.8.7.6.3.4.5.2.1.2.1.e164.arpa. Interestingly, each digit is treated as a subdomain. This allows ENUM to ignore the nuances of country codes, city codes, etc. that vary broadly worldwide. When this address is queried, the DNS can return a specific IP address corresponding to the telephone number, or it can return a rule for rewriting the original number into some other form. For example, rules can be returned to rewrite 1.212.543.6789 as sip:36789@nyc-gw.example.net, sip:caryfitz@service-provider.com. ENUM offers the possibility to reuse the worldwide DNS for VoIP. ENUM is a standard set by the IETF as RFC3761.
Me: Beep boop boop beep boop boop boop beep "Hello, refrigerator? Yes, can you please order a jug of milk, some butter and a box of Kraft macaroni and cheese for delivery this afternoon? Thank you!"
Beep boop boop beep boop boop boop boop "Hello, stove? Yes, at 6:15, please turn on the gas burner in the lower left hand corner and place a pot of water on the burner. Use the recipe for mac and cheese for further instructions. Talk to the refrigerator for the necessary ingredients. Thanks."
+1 Insightful, -1 Troll. What can I say, I'm an Insightful Troll.
Where every Cisco VoIP system falls down is on the ammount of bandwidth required to support VoIP. From a telco operator perspective (voice or data) your greatest operational expendature is your bandwidth. Using IP or SIP costs you far more in bandwidth than is economic (when compared to alternatives). Yes you can multiplex voice and data but that takes even more bandwidth than doing it seperately! GSM is probably the most efficient way to carry voice over a digital channel. Does very well at 22kbit/sec. You even can do voice over GPRS at 33kbit/sec (the latency sucks, but you can do it). But try running a SIP session and it simply doesn't work. The protocol to establish the session and the overhead cannot be done on a low bandwidth channel. VoIP makes sence only when bandwith is free, but in the real world it isn't and the commercial imperative is to make the most of it.
First off, I found the article kinda ho-hum (yeah, I read it...what was I thinking RTFA?). I've come across better articles on the net from such odd sources as USA Today and such. I will admit though that the last page had a few things that made me go "hm", in particular how they remind us that monitoring this system will cost a pretty penny.
I've been doing research for a client that is wanting a VoIP/call center solution. When I started, it was fairly simple. Looking at the different services offered (Vonage, Broadvoice, Broadvox, Packet8 looking like the best solutions so far) and then I went to look at Cisco gear and see what it would be to set everything up yourself. And then I looked at some IP PBXs from 3Com, Avaya, Siemens, and Zultys. You know what guys? We've got quite a load of solutions out there for someone who wants VoIP. And these were all hardware-based...I didn't even bother looking at software-based solutions.
I'm finding this whole VoIP thing to be just as interesting as WiFi...a wild new market with everyone trying to establish a foothold (remember the dotcom days with everyone trying to grab as much marketshare as possible?) and then weather the storm and see who survives. Interesting indeed.
"He uses statistics as a drunken man uses lampposts...for support rather than illumination." - Andrew Lang
know Jack Schitt
Sweet so when my city will have 100% wireless coverage i could use my PDA to do VOIP basically for free. So in the far future we could all use those star trek chest thingies to communicate for free. Cant wait for the spammers to call me about penis enlargement pills. Captains log, earthdate december 14 2040 microsoft has a new discount on penis enlargement pills.
http://shit.slashdot.org/article.pl?sid=04/09/21/1 649216
Holy oxymoron batman!
Skype Is working very well for me. Free, VERY high quality calls to other people who have the program installed, and cheap, local rates to any land line/cell phone in the world!
I wrote a post some time ago about how I thought VoIP was not ready for primetime. I was subequently trashed by a self-affirming moron and rated down. Whatever.
Voice communication relies on time sensitive delivery of very small bits of information.
IP networks are designed to deliver large gulps of information in a non so timely fashion. Wht I mean by this is that in an IP network, equipment will deliver information as quickly as it can, but there's nothing the 802.x quite of protocols which inherantly facilitates predictably timely delivery of data. Timely delivery is governed by network and infrasturcture "health". Sure, there's QoS, but that ultimately gives very little benefit unless the network is under heavy load anyways, in which case, VoIP is a bust regardless.
Conversations can seem decoupled. Calling someone 1/2 mile away can introduce the latency that can be expected when calling overseas. It doesn't feel "like a phone" to many end users.
Jitter, latency (huge), and the general difficulty of "simulating a telephone" over IP services is what will prevent VoIP for taking hold until several generations of technology and a generation or two of home connectivity methods is introduced.
Contrast ATM networks, which are designed specifically to deliver small bits of information very quickly. These networks are ideal of VoIP.
Poeple don't have ATM to their houses, they have DSL or cable services which offer NOWHERE near the reliability of a typical voice network.
Someone can fairly realistically expect 1/2 of a building to be blown to pieces, while a phone in the other half will work. This is how reliable voice networks have been.
Within a company on a controlled LAN, VoIP can work because you have some control over the quality of the service. To the home, we are not close to being ready.
I've implemented VoIP switches since their initial introduction, I have spoken at international conferences on the merits and pitfalls of VoIP. I'm not trying to toot my own horn, I'm just saying.... I've used and abused these switches, phones and protocols, and I find them lacking outside tightly controlled environments. Across a vendor's backbone? Sure, no problem. Will I use it exclusively in my home? No freaking way.
From the article: "both data and voice could be carried on a common, packet-based network. This would simplify management by reducing the number of networks to manage, and lowering network facility and hardware costs."
This is total bullshit. When you integrate IP services into the core, your IP core suddenly goes from being a best-effort delivery system with maybe some packet priorization, simple and easy to understand, to a system that has to implement QoS
and tagged switching. In addition to the added complexity, this means all your routers, which are not nearly as stable as TDM devices, are now so critical that the amount of planning that surrounds maintaining them skyrockets. It also means your IP techs have to know about voice, and visa versa.
It doesn't reduce your hardware costs unless you cheap out on the implementation. Current DWDM and first-rate TDM switching technology is getting incredibly cheap compared to just 5 years ago, so ripping out old TDM gear and creating multiple networks over the same fiber infrastructure is starting to become cost-competitive at a hardware level. When people think about all the "savings" of not wasting bandwidth on the wire, they never seem to consider the recurring HR support costs that come along with that. Compare the price of just adding more bandwidth, people!
Using VoIP only seems to make sense if you've neglected your voice infrastructure like some carriers have, clinging to obselete equipment rather than upgrading to newer, easier to use, cheaper to upgrade, more standardized, systems, and you are up against a wall and have to rip it all out.
Some CEOs in some boardroom obviously decided that they wanted to see if they could ditch their human resources in the TDM area, and make their data folks do double-duty. Lured by the apparently cheap IP gear, they are now being surprised by the cost of making IP gear actually meet sane telecommunications standards for stability, and will be even more surprised when those retrained LAN administrators they hoped to replace their TDM professionals with start to be in higher demand, and they have to start paying them just as much and worrying about turnover.
I know, I've been there. Anyone who trusts a CISCO sales representative (or any sales rep) to sell you something that actually meets your requirements is a fool. Only now have CISCO finally realized that selling boxes that cannot do full line rate with any amount of filtering turned on is getting customers irked. Sigh. If only ATM had taken off enough to reach economy if scale in the 90's. That was a technology that was worthy of the integration concept.
(The precarious pile of standards and patches to standards that these VoIP people have exuded to make packet-based networks do something they were never intended to do also bears mention.)
Someone had to do it.