This is by far the most overlooked feature of being a part of the Google ecosystem. I don't think Microsoft or Apply can match the ability for me to give out a number to someone, not a special username and not something that requires you to have a particular device, Just a plain old phone number. People can call me on that number and I can pick up on
1. My Phone
3. My Tablet
3. Any Computer I am logged into <- This one is special
In addition I can call other people using any of those devices and pay nothing for the privilege.
Look. Apple can go buy DID's from any number of vendors and become a phone company tomorrow.
They don't want the regulatory exposure, and frankly, Google doesn't either.
Buying a DID and doing call forwarding is ridiculously trivial to do. You can replicate Google Voice on your own servers in 10 minutes using Asterisk or other Open-source projects.
Disclaimer: I'm the community manager for 2600hz, a company that builds an open-source stack for Telecom applications. You can mirror this functionality in 10 minutes using our stack. You can also do this with Plivo easily. It's not rocket science.
How much would it cost? And would I be able to transfer my Google number to my own setup (I have a really great vanity number and worked hard to get people to switch to it)? Also, would it let me do dial out with that number as well? How would the OS integration look like on iOS/Android? Could I get voicemail transcripts? Would setting this up use my "minutes" or "megabytes" on my mobile plan? If it's using megabytes how much would that be a per minute?
I am very happy with Google Voice so far, at least on an Android phone, but I from what I remember, iOS doesn't integrate with it as deeply. Also, I would love to be able to somehow ditch my "minutes" and only use the data connection for everything, but I am not sure if that's even possible today without carrying around a mobile hotspot.
Edit: also, GV gives me free text messaging. Can I replicate that with my own setup and how much will it cost? Do I have to use something like Twilio or Tropo for that and pay per message? What about MMS? Can I have those forwarded to my email?
If you choose to answer these, thanks in advance.
Edit 2: Looking at AT&T, they have decently priced data plans for things like tablets and gaming devices [1]. I understand that I can get a SIM card from them telling them I'll have a tablet on their network. Can I then buy an unlocked iPhone from Apple and do this type of setup you are talking about using SIP and something else for SMS and save money?
You can port the number to your DID provider. Most people use bandwidth.com but flowroute and voip.ms are also ones that are thrown around a lot. Most of them charge <$5 for porting.
OS Integration is possible only through 3rd applications (unless you have android which has a native SIP stack on the device) like Groundwire or Bria. It uses megabytes and I believe, if you're using G.711 (the normal cell codec) you'll consume data at 64kbps (peak). I've never gone over on using it but YMMV.
iOS integration is tricky because of the "keep-alive" strategies designers have to use. The VoIP app has to be able to notify on inbound calls, but that is usually only possible when the app is in the foreground. There are some cool hacks to get around this (I like the groundwire message personall).
Text messaging is dirt cheap; many providers will give you unlimited usage for very cheap (I've seen it as low as $.40 per DID per month). You could use twilio, but you'd be better off with tropo since it's more open (sort of, not really the time or place for that sort of debate).
Edit 2: Yes you can.
I think I answered everything :). You can do all of this for free (except for usage) if you install the 2600hz kazoo stack (here's the Chef-Solo install instruction set: https://2600hz.atlassian.net/wiki/display/docs/Deploying+Kaz...). You can also do a hell of a lot more than just call forwarding :).
We're fully open-source, it's free, and it has all of the APIs you're looking for when building kickass mashups.
A hosted version should be more widely available in Q1 next year.
Thanks! The only question I'm still wondering about is whether I can get AT&T to sell my just a data plan and then use the SIM card in an iPhone (or similar) without them forcing me to get a voice plan as well.
Forgot to mention, we wholesale minutes at $.0099/minute, and you can find lower rates from less trustworthy providers with a little googling. 99 hundredths of a penny per minute is cheap, but it's not call center cheap, if you dig my drift.
I saw the pricing. So, for a hypothetical 500 minute plan we are talking about $4.95 for these minutes + 235 MBs of data, plus $5-10 for a DID + $30 for the AT&T 3 GB plan + $0.40-1 for SMS. So, $45/month I can have complete control over my phone? That's not bad at all.
Two more questions: First, does any of this support E911 or whatever is used for cell phones? I know Google Voice doesn't do 911, but if this is going to be my only phone with only a data plan, that becomes more important.
Also how is the reliability of this setup compared to conventional networks? What are the chances that I'll be stuck somewhere and need to call 911 without the ability to do so?
E911 is complicated when it comes to VOIP. With providers such as Flowroute or Vitelity, you basically fill out a form with your current address/location and it takes some time to propagate to whatever system the dispatchers use. Basically, you can kiss E911 good-bye if you're using VOIP on the go.
Re reliability: VOIP, especially over a data plan, is not as reliable as conventional networks. I wouldn't rely on it for emergency situations. I switched from landlines a number of years ago at home and I've been at the mercy of my cable company outages.
FYI, I run a dedicated Freeswitch box and my providers are Flowroute and Vitelity.
We <3 freeswitch and run it at the core. On Kazoo we do Dash e911 via API so you can program all the DIDs in one fell swoop.
There are a variety of reasons why e911 sucks, not the least of which is the fact that if you don't populate it, the calls go to the Nation 911 hotline, which is slower due to not being local :/. Don't even get started on the topic of SIP based e911 on Softphones, it's a joke.
Flowroute and Vitelity are awesome; in particular the CEO of flowroute is a smart dude. Bayan was on VUC last week talking about the future of service provider networks (specifically that he envisions a world with SBC's). That's interesting, to say the least.
VoIP over a data plan is as reliable as your connection. That being said, use G.729 if you can as the lower bitrate holds up better over GSM/LTE networks. That being said, don't rely on it solely as there is no SLA on packets carried over the public internet.
I love Flowroute. Their web interface is much cleaner than Vitelity's. However, I wish they would take the initiative in supporting more open codecs other than G.729, G.711, etc. I dislike the G.729 licensing scheme. On another note, I was surprised about a year ago that Flowroute added support for T.38, yet there was no announcement made. In spite this, I haven't been able to get faxes to properly transmit and receive.
By the way, Kazoo screenshots look nice, definitely going to check it out.
T.38 In VoIP is tricky. We run almost every codec in Kazoo that FS supports (711, 722, 729, all the way up to Siren64Khz and h.263 for video), and I can tell you that t.38 has been the trickiest.
A lot of the transformations layer7 firewalls apply to packets, particularly SIP packets, can cause unexpected side effects with an extremely time sensitive application like fax. In short, moving packets involves buffering on some level, and fax machines are intolerant of any jitter or delay. Therefore, delivering media in t.38 becomes an exercise in conflict mitigation.
In summary, t.38 is active on a lot of providers who don't talk about it because heterogenous equipment might not play nice.
> I don't think Microsoft or Apply can match the ability for me to give out a number to someone
If they're willing to burn the cash (and Apple certainly has a lot lying around), they could. Google isn't a telco and they're paying CLECs (competitive local exchange carriers) for every phone number and for terminating every minute of every call that a Google Voice user makes. This makes Google Voice quite different from their other products - unlike gmail for instance, most of the Google Voice product actually comes from other companies and most of the per-user cost is probably going directly to these companies rather than to their own infrastructure and software development costs. Yes, the Google Voice software and infrastructure is still important but just not to the same degree as their other products. I often wonder if this is the reason why they only commit to making Google Voice free for one year at a time.
Too bad Google has never been that serious about these VOIP options. I mean, where's my "VOIP Call" button next to the normal Call button in my Android phone? Why isn't the "Video Call" button there as well, like Facetime is in the iPhone? I've been expecting this to happen since Froyo in 2010, and I'm still waiting for it to happen.
Google only keeps them deep in the Google Talk app, and I bet most people using Android phones have no idea they are even there, so it's almost like they don't even exist. Google needs to do a better job of deeply integrating VOIP into Android, and I don't care that carriers won't like it. Same goes for their "Google Voice" service.
The difference is that Google provides ad based services. It's important that users remain loyal to use their services. Both Microsoft and Apple are predominantly product based companies. They sell products and all that matters is whether the product is good enough to pay for.
1. My Phone 3. My Tablet 3. Any Computer I am logged into <- This one is special
In addition I can call other people using any of those devices and pay nothing for the privilege.