November 27, 2005

Regulating the Dialing Plan!

Recently there was a news item that Indian DoT has issued “notice” to Bharti, alleging that Bharti is conforming to the National Numbering Plan 2003 decree. Apparently, Airtel offers both postpaid and prepaid service options on the same phone number. The user will select the option by dialing “#1” prefix. DoT claims that this violates its NNP. Airttel claims that it is abiding by the licensing agreement.

It is astounding that this will be a point of debate. So I tried to locate NNP 2003, but failed. But there is a position paper from TRAI and it seems to suggest that dialing plan is part of the ruling. Since the dialing plan is an understanding between the users and their service providers, what is the social benefit in standardizing it? Does spped dialing violate this ruling?Why would TRAI, which has made rational ruling on VoIP matters will develop this kind of regulation?

Posted by aswath at 11:00 PM | Comments (0)

November 26, 2005

Inexpensive RAID Software

Washington Post has a story about RAID hardware and software for personal use. Of special note is a software:

Shadow 2 ($30) from NewTech Infosystems Inc., or NTI can also copy selected directories of your PC as soon as they are created, functionally mirroring them onto an external device such as a hard drive, USB key or even an MP3 player like the iPod. Conventional backup software copies data only once a day or once a week, so the backup is rarely up to date with newer, and often the most essential, files.

Shadow 2 also makes it easier to restore your data. Other backup programs store layers of incremental backups in a proprietary format that only they can unlock and recreate. Shadow 2 preserves data in the original format, simplifying and speeding the process of copying your data back.

Through Dec. 31, Shadow 2 software is being sold for 99 cents, making it both inexpensive and easy to make data backup one of your New Year's resolutions.

Posted by aswath at 09:12 PM | Comments (0)

November 24, 2005

Enough Already with Pay Per Call

Om has posted some information on Google’s Click-to-call beta service. Alec and Venkatesh have posted their thoughts and have speculated how Google could have implemented it. I am decidedly not excited, though I am supposed to be in a festival mood.

In a nutshell, when a user clicks on the phone icon that is displayed in a text ad, a popup window requests the user to enter the phone number at which Google can initiate a third-party call connecting the user and the advertiser. Not being evil, Google assures us that they will maintain the user’s privacy. We are also informed that the call is their “dime”. Now it is not clear how long will these two last. It is reasonable to expect that eventually Google will charge the advertisers for the call. When that happens, wouldn’t the advertisers want to get additional information on the callers? After all, even the Bells have to share callers’ information on toll-free calls.

Om, Alec and Venkatesh have speculated on the specific mechanism used by Google. It could be native PSTN system based on TAPI or it could be some sort of VoIP platform. Whatever it is, Google has inserted itself in the middle and as such it is violating the end-to-end principle. That is why it is not so exciting. Now instead of this Click-to-call icon from Google, consider the case where the advertiser places a “Call-me link” as suggested by FWD. Same click-to-call feature is realized. But Google gets only pay-per-click revenue instead of the inflated pay-per-call. No user information needs to be exchanged. And best of all, end-to-end principle is upheld. Now it is time to be jolly.

Posted by aswath at 02:17 PM | Comments (1)

November 18, 2005

Simply Relevant

A couple of days back, fellow blogger and friend Alec told us that his company iotum started beta testing of their first offering relevance engine. In case you have not read about it it collects information about your reachability status and using historical data, makes a decision on the importance of an incoming call and directs it appropriately. Once this basic capability is available, one can envision, as they also suggest, them enhancing the engine with AI capability for handling the proper way of handling an incoming call. If you live in 514, 416, 613, 905, or 519 area codes you may be eligible to sign up for the beta.

They need to restrict the trial to these area codes because the Class 5 switch and its call processing software need to consult the relevance engine and that mean cooperation from the carrier is needed. They can extend the trial to VoIP providers. Given the dearth of meaningful applications in the VoIP domain one would think VoIP providers will be more than willing. After all isn’t FWD billed as a “community lab”?

All these logistical and business difficulties suggest that we have not really moved away from the PSTN paradigm. But a device like PhoneGnome allows us to do just that. Since PG behaves like a Class 5 switch AND a SIP Proxy, we could make the PG consult the Relevance Engine is consulted. Given that iotum and PG are known to each other and that they are related through Andy, it is very likely that they may be working on this as we speak.

Posted by aswath at 07:01 AM | Comments (0)

November 07, 2005

Memo to Mr. Whitacre

Dear Mr. Whitacre:

You are probably surprised to see the memo pass through your filters and land on your desk. I am sure you will be more surprised when you find out that it is related to your bravado comments about owning the pipes and that those who profit from using these pipes should share some of their revenue with you. I am sure you are aware that your comments have been vociferously decried by the same service providers and also in the blogosphere. My aim in this memo is to suggest what should your position be that is reasonable and more importantly, be consistent to the technology.

In the Business Week interview, you have said that others can not use your expensive pipes for free and that you need to be compensated for using those pipes. You seem to have in mind companies like Google who probably may be benefiting economically. It is not clear what your disposition is if a non-commercial entity or a government agency is the one who is using your pipes. Logic suggests that you will demand financial compensation from them as well; but the context suggests that you are not concerned about them. It is not clear how one can justify this dichotomy. In the telephony world, you do not charge differently based on the “importance” of a call. A call is a call is call. So how do you justify a different model in the Internet?

Predictably, the affected service providers were quick to express their objections. Vonage’s Citron concludes that you are either going to block his traffic or are going to charge him for carrying his traffic. Skype’s spokesman remarks as do many in the blogosphere that the customers are paying for the access and that should cover for all the uses. Amazon spokesperson is more charitable. He recognizes the need for SBC et al. for their own existence. But, he adds, it is an exercise of market power if they extort fees from internet companies. We all know that Internet is really a concatenation of many networks (after all the original name was “Catanet”) and that free exchange of traffic between these sub-networks is an axiom. As Internet was commercialized, a natural order arose and some rules were established. Occasionally these peering arrangements disputed as evidenced by the recent fracas between Level 3 and Cogent. But as Tom Evslin points out there are mechanisms in place. But it is worth pointing out that none of them take into consideration the criticality of the application or the source of the application.

At this point you might think Skype/eBay talks about pay-per-call (presumably higher than par-per-click) and wonder why application independence does not apply to them. You may also wonder the deafening silence from the pundits on this matter. Sometimes I wonder too. But I have concluded that if they truly believe in this, then they are going to be sorely disappointed.

Let me also tell you that I can not totally dismiss your desire for a return on your investment. It is a legitimate one. But it has to be done in a manner consistent to the technology. The business you have chosen is that of providing access. This means, as far as you are concerned, your users are consuming resources at the network layer. You are free to charge for the resources consumed at that layer and no more. You can levy fees based on the amount of bandwidth consumed. If you provide QoS, then you should be able to charge based on the bandwidth consumed according to the level of QoS. For example, in many countries, ISPs place a cap on the amount of downloaded bandwidth per month. In Hong Kong, an ultra high-speed access provider places different limits on bandwidth consumption for domestic and international end points. On this matter you will find friends in unusual places (Geddes, Robertson).

But unfortunately, in US such limits have not been placed. It looks like you want to change this. Then you said state this explicitly. I should warn though that it is very likely that the market may not go along, especially if there is at least one alternative service providers who has a different revenue model.

Also don’t try to spin your way out of this mess by suggesting that this is applicable only to the new network that you are deploying. This will get you in more trouble. You know in Cricket, spinners are effective in getting the batsmen out. But this assumes that your team is good at tight fielding positions. Even then, a batsman in good form can score high runs. The Internet is full of batsmen in good form and trying to spin your way out will not be effective. So you need to bring in the pace bowlers and deal straight. My concluding advice is that you should pray at the alter of Layering Principle, keep cost based tariffs and hope that the Invisible Hand favors you.

By the way, you will benefit from reading what other bloggers (Saunders [a], Saunders [b], Stastny [a], Stastny [b], Pulver [a], Pulver [b] and references therein) have written on this topic.


PS: After I drafted this note, I read your interview with Newsweek. It looks like you are backing away from your previous claim. By the way, it is nice you enjoy iPod. But I am surprised that you use VoIP. Why? What did you like/dislike? Have you thought of porting the things you like to the PSTN world?

Posted by aswath at 05:22 PM | Comments (0)

November 05, 2005

Wild Mushrooms

Only yesterday Om told us about WaxMail, a plugin for Microsoft Outlook lets you add voice messages to your email messages. Today there is a story about NowPos. It is an email service that allows one to send voice recordings to any email recipient. There is a big difference: the recording is stored with the service provider and the recipient has to download it from there. This being the era of Web 2.0, an advertisement announcement will be played. Don’t worry it will last only 10 seconds.

Now GMail allows 10MB (which is like 10 minutes of MP3) attachment. Why would I need yet another email just to send 3?

To use one of Om’s lines, monsoon has arrived. Will we see more wild mushrooms?

Posted by aswath at 02:23 PM | Comments (0)

November 02, 2005

Evolving Gnome: A Model for the Incumbents?

Yesterday, TelEvolution the company behind PhoneGnome announced an enhancement called SoftGnome. If you recall, PhoneGnome is a soupedup ATA that can connect either to PSTN or to a VoIP network. A clever aspect is the VoIP id is derived from the PSTN number. So PhoneGnome users always dial the PSTN number to reach the intended party and the device picks (based on the configuration) the appropriate network. But it is known for its ease of configuration – plug and play; no need to administer the phone number or anything else.

It has been widely reported that PhoneGnome is based on Sipura 3000. Apart from its ability to support the dual network connection, 3000 can also take an incoming call from PSTN and switch over to the VoIP network or vice versa. SoftGnome takes advantage of this is able to present the PSTN call to any complying SIP-based softclient that is connected to the Internet. From all accounts (by Alec, Andy, Phoneboy and Voxilla), SoftGnome is just as easy to setup as PhoneGnome is. But unlike PhoneGnome which is billed and marketed as a product, SoftGnome is a service and there is a monthly subscription fee (after the free one month trial period, the fee is $4.95 per month).

I am not able to figure out why they have moved to a service model. If my understanding is right, most of the function is performed by PhoneGnome, except possibly for the initial assignment of credentials. It has been reported that eventually an API for PhoneGnome will be made available. So can’t a third party develop an application that duplicates this?

There is an additional observation I would like to bring to your attention. PhoneGnome, just like any other ATA is really performing the functions of a Class 5 switch. PhoneGnome is able to add additional functions like voice mail, voice mail to email and now external notification of PSTN call activity – all for one low price of $120 (I am ignoring the service fee for the reason I have already stated). This means that PSTN operators can easily offer these services as well and due to engineering advantage, the capex should be much less than that for a standalone device. If only the incumbents follow this path, instead of talking about blocking VoIP or introducing VoIP in stealth.

Posted by aswath at 03:34 AM | Comments (0)

Copyright © 2003-2009 Moca Educational Products.