shyam somanadh

shyam somanadh

20p

13 comments posted · 1 followers · following 1

14 years ago @ MediaNama - Virgin's New Data Card... · 0 replies · +1 points

You don't need FUP on data cards because they have such awful bandwidth (average 2-3 KBps and peak at about 11 KBps) that it is simply impossible to download any decent measure of data on it. Their concern would be more about provisioning capacity on the network itself.

14 years ago @ MediaNama - Why The SIM Card Shoul... · 0 replies · +1 points

It is an interesting viewpoint on the NIDs, would love to see a more detailed take on this, more on the lines of pricing, memory and existing solutions.

14 years ago @ MediaNama - Updated: Facebook Usag... · 0 replies · +2 points

Orkut is a placeholder for Google for the time being. It has an abysmal market share in regions that count for a social networking product. I think they are just waiting for Wave to mature and letting go of the immediate gains for significant future gains. Of course, that logic does assume that Wave will do well, which we have to wait and see.

14 years ago @ MediaNama - Updated: The New Redif... · 0 replies · +1 points

Pretty decent idea to drop the indic language features. It is very likely that the uptake is mico-minimal; not worth the effort that needs to go into integrating the product into the main email product itself. Think about it: of the daily email traffic, what percentage would actually be represented by composition in local languages? Email is really the wrong end to tackle the problem.

14 years ago @ MediaNama - Q409 Call: Rediff Buys... · 0 replies · +1 points

That is an interesting angle to the API story, that switching from F8 to a REST-based API will somehow increase adoption. Shows how precious little integrated thinking happens in our online space. If you are in the apps space (those work using the API), there are only two platforms, F8 and Opensocial that are worth targeting. It does not matter whether your target developers are in India or if they are abroad. Why?

Because of the manner in which the app economy works. The app companies make decent money only at massive scale, which is obvious because they use CPM to generate revenue. There are companies out there which make about $200 or around that on a daily basis from the apps. But they also fall under the top 10 apps in either of the platforms. Developers simply won't bite if they have to write three versions of the same app. And the lack of traction just adds to Rediff's woes.

A tiny percentage of these companies actually make money, the rest die trying. And even those that make money have a hard time sustaining it. And running the apps are not cheap either, they require significant infrastructure and technology backing that is rapidly turning out to be very specialized a field (message queues, massively sharded DBs, key-value stores etc) now.

Coming back to the integrated thinking, I don't think anyone has a clue about how this will actually help the companies themselves in the longer run. If you do a cost-benefit analysis of it, I'd be very surprised to see if there is much filling up the benefits column other than sounding cool. Do they have any metrics on what usage does it require for it to be worthwhile the company's time in running it? What is the intended end-usage? Most importantly, most of them have trouble pulling off simple consumer-facing products in the first place. An API is a nice toy to have, but it is a toy all the same.

14 years ago @ MediaNama - MapmyIndia Compatible ... · 1 reply · +1 points

It is not surprising that MMI has started selling maps for Garmin. If you compare the current number of Garmins out there and the current number of MMI devices that have already been sold, even if MMI manages to sell the data in a year to even 1% of the Garmin user base, it would have a sweet deal on its hands.

INR 7K is very steep, if you consider the fact that this is just map data and not the device itself. Would be interesting to see what kind of agreement they have with Garmin on this. Incidentally, I have seen a Garmin showing a decent level of detail within Delhi (http://tr.im/ij4N). It is cheaper to buy a 180-day license of the phone software at less than INR 2K, though you would wind up paying for data if you are on roaming in India, which you can still work around by preloading the maps.

MMI has a major license-related problem on its devices, which makes them really expensive. Their cheapest device starts at INR13K, while Satnav's most expensive device is at INR 16K. So they can't really compete on price.

From what I understand, the question of mapping data being interoperable does not really arise. Underneath the surface they all share the same shape files and tiles to overlay the details of the location. It just boils down to how do you want to do your distribution of your map data.

15 years ago @ MediaNama - It's Official: Be... · 0 replies · +1 points

It is a smart idea to not spend much on marketing in these products in India. There is simply no market for these things out here. You can pump in a boatload of cash in SEM and drive traffic, but that kind of traffic acquisition cost can't be sustained in the long run, even if online advertising had not fallen off a cliff. The market size plateaus off at around 3MM page views per day for the properties at reasonable cost/traction and that is way too tiny a number to monetize, especially when the inventory quality is pathetically poor for such properties compared to content inventory.

15 years ago @ MediaNama - Editorial: Should We M... · 0 replies · +1 points

As a publisher you are already accountable to what you publish on your site, with or without moderation. But being answerable does not mean you will get thrown into the slammer for allowing comments, but allowing anything and everything to be said is an issue.

From previous experience in the matter (where we got dragged into a cybercrime cell investigation into some malicious comments), what I remember is that as long as you log the data (unpublish comments, don't delete them) and provide them if the court asks you for it, you are pretty much in the clear. If, as a publisher, you allow comments to be carried on your system that will get you into trouble as any printed material, the issue is not really with blogging or freedom of expression.

What stands out about the case is that the community owner is being shown as the intermediary (is this accurate, especially when the court has given enough leeway to the intermediary?), than the platform itself. This will, at least in the short run, open the doors for needless litigation regarding the internet and some sense will prevail, sooner than later. Incidentally, the court has only said the person has to defend himself, than pronounce him guilty. In all probability, with decent enough lawyers, he should be able to walk away from this.

An expensive irritant? Certainly it is. Death to freedom of expression? It certainly is not.

I guess an ideal solution would be to have a first level of redressal that does not involve the courts. Something like a mandatory offensive content contact link on every website, using which you can notify the platform/community owner about content you may think is offensive. If the requests received on such mechanisms don't get a response or does not get acted upon, it is only then then road to the court is available to the plaintiff.

Of course, this means you will need to put away even more resources to monitor that channel of communication and be swamped with requests, but it can be much better than being dragged to court every time.

While on topic, it is a good time to point out that you don't have a ToS anywhere on the site. Get one ASAP and ensure all content/comments are in compliance with that :)

15 years ago @ MediaNama - NDTV Convergence Launc... · 0 replies · +1 points

The site itself has been built by Tekriti (the same guys who did my.indiatimes.com and itimes.com for TIL?) and it is pretty much a souped-up white label deployment of Tekmedia, as seen at glued.in.

NDTV is very much an ASP.net shop, so anything that is on LAMP is a vendor deployment. Rest you can put people and places together and piece together the story.

What is even more interesting is that the videos are actually being streamed using Akamai's CDN than the traditional progressive download route that everyone is using.

I will confess that I don't know what kind of deal NDTVC has with Akamai, but in most cases, serving streams wind up being a whole lot more expensive than serving clips through a progressive download.

The clips on NDTV.com are also streamed the same way. Unless they have a huge issue with partial play outs (ex: only 20% of viewers are watching 90%+ of any clip), streams will be more expensive.

Pre-rolls are done via Vdiopia (looks like different inventory from ndtv.com), which should get them decent (but lesser) money, but not enough probably to cover cost per stream served.

It is one of those crazy situations where you really don't want to be a runaway success and you don't want to fail either. What exactly is the sweet spot? Your guess is as good as mine.

Don't be too surprised if all NDTVC network sites soon start playing clips only from Tubaah, they share common video IDs in any case right now.

15 years ago @ MediaNama - NDTV Convergence Launc... · 1 reply · +2 points

Quick, pass me some shades, the innovation light is blinding me :-)