Rethinking the Indian Telecom Sector

I have been observing the Telecom/Internet Infastructure in India both as a former vendor to service providers and also as a customer. I am of the firm opinion that in its current state, the sector will not grow and also shall hinder the progress of the rest of the country by not fulfilling its role as a connectivity provider.

I feel it is time to break the sector proactively and nudge it to bring in disruptive innovation.

Run-down Telecom Infrastructure

The Indian telecom sector is bridled with so many inefficiencies. The telecom companies have no incentives to create innovative offerings, let alone raise their quality of services.

Entangled sector
Entangled sector

Part of the problem might lie in the fact that creating infrastructure for a pan-India telecom service from the ground up is a huge task. Maintaining whatever infrastructure that has been created is even more of a nightmare.

Every telecom company in every circle needs to dig ditches across cities, towns and villages. Along highways, waterways, railways but mostly in the air on tree tops, roof-tops and even on electricity poles. The CapEx sunk on this infrastructure goes waste many a times, due to the whims and fancies of individual local bodies who decide to fix a road, dig a tunnel or just embark upon some ‘developmental’ work. The fuel costs that goes into powering the diesel generators, inverters for the base stations and repeaters easily consume most of the OpEx.

With the infrastructure in such a mess, no wonder consumers are short-changed or presented with a lack of alternatives.

There are other issues that contribute to the mess too. The state-owned telecom companies serve more as India’s equivalent of a Social Security program rather than efficient and innovative Service Providers. Even ignoring probable corruption, the very nature of their slow and staid operations does a huge disservice to the young and dynamic nation.

Suggestions for a viable and efficient sector

Now, after a thorough panning of the existing situation, here are some suggestions for fixing the sector.

Photograph by Ken Banks
Efficient Infrastructure

Break the state-owned operators into two parts initially – one, to pool all their infrastructure assets, the second to hold all their operations including subscriptions, plans, etc. The operations part could be further broken into smaller units – a la Baby Bells.

The infrastructure, including highly valuable land, buildings, country wide fiber, last mile copper, etc should be vested with a new company probably called – The National Communications Infrastructure Company. All the private operators should also be asked to hand over their infrastructure assets for a corresponding stake in the new company. This company to be run by professional management drawn from the government, private businesses, academics should be the sole custodian of creating, maintaining and innovating country wide communications infrastructure.

From no MVNOs to only MVNOs

A Mobile Virtual Network Operator (MVNO) builds consumer services on top of leased infrastructure – both last mile and upstream. Currently, India has no MVNOs by design.

The Infrastructure Company should be able to lease out the infrastructure to all the existing players under a compulsory mandate. All existing telecom operators should transition to become MVNOs.

With the costs of capital coming down, there could be new special purpose MVNOs who provides only B2B services, M2M services (very relevant in the context of 100 Smart Cities/IoT),

P2P services. With decreased responsibilities, and increased focus, the new Infrastructure Company as well as the new MVNos are bound to innovate and create valuable products and services for their users. The consumers benefit from a decent competition and reliable services.

The society at large is benefitted by a decrease in the number of road-cutters. Mother earth will be happy because of reduced pollution.


  • McKay Savage (
  • Ken Banks (
  • All logos for representation purposes only

Bazaar in the Cathedral – Changing faces of Open Source Software

Bazaar and the Cathedral

Some time ago – not too long ago – contributors to Open Source Software – nay Free Software – used to be highly motivated geeks mainly driven by the need to scratch their personal itch. Swarms of such geeks got together in the nascent stages of the Internet and built solutions such as GNU tools, Linux, Apache, Sendmail, Perl, Python, Ruby, GNOME, Debian, Slackware, etc. More importantly they built communities. LUG meets, Install fests, mailing lists, IRC were the places that inspired and helped Free/Open Source Software take roots.

And that was when individuals were looked upon with great respect and celebrated amongst these communities as heroes for creating cutting-edge software. A Marcelo Tosatti, who at a ripe age of 18 would decide what features would go into the Linux kernel, an Alan Cox, whose decision to take a sabbatical to study management caused a few ripples, Miguel de Icaza whose influence and contributions were stuff that made legends, Rasmus Ledorf, Monty Widenius, Brian Behlendorf, Richard M Stallman – names among others who were recognized and who inspired thousands others to contribute their mite to Open Source. This was what Eric S Raymondtermed as the Bazaar as opposed to the proprietary and big company Cathedral model.

Cut to the present. Open Source Software is mainly driven by commercial interests. Contributions based on an individual’s interest has decreased radically. Developers responsible for path-breaking tools and packages are no longer recognized by name. Instead, a Hadoop is Yahoo’s Open Source project, Go was written at Google, Docker, Chrome, etc where the individual’s name is lost in a marketing scheme of things.

So, is this change a bad thing? Has the community interest lost out to commercial interests? Are individuals no more motivated to contribute? Do idealism driven interests matter anymore?

I believe that this change has benefited the community as a whole. When there was minimal or no commercial interest in the open source software that one contributed to, the individual’s interests had to be subsidized by the organization he worked at. Hence the profiles of the contributors were more like academicians (eg Alexey Kuznetsov), students (eg Linus Torvalds), hobbyists (eg Bruce Perens) or committed idealists (eg Richard Stallman).

With companies, corporations and non-profits backing newer open source projects, there is a slew of cutting edge technologies available for everybody to use and improve.

  • This has ensured that the same technologies that helped Google gain scale and solve problems are available to some students in a remote town in India to cut their teeth with.
  • This has helped seed thousands of startups who wouldn’t need to license expensive technologies just to attempt and realize their ideas.
  • This has helped lay consumers who use such open source technologies without knowing its technical intricacies. (eg Android).

This has also helped companies and corporations benefit from the value additions other companies/individuals make to the projects. Corporations also stand to gain from a huge talent pool that have already demonstrated their expertise using the same tools that their engineering teams use. It also helps that these open source software face almost no entry barriers into billion-dollar companies, which earlier was the case.

Nowadays, it is a given that the platform of choice for any application to be developed has to be invariably open source (Linux, Chrome, MySQL, NodeJS, etc). Free/Open Source Software also levels the playing fields for different sized companies alike.

So, the changing face of Open Source Software development from a Cathedral to a Bazaar to a Bazaar-in-the-Cathedral model has had a profound impact and is definitely benefiting the community as a whole.

Photo Credit:

First posted on LinkedIn at

Comments on the Government’s Draft Open Source Policy

It is heartening to note that the new Government is trying hard to involve people in more and more of their policy initiatives[1]. One such initiative is their “Policy on Collaborative Application development by Opening the source code of Applications“.[2]

Since I have extensive experience working with Government Organizations to deploy various solutions, I take the liberty to comment and suggest some points that augment the central idea in their draft policy.

Need for standardization and simplification:

The foremost problem I see in order to create a culture of developing and using open source components are complex rules. Though there has been a start, a lot more simplification of rules is the need of the hour. In my experience, though we started out with a good set of documented processes, we finally ended up translating Swamy’s Handbooks and Manuals to Software Code. Very obviously the software turned out to be heavily customized for the usage of the organization.
Simplified and standardized processes across government organizations help in developing reusable and easily maintainable code. This, I think is a must do to encourage proliferation of open source culture.

A cloud based platform for apps
One of the main problems, most Government organization face is that they do not usually have dedicated IT personnel and even the ones they have are inundated with a lot of work. This leads to under-maintenance, support starved and no motivation to use scenarios.

A cloud based platform similar (in concept) to Salesforce[3] could go a long way in helping developers integrate components and solutions quickly, while easing maintenance nightmares for the organizations. Of course, the platform should be based on Open Source stacks.

Style Guides, Aesthetic Sense and Intuitive Design
Currently User Interfaces, most documentation (including this Draft Policy) suffer from just a “adequate-is-enough” syndrome. Recently there was an instance which made the rounds on social media, where a printout of a spreadsheet was pasted in another spreadsheet (for compliance) and circulated. As long as the real benefits of “digital” content is not realized by the user community, most of the efforts go useless.

Towards this, educating users, compiling and compelling adherence to style guides, inculcating an aesthetic sense towards creating content should be invested upon. Developers need to prioritize intuitive design, localization from the ground up for gaining maximum value from these efforts.

The Policy should consider these factors and put in a set of holistic guidelines, so that the entire program can be sustained and expanded.

I worked as a contractor for BARC, CSB, University of Mysore, BSNL to develop their e-Governance and other software for almost 10 years since 2001-2011. For most of them, I began by convincing them of the need to use open source software. I also committed to deliver and delivered the source code for software I developed.


Photo Credit:

Frugal Innovation and Android One

Frugal Innovation

Android One looks to be the perfect example for global companies pondering overFrugal Innovation. The project seems to have got quite a few things absolutely right.

  • The consumer price point seems to be just right to target the lower and the middle income population which is a significant chunk of India’s whole.
  • The assembly is done locally by local partners.
  • The specs look highly competitive.
  • The software and some features are tuned to Indian needs and sensibilities.

Though they do not make money directly, Google’s monetization strategy from services like Search, Youtube, Mail look pretty sound and they should add significantly to their bottom-line from these markets. Especially since they could have a first movers advantage in the nascent vernacular markets.

There is an obvious lesson here for tackling the ‘Emerging Markets’ conundrums that lot of product teams in global companies face.

  • Deciding what features to strip so as to de-value their cutting edge innovations to make it affordable in these markets (eg Car manufacturers)
  • Bringing their innovations as-is with a significant cost reduction and risk “brand dilution”.
  • Sell products targeting a premium and niche customer base and fail to gain suitable market traction.

Google seems to have found a solution for all these difficulties and might just go on to create a customer base of the next 5 billion smartphone users.

Can this be a template for others? Only time can decide.

Photo Credit:

Originally posted on LinkedIn at

Apple should buy Amazon. Here’s why!

Well, not the whole, but their entire technology division – including cloud and hardware.

‘Apple is a great hardware company, but are clueless when it comes to software’ – is an opinion that is shared across the industry and has almost become a cliché. This may not be really true as they were the ones who figured out how to adapt Unix for desktop usage.

However, when it comes to building data-driven products and services, their capabilities seem to be woefully inadequate. Even their cloud engineering prowessseem to be still lacking.

OTOH, Amazon is almost a monopoly when it comes to IaaS and PaaS. Also, their technologies are now deemed standard, with rival efforts like OpenStack copying their API, heuristics, etc. But, Amazon seems to be hitting quite a few blockers when it comes to consumer hardware. Though Kindle is by far their most popular hardware, the other devices haven’t captured consumers imagination like it has for Apple’s products. Amazon’s efforts at replicating Google’s services like Play, Store, etc have been pathetic with very few takers.

Amazon’s cloud services are the first choice for developers and start-ups when it comes to the cloud/backend services. Apple’s iOS platform is the first choice for developers when it comes to building apps and MacOS is increasingly the personal OS of choice for both developers and casual users.

Technical synergies aside, Amazon has been having quite a lot of problemsincreasing their profit margins. While most of their costs are attributed to the investments being sunk into developing new products, their retail operations seem to be making quite some money.

As on date, Apple(AAPL)’s market cap is around USD 640B and had cash reserves of about USD 160B last summer. In comparison, the entirety of Amazon (AMZN)’s market cap is around USD 142B. Though nobody knows the valuation of their technology business, some estimates say it might be around USD 50B by 2015. So Apple can easily afford to buy Amazon’s technology division and still have cash to burn.

Jeff Bezos could use some other muse. Not that he is short of them already!

When the chickens come home to roost!

14 March 2025

Hi, I’m Gyanesh Harindra – a 36 year old investigative journalist with The Hindustan Dak. Of late, my efforts at digging for the truth are throwing up a lot of uncomfortable facts and questions. Let me take you through the journey that has left my brain think it is at the wheels of a rudderless boat.

12 January 2025

It was a chilly Sunday morning. I was getting ready to meet the mysterious caller who had called me last night. The caller had sounded very excited and worried at the same time. He had mentioned about some important piece of information he had in his possession. I thought it was odd that he wanted to meet me on the 0710 Chennai Express. In my career so far, this was the first time somebody wanted to meet me while travelling.

I quickly finished my breakfast Soylent and hopped on my bike. It was a good thing, I remembered to plug the bike for charging as the first thing when I woke up that morning. I was just in time to catch the 0645 metro to the railway station.

I was wearing my red pullover that served two purposes – to beat the cold and to identify myself to the mystery man. My next seat was occupied by a meek looking teenager, who was engrossed in watching a kannada movie on his Personal Access Display Device. As the train started to move, a short and stout bespectacled middle-aged man – with a prominent religious mark dotting his forehead and temple making him look very spiritually inclined – came and asked the boy sitting next to me if he could switch seats. The boy though taken aback, silently nodded and just as silently went and sat in the man’s seat a few rows away. My new neighbour made himself comfortable and slowly started searching through his pant pockets.

He took out his mobile phone – the model looked at least a few years old – and proceeded to check his messages. Quite suddenly, he turned to me and asked me if I had a spare memory pin. I understood that this was the mystery man who called me yesterday and had asked me to get a couple of memory pins. I handed him an empty memory pin, I was carrying. He inserted that into his mobile phone and started tapping furiously. He kept the phone in his shirt pocket, turned to me, gave back my pin and said – “Namaste, you must be Gyanesh”. I replied “Yes”. He followed:

“I’m Yethiraj Shodhan. I’m working at the UIDAI as a Project Specialist for the past six years. I know about your RTI applications seeking information about the recent Guttahalli property disputes.“

My eyes lit up expecting a stash of information about the most famous case in recent times. However he said “I’m sorry I do not have anything specific to the case, but I do however want to tell you something more sinister that is connected to that case. What you do with the information I’m about to reveal is upto you.”

“As you know for the past few years the UID as a platform gained quite a bit of steam and has established itself in an unassailable position. Every government body uses UID to authenticate and authorize, every financial institution tags every transaction with the UID, telecom companies use it to authorize their services, transport companies use it to authenticate online and book seats, toll plazas use it with their pre-paid cards to allow vehicles through their poll booths unattended, entertainment centers like multiplexes, amusement parks use it to verify access. UID is so ubiquitous, that people have long forgotten all the opposition and brouhaha it faced in its initial stages.

“UID has proved to be so effective for law enforcement that agencies have forgotten investigation techniques without using UID.”

At this point, I was a bit apprehensive about his credibility. However I asked him “Isn’t UID not supposed to track people? Wasn’t that guarantee provided when this was approved by the parliament way back in 2014?”

“You’ve been asleep at the door, my friend. Four years back parliament approved a one line amendment that allowed tracking of UID transactions, of course, to be closely monitored by an oversight committee. This was followed by the establishment of a new specialized agency called Internal Intelligence Unit. They are eponymously called just In2. Now, In2 owns a data center that has a peering arrangement with the UID data centers. Actually I’m not sure if there is only one or many, though.”

“I act as the liaison for UID and In2, and I know first hand how the UID has become a boon for law enforcement. For instance, the police identified the perpetrators of the bear run on the stock market within minutes and arrested them almost red handed at their own premises. Trailing and IDing suspects has become a no brainer now. The general law and order has improved quite a lot because of the UID, though the government doesn’t credit it publicly.”

“However, I’m not here to advocate for UID. I’m here to do just the opposite. There is a cabal of politicians, intelligence officers, businessmen, bureaucrats who are subverting the whole process and systems for their own benefits. Though there is a parliamentary oversight committee that meets once every month and goes through the summary of the logs prepared for them, they aren’t equipped to cross examine and go deep into the workings of In2.”

“The additional director and quite a few others are fixing these loopholes to their benefit. They have systems in place that ride on top of In2’s systems and obtain comprehensive information about any person who has a UID number. They also cross link the bank account numbers, PAN number, Voter number, Insurance, Licenses, business permits, Professional membership numbers, private clubs memberships, credit/debit card numbers, employee numbers and what not, to create fingerprints of private citizens and businesses. This information is then sold to whoever comes looking for it including political opponents, business competitors, stalkers, celebrity hunters, underworld goons, marketers and even foreign spies.”

“The way the cabal have stitched these systems, it is very easy for them to monitor the movements of any person entirely remotely without the tracked person coming to know about it in anyway. In fact the case you’re pursuing has the potential to blow the lids of these operations. I actually tried to piece together as to what would have happened on the case. This is my theory.”

“The Professor who had inherited a large property had no heirs. He wanted to give away the property to some charitable organization. He in fact had registered a will to that effect. But once he died, a new will was found which was linked to the property at the registrar’s office. Of course it was authenticated using his biometrics. This ‘will’ bequeathed his property to some individual who claims he had served the Professor as his secretary. But this is hotly contested by the charitable organization who says the Professor had indeed reiterated his intent to leave his property for the organization just a few weeks back before his death.”

“There is no doubt that there is some foul play in the whole story. I recently discovered that the same cabal had hired some underground hackers to create a nano device that grabs signatures from the wire while they are being sent to the UID authentication servers, which they can reuse for various forgery purposes. They’re very discrete in using this and I believe they’re using it only for a limited number of high value frauds like this one.”

“There is no way these people can get caught as they are always a step ahead of anybody who might want to investigate them. They are constantly monitoring the airwaves and the cables for communication about this facility. Once alerted,they simply throw them off trail or eliminate their threat.”

I was stunned by this deluge of information. I realized that I was not prepared for this at all. I asked him as to how did he expect me to proceed. He said;

“The UID is doing quite a lot of good for a large number of people, while it has the potential to bring upon a great deal of suffering because of the actions of some people. In my opinion stopping UID right now might not be a great idea. It should have been stopped way back when it was being conceived or was in its nascent stages.”

“I’ve transferred whatever I had with me about these activities into that memory pin I gave you. I’ve been spiritually inclined for the past couple of years and I have quit my job. I’m taking up Sanyaasa Deeksha. I leave it entirely to your discretion, as to what to do with the information. In some ways, I just shifted my wheel of burden onto your head.”

“My stop has arrived. I’ll alight. May you be blessed with the knowledge of righteousness and may you follow that path. Goodbye.”

As soon as he left, I got a message on my mobile phone – “Congratulations! Your data plan has been upgraded to an unlimited plan. Please swipe your index finger across the UID Validator™ in the device to accept and activate”.

On predilictions of predictions

So, Princeton University researchers claimed that they found strong evidence that Facebook will be largely abandoned by 2017, like an infectious disease.

Engineers at Facebook were “intrigued” by the article and proceeded to find out what’s in store for Princeton …

… and they found out that “Princeton may be in danger of disappearing entirely”.

So, as they say “Correlation is not Causation” –


The missed slingshot

Here is a sketching attempt I did out of boredom!

Missed slingshot

It depicts a scene that shows a man with a bump on his head shouting at a naughty boy. The naughty boy – whose shot at a mango from a catapult has missed and has hit the man on his head, is seen frightened.

I have uploaded it onto Open Clip Art for your usage pleasure.

Critique, comments welcome 😀

Migrate OpenStack Keystone to a different IP address

This hack describes how to change openstack service end-points after there is a change in the host IP address.

When you initially install and provision OpenStack, Keystone (the identity service) creates a set of end-points for all the services. You can get the list of service end-points by typing

# keystone endpoint-list

Particularly of interest are the “publicurl”s. A typical publicurl of a service end-point would look like$(tenant_id)s

If the IP address of the host changes, the end-points aren’t migrated automatically. This may cause quite a lot of dependent services to stop working.

One such service is “Horizon” which is the web interface and dashboard for OpenStack. Once the IP address is changed, you won’t be able to login to the admin dashboard as well.

To fix this, the service end-points needs to be updated. Unfortunately, the keystone tool supports only creation of end-points at this time. You’ll have to manually update the end-point URLs in the database.

In this example, I’m using mysql as the database and I have changed the host IP from to

Login to mysql (probably as root), select the keystone database and query the URLs.

# mysql -u root -p keystone
Enter password:
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A
Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.

mysql> select id, url from endpoint where interface='public';
| id | url |
| 384ee4e74ad19a71bc56bff55e18e25c |$(tenant_id)s |
| 675affbfacf34c0899a994d72eea7d57 | |
| 34818b5b99394e4d9894e932403696cf | |
| 8b363857ce294150567571dfc476a83 |$(tenant_id)s |
| b4041634dc614f8876baa119769e784e | |
| c6edf51290e34b84995bccacbc2a2454 | |
6 rows in set (0.00 sec)

Now that we know what to change, update the URLs (as suggested below)

mysql> update endpoint set url='' where id='c6edf51290e34b84995bccacbc2a2454';

Now doing a keystone endpoint-list should show your new URL.

To change the IP address in the configurations do the following;

# cd /etc/
# find {keystone,nova,cinder,quantum,glance} -type f -exec grep -nH {} \;
# find {keystone,nova,cinder,quantum,glance} -type f -exec sed -i s/ {} \;

Restart apache2, memcached and try logging into Horizon.
Of course, YMMV!