Takshashila Position - On Developing an Indigenous Mobile Operating System
Top government officials have revealed that the Indian government is promoting an indigenous mobile OS to rival Android and iOS. Many failed attempts to build a rival OS (OS) by different entities teach us that the government is unlikely to achieve the stated objectives. The government or its agencies should not adopt a do-it-yourself approach to building OSes even if they can do an outstanding job. A better alternative would be to back an open-source OS with a strong developer community.
The indigenous OS, developed by an IIT Madras incubated startup, is expected to inject competition into the market. This decision comes from antitrust judgments by the CCI that Google abused its mobile OS market to charge exorbitant commissions on purchases made through the Play Store and gain an undue advantage in other verticals. Android dominates the mobile OS market in India, with a market share of over 95%.
One of the critical reasons for Android’s widespread adoption by original equipment manufacturers was Google’s decision to open-source it. It is licensed under a permissive Apache 2.0 license, allowing derivative works with or without modifications to be created. These actions have created robust developer and user communities. However, the Google Play app ecosystem is a proprietary product, and only licensed devices are allowed access to it. The widespread adoption of the OS and the vibrant app ecosystem create network effects that make it very difficult for new entrants to penetrate the market.
Even if the government can build an excellent OS, creating the app ecosystem will still be challenging. Platforms have come to dominate our Internet usage. Communication, social media, reading books, and listening to music are all done over platforms, and apps are the primary gateways to these platforms. Without a large user base, the incentives for developers to create apps for a new OS are lacking.
If the goal is to have an alternative to the dominant mobile OSes in the market, the government is better off backing an open-source OS with a strong developer community. It can help create a sustainable open-source ecosystem by providing grants to developers supporting these projects or by creating demand for it through preference in procurement.
Several open-source mobile OSes already exist. Some of them, like LineageOS and ReplicantOS, are even based on the Android Open Source Project. Ubuntu Touch is another mobile OS that is based on the popular open-source desktop OS Ubuntu. While these OSes held much promise, most of them have not seen widespread adoption.
A community of developers already exists for popular open-source projects who can respond to users' needs and market incentives. However, the challenge is the app ecosystem which allows a smartphone to do things beyond making and receiving calls. Without the network effects of a widespread adoption base, there are no incentives for app developers, and without the app ecosystem, there will not be widespread adoption.
Pro-competition legislation, such as the Open App Markets Bill being deliberated in the US, seeks to mandate Apple and Google to allow third-party app stores and alternate payment channels for in-app purchases. While there are concerns around security, there is a good possibility that app stores that take a solid approach to security and privacy without charging a bomb in commission will also come up. The delinking of the vertical integration of app stores and operating systems will create a level playing field. Given a competitive ecosystem, third-party app stores can address users' needs, whether it is low commissions, security, privacy, or payment options.
Under C-DAC, Chennai, the government had earlier built an indigenous OS for desktops called BOSS (Bharat Operating System Solutions), a Linux distribution based on Debian. Very popular Debian Linux distributions, such as Ubuntu, already exist and have a vibrant developer community. With a broad user and developer base, tech support is more readily available, apps are more widely available, and the software tends to be more stable.
Another reason the government should not be building OSes is that citizens don’t want the government inside their phones. Access to location, contacts, and communication on a system created by the state raises questions about surveillance to a whole new level.
Building an operating system and the ecosystem around it is a task best left to the market or the open-source community. It requires technical skills and mobilisation of a broad developer community which are not the forte of a government or its agencies. A do-it-yourself approach should be the last of the government’s options.
