Fragmentation?

When I read about fragmentation of devices on Android, I mentally nod, agreeing with the author of the article. When I see the thesis that device fragmentation interferes with software development, I raise my eyebrows perplexedly. But when they try to prove to me that Android fragmentation is the main problem of this platform and the main minus for mobile software developers, I frankly start to laugh. Sorry, but this is either purely marketing intentional nonsense, or a complete misunderstanding of the situation.

Imagine that you are a furniture manufacturer. You work in the furniture market, you compete with other manufacturers, dividing the entire market into certain parts. Your task as an entrepreneur is to grab the largest possible radial angle on a pie chart with the title “Furniture market share”. The pixels in this diagram are people who buy furniture. These are different people and they buy different furniture: from unpainted block beds from Ikea for 2,000 rubles, to sofas covered in leather of young combed crocodiles from the Caroline Islands, each of which stands as your apartment.

Each of these price segments has its own competition. You can make cheap furniture, sell it in cheap stores, don’t even know about the existence of the Caroline Islands, and at the same time do not in any way intersect with manufacturers of luxury furniture who will also not know about your existence. And this is completely normal. No one calls it fragmentation. This is called another word “segmentation” - market segmentation into different market sectors that are isolated from each other, which, although they are part of the same type of product, depend on different sources of supply and demand.

What do we have on the modern mobile software market?

There is Apple, which makes high-quality, top-end smartphones, which at the time of their release are always on the verge of fantasy. This is an elite segment. Yes, it is very large even in its absolute value, but this is not the point; it is already changing now and will continue to change over time.

There is an Android. This is not a product in itself, but a platform for creating products. What smartphones are there on Android? There are the same powerful top-end units, flagships, the dream of every geek buying everything that brings him closer to the ideals of the cyberpunk future. They (top-end units) compete with the iPhone and have approximately the same similar configuration. But, it so happened that on Android from the very beginning there are non-flagships. And it is precisely this circumstance that has started since the start of the modern round of the smartphone race and does not let people fall asleep for whom the word “fragmentation” seems like a terrible minus. My friends, device market fragmentation is a plus !

What do these strange people call fragmentation? Answer: the presence of various devices running on the same system, but with different parameters: resolution and screen density, power and number of cores, the presence of hardware features, and the version of the system itself. The existence of several dozen such combinations is annoying for some developers. Here's how for them, if for example we’re talking only about screen resolution, the comparison of iPhone and Android looks like:




This is not true.

All of the above parameters together determine the device in a certain sector of the smartphone market, position it on a scale of “megaflagman - budget phone”. Drawing an analogy with our furniture example, we can confidently say that in the case of smartphones, these sectors do not overlap in the same way, because real gadgets are bought only by flagships, and those who buy budget smartphones do it because of price considerations, therefore it’s unlikely whether they would even buy a smartphone if they hadn’t been in the market for their budget options.

In addition, the freedom of choice allows different manufacturers to issue their own different solutions, thus increasing the total absolute number of devices on the market, investing in marketing independently, and, accordingly, increasing the total number of users.

Here's what the diagram above should actually look like:



The iPhone circle turns into a pacman - there simply aren’t those various configurations, the absence of which provides a small degree of fragmentation. But there are not those users who like the mouth of the pacman, but do not like its carcass.

Hence thesis number 1: the presence of various configuration options that allow you to make budget and just different devices, expands the market, and not splits it. More options - more people buy smartphones, more absolute values ​​in a pie chart. Yes, in return for more sectors on it.

Imagine a world that is ideal for imagining complaining about fragmentation, in which there is no division into different device configurations in the Android camp, and all Androids roughly represent competitors to the iPhone, that is, they are in the higher price segment. Then, let's say we would have a line of different HTC Desire and Samsung Galaxy S with approximately the same basic configuration. There would be beauty, an almost complete absence of fragmentation, substantially less sectors in the diagram. But there would be no cheap budget devices for 8 thousand, there would be no sliders, there would be no horizontal form factors: which means that the absolute values ​​of the number of users and devices would be less. By removing fragmentation, we simply cut out the mouth of the pacman, and do not smear the two largest sectors in a full circle.

Further. In fact, if it is physically unpleasant for you to have several subdirectories in the / res / directory, you can personally destroy any fragmentation. To do this, the Android Market has filtering mechanisms for features of devices that are designed specifically for such lazy people.

Thesis number 2: Do not want to fight fragmentation - do not fight. Do not want to make an interface for ldpi - do not. Do not want to support navigation without a touchscreen - do not support. You can generally release the application exclusively for one device, and it will be a complete absence of fragmentation!

All this concerns fundamental fragmentation - when there are various screens, form factors, and so on. There is also technical fragmentation - when the system becomes obsolete and does not update on some devices, different video chips, etc. It’s already quite possible to deal with this to Google itself, which it is doing quite successfully - at least a serious step has already been taken to oblige all manufacturers to issue mandatory system updates within 18 months of the device’s release. Technical fragmentation is minimized, and fundamental fragmentation is a plus for the developer.

You know, talking about the minuses of fragmentation is like if a furniture maker started saying: “It's awful, I can't live like that anymore, I have to think about how to make different sofas for the poor and the rich, and what’s the most scary - I have to do not only sofas, but also armchairs and double beds! What a terrible market. ” Dear, you do not want to make double beds - do not, where does the market?

Final thesis # 3: Apple made a decision for you and got rid of fragmentation by simply not releasing the class of devices that forms the most insignificant part of the diagram and is needed by the most insignificant part of users. Android gives you the opportunity to independently make this decision, personally choosing the level of fragmentation for yourself.If iPhones had fragmentation involving the currently unused budget segments of the market, would the developer’s profit be more or less? Of course, more. Since then optional fragmentation as an opportunity to select different supported segments can be a minus?

Therefore, friends, the next time you will be told about how difficult it is to live in a fragmented android world, advise you to cut your mouth out to your android pac-pack and just look at the results.