How to refuse to serve goods and reclaim absolute control of the data

While enjoying the convenience of various services, you also face concerns about the data being managed. And the shutdown of a service, like you need to recycle scrap car left residue, just as Google reader recently shut down-for whatever reason, Google's longer-term layout or prevent privacy issues, it suddenly shut down, Leaving behind is a mess of XML files. So how do we stop this problem from happening again? Mikio L. braun--, a PhD in machine learning at the Berlin University of Technology, Twimpact, co-founder and chief data scientist, published his thoughts on the blog.

The following is the translation:

Mikio L. Braun has always believed that the key to the problem is to use the user as a commodity, not God, and in this article in Derek Powazek, he realizes that even a paid service does not guarantee you the right to dominate. Ultimately, to gain absolute control over your data, you must reclaim your data and build your own services. Like the WordPress blog system on your own server: As long as you pay for it, it will always be in your hands. So whether it's software or data, you have to keep them running on a computer that you can control to keep them alive.

But how to put an end to this kind of Mikio, L. Braun shares the network calculations he used:

In the mainframe era, computers were clumsy, and only large institutions had rights. A time-sharing operating system has been invented, allowing computers to be used by more people, via text-type dumb terminals (dumb text-terminals) to connect to giant machines. Most of these terminals work in a block-oriented way, using tabular submissions to get results, something similar to how Web pages are operated? Obviously the service is hosted on the mainframe, and your job depends on the center machine.

But with the advent of interconnection and home machines, everything is changing. Instead of very rare mainframes, you work with minicomputer. Mail, FTP, and even HTTP are designed to allow a loose run, and basically anyone can connect the computer to the network and run the service on the server he's interested in.

Of course, you still need to address a package of technical issues: bandwidth issues, using a dynamic DNS service to map changed (dial-up) IP addresses to a DNS entry, Linux, or other forms of Unix-derived system-related knowledge. However, these are achievable.

(Responsible editor: Fumingli)