Shared publicly  - 
 
A response to +Steve Yegge's rant

Hello Stevey,

Only yesterday did I really read your famous rant. I think your point on platform-making isn't totally correct either. Here's my criticism:

1. There is a reason why Google seems not so interested in becoming a Platform company like Microsoft and Facebook.
Building and maintaining a platform means you don't create solutions to concrete problems; instead, you create an "app marketplace" that lets 3rd party app developers do the real job. It's like you're a landlord and you rent your lands to businesses and charge them for rents. Google has a culture to make innovative, concrete things that solve real problems, and I hope Google not degenerate to a purely landlord-style business.

2. The right policy on "platform" is to persuade app developers to develop for an open, neutral platform.
For example, for web apps, use OAuth/OpenID/HTML5/JavaScript instead of the Facebook platform; for local apps, use Java/GTK+/Qt/wxWidgets or even "Chrome Web apps that run locally" instead of proprietary OS-specific platforms such as Windows Platform SDK and Mac OS X's SDK.

3. A virtual platform or a galaxy of real products? I hope Google learn from General Electric.
Both Google and GE started with a good product/service, Google with Page Rank-based Web search and GE with the light bulb. GE then expanded to a wide spectrum of industries, even including nuclear reactors. So GE is a typical example of a galaxy of real products and services, instead of doing the "platform business". What's even worse than doing a platform business is if you want to do an "online platform" business, as that can easily become bubbles. I'd rather be a real landlord than an online landlord...

4. No one will win a war of proprietary platforms.
Another problem with building a proprietary platform is no company can ultimately win such a platform war. Every company seems to want to "build a platform" and become an online landlord after it has some good services. Amazon wants it; Facebook wants it; Download.com wants it; Google wants it too. But ultimately, only those companies who champion open platforms and standards will win. Google is good in this aspect with its HTML5/JavaScript-compliant Chrome Web apps initiative.

5. Who will win the war of social networks?
As for the war of social networks, I believe ultimately an open standard called "Next-Generation Blogging" (NGB) will win. It's an upgrade to the current blog, allowing near real-time delivery of your new blog posts to your subscribers (so they see them immediately in their NGB user interface), and post-specific, circle-based access control (as in G+). See this vivid illustration for what NGB is like (https://plus.google.com/u/0/102291835965130378165/posts/BZgKHy2NsYc) and join Googler +Brad Fitzpatrick to develop it and beat Facebook!

6. Closing remarks and a wish for +Larry Page and +Sergey Brin
You see those decades-old communication protocols like SMTP (email), NNTP (newsgroups) and IRC are still in active use today, so will their younger brother RSS. Why? Because they're all free, open and public protocols that model certain forms of communication. In contrast, proprietary communication systems like MySpace, Skype, Facebook and even G+, their impact on human history won't be as profound as those open protocols.

Those open protocols were mostly invented by scholars in the academia, or by people in the industry who have farther sight beyond immediate commercial gains.

You know, really great companies, they invent great things like the hard disk, and found great industry standards like PC and Unicode.

Larry and Sergey, I hope you can do more in things that will have really long-lasting impact on the human civilization, rather than things that drive the stock price up and down so erratically.

7. "Platformology" (newly added)
But we do see good examples of successful, long-standing proprietary platforms, such as Microsoft Windows, Mac OS X (and other Apple platforms) and Facebook. And we see how difficult it is for a new proprietary platform to overthrow a current dominator (e.g. G+ vs. Facebook), even if the new challenger is an industry giant like Google. So is there a set of laws that govern platform wars? There seems to be.

Law 1: The market rewards the first generation of explorers commercially, just like the patent system does.
Microsoft and Apple make the most and second most widely used desktop operating systems today largely because they were early players that pioneered graphical user interfaces. Likewise, Twitter and Facebook are still the top two spots in the SNS market for the same reason.

Law 2: Later challengers can only win with open platforms and standards.
Although Windows and Mac OS X are the most and second most widely used desktop OS's, the third is Ubuntu and the fourth is Linux Mint, a variant of Ubuntu. See? Only open platforms or standards can challenge early, proprietary dominators. For another example, Britannica and Microsoft Encrata used to be two dominating electronic multimedia encyclopedias, but now Wikipedia takes everything. For a third example, Microsoft Office faces increasingly serious challenge from OpenOffice (and now LibreOffice). As for social networks, Facebook won't continue to dominate. Facebook and Twitter are actually on the wrong side of history. Can you imagine if telephone companies block calls from each other? Then you had to subscribe to multiple telephone companies and get a different telephone number from each company in order to be able to make calls to all your friends. No, you should just need to own one telephone number to call all friends and be called by them. That ridiculous scenario is exactly happening in today's social networks: you have to own multiple social network accounts (one Twitter, one Facebook, one G+) and visit all of them every day to exchange updates with all friends. This is the wrong side of history. Even early telephone companies didn't do this evil. Maybe +Eric Schmidt should meet his Washington friends and let FCC intervene. The real future of social networking will be like email: everyone only needs to own one account to be able to interact with all friends, and this will be based on an open protocol like SMTP is for email.
5
3
Dave Gray's profile photoZiyuan Yao's profile photo
3 comments
 
Just added a seventh part :-)
 
Update: Linux Mint now becomes the top distro on distrowatch.com's distro popularity ranking list (therefore the third most popular disktop operating system behind Windows and Mac OS X). This is largely because the newly released Ubuntu 11.10 continues to pisses off its users with its Unity user interface.
Add a comment...