The Audacity of Copying Well – Stratechery by Ben Thompson

A year later Facebook founder and CEO Mark Zuckerberg did what he should have done instead of releasing Poke, offering $3 billion to buy Snapchat outright, but it was already too late: Snapchat founder and CEO Evan Spiegel, perhaps emboldened by the then-widespread realization that Instagram had sold too early, turned the offer down, leaving Facebook to try its hand at Slingshot, another Snapchat clone; that failed too.

Source: The Audacity of Copying Well – Stratechery by Ben Thompson

BYVoid/OpenCC at 7fdaa43f1c548cc53ab9c7b59a697851060f4f46

Open Chinese Convert (OpenCC, 開放中文轉換) is an opensource project for conversion between Traditional Chinese and Simplified Chinese, supporting character-level conversion, phrase-level conversion, variant conversion and regional idioms among Mainland China, Taiwan and Hong kong.


Features 特點

  • 嚴格區分「一簡對多繁」和「一簡對多異」。
  • 完全兼容異體字,可以實現動態替換。
  • 嚴格審校一簡對多繁詞條,原則爲「能分則不合」。
  • 支持中國大陸、臺灣、香港異體字和地區習慣用詞轉換,如「裏」「裡」、「鼠標」「滑鼠」。
  • 詞庫和函數庫完全分離,可以自由修改、導入、擴展。
  • 支持C、C++、Python、PHP、Java、Ruby、Node.js and Android。
  • 兼容Windows、Linux、Mac平臺。

Source: BYVoid/OpenCC at 7fdaa43f1c548cc53ab9c7b59a697851060f4f46

The “hello world” of neural networks

Neural networks — also known as “deep learning” — are hot!

And now iOS 10 and macOS 10.12 come with the BNNS framework, or Basic Neural Network Subroutines, that lets you put neural networks into your own apps.BNNS runs on the CPU and is heavily optimized to be as fast as possible. There is also a version of these routines that use Metal and the GPU (part of the Metal Performance Shaders framework).

In this post I’ll show you how to get a really basic neural network up and running with BNNS.

Source: The “hello world” of neural networks

从米家到 HomeKit,你只需要一个树莓派 – 少数派

HomeKit 相信各位都非常熟悉了,得益于它原生自带,在 iOS 上可以说是十分优秀的。无论是你可以告诉 Siri 我回家了,早上好等这样的文字激活你设置好的场景,还是可以直接从控制中心看到和控制你所有支持的设备,比起其他第三方智能家居 App 使用体验都更佳。

Source: 从米家到 HomeKit,你只需要一个树莓派 – 少数派

Forge: neural network toolkit for Metal

However, using the GPU also has a downside: Metal only thinks in terms of textures and images.

\Using images makes a lot of sense for convolutional networks but you’ll still have to convert the output from your network into something you can use from Swift. And what about neural nets that work on non-image data such as audio or text?

Forge makes it much easier to convert your data to and from these textures.

Source: Forge: neural network toolkit for Metal