So we beat on, boats against the current, borne back ceaselessly into the past.
於是我們繼續往前掙扎，像逆流中的扁舟，被浪頭不斷地向後推入過去。
多年後的 Wretch , Pixnet 以及 yam (根據 alexa 70%+ 流量在 blog)
I’m delighted to announce that we’ve reached an agreement to acquire Tumblr!
We promise not to screw it up. Tumblr is incredibly special and has a great thing going. We will operate Tumblr independently. David Karp will remain CEO. The product roadmap, their team, their wit and…
大亨小傳主題曲，晚上廣播一連聽到好幾次！
每個人都會看到別人表現出來的那一面，並且羨慕它；但實際上通常沒有想像中那麼美好。
要裝什麼桌面？有哪些 TV 優化過後的 App。
坦白講，看完 Netflix 的 IR Letter 以及本週聽到的 Pixiv 演講，我比較認為是別人用數百倍的努力與金錢堆疊的努力，台灣敗的是方法不對、資金不足，還可能沒有比別人努力。
Improving our service
We are currently spending about $350M per year on a wide range of efforts to improve our service and app, and we are constantly getting better.
One area is our core services: our streaming delivery, sign-up, billing and customer service, across more than 1,000 devices being used in more than 40 countries. Here we strive for operational excellence to avoid any problems. Members want Netflix to just work — flawlessly. On this front, we’re well ahead, but we have plenty of room to improve. We continue to invest heavily to ensure that our service is always available, our streaming has a minimum of buffering, and has great audio-video quality.
Another area of focus is personalized merchandising, which drives what content we feature on a given member’s initial screen. Google search is an example of a ranking system, where results
are automatically computed to show Google’s estimate of the most relevant answer to the query. For Netflix, the user’s home page is the personalized ranking of what we think will be most relevant for that specific user at any given time. By analyzing terabytes of data from every recent click, view, re-view, early abandon, page views and other data, we are able to generate a personalized homepage filled with the content most likely to please. Our aim is to keep inventing and tuning algorithms to generate higher satisfaction, viewing, and retention, for whatever the level of content we can afford in that territory.
All of our algorithm work, like with Google search ranking, is proven or disproven by A/B testing. Only algorithms that lead to an improved experience get rolled out to everyone.
There are a hundred other areas we also are improving, like how smoothly our scrolling works on an iPad, or how well our kid’s area works on a PS3. Most of this work is guided by A/B testing as well.
The core metrics we use to evaluate A/B testing are signup rates, viewing, and retention. While our app is much better than it was 5 years ago, it is nothing compared to what it will be 5 years from now.
如果 Facebook 的貼圖小舖是像 MSN 或者是 Plurk 一樣，讓使用者自己上傳，免費或者是賣上傳數量，應該在亞洲的使用量會大增，而且會對其他傳訊軟體進行排擠（？）
如果談起越南，你想到的會是越南新娘？胡志民小徑？還是牛肉河粉呢？
在前陣子，Google 工程師Neil Fraser 的一篇部落格（中譯版）裡頭，他發現越南11年級的學生可能可以在 Google 的面試流程中連闖數關，這想起來，是非常令人驚奇的。（雖然我懷疑這可能是跟台灣類似，學生在更早就透過背頌的方式就理解這些數理題目。）
四月初，有幸拜訪日本的一個遊戲公司，在裡頭的 Cho San 向我們介紹他們公司的開發者來源，除了我們熟知的日本名校外，他們也與越南的胡志民大學建教合作，招募最優秀的 CS 學生，來為他們的公司開發遊戲。
在前幾天的Echelon 台灣站，其中上台 Pitch 的 Seconds 團隊，是一支來自新加坡的團隊，在活動的前一晚，跟他們聊天，發覺他們已經成功出場過兩次，並將產品賣給日本的公司，其最新的產品 Seconds，主要的開發團隊位於越南河內，並擁有 9 位的開發者。
他們甚至透露，在當地招募開發者，其月薪為一百美金，如果需要通曉英文的，則是兩百美金，這則是另一個相當令人驚奇的數字。
根據世界銀行在2011年的數字，越南的上網人口比例為35%，人口數為 8700多萬（是世界上第十一大人口的國家），代表在現在，越南的上網人數可能已經遠超過台灣的人口數，那麼，會有不少人有能力上網且寫出好程式，也是不太令人意外的。
PayPal 幫
