With the new change to the AdSense referrals, it would seem the AdSense team also did some cleaning house and dropped the Picassa referral program for publishers. Previously, publishers could refer potential Picassa users to the program and earn up...

e龙采用互动维客WIKI技术推出新频道 把原来的目的地指南改成目前的这个形式肯定还是需要花费不少精力的，原有数据的导入，整合是件很麻烦的事情。看来Elong也希望在生产、组织、表现形式上做一些创新。不过我个人不是很看好Elong与互动维客的合作，毕竟互动维客是个非常通用的平台，无法针对旅游目的地做很大的改动，从而无法很好的实现Elong这次合作的初衷。...

Have you noticed many $5 AdSense or AdWords referral popping up on your AdSense reports yesterday? Don't worry, you aren't losing out on the potential $100 you could earn when that referred publisher earns his or her first $100 with...

1) Blog PageRank 被清空了，原来是 5 ，现在是 0 。难道海底光缆断掉影响会这么大么 ? 我还指望今年能升到 6 呢。现在从头再来。

2) Google Adsense 收入"锐减"，其实原来也没多少。 不过从 Dreamhost 收到一笔引介费，明年的托管费用差不多了。从二月份开始，在页面上测试 Yahoo! 中国的 P4P CM 广告，与 Google 对比一下。

3) 因为海底光缆的影响，07 年 1 月份的访问量还是下降了许多。网络不快，所以 1 月份的更新也比较少。

4) 没有 4 了。

--EOF--

发表者：谷歌“精神文明办”

在外界想来，一个搜索引擎公司的反作弊团队，不同于 Google 其他工程师组合，多少让人觉得有些严肃和另类，当认定某个搜索结果是作弊后的产物时，他们似乎有能力做出并执行冷酷的判决。但实际上，在 Google，他们是一些既有原则又充满创意的工程师。

反作弊天条一：善意原则
“假设网站设计意图是好的”，很像英美法系——“无罪设定”模式在 Google 内部反作弊的复制，我们称呼它为“善意原则”，这是我们在施行反作弊技术时的起点。在确定某一网站是否有作弊嫌疑时，Google 会花很长时间，走很多流程，收集证据，来辨别这些网站是否真有所谓“恶意”。没人会武断地给某一个网站下结论，轻易冤枉任何网站。这是一个挺让人愉快的行为方式——不要错杀。因为，我们相信，错杀一个网站，就是对用户有价值信息的损失。

反作弊天条二：坚持自动、机器判断的原则
这一点同上面讲到的“善意”原则是相辅相成的。人们总愿意相信自己是细心与理性的，但很多时候，机器也许比工程师更加理性与细心。
在搜索结果作弊上，也许有一万种方法，而我们能做的，就是把其中的规律总结出来，然后搞出一些智能算法并不断完善，剩下的，我们希望交给机器。似乎也只能这样做，如果反作弊要纯粹依赖手工，我们的员工人数恐怕将达到天文数字。
统一标准，从而保证一个高质量稳定的网页结果，这是多数人都愿意看到的。因为，作弊结果比率甚至百分之零点几的升高，对用户体验都是很大的伤害。在谷歌中国工程研究院从有了第一个工程师，就开始了反作弊的工作。

比例尺搬家风波
Google 内部，有一个比较世界各地反作弊水平的图标。这是个很容易让各国技术团队紧张的图标，就在前不久 Google 中国反作弊小组在谷歌全球都引起一个小小“比例尺搬家风波”。
原来，坐标的比例尺在靠近横轴的右下角，半年来，经过中国团队发作弊团队的共同努力，中文作弊结果比例下降很快，曲线与坐标横轴太接近以至被比例尺挡住，为此，这个通用多年的图标不得不将比例尺从图标右下角搬家至左下角才解决了问题。
然而，与网络搜索作弊的博弈，漫长得看不到尽头，我们仍需努力。通常，Google 内部一个产品研发周期是 6 个月，但是，在这个快速创新的氛围中，却有一个恒久不变的内容，那就是反作弊技术。

最后，大家猜猜看这是什么意思



其实，这是 Matt Cutts 挂在办公室外被“篡改”了的名牌。Matt Cutts 是 Google 最早发现并研究搜索引擎作弊的“教父级人物”。（庆幸一下，能跟这样许多的大牛一起讨论问题）
在总部，Matt Cutts 和他团队的成员们分享一间带隔板的写字间。在写字间的隔板上，贴着主人们（照片中暂时擦去其他同事的名字）的名字。恰好 Matt Cutts 旁边的工位一直空着，有一次，一个团队成员突发奇想，将 Matt Cutts 的名字同时贴在两个CD 盘中（Google 用 CD 盘放名牌），并在上边大大地标注“Keyword Stuffing”（关键词填充），以表示“看，Matt Cutts Spam 了两个办公位”

注：关键词填充（Keyword Stuffing）是网站作弊中一个很常见的方式。

刚看了老康的Web2.0时代RSS信息过载问题，忽然想起原来在博客网时做了一个rss阅读的小产品“RSS博览”，看了一下竟然还在运行，心里真是高兴。 我一直还是喜欢那种一页式的RSS浏览，只要开一页就可以看到所有RSS源的最新更新，看起来一目了然，不需要像现在在抓虾里面一个源一个源的点，累死了。 http://rss.bokee.com/portal.102746.html 所以到现在我还在看自己搭的lilina上面的聚合，而不选择去抓虾，让我搞不懂得是，难道大部分人都喜欢一个rss一个rss的点？真不可思议。...

The best way to describe feed subscription trends is that they are feisty. Some days your numbers are up, some days (like weekends), they can dip down. This is all normal behavior (read more about why feeds behave the way they do in our TechCrunch case study). But sometimes, other things can happen that result in erratic behavior and we do our best to communicate to you in these instances.

In the last couple days, some of you may have noticed a bump in your subscription numbers, and others of you may notice this bump in the coming days. Why, praytell? MyYahoo!, the oh-so-popular web-based feed-reader (hyphen-mania!), is again reporting subscription numbers. Neat. (You can catch up on the full story here). While we celebrate the closing of another forum thread, we hope some of you are pleased to see your subscription numbers on the rise. NOTE: Your subscribers were not lost during this time, they just weren't being reported by us on your Analyze tab.

Stay tuned to this blog for more gripping updates. This post was brought to you by People for the Excessive Use of of Hyphens and Parentheticals.

We've been busy: since our announcement in september of our millionth signup, another half-million users have joined the fun. And because of this, we've been more quiet then I would like, and I want to apologize. It feels that we've...

a seemingly normal visualization graph on a box of Cheerios, which compares the effects that a bowl of Cheerios, a glucose drink, & a skipped breakfast have on you "power of concentration" over time.

however: "First up, a glucose drink! The breakfast of champions! Who hasn't left the house of a morning, pausing only to swallow down a couple of cans of Tango or Lucozade? ... But wait! Sugary energy drinks aren't the only competition! The other condition is.. no breakfast! Which actually beats Cheerios in the first half hour! Clearly, the subjects were still mulling over the pseudo-scientific crap they'd just read on the Cheerios box and couldn't concentrate on.. whatever it was they were given. In the end, of course Cheerios come out on top but it hardly tells you anything you didn't know before - as the only solid food in the experiment you might equally read the result as, Cheerios - better for you than starvation."

see also casualties as candy & news controled rice cooker & visualization with cookies.

[link: pappenheim.net|via boingboing.net]

I was reading Slashdot yesterday and came across an interesting tidbit. It turns out that, according to data from Google Trends, Yahoo! passed sex as the number one search term on their search engine last year. When we heard that...

2006年正月初七到的上海，初八到的若邻网络。到下礼拜一（2月5日），正好一年了。没有理由再继续下去，却有许许多多的理由应当离开。 非常高兴认识了一些不错的同事，很感谢这一年中他们给与的友情和帮助。希望若邻这个团队中的每一个人能走好。 我还会在上海，还会在社会性网络中，还会追逐自由与空间，作着自己想要做的事情 ： reloadding:// u r in socialnetworking

Posted by Tom Galloway, Technical Writer

Yesterday was National Gorilla Suit Day (NGSD). No, I'm not making this up--MAD Magazine cartoonist Don Martin did, back in 1964. Naturally, there's a full story behind this holiday, and it's been chronicled by my friend, writer Mark Evanier. If you've visited the right websites over the last week or so, you'll have seen various banners promoting National Gorilla Suit Day, which were provided by Mark.