对2008再说点什么——质量

2008年，最大的质量门事件，应该非“三鹿奶粉”事件莫属，三鹿集团：害人终害己 ，一个企业彻底破产，还欠下十几亿的债务，该由纳税人来买单。

软件质量实际比传统行业更糟糕，只不过暂时还没有出现巨大的危机。以前的质量事故，数不胜数，这里暂且不表。就说2008年，这里有一份2008年web发展十大失败的名单，其中90％的失败都来源于质量，而且都是和“宕机”、“停机”、“崩溃”、“恐怖的故事”等联系在一起。

质量永远是成熟行业的象征，是企业的生命线。

Top 10 Web Dev Blunders of 2008

10. 电子商务站点在黑色星期五宕机

这是每个在线零售商的噩梦: 一群人聚集到你的站点，希望花出大笔大笔的钱，而你的站点却宕机了。不幸的是，在之前的那个星期五，它让那些在线零售商的噩梦变成了现实，不管是大的还是小的网站。

黑色星期五是今年美国的最大购物日，网上购物也是如此。许多主要的站点，像 Amazon, Sears 和 Live Cashback 都崩溃了，迫使这些购物者转向其他地方完成他们的黑色星期五购物。

事实证明，这种大面积的宕机是多年来持续发生的情况。黑色星期五发生在每年的同一天，但仍然有许多网店不能解决，他们需要在服务器部门留有重要的备份。在线购物持续增长，会导致这种流行病逐年增加，除非这些电子商务网站可以未雨绸缪以应对激增的流量。

9. 蜂拥克隆 Twitter

2008年似乎是Twitter的克隆年。在过去的几年里Twitter引发了web风暴，因此许多web开发者去尝试做这个流行web服务的克隆品。并且看起来这股Twitter的克隆风远未结束。

虽然一个网站在某些方面的独特性更具有竞争力，但是一个服务的完全复制从来不是一个好主意。然而，过去的一年中，涌现出很多的Twitter克隆品，而且，他们几乎全部都和Twitter没有任何区别，甚至更糟；几乎所有的都不如原来的好。

人们应该长远地思考如何克隆一个站点。除非两个服务之间有重大区别，否则克隆品根本找不到任何吸引人的地方。

8. MobileMe的糟糕发布

当一家公司的头儿承认产品的发布确实糟糕时，那么这个产品的推出就是真的差劲了。苹果的MobileMe发布就是这样一个例子，尽管它结合了新的Mac Leopard操作系统。

随着Apple推出它万众期待的iPhone 3G，它也宣布一个新的同步服务器MobileMe会同iPhone 3G,iPhone 2.0软件以及Apple商店同时推出。MobileMe 是一个同步服务，它允许Mac和PC用户通过一个web界面去同步他们的联系人，日历，电子邮件，相片和更多的东西。

推出的第一天便充满了大量的问题，性能缓慢，宕机，像用户随机注销这样奇怪的问题，还有一个致命的问题：整个第一 天里，服务都无法同步日历和全部联系人。就像苹果CEO Steve Jobs在一封内部邮件里所写的一样——这不是苹果的“光荣时刻”。自那以后，苹果修复了那些漏洞，并且承诺所有的MobileMe用户可以用两个扩展来 免费使用90天。

7.Gmail宕机30小时

Gmail 是最受欢迎的电子邮件提供商之一， 有许多来自Google Apps 服务的用户。许多企业使用 Gmail (不管是免费还是付费的用户)，所以当这个服务部分暂停高达30个小时的时候，这种丑事立马众人皆知。对Google而言这次宕机的不幸在与，许多受影响 的人是小企业，超过24个小时无法访问他们的gmail，在这个过程中流失了很多利润。

6. Digg 脚本

Digg 上看到一些整个2008年被禁止的用户列表（还有未列出的许多其他用户），是因为他们使用脚本来自动给他们朋友的意见进行投票。虽然许多人认为不给这些人激活帐户的第二次机会是不公平的，但是其他人也认为这种零容忍政策必须被规范。不管你站在哪一个阵营，很少有人不同意Digg脚本危害了社区。

为什么这是一个web发展的失败？ 如果你想做一个Digg脚本，你最好确定它不会公然的容易被检测到。NETTUTS建议这些脚本背后的编码者把他们的时间放在易用方面。例如，编写一些实际帮助社区而不是危害它的应用。

5. Facebook Beacon

当Facebook推出他们的创新广告平台Beacon的时候，受到了极其严厉的批评。 事实证明，Facebook的用户不喜欢让web上的每个人知道他们的交易记录。在Facebook的Beacon伙伴站点上有许多人们购买物品的恐怖故 事，包括一个在Overstock.com上买了一个订婚戒指的哥们，立刻所有的交易详细记录被显示在了他的Facebook资料里，毁坏了他的订婚惊 喜。

Facebook之后在Beacon里增加了opt-in选项，但是很多不良的影响已经造成了。 结果Facebook beacon导致一对夫妇集体诉讼Facebook和其合作伙伴的相关服务。该诉讼称Facebook不仅仅是从Facebook自己的用户，而且也从其 合作伙伴站点的用户那里悄悄获取数据，此外，Facebook承认收集数据，甚至于用户选择退出交易共享选项之后也无法避免。

4. Cuil的 Sub-par发布

大肆宣传安全模式的 Cuil,这个创新的搜索引擎在高期望中推出了。被前Google员工创立的Cuil，保证它是比Google更好的搜索引擎，有和关键字相关的更多的结果，以及一个更好的搜索结果界面。 Cuil给人的印象深刻，（最终）可能是Google的竞争对手。

然而,Cuil的推出是让人十分失望的。 首先，该网站宕机数个小时，这对于拥有巨大流量的网站来说还情有可原。更重要的是，Cuil并没有履行其在宣传期的账单。搜索结果甚至连Google的相关性也不如， 并且他们的索引机器人使网站崩溃。

3. Amazon S3 停机

2月15日,亚马逊的三个物理服务器之一宕机数小时，扼杀了亚马逊S3云存储系统。许多像Twitter, SmugMug这样的创业公司，甚至纽约时报都在站点上看到了错误消息。许多最流行的网站都受到了S3宕机的影响。

使宕机带来麻烦的是开发团队缺乏沟通。亚马逊S3开发团队花了一个小时去查证宕机事件， 并且他们直到问题被修复数小时之后才再次回应。许多S3的客户不能接受在宕机期间关闭整个业务。

当开发危机发生的时候， 透明度和大量的沟通是关键。

2. Twitter Fail Whale（Twitter宕机标志） 的普及

随着今年Twitter持续的发展， 这个流行的微博客站点频繁的出现了臭名昭著的 Fail Whale. 感谢Twitter的恐怖的正常运行时间， 这个错误页面出现的如此频繁以至于Twitter用户都开始喜欢这条睡觉的鲸鱼了。不久这条Fail Whale就有了它自己的 fan club、商品甚至是一个 Twitter account.

不是多数web开发者都可以声称他们的错误页面有一个fans club的， 也没有多少人希望这样。错误页面是早已存在的东西了，是任何开发团队都想让他们的用户去看到的。幸运的是，Twitter的错误页面是有趣的，很有吸引 力，所以在出现错误时它给人的感觉最好。然而，我在怀疑 Twitter的这些家伙是想推广这条睡觉的鲸鱼。

1. Site Meter导致IE用户崩溃

2008年8月1号是 Site Meter 开发者会记住很长时间的一天。这是有效关闭大量互联网上IE用户的一天。实际上，任何安装了Site Meter tracing的站点都会给IE用户抛出一个“操作故障”信息。
虽然部分责任在于在Internet Explorer （他们有一个众所周知的错误，引发了错误），但大部份的指责指向SiteMeter开发者，因为他们没有彻底的测试就把一些修改放到产品里了


看一送一：

Top 10 Web Platforms of 2008

Written by Richard MacManus / December 21, 2008 7:00 PM / 17 Comments

« Prior Post Next Post »

2008 has seen a proliferation of new Web platforms, including a few major ones built using open standards. In this final instalment of our series of top products of 2008, we choose the top 10 Web platforms of the year.

digg_url = 'http://digg.com/software/Top_10_Web_Platforms_of_2008';digg_bgcolor = '#ffffff';digg_skin = 'normal';<iframe src="http://digg.com/tools/diggthis.php?u=http%3A//digg.com/software/Top_10_Web_Platforms_of_2008&amp;s=normal&amp;t=Top%2010%20Web%20Platforms%20of%202008%20-%20ReadWriteWeb&amp;k=%23ffffff" scrolling="no" width="52" frameborder="0" height="80"></iframe>We've written a lot of times about platforms for the Web. A web platform can be as simple as an API, like the one offered by Twitter, which allows external developers to tap into a company's data. It can be software and services, like Amazon's Web Services. It can also be a fully fledged development platform, such as iPhone SDK and Adobe AIR. Whatever the case, platforms on the Web mean allow people to build on top of another company's product, so we think it's an appropriate way to close our Top Web Products series.

This is the tenth and final post in ReadWriteWeb's series of top products of 2008. Here are the previous nine:

• Top 10 Semantic Web Products of 2008
• Top 10 International Products of 2008
• Top 10 Consumer Web Apps of 2008
• Top 10 RSS and Syndication Products of 2008
• Top 10 Mobile Web Products of 2008
• Top 10 Enterprise Web Products of 2008
• Top 10 Real World Web Apps of 2008
• Top 10 Digital Lifestyle Products of 2008
• Top 10 Alternative Search Engines of 2008

Note: unlike some of our previous lists, we've attempted to order this one according to impact in 2008.

1. iPhone SDK

ReadWriteWeb named Apple as our Best BigCo of 2008, largely due to the iPhone platform. The iPhone truely reached a mainstream audience in 2008, when the 3G iPhonelaunched to much acclaim in July. But more significant than the phone itself was the simultaneous launch of the Apple App Store. There were 552 applications available at launch and at time of writing that's risen to over 10,000 applications. So in less than 6 months, the number and variety of iPhone apps has expanded greatly. Indeed, any startup worth its salt has an iPhone version of its web app - the ultimate proof of a successful Web platform. was

For many years now everybody (including us) has been predicting that the next year will be the one that the Mobile Web breaks through. 2008 was finally that year - and it's mostly thanks to the iPhone development platform.

2. OpenSocial

November was the first anniversary of Google's OpenSocial, an open API framework for social networks and websites. During 2008 OpenSocial gained a lot of traction; statistics released in November stated that OpenSocial had reached 675 M registered users at that time and there were 7,500 applications.

Most impressive is the list of organizations who signed onto OpenSocial and are actively developing apps for it. That high powered list includes MySpace, AOL, Bebo, hi5, LinkedIn, Ning, Orkut, Yahoo!. Of course still missing from OpenSocial are Facebook and Microsoft, but at the rate OpenSocial is ramping up - they may not need them.

3. Adobe AIR

As we noted in our Best BigCo 2008 post, this was a year in which Adobe's 'Rich Internet App' strategy bore some juicy fruit, with many compelling apps released that were built using AIR (Adobe Integrated Runtime). AIR is a platform that makes it easy to build attractive Internet connected applications that live outside the browser. Last year we noted its potential - AIR was called 'Apollo' for much of its beta - but in February 2008 AIR was officially launched and soon we saw a lot of stunning apps emerge. For examples, check out:

• 6 Adobe AIR Apps to Check Out
• 10 Adobe AIR Apps Bloggers Will Love
• Adobe AIR Goes to Work: 6 Apps for the Corporate Desktop

We've named it our number 3 Web platform of the year, because it breathed new life into Rich Internet Apps this year.

4. Twitter API

The importance of Twitter's API to its success this year can't be overstated. The number and variety of Twitter reader clients alone is amazing, let alone the many other ways Twitter's data is being used (e.g. for e-commerce purposes). It must be said that Twitter has experienced much downtime and many technical glitches throughout the year, but even so it has continued to expand its API service. As Twitter co-founder Biz Stone said a ReadWriteTalk interview last September:

"The API has been arguably the most important, or maybe even inarguably, the most important thing we've done with Twitter. It has allowed us, first of all, to keep the service very simple and create a simple API so that developers can build on top of our infrastructure and come up with ideas that are way better than our ideas ...

So the API, which has easily 10 times more traffic than the website, has been really very important to us. We've seen some amazing work built on top of it from tiny little mobile applications like an SMS timer that just allows you to set a reminder over SMS to call your mom or something like that, to more elaborate visual recreations of Twitter like twittervision.com, which shows an animated map of the world and what everyone's doing around the world with Twitter. ... The [Twitter] API has really been a big success for us, and it's something that we want to continue to focus our efforts on, looking forward."

5. Facebook Platform

Facebook continued to become more popular throughout 2008, reaching 140 million active users by end of the year and growing at a rate of 600,000 users each day. For that reason we gave it an honorable mention in our Best BigCo 2008 post.

However in terms of the Facebook platform, overall we felt that Google's OpenSocial overshadowed it in 2008. As we wrote in our mid-year review, Facebook Platform: The Fanfare Revisited, when the Facebook platform debuted last year it was touted as the next big thing. You no longer needed to bring the audience to your app. Instead your app could be delivered to one of the largest audiences around the web. And not just delivered, but injected into a massive social network. But while it started great, it turns out things are not that simple. Three fundamental issues surfaced:

1. Technical: Should the app be just a teaser that leads users to their site or should it be a duplicate and have full functionality?
2. Business: If e.g. New York Times builds a Facebook app, will it be economic for them (since there's little revenue in Facebook)?
3. Provider costs: Does it pay for Facebook to maintain the platform? As a business with a huge valuation, Facebook needs to maximise profit.

Despite these issues, Facebook's platform shows no signs of slowing and many startups have ported their web apps to Facebook. It may not be the bee's knees anymore, but it's still a very effective platform for startups to utilize.

6. Android

Google's open mobile OS platform, Android, burst onto the scene this year as a rival to Apple's closed iPhone platform. Google spent a lot of time this year encouraging developers to create applications for Android - and rewarding them for doing so with cold, hard cash with the Android Developer Challenge (see our previous coverage here). This led to many third party apps and multiple App Stores. The first Android phone - the "T-Mobile G1 With Google" - was launched in September, followed by a second Blackberry-like phone in December. Android apps are showing steady growth and we can expect to see this ramp up in 2009 as more handsets come on the market.

Android went open source in October, which starkly set Google apart from Apple's controlled platform. While iPhone was our top platform in 2008, Google has the opportunity to challenge for this mantle next year.

7. Amazon Web Services

Amazon's leading edge Web Services stack was first introduced to the world in 2006 and it continued to impress in 2008 - albeit with more of a business focus. Amazon Web Services basically became a more mature offering in '08 and it shored up its support services.

Amazon Web Services, led by Simple Storage Service (S3) and Elastic Compute Cloud (EC2), was the first major cloud computing platform and today it powers the backend for many startups. At the beginning of the year we noted that Amazon's web services now accounts for more bandwidthpremium for-pay support packages for some of its core infrastructure services. S3, EC2 and Simple Queue Service (SQS) each received the gold and silver level support treatment. In October Amazon announced that EC2 was coming out of beta and that it now supported Windows Server 2003 and Microsoft SQL Server (i.e. expanding beyond Linux distributions and OpenSolaris). At the same time Amazon offered a Service Level Agreement for EC2 and promised an availability of 99.95%. than all of Amazon's global web sites combined. Then in April, Amazon announced

8. Live Mesh

The Live Mesh service launched in April as an invite only "technology preview". It is Microsoft's attempt to tie all of our data together. Live Mesh synchronizes data across multiple devices (currently just Windows computers, but theoretically it will extend to mobile and other devices in the future) as well as to a web desktop that exists in the cloud. It can sync data across devices used by a single users, as well as create shared spaces for multiple users.

Essentially, Live Mesh is a collection of feeds (which can be expressed as ATOM, JSON, FeedSync, RSS, WB-XML, or POX). Every piece of data entered into a user's Mesh -- be it a file, a folder, a message, a user permission, or a new device -- is rendered as a piece of information in a feed. The feeds are then synced with other devices that are part of that Mesh following rules for how to sync each particular piece of information (i.e., File A may sync with Users 1, 2, and 3, while File B may only be told to sync with Users 1 and 2).

9. Fire Eagle

Earlier this year Yahoo announced that the closed beta period for its location platform Fire Eagle had ended and that the service was now open for everybody. A number of high-profile services, including Brightkite, Movable Type, Dopplr, and Pownce have implemented Fire Eagle through the numerous APIs Yahoo provides for accessing the service.

As we wrote about Fire Eagle when the beta was first announced, it offers API kits in five different programming languages, it's got user authorization protocols already available for web, desktop and mobile apps and it's using the open standards community built oAuth to facilitate faster, more secure mashups. So this platform is leveraging universal open standards.

Note: also see our coverage of the Yahoo! Internet Location Platform, a collection of in-depth geo-location based APIs.

10. Mozilla Weave

This year Mozilla announced Weave, a new web platform that will store users' browser metadata in a cloud environment for access anywhere. Weave is a "framework for services integration" that will, according to Mozilla, "focus on finding ways to enhance the Firefox user experience, increase user control over personal information, and provide new opportunities for developers to build innovative online experiences."

The basic idea is that browser metadata (things stored in your Firefox profile like bookmarks, history, RSS feeds, usernames and passwords, etc.) is pushed into the cloud and stored on Mozilla's servers. The data is available to users from wherever they get online and users can share information with friends, family, or third parties while retaining control over how, when, and if the info is shared.

Unlucky Not to Make Top 10

There were many other Web platforms that impressed us during the year. Google's Chrome browser is highly promising (but we felt it was too early to be in the top 10), Google Gears helped take the online world offline, meebo created an intriguing platform based on its core IM capabilities, Bungee Labs developed a great mashups platform, Salesforce.com had a strong year again, and on and on. It seems like every major Internet company nowadays has a platform, which is great for developers and users alike. Check out our earlier post listing 10 promising web platforms for other examples.

We hope you agree with our top 10 list, but we're sure there are one or two platforms you think should be here instead. So let us know in the comments.

Also for more about the theory and practice of platforms, check out these RWW posts:

• APIs and Developer Platforms: A Discussion on the Pros and Cons
• So You're Launching a Platform: After Ubiquitous APIs - What's the Next Frontier?
• Picking a Platform: 5 Issues to Consider
• When Web Sites Become Web Services