20070319 科学幻想 Philipp Lenssen Illustrated Google Drive

海量数据存储、授权、高速访问（下一代的Amazon AWS S3），确是网络操作系统(Web OS)的关键之一。 海量存储空间、极速传输、精确授权、单一URL(所有目录及文件皆对应永久短静态URL)、管理界面UI、全球范围分布式存储(类似Yahoo! Mail 的 CDN)、多版本存储、每URL电子邮件集成、精确搜索、、、、、、 资金上、技术上的挑战到底多大（宽信道、大存储、密集查询、URL访问授权、版本存储），这还不等于是由一 家商业性公司重建全球整个互联网！，到底要花多少钱？有底？没底？？？

Google Drive是什么? Google Drive(不要跟Google现有的代号为"鸭嘴兽"的内部Gdrive项目混淆了,)是一个网络应用,允许Google的用户――那些有Google 帐户的人――储存文件,并可以从任何地方获取.让我们先给新用户100GB,只是想有些像Gmail发布时的那种轰动效应(而且跟Gmail一样, Google Drive应该也会在测试阶段采用邀请制). 这项服务的其它潜在的名字有:Gdrive、Google Files、Gfiles、Google Store、Google Zone、Google Data、Gvault、G:、或者 Google Storage. 主屏幕 用你的Google帐户(跟Gmail的用户名和密码一样)登录到Google Drive以后,你会在左侧看到一个文件夹列表,在右侧看到文件列表,上下分别是搜索框和广告. 文件夹系统 Google Drive能够重新使用Windows资源管理器的概念,但会给网络带来一场变革.不需要在这里重新发明一下车轮,因为人们已经习惯了这个界面,因此他们 将会很容易理解这个新工具的概念.你可以展开和收缩文件夹,打开一个文件夹和它的内容,还可以上传新的东西,并在线浏览(或者下载后离线浏览). 文件夹有三种安全设置:向世界分享(人头图标)、隐私(钥匙图标)、和向被邀请 的朋友分享(人头+钥匙图标).在默认设置下,所有东西都是隐私的,因此只能通过你的Google帐户访问到.(当有人盗取了你的密码,你会比以前有更大 麻烦!) 另外,还有一个Google系统文件夹的区域.那不是被用户创建的,而是由 Google来自动填充.例如,当你有一个Blogspot博客并写了一片新日志,一个新的子文件夹会被添加到你的"Google services/Blogger/"目录里.这个新文件夹包含了这篇新日志的图片和文字内容.对Picasa网络相册(Picasa到那个时候可能会被 叫做"Photos")或Gmail等等也是一样.届时你就能够备份你所有的Google数据――或者部分数据――通过点击在任何打开的文 件夹下的"下载文件夹"链接.当然,你还可以搜索储存在你所使用的所有Google网络应用里的数据. 要注意的是Google系统文件夹本身不能被编辑――内容和安全设置都不行―― 并且在里面的文件会自动地被链接到你在其它地方的Google数据.因此当你编辑了一片就日志以后,在这个文件夹里的对应内容也会改变.还有,Google 系统文件夹将会占用你的100GB免费空间,因此其它的Google服务将不会再有它自己的储存限制……也就是说如果你愿意的话可以用55GB来 存储Gmail数据,只要在Google Drive上有55GB的剩余空间. 管理文件 当你打开一个文件夹的时候,会有一个包含至少10个文件的列表出现在眼前.你可 以选择一堆文件然后执行一些操作,比如删除、复制、移动等等.你还可以给文件加星标来更好地记住它们,或者是加标签,像"工 作"、"娱乐"、"骑马"之类的关键词. 当你点击一个文件的时候,会出现几个选项,如: 在线查看这个文件,例如一个Word文档可以在Google文件中打开(这项操作会打开一个新窗口,如果这个文件夹是公开 的话,还会有一个跟别人分享的永久链接――对了,新的窗口也会有广告) 把文件下载到本地PC 如果文件类型是图片,当你点击这个文件的时候会马上出现这个图片的大图(另外, 图片在文件预览中会有一个缩略图).其它类型的文件或许也可以直接预览,比如文本、MP3、WMV、XLS、PDF或者富文本. 文件的其它信息可能也可以看到,像大小(KB)、Google的病毒检查状态, 如果是共享文件,还会有一个"查看次数"统计和文档的反向链接. Google系统文件(也就是储存在其它Google应用上的东西)还有一个特 殊模式:当你点击Google系统文件的时候,你会看到文档的源文件,比如在Gmail帐户中的一个Email,或者Blogger博客里的某篇日志的编 辑模式. 搜索数据 在Google Drive顶部的搜索功能,很引人注目,因为它不仅能找 到你上传的任何文件(在短短几秒钟内),还可以搜索你现有的Google数据.如果你在Orkut 里建了一个档案,你的档案和你的朋友的名字都能被找到.当你在Gmail里写了一封Email,也同样能被找到.还有你的Google Base数据、搜索历史(如果开启了的话),博客日志、电子表格、你上传到YouTube上的视频的评论、Gtalk聊天记录、应用在个性化主页上的小工 具、你在别人的Blogspot博客上的评论(如果你发表评论的时候Google帐户处于登录状态)、Google Checkout的付款历史和详情, 等等. 在另一方面,高级搜索选项允许你去指定具体想找什么.你可以搜索 最近修改过的文档,或者只搜索非Google系统文件,或者只搜索图片,或者只搜索你分享的东西,等等. 顺便说一句,由于Google的自我审查功能,在中国的用户将不能够在 Google Drive里搜索特定的敏感词汇,像"×权"或者"××功"(包含这些关键词的文件也不能被储存在drive.google.cn上).Google将会 争辩说审查Google Drive违背了他们的使命,但不给世界五分之一的人口提供存储服务是更严重的事.其它国家可能也有类似的选项内建在 Google Drive里,取决于什么是违法的或者在当地不受欢迎的. 上传文件和备份 无论何时你打开一个文件夹,都会在下面看到一个"上传"按钮.你能够通过点击它 来选择一个本地PC上的文件并上传,比如一个PDF文档.可能还会有输入过滤器,过滤所有储存在外部网站上的文件.此外,Google会提供给你一个同步 工具――独立的,或者作为Google桌面套件的以部分――因此你能够选择将文件夹的创建/更新自动化. 你不仅能上传文件,当然还可以下载.当你点了"下载"以后,一个ZIP文 件下载进程会被初始化.这个ZIP会包含在一个特定文件夹中的所有文件(以及子文件夹、子文件).例如,现在备份你所有的Gmail邮件就像备份 Blogspot博客一样容易了,因为你不用再去学习具体的备份过程. 而且因为 "Google服务"就是里面的一个文件夹,所以你基本上可以通过一次点击来备份你的整个Google OS.另外,使用Google Drive同步工具,你可以把这个过程自动化(比如,把同步工具设成每24个小时下载/上传你的数据――当然是在你开机的时候). 现在,备份功能不只是一个让人们使用Google Drive的有吸引力的功能.突然间,所有Google连接的应用对用户来说都会变得更有价值,因为封闭的东西更少了.讽刺的是,这个交叉集成服务涉嫌 "Google垄断",因为人们可能会这样说: "我想开我的第一个博客,我该用哪个呢?Blogspot还是 TypePad?" "TypePad有更酷的功能,但你最好还是用Blogspot…… 因为它集成在Google Drive里." 我没有在这里说TypePad比Blogspot更好还是更差,但你明白我的意 思――当Google Drive真正发布的时候,所有其它网络应用竞争者突然都有了跟Google OS"不兼容的文件格式",这让他们的服务更没用了! 相关广告 我们不能期望Google Drive把那些存储空间和服务白白送人,对吧?广告将会成为Google在这项服务中挣多少钱的关键.跟传统的广告条不同的是,这些广告将会跟 你正在看的文件和文件夹相关.如果你的文件叫做"my-web-host-terms-of-service.txt",那么下面的广告可能会是 "$1.99域名注册". 就像在Gmail身上发生过的一样,一些人会误解这项功能,并且很害怕: "Google是怎么知道我的隐私文件的内容的?"答案是你向Google Drive上传任何东西,它当然都会被放到Google的服务器上,因此从技术上讲,每个有接入权的Google员工都能够看你的文件.在另一方面, Google的动机不是检视你,而是从你身上挣钱,而且他们的广告投放完全是自动的,所以没什么可怕的,倘若你信任Google的话……还有倘若你 信任你 的ZF的话,因为他们可能会要求Google在背地里提供你的隐私数据(比如美国). 集成到其它Google应用中 Google Drive扩展了drive.google.com(不管子域名将会是什么)这一网站的范围.因为现在,无论何时你向Google上传东西――比如 点击Gmail里的"添加附件"――你都会有两个选择……从本地磁盘上传还是从Google Drive上传.也就是说当你在Google Drive上存了什么东西的时候,你可以在多种Google产品中调用它. 从Google Dirve"上传"东西的好处是那个文件已经在Google服务器上了,你不用再等待上传的过程.如果你以前把一个视频上传到了Google Video上,那再往YouTube上传的时候只要一秒钟就能完成(因为你可以通过Google Drive接入到Google系统文件).Google只要在自己的服务器上处理一下本地资源就行了,你不用发送任何数据……只要你在Google OS上.

What's Google Drive? Google Drive (not to be confused with Google's existing internal Gdrive project, code-named Platypus) is a web application to allow Google users � those with a Google account � to store their files, accessible everywhere. Let's hand out 100 Gigabytes for starters, just to have some wow factor like when Gmail was released (and like Gmail, Google Drive ought to start out as invite-only Beta). Other potential product names for this service are e.g. Gdrive, Google Files, Gfiles, Google Store, Google Zone, Google Data, Gvault, G:\, or Google Storage. The main screen After logging into Google Drive with your Google Account � the same user name and password used in e.g. Gmail � you'll see a folder pane to the left, a file view to the right, a search box on top and some ads below. The folder system Google Drive can reuse the concepts of Windows Explorer, but bring the whole cabundle to the web. There's no need to reinvent the wheel here, as people are already used to this interface, so they will have an easy time understanding the concept of this new tool. You can expand and collapse folders, open a folder and its contents, and upload new stuff to it, and view this stuff online (or offline, by downloading it again). Folders come in three security flavors: shared with the world (the people icon), private (the keys icon), and shared with a group of invited friends (the people + keys icon). By default, everything is private, so it can only be accessed via your Google Account. (When someone hacks your password, you're in even more trouble than before!) Additionally, there's a section for Google system folders. These are not user-created, but automatically filled by Google. For example, when you have a Blogspot blog and you create a new post, a new sub-folder will be added to your "Google services/Blogger/" directory. This new folder will contain all the post's images as well as the content of the post as Rich Text file. The same for a Picasa Web Album (Picasa is likely called "Photos" by then), or your Gmails, and so on. The neat thing is that you will thus be able to backup all of your Google data � or parts of it � by hitting the "download folder" link below any opened folder. And of course, you will now finally be able to search for all your data stored across the different Google web applications you use. Note that the Google system folder itself cannot be edited � neither its contents, nor its security settings � and files in it will be automatically linked to your Google data elsewhere. So as soon as you edit an older blog post, it will also be changed in this folder. Also, the Google system folder will count against your 100 Gigabyte free storage space, though individual other Google services will now have no more (specific) storage limitations on their own... meaning you can have e.g. 55 Gigabytes of Gmail data if you want, as long as there are 55 Gigabytes left on your Google Drive. Handling files When you open a folder, you're presented with a list of the latest 10 or so files in it. You can then select a bunch of files to apply actions to them, like deleting, copying, moving and so on. You can also star files to better remember them, or tag/ label files with keywords like "work," "fun," "horse riding" and so on. When you click on a file, you're presented with a couple of options, like: View this file online, e.g. a Word document can be opened in Google Docs & Spreadsheets (this will open in a new window, which has a permalink you can share with others, if the folder is public � the new window will include ads too, by the way) Download this file to your local PC If the file-type is an image, then you will also be immediately presented with a larger version of the image upon clicking the file (additionally, images will contain a small preview in the files overview). An instant preview may also be available for other file-types, such as text, MP3, WMV, XLS, PDF or Rich Text. Other information may be available for files as well, like its size in kilobytes, its Google's virus check status, and if it's shared, a "times viewed" statistic and a list of backlinks to the document. There is a special mode for Google system files (that is, the stuff stored in your other Google applications): whenever you click on Google system file, you will be additionally offered to be taken to the source of the document, e.g. an email in your Gmail account, or the edit mode of your Blogger blog for a particular post. Searching your data The search feature on top of Google Drive is quite remarkable because it will not only find any file you uploaded (in split-seconds), but also search across your existing Google data. If you created a profile in Orkut, your profile and the list of your friend's names will be found. When you wrote an email in Gmail, it too will be found. The same goes for your Google Base data, your search history (if enabled), your blog posts, your Google spreadsheets, descriptions of videos you uploaded to YouTube, your Google Talk history, names of gadgets you included on your personalized Homepage, comments you wrote at Blogspot blogs of others (if you were logged-in to your Google Account at the time), your Google Checkout payment history and details, and so on. The advanced search options on the other hand allow you to specifiy exactly what you're looking for. You can search for documents that were recently changed, or search through your non-Google-system files only, or search for images only, or search for stuff you shared only, and so on. Users in China, by the way, thanks to a Google self-censorship feature, will not be able to search their Google Drive for certain sensitive terms, like "human rights" or "falun gong" (neither will they be able to store files containing such terms on drive.google.cn). Google will argue that censoring Google Drive compromises their mission, but that failing to offer storage at all to a fifth of the world's population does so far more severely. Other countries may have similar options built into Google Drive, depending on what is illegal or unwanted locally. Uploading files, and making backups Whenever you open a folder, you'll see an "upload" button below. Clicking it will allow you to select a file from your local PC, e.g. a PDF, and upload it. There may also be import filters, say for all files hosted on an external website. Additionally, Google will offer you a synchronization tool � standalone, and as part of the Google Desktop suite � so you can opt to automate the folder creation/ updating process. You can not only upload files, but download them as well, of course. When you hit "download," a ZIP file download process is initiated. This ZIP will contain all files (and sub-folders/ sub-files) in this specific folder. Backing up all of your Gmail emails, for example, is now the same process as backing up your Blogspot blog, which is great because you don't have to learn about specific backup procedures. And because "Google services" is a folder in itself, you can pretty much backup your whole Google OS with a single click. Also, using the Google Drive Synchronize tool, you can automate this process (e.g. set the synchronizer to download/ upload your data every 24 hours � when your computer is on, that is). Now, the backup feature is not just a nice argument to convince people to use Google Drive. All of a sudden, all kinds of Google-connected applications will suddenly become more valuable for users, because there's less of a lock-in. Ironically, this cross-integration has some interesting "Google monopoly" implications, because people will go like this: "I want to start my first blog. Which should I use, Blogspot or TypePad?" "TypePad has cooler features, but you better go for Blogspot... it's integrated into your Google Drive." I'm not saying TypePad is superior or inferior to Blogspot, but you get the point � with the release of a true Google Drive, all other web application competitors suddenly have "incompatible file formats" for the Google OS, making their services less useful! Related advertisement We can't expect Google Drive to give away all that storage and service for free, right? The ads will be how Google makes money from this service. As opposed to traditional ad banners, these ads will be targeted to whatever files and folders you're currently looking at. If your file is called "my-web-host-terms-of-service.txt", then the ad below may read "$1.99 Domain Registration". Just like it happened with Gmail back then, some people will misunderstand this feature and be scared of it: "How does Google know the contents of my private files?" The answer is that as soon as you upload anything to Google Drive, it will of course have to be transferred to Google's servers, so technically every Google employee with access will be able to read your files. On the other hand, Google's incentive is not to spy on you, but make money from you, and their ads are targeted completely automated, so it's nothing to be scared of provided you trust Google... and provided you trust your government, as they may be able to query Google for your private data behind your back (e.g. in the US). Integration into other Google applications Google Drive extends the scope of just that single website drive.google.com (or whichever sub-domain it will be). Because now, whenever Google offers to you upload something � say, you hit "add attachment" in Gmail � you will now get two choices from where to take the file... from your local drive, or from your Google Drive. This means when you store something in Google Drive, you can can access the resource from within a variety of Google products. The nice thing about picking files from your Google Drive for "uploading" is that as the file is already on Google's server, you don't have to wait for the upload time to progress. Uploading a video to YouTube will be a matter of a second if you previously already uploaded it to Google Video (because you can access your Google system folder using Google Drive). Google just has to handle sharing local resources among its servers, and you don't need to transmit any data... as long as you stay on the Google OS, that is.

20051116 21世纪经济报道 周默鸣 Craigslist 美国最火的分类广告网站

Craigslist是美国最高访问度的本地化、分类广告网站 商业网站，客户功能第一，界面简洁第一，信息的高针对性第一，数据的高度结构化及查询的高度智能化第一，超级数据仓库的检索高效率第一，项目规划中就应考虑多点大规模、瞬间、海量爆发式访问。 利基行业商业网站的核心业务就是提高网络数据的结构化程度。 对于中国大陆目前市场，由于社会、政治民主化进程严重滞后，网络可到达性暂时是是很难的目标（可预计的将来，世界上最高可用性的DNS服务，如ghs.google.com仍将会因为政治原因被刻意屏蔽），所以，商业网站不可能由商业企业自行部署在托管的本地IDC机房，合理、可靠、经济的解决方案是必须借力业界领导企业的 Web OS （存储、数据仓库、计算、数据库、电子邮件）解决方案平台，如Google AppEngine， Amazon AWS，Akamai CDN。 网站应用架构前，直觉虽然最最重要，但是及时完成家庭作业仍必不可少，专注于细节会坏事，完全没有细节也无法成事，平衡感很重要。 首先，必须完整、精确、及时理解、解构目标客户（市场）需求。不现实的、不合理的、非常急迫的、看起来好像短时期内很难实现的客户需求就是市场最大的利基，根本无需考虑实现该功能所需的带宽、存储及网络计算资源。 中国大陆人民信赖盗版，国家权力机关宽容盗版，这意味着降低知识产权及其实现方式的分销渠道成本、海量扩大销售规模、降低正版用户购买成本有极为急迫的市场需求。 面对中国大陆市场，网络作为现当代、及未来很长时间最最重要的通讯手段（也是商业工作流平台），最迫切的市场需求就是网络的高可到达性、高稳定性、数据的高针对性。 毋庸讳言，GFW是很好的利基市场，可以导致巨大的商业收益。 Craig在45岁创业才成功，初始仍是以本地化社区免费广告服务(Google Base)为导向，功能单一，界面简洁。

Craigslist：探访美国最火的分类广告网站     作者：周默鸣 | 出处：21世纪经济报道 | 2005-11-16     特约记者 周默鸣 旧金山报道     Craigslist的网站上没有图片，只有密密麻麻的文字，标着各种生活信息，是个巨大无比的网上分类广告加BBS的组合。虽然看上去颇为乏味，可是 Craigslist却是美国人最喜欢的网站之一。有人在这里卖掉自己的旧车，有人在这里租到中意的房子，有人在这里找到工作，有人在这里找到帮自己修电脑的人，还有人在这里找到女朋友。 　　"没有商业模式" 　 　Craig Newmark今年52岁，单身，美国有媒体把他列为旧金山湾区(Bay Area)最受欢迎的单身汉之一。Craig Newmark是全球最大也是最早的分类广告网站Craigslist.org的创始人，可是他现在的头衔既不是总裁，也不是CEO，他仅仅是公司的客户服务代表。他每天要花8个小时以上的时间，坐在电脑前，处理各种投诉。 　　在11月11日接受采访前，记者与Newmark通过三封邮件，他回邮件的平均时间是10分钟。考虑到他每天要处理数以千计的邮件，这个回复速度近乎惊人。 　 　在远离旧金山闹市的Sunset区，Newmark租了一幢三层楼的公寓作为办公楼。他特意挑了一个远离市中心，但是交通颇为便利的区域，办公楼的前面 就有有轨电车通过，站在街对面，很难辨认出这就是家喻户晓的Craigslist网站的所在地。Craigslist.org的招牌混在其他招牌之 中，仔细搜索，方能看到。Newmark和CEO Jim Buckmaster共用公寓第三层的一个房间，楼道狭窄，房间凌乱，除了几台电脑，书和杂物堆满了桌子。 　　 Craigslist的网站上没有图片，只有密密麻麻的文字，标着各种生活信息，是个巨大无比的网上分类广告加BBS的组合。 虽然看上去颇为乏味，可是 Craigslist却是美国人最喜欢的网站之一。有人在这里卖掉自己的旧车，有人在这里租到中意的房子，有人在这里找到工作，有人在这里找到帮自己修电脑的人，还有人在这里找到女朋友。 　　现在，每月有1000万人在使用Craigslist网站，月浏览量达到30亿次，网站的网页数量在以每年近百倍的速度增长。按照Buckmaster的说法，以目前的增长速度，Craigslist的页面数量将很快超过 Amazon、 Yahoo！甚至Google，而Craigslist至今却只有18名员工。 　　Craigslist每月新增的帖子近10亿条，其中只有不到1%是需要付费的，尽管如此，外界估测 Craigslist近年的年收入已经达到1000万美金。究竟收入到底有多少，Newmark一直拒绝透露。 　　"我们的首要任务不是挣钱，服务社区是我们的首要任务。"Newmark对记者说，"我们不会上市，不会引入风险投资，不会在网站上刊登广告，不会收取会员费，大部分人的帖子将是免费的。" 　　据Newmark介绍，近几年，为数众多的风险投资公司找过他，就连他自己也记不清具体有多少家，但是他从来没有动过引入风险投资 的念头，他说自己并不缺钱，更不想被风险投资所左右。 　 　"1997年，我们网站每月的浏览量突破100万，当时微软来找我，要在我的网站上登广告，被我拒绝了。"Newmark说，至今， Craigslist上没有任何形式的广告，网站所有的收入来源，都来自于旧金山、纽约和洛杉矶三个城市的招聘广告，旧金山的招聘广告每条收费75美元， 纽约和洛杉矶每条收费25美元。 　　事实上，Craigslist已经在全球120个城市开通，包括在中国的北京和上海，南非的开普敦，还有位于加拿 大、南美、欧洲等多个国家和地区的城市。 　　"有人说我们是一个没有商业模式的公司。"Newmark说，"确实，我们没有商业模式。" 　　"不收费"也盈利 　　从1976年到1993年，Newmark一直呆在纽约IBM工作，是一个颇为成功的软件工程师，但是他的性格却与众不同。 　　11月11日中午，记者走上狭窄的楼道，来到三楼，Newmark从办公室应声而出。他略显羞涩，说话不紧不慢，回答问题总是没几句便戛然而止。直到下午3点钟采访结束，Newmark才对记者说，他还没有吃中饭。 　　Newmark说自小他就容易受同学捉弄，很少有同学邀请他一起玩，他在"社交能力"这一项的得分一直很低，六年级的时候，经多次教育无效，他的老师无奈之下只好把他交给了校长处理，对于自己的不合群，校长也束手无策，最后只好教他下象棋，因为象棋是最不需要社交的游戏。 　　除了象棋，世界上还有一个不需要多少社交的工作，那就是软件设计，于是Newmark在IBM一干就是17年。 　　1993，Newmark终于对自己的工作感到厌倦了，他从美国最东部的纽约，来到最西部的旧金山，试图改变自己的生活。不过他找 到的工作，仍然是软件设计，在一家叫Charles Schwab的公司，他一做又是两年。 　　"20年的积蓄，至少可以让我失业几年，我想做一件自己喜欢的事情。"1995年，Newmark辞掉了工作，"我当时根本没有想到，我会拥有一个网站，更想不到网站可以变成现在这个样子。" 　 　Newmark开始设计一个自动发送email的软件，通过这个软件，Newmark可以通知他的朋友将要在旧金山举行的各种文化和艺术活动的消息。随着越来越多的人要求加入email发送名单，很快他就感到这个email发送系统无法承担这么大的发送量。于是，他设计了一个网站，感兴趣的人可以随时上网查阅，比发送email方便很多。 　　网站雏形出现后，Newmark原本想把给他命名为"SF Events"，但是他的email发送系统，早已经被人们习惯叫做"craigslist"，有人提议沿用这个名字。Newmark最终选择了 Craigslist.org，而不是Craigslist.com。 　　"因为我并不想把网站变成一个商业网站，我想做的，是一个社区服务网站，一个志愿者网站。"Newmark说。 　 　网站成立以后，Newmark与几个志愿者共同维护网站。Craigslist上的内容也逐渐扩展，在当初的文化艺术活动信息中，慢慢加入了旧货买卖， 应聘招聘等信息。网站越做越大，Newmark依然坚持不收费原则，也依然把网站定位于旧金山以及周边城市，因为旧金山是一个信息需求非常巨大的城市，旧金山湾区也是美国最活跃的地区。 　　 两年后的1997年，Craigslist的月浏览量达到了100万。 　 　然而网站的扩张并没有留住志愿者，因为很多人无法长期坚持无偿工作。虽然不断有新人加入，但是离去者甚多。1999年，Newmark决定停止志愿组织的形式，组建公司，雇佣员工。同时，Newmark也跨出了收费的第一步，他决定对在旧金山及周边地区公司的招聘广告收取每条25美元的费用。这一收费后来涨到了45美元，并且最终涨到了目前的75美元。他随后增加了网站维护人员和客户服务人员，并新增了会计人员，专门追踪招聘广告的收费。 　　"互联网泡沫危机中，95%以上的同类网站纷纷倒闭。而让别人不解的是，一直不以赚钱为目标的我们，却从1999年开始一直保持盈利。"网站如此受欢迎，连Newmark自己也感到吃惊。 　   Newmark打假 　 　1999年底，一个叫Jim Buckmaster的人开始在Craigslist上发帖子。除了通过Craigslist卖掉不少旧货、租到一间房子之外，Buckmaster还发布了自己的应聘简历。而第一个给他打电话表示想招聘他的人，正是Newmark。Newmark当时已经越来越穷于应付日渐庞大的网站，他需要找一个专门管理公司的人。 　　2000年，Buckmaster正式成为Craigslist的CEO，Newmark本人则退居为客户服务代表。 　　Buckmaster上任后的第一件事，是把Craigslist扩张到了湾区以外的第一个城市波士顿。在Buckmaster逐渐把Craigslist推向全球各大城市的时候，Newmark一头扎入了自己的客户服务。 　　"我不再是一个工程师。"Newmark说，"我现在最重要的工作，是清理我们的网站，帮人们解决各种问题。"现在经他的手清理的信息，每天将近1000条。 　　随着Craigslist上的内容爆炸式增长，恶意信息也逐渐出现。有人在上面张贴商业广告，Newmark和另外4个客户服务代表的任务是把商业广告一一删除。 　 　"Craigslist是一个免费自由平台，所有信息的发布都是自由和免费的，对于恶意信息的监控，我们主要倚赖的是客户的反馈。"Newmark说， "这是一个良性循环，我们的反馈越及时，人们就越信任我们的网站，也就会有越多的人自愿向我们汇报他们发现的虚假信息。" 　　"有一天，有人向我们汇报，纽约一个房屋卖主在网站出卖产权不是他自己的房子，我们很快就删除该帖子。" Newmark说，"如果有人不断散发虚假信息，我们还会停止他们的账号。" 　　不久前，Newmark又接到了一个投诉，说在网站上有人张贴招揽妓女的广告。他马上向旧金山警察局报了案，"警察还在追踪，他们经常给我打电话反馈。"Newmark说。 　 　"你必须非常耐心，专心处理好所有细节。"Newmark说，在他看来，客户服务是Craigslist得以生存到今天的最重要原因，美国并不缺少同类型的网站，但是很少有人能坚持10年，并且人气越来越旺。去年，旧金山市破天荒地把10月10日定为Craigslist Day。 　　Craigslist还在不断向其他城市扩展，5个客户服务人员显然不够。Newmark正在招揽更多的人，但是为了控制成本，他并不打算招揽过多的员工。他认为Craigslist需要的是上网者共同的信任和维护。 　　"我们的首要任务，并不是赚钱，而是社区服务。"Newmark说，"很多其他的网站把赚钱当成第一目标，这就是为什么他们的网站上有这么多收费项目，这么多跳出式广告，甚至这么多有偿信息。 　　联手eBay 　　2004年8月，eBay买走了Craigslist 25%的股份，这些股份来自于一位Craigslist的持股者。Newmark一直不肯透露任何与股票出卖相关的信息。但是他坚称，自己不会卖掉 Craigslist。 　　有美国媒体估计，如果Newmark卖掉Craiglist，价格将在1亿美金以上。 　 　eBay的发言人曾经对美国媒体说，eBay购买Craigslist是想积累更多有关分类广告的经验。Buckmaster则表示，eBay的入股可以让Craigslist在如何防范欺骗性信息、保护网络用户上做到更好。另外，他认为，eBay的入股，还有助于Craigslist的进一步国际化。 　　"eBay不会参与我们的管理，我们也不会改变既有的免费方针。"Newmark说，虽然Craigslist也在逐渐扩大自己的收费项目，比如说，从明年开始，他们将向纽约的房屋出卖广告收取费用，但是永远也不会对个人张贴的信息收费。 　　Newmark告诉记者，11月19日，他将和Buckmaster去牛津大学参加一个互联网研讨会，会后Newmark将回到位 于新泽西州的家乡，与父母团聚，他还将去见女朋友的父母，"她和我是自小一同长大的伙伴"。

Stopdesign Redesign the Blogger.com for Google

Blogger界面模板动一下就是上千工时。

http://www.stopdesign.com/log/2004/05/09/blogger.html

For those of us involved in the project, we’ve been waiting months for this day to come. At long last, I’m proud to announce the launch of a project representing the latest collaboration between Stopdesign and Adaptive Path: the redesign of Blogger.com. Congratulations to the entire Blogger team on completing hundreds of hours, and expending tremendous effort to fit so much into this launch. This is Blogger’s first major overhaul since getting acquired by Google in February 2003, and it’s a biggie. The redesign includes a major reworking of Blogger’s home page and sign up process, and a new dashboard for logged in users. With this redesign, we focused on making the Blogger brand and interface much more friendly and approachable. The design features rounded corners, large icons, direct concepts, and helpful directions. Adaptive Path and Stopdesign worked with Google to pare down the sign up process (account creation), making it short, intuitive, and effortless for first-time visitors. We aimed at creating a structure and design that has as wide a range of appeal as possible. From the large base of Blogger’s existing users, to the multitudes arriving from Google on a regular basis who have no idea what the word blog means. The Blogger home page now promotes the simplicity of creating a new blog by highlighting the short 3-step process. The page also presents a simplified definition of the word “blog” right up front, and directs new users into a short tour if they want to learn more. In addition to the site redesign, Google ups the ante for this relaunch by adding a slew of new features for their users: comments, user profiles, individual post pages, and conditional template tags. On top of that, they’ve significantly revamped the application interface to mirror the new look and feel of the site. A few additional notes about the redesign: New Set of Blog Templates One of the new enhancements Stopdesign worked with Google to complete and include in this launch is the creation of a whole new set of user templates available for immediate use. Rather than have all templates created by Stopdesign, I thought it would be fun to expand the team involved. What if I asked other talented individuals I respected and wanted to work with to join me in designing brand new templates? I wanted to share in the responsibility of template design and creation, providing even more variety to Google and Blogger’s user base. You may know most (or all) of the names of those who graciously accepted to help me create these new templates: Dan Cederholm Todd Dominey Dan Rubin Dave Shea Jeffrey Zeldman A peek at a few of the new blog templates: The full selection of templates (along with their full-size previews) are visible inside the Blogger application. You may see more from the individual template designer on their respective sites about the design choices and techniques each of them used. As for the impact of this part of the project, not only will thousands (millions?) of blog templates now be generated with more flexible, scannable, indexable, and standards-compliant code, but anyone throughout the world can now create an account, start a blog, and have a personal site designed by any of the talented gentleman listed above, all for free. The Evolved Identity  Sharp-eyed party-goers at the Blogger party in Austin during SxSW may have noticed the signage and multitude of free t-shirts they handed out which featured a new identity with rounded corners and a new type treatment to match. When we set out to establish the new visual identity of the site, we kept running up against the hard corners of the old Blogger logo, and the somewhat impersonal type treatment which limited the friendly qualities we were trying to inject into the brand. We didn’t want to push the logo far from what everyone knows — after all, the big orange B is widely known and loved by many a folk. Instead, we chose to evolve the logo, so it could take on a few new attributes. We sanded down some of the sharp edges and refined a few of the proportions, then carefully selected a typeface more suited to a product targeting the masses. Home Page Visitors without Blogger accounts, or users currently logged out will see a different home page at blogger.com than users logged into the system. The logged out page provides answers to the question, “What’s a blog?” and highlights how short and simple the new account/blog creation process is. The logged in page presents a new Dashboard view. The Dashboard acts as an enhanced intermediary between the site and the application interface. It provides a customized view of the Blogger home page by displaying a quick overview of the user’s blogs, related profile information, updated blog content, and Blogger news. Can’t say what else is in the pipeline for additions to the dashboard, but users can most likely expect many more changes and new features in the near future. The new icons and buttons produced for Blogger (visible on the logged out home page and throughout the site and application) are intended to complement the amiable nature of the new brand. They pick up where I left off on a style I started using about six years ago for the Webmonkey Kids design. Some of the icons were sketched on pen an paper (ahem, numerous times). After finalizing an approximate look, the icons were drawn in Illustrator using a pressure-sensitive Wacom tablet and pen to create the base artwork, then cleaned up by manipulating paths and Bézier curves within Illustrator. Design Execution As true to the Design Process article, this project started with discovery and competitive analysis phases, led by the Adaptive Path team. I’ll let them go into as much detail as they want about the early process for this project. Freeform exploration and brainstorming produced descriptive words, qualities, and elements to be used in the design. This was followed by tiny thumbnail sketches exploring loose layout and composition ideas. A few rough ideas sketched on paper made use of bold lines and thick rounded corners, which eventually led to the final design. I’m using Adobe Illustrator more and more often for comping designs, and the Blogger design was no exception. It allows for easy object creation, manipulation, and resizing since everything is in vector format. I can easily move whole sets of objects, guides, and paths. The recent addition of the Effects menu allows me to apply a rounded corner appearance to each block without actually rounding those corners, as seen below: (One gripe about using Illustrator for web stuff: after 11 versions, there’s still no easy way to simulate underlined text without manually drawing lines with the pen tool, then moving them every time the text reflows.) A few clickable prototypes, several days of user testing, more tweaks and changes, and the down-and-dirty CSS work began. (Note: CSS didn’t really even enter the process until this final stage.) Technical Insight I won’t go into much detail here on technical details. But I will confess that this project is what finally prompted me to write about the IE Factor. The frustration of building a design in CSS and getting it to work across all browsers is certainly more apparent when you’re trying things you’ve never seen done or explained before. IE was the main culprit. Both Windows and Mac versions proved exceptionally difficult to bring into submission, but at least the Mac version had known workarounds and filters. Some of the features of the design (like the subtle background changes when hovering over the icons on the home page) were completely disabled for IE because they produced unintended and unacceptable results. The rounded corners seen throughout the Blogger redesign (and in several of the user templates) make use of an expansion of the Sliding Doors technique written for A List Apart last year. The Blogger design is a fixed width, which means most of the modules of the site exist at pre-defined widths. Since the width of each module is known, one image is used for the top-left and top-right corners of a module, and another image is used for the bottom-left and bottom-right corners. The images are called in as background images for two nested elements. Since these two elements contain all the text of the module, they expand infinitely as the module grows in height. Think of it as Sliding Doors turned on their sides. For modules requiring a border, the two images are modified to include top and bottom borders connecting the two corners. A third element gets nested in the HTML that uses left and right borders which connect top and bottom corners. This design posed many other challenges when building it out, specifically because we wanted to allow the text and each of the design elements (header, modules) to be as flexible and scalable as possible. The markup construction was tricky and required compromises in several places. As is evident with the rounded corner modules, extra divs were necessary for each background image called in. In CSS3, border images will certainly help eliminate the need for extra elements. And I’ve been pressuring Tantek to get the CSS Working Group to consider allowing us to set multiple background images on one HTML element. Wrap Up As I have the habit of doing on product launch day, I’ve provided more than enough information for anyone to digest. I’ll close it off by saying that it was an incredibly fun project to work on. From the pleasure of partnering with incredibly smart folks at Adaptive Path, to the fun-spirited team at Google we got to work with, to the designers who agreed to help participate in a template project with little knowledge of the entire picture and, initially, not knowing who the client was. This launch is a huge deal for Blogger. They may not have gotten everything right out of the gate, so I’ll cut them plenty of slack, and I would hope others do the same based on the number of improvements and enhancements they’ve made all at once. For instance, those who know my style and promotion of web standards may immediately try to validate the home page. I’ll save you the time and tell you it won’t validate right now. (If you’ve already visited Blogger and hit your Validate HTML favelet within a minute of seeing the new design, shame on you. Don’t you have better things to do? Aren’t there other things more interesting than making sure a new site launches with valid code the very same day?) A few simple omissions of closing brackets, and use of the target tag within a XHTML 1.0 Strict doctype is preventing the front page from validating right now. They’ll get to it eventually. Bear with them as they try to make sure everything else is working correctly for all the users who depend on the service to publish their thoughts and experiences to the world. Fixes and new features will most likely continue to roll out in the weeks ahead, so expect changes over time. Overall, what a great project. Despite the frustration with browser rendering consistency that plagued us on the tail end of this project, it was still fun. Good client, good team, good product. What more could you ask for? For the rest of the story on new features and changes, check out The Great Blogger Relaunch, which tells the story from their perspective.