写在云端

反思

Sat, 26 Jul 2008 19:53:54 +0800

看了我们学校傅海阳教授在新浪博客上发表的《 必须正视TD-SCDMA可能存在的重大技术错误 》一文，看了后面那些评论，我有说不出的感觉，中国现在互联网的网民数目是全世界第一，但是素质有多高呢？有些网友动不动文章没看全，没仔细看就开始骂，而且都是匿名骂，有本事就亮出身份来，好歹教授以自己的真名来关注TD-SCDMA，提出自己的见解。动不动就开骂，这已经从天涯论坛和MOP转移到了新浪和搜狐博客中，我们的网民数目和网民的素质根本不是正比例发展的。真可悲！

看了那些谩骂的评论，有几个是专家，有几个真正的研究了3G，有几个是通信专业的呢？一群.........

还是要完全看完傅博导所有的博客文章在评论，起码你得知道傅博导为什么写这篇文章。

或许这也是很多个人博客要加上验证机制和审查机制的原因，网络上充斥着太多的另类。

此文有感而发，不针对某人，不针对某事，仅是对现在网络上的某种现象发表感慨！并不是因为傅博导是南邮的教授为我就出来为他讲话，而是现实的网络中，这种动不动就谩骂的人太多了！！

相关文章|Related Articles

本文若未特别注明则采用知识共享署名-非商业性使用-相同方式共享 2.5 中国大陆许可协议进行许可.转载或引用本站的文章请遵循知识共享协议.

[v] BBC 的奥运宣传片

Sat, 26 Jul 2008 18:08:28 +0800

Shared by 伊兹拉莱
没法说感觉

这个动画作品由伦敦RKCR/Y&R + Jamie Hewlet(街头霸王原设) + Damon Albarn(虚拟音乐团体Gorillaz 的灵魂人物 / 前英国Blur乐队主唱) 共同打造，目的是为BBC的奥运报道造势。借用煎蛋的话：“除了歌词有点雷以外，还是很牛逼的”。(视频↓)

奥运期间博客应当实行单双号限写

Sat, 26 Jul 2008 14:20:55 +0800

Shared by WilliamLong
我觉得, 应该发起单双号写博客运动: 从今天起, 互联网的所有中文博客应当全面实施奥运管制. 所有博客将按照单双号决定内容发布. 一方面是为了奥运期间为外国友人让出国际互联网的宝贵带宽, 另外也是减少尾气排放, 改善言论空气质量的一个好措施.
某专家做了一个统计, 从7月24日至9月20日, 限写措施将使中文互联网停止博客约20万个, 在北京的外国友人的网络条件和言论空气质量都将得到极大改善.

我觉得, 应该发起单双号写博客运动: 从今天起, 互联网的所有中文博客应当全面实施奥运管制. 所有博客将按照单双号决定内容发布. 一方面是为了奥运期间为外国友人让出国际互联网的宝贵带宽, 另外也是减少尾气排放, 改善言论空气质量的一个好措施.

某专家做了一个统计, 从7月24日至9月20日, 限写措施将使中文互联网停止博客约20万个, 在北京的外国友人的网络条件和言论空气质量都将得到极大改善.

我建议可以执行如下措施:

7月24日至9月20日, 不和谐的博客须按照单双号行驶措施规定写博客; 没有在国内备案的博客奥运期间禁止写.

较为和谐的博客和BSP托管的须按照单双号措施规定写作. 奥运期间禁止博客将携带易燃, 易爆等危险品关键词进入北京. 奥运时间早6时-晚22时禁止使用敏感词穿越防火墙, 以免给外国友人带来不必要的网络交通负担.

散兵游勇类博客(包括已经备案的博客)须按照单双号行驶措施写作. 同时, 非主旋律型博客奥运期间也注意减少书写.

和谐类博客, 新华网, 人民网, 千龙网等不受单双号的限制.

如果有违反, 拔线处理, 并罚款一百元.

其他细节:

在0点前写文章而0点后发布的, 仍要接受处理; 单双号期间，如果一市民拥有单号博客, 零点之前开写却需要在零时之后发布博客的话, 因已是次日凌晨, 是双号时间, 所以不能在互联网上发表. 如果不能赶在零时之前把博客发出去的话, 要么干脆放弃写博客, 要么存着草稿, 不能违反规定.

博客留言也同理, 单号的留言者不能去双号的博客抢沙发. 同时不能在单号日留言.单双号划分方法也很便民: 已经备案的博客按照备案号码的奇偶性确定博客的奇偶性. 没有备案的网友请按照姓氏比划. 使用英文ID的用户和留言用户请使用首字母的序数(A=1, B=2, 以此类推).比如张三, 张一共7划, 就只能在奇数日子写博客. 王四的王一共四划, 就禁止在张三的博客上留言. 奇数日也禁止留言. 用户 Dick 不能在 Cao 的博客上留言, 并只能在偶数日写博.

工业与信息产业部和网监部门将会加大执法力度, 如有违反, 严惩不贷.

哪位同意我这个措施的, 请随喜转载此文. 不必署名, 写不必提供出处. 我只希望把这样的国际主义精神散布到互联网的每一个角落. 整个互联网一起支持我们的国家, 把本届奥运会扮成史上最团结, 最和谐, 最让外国人满意的运动会.

另外, 我的号码是偶数, 我也在这里做个表态: 奥运期间, 我奇数日坚决不写博客.

乐高：北京，北京

Sat, 26 Jul 2008 11:30:02 +0800

乐高奥运北京，由香港乐高爱好者使用30万块乐高积木，4500个乐高小人组成，24平方米的乐高奥运城包括了鸟巢，水立方，还有民居等等北京奥运的特色建筑。

[link]

... 猛击这里参与该话题的热烈讨论，目前已有10人登录火星 »

Knol and void: The day I became a published Google 'expert'

Sat, 26 Jul 2008 03:00:00 +0800

I'm an expert on pit bulls! Really.

I just wrote a knol, a Web article akin to an encyclopedia entry, using Google's new Knol publishing platform launched publicly on Wednesday.

With Knol, Google is encouraging people to create more authoritative content that can be indexed by its search engine and monetized with ads. Unlike blogs, which tend to be casual and opinionated in tone, knols are supposed to be fact-based, informative, and well-sourced articles on a specific subject.

The author's knol on pit bulls.

Google is dismissing the notion that Knol is its Wikipedia killer, but both operate under the premise that Web users can collectively create a knowledge base that can be searchable and vast.

The difference is that while anyone can edit a Wikipedia entry, which can lead to pages and pages of edits and contradictory revisions and accusations of bias, knols have an author's name attached. A knol author is responsible for the content and can choose to allow others to edit it, or filter suggested edits or even block public editing entirely.

I decided to try Knol out. First I had to figure out a subject I felt I knew enough about. I walk dogs as a volunteer at the local animal shelter where there happen to be a lot of pit bulls. I've learned a lot about the dogs and have become disturbed by the amount of misinformation that circulates about them. So I did some research and wrote a knol entitled "The pitfalls of stereotyping pit bulls."

The hardest part was the research. But given that I do that every day for my job, it wasn't all that tough. I wrote the item in Microsoft Word and then cut and pasted it into the Knol page. It was easy to use the editing tools and add images. However, I think the page looks rather simple and dull. The system lacks the ability to add background colors and other stylistic flourishes that give blogs that individualistic panache.

Once I published the knol using the default "moderated collaboration mode," a colleague logged into his Google account and suggested an edit to my knol. I, in turn, rejected that edit (it's irrelevant that ex-Atlanta Falcons player Michael Vick, who I mention in the knol in reference to his dog fighting charges, wasn't that great of a quarterback). The system didn't notify my colleague that I snubbed his edit; he had to go to the page and keep checking the site for himself. It would be nice if the system were to notify people of the status of their suggested edits. Later, I found out that when an edit is accepted, the person who suggested it will be listed as a contributor in the contributor's list on the page.

My colleague, Tom Krazit, suggested an edit to my knol, which I subsequently rejected.

It also took a few hours for the system to index my knol so it could be searched via the main Knol page and even then, it only initially showed up when I searched by subject (pit bulls) but not by author name. By the next morning, I could search also by author name. The knol has yet to show up on the Google search page using both subject and author.

A Google spokeswoman said it takes time for the company to index new knols, but didn't say how long.

After some digging around I figured out how to add advertisements through Google's AdSense program, but I won't see any on the page for awhile (it can take up to two weeks, the system said).

Adding a New Yorker cartoon was simple. I was directed to the online New Yorker Store where I searched for cartoons dealing with pit bulls and found one. But when I added it into the blog it automatically inserted it at the top of the text and above the other image I had chosen. It didn't look right, so I removed it. If I had had the ability to determine where on the page the cartoon should go, I would have used it closer to the bottom of the page.

This New Yorker cartoon, while it was appropriate to the subject matter, was removed when I learned I couldn't control where it appeared on the page.

(Credit: The New Yorker)

I felt an odd sense of power, and responsibility, creating my knol. It gives me the ability to publish anything I want, without having to run it past an editor like I do at CNET News. And once it is published, it is a permanent record and has an air of legitimacy that editorializing and gossipy blogs don't have. It's a Google knol--"a unit of knowledge" as the Web site describes it, lending it at least the illusion of propriety.

But what if I wanted to write something inaccurate or defamatory? Already that question has been put to the test with a knol written by Rachel Marsden, the ex-girlfriend of Wikipedia co-founder Jimmy Wales.

Her knol is entitled "Jimmy Wales (Jimbo Wales)" and the summary describes Wikipedia as an "online libel board," that "any loser can use to smear people who are more successful than them."

I asked the Google spokeswoman about this situation and her response was: "Knol will be subject to our general content policies and terms of service, and knol content will be treated under those policies like any other user-generated content for which we provide a distribution platform. In particular, we will provide community flagging tools and the usual legal notification processes, so that we can comply with applicable laws and regulations. In addition, because knols are attached to verified author names, we think that the structure of Knol will actually provide something of a disincentive to defamatory or other harmful content."

It will be interesting to see how the Marsden-Wales fracas plays out on Knol. Google's response didn't give me any confidence that the system won't be widely abused. And it's likely that people who disagree with my knol will create one of their own with contradictory conclusions.

In an interview on Wednesday, Knol Product Manager Cedric Dupont said Google won't be determining the legitimacy of knols or verifying the authority of their creators. "We are not editors in any way," he said.

"We think we make it very easy for the user to determine the trustworthiness of the content."

I've deemed myself an expert on pit bulls by writing the knol. We'll see if the reader reviews and ratings suggest otherwise.

BMW 7系2009款,第5代宝马7系车

Fri, 25 Jul 2008 23:35:25 +0800

宝马公司（BMW）推出了第5代7系车，即2009款，将在10月份的巴黎国际车展上向公众展出，明年初国内开始销售，不久前已在莫斯科红场的巨型沙漏中揭开面纱。第4代的 BMW 7系是在 2001 年法兰克福国际车展上亮相，当时给不少人带来了惊讶，大众对 Chris Bangle 的争议达到一个高潮，相对于第4代的惊天动地，第5代 BMW 7系平静地延续推进了 Chris Bangle 带来的设计理念（revolution 到 evolution）。对 BMW 7系2009款的硬件参数这里不再详述，请搜索查阅。在外观上延续深化了第4代带来的华贵、动感、强健、驾驶乐趣等，并吸收了 BMW CS 概念车的一些设计元素，这两辆车的外观设计主持设计都是 Karim Antoine Habib ，更大（纵向）更强的鼻孔（进气孔），前大灯以及尾灯的变化等，从前脸看有点5系的影子，横贯车把手的腰线已经是流行的设计。下面是图集。
BMW 7系2009款（第5代）

[Show as slideshow]

BMW 7系2007款（第4代）

[Show as slideshow]

BMW 7系2009款在莫斯科红场的巨型沙漏中亮相。关于2009 BMW 7系详细的介绍可以参见 Paul Tan 的文章（马来西亚著名的 blogger ）。

BMW Concept CS 的视频：

2008.07.07 — 点线面,BMW活动雕塑 (10)
2008.06.11 — BMW GINA 概念车,布料表皮 (16)
2007.07.24 — BMW调研报告：汽车中的生活秘密 (1)
2007.04.14 — TED Talks- Chris Bangle: Great cars are Art (0)
2005.09.19 — BMW Z4 Coupé 概念车亮相法兰克福车展 (2)

A Small Design Study Of Big Blogs

Fri, 25 Jul 2008 09:33:00 +0800

We are quite curious; particularly when it comes to design and web-development we want to know just everything. Therefore, following our web form design survey, we have decided to take a closer look at blog designs, analyze them and find out which design solutions are common and which solutions are not used at all.

Since we wanted to make the survey as objective as possible, we used Technorati Top Blogs and analyzed 50 most popular blogs which appear there. It doesn’t really matter if the Technorati list is correct or not — we wanted to find out what design solutions big players (aka most popular blogs) prefer. Popular blogs are often considered to be examples of effective and functional design (which is not necessarily the case).

Blog Design Survey

We have identified 30 design problems and considered solutions for each of the problems separately. We have posed 30 questions which we would like to answer with our blog survey. Below we present findings of our survey of popular blog designs — the results of an analysis of 50 popular blogs according to Technorati’s Top 100.

Considering Technorati’s top 50, we have filtered out social networks and blogs which have artificially become popular — e.g. via backlinks in released Wordpress themes. The prize for the most cluttered CSS-code goes to Smashing Magazine — however, we are going to do something about it over the next weeks.

Please notice: the results presented below should not be considered as guidelines for an effective blog design. They are supposed to give you the intuition of which solution may be better than the other one. However, you need to consider the context you are working within as well — following our findings blindly won’t necessarily improve your design. Still it is useful to know what big players do and, more importantly, what they don’t do.

1. Layout

Let us start with the analysis of the layouts used on the most popular blogs in the blogosphere. Are there more 2-column-layouts than 3-column-layouts? Are layouts centered? Are they fixed, fluid or elastic? Are tables still used? In this section we are going to answer all these questions.

1.1. How many columns?

The question whether one should use 2 columns or 3 columns in a design layout is almost philosophical. Unfortunately, we weren’t able to find any results from usability studies which would prefer one solution over the other one. As usual, it depends on the context and on the target group you are dealing with. In some cases it is just impossible to find a proper balance between primary and secondary content with 2 columns. In these situations you may need to divide the second column (sidebar) into two parts — in fact this solutions is used quite often.

In both cases it is necessary to make the structure as transparent and clear as possible. A layout with 4 and more columns is often not a good idea.

According to our findings,

58% use three and more columns in layouts
(TalkingPointsMemo, CopyBlogger, Mashable, Lifehacker),
42% use 2-column-layouts
(Zen Habits, GigaOM, Google Blog, Seth Godin, Boing Boing).

TPM uses a multi-column layout. 58% of top blogs use the same design approach.

Apparently, 50 blogs are enough to find unusual layout solutions. Drudgereport uses something which we prefer to call “anti-layout”. And TPM uses sometimes 2, sometimes 3 and sometimes 4 columns. In the latter case we just don’t get why the titles of the articles are not clickable.

1.2. Centered or left-aligned layout?

Actually, one could observe the rising trend toward centered design layouts few years ago — often it was the attempt to find an optimal solution for readers with large and small screen resolutions. Since more and more users tend to switch to higher resolutions, one tries to balance the amount of space surrounding the layout by providing the same amount of white space on the left and on the right of the layout and thus focusing reader’s attention on the content by placing it in the middle of the browser window.

ReadWriteWeb with a centered layout. According to our survey, 94% of top blogs have a centered layout.

And apparently this trend has managed to become the standard approach over the last years. In fact, according to our survey, 94% of top blogs have a centered layout. Among them are CopyBlogger, Problogger, ReadWriteWeb, Mashable, Ars Technica, Techcrunch and Huffingtonpost.

1.3. Fixed, elastic or fluid?

To be honest, we haven’t expected such a strong trend toward fixed layouts (px-based layouts). It is truly remarkable that among 50 top blogs not a single one used an elastic layout (width of layout grows with the growing font size) and only a small fraction uses fluid elements (layout changes depending on the size of the browser window). Here are the exact findings:

92% of top blogs used a fixed layout,
8% used a fluid layout or a hybrid layout with fluid layout elements
(Engadget, Smashing Magazine, Gigazine, Coorks and Liars).

Fluid layouts can better adapt to user’s preferences while fixed layouts make it easier for the designer to make sure that a certain design decision holds independent of the font size and the browser window size.

The main disadvantage of fluid layouts is their growth in width when used with a wide screen resolution: the line length becomes to big (here in the example Engadget with 150 characters per line). One can counter it using the max-width-attribute.

1.4. Width of the fixed layout?

As we have observed a strong trend toward fixed, pixel-based design layouts, we have decided to take a closer look at these layout and try to detect common attributes of such layouts. In particular, we have considered the width of the fixed layout which usually corresponds to the width of #container or #wrapper. Apparently,

9% used ≤ 800px
(PostSecret, Seth Godin, Google Blog, BeppeGrillo.it),
15% used 801 – 900px
(Neatorama, Kottke, DailyKos, Perezhilton, TUAW, Yanko Design, Scobleizer),
20% used 901 – 950px
(Huffington Post, BoingBoing, TreeHugger, Dooce, Blogoscoped, SearchEngineLand),
56% used 951 – 1000px
(ars technica, Lifehacker, TechCrunch, ProBlogger, A List Apart, TMZ, Wired, GigaOM, Joystiq, Zenhabis, Copyblogger, Consumerist, Slashfilm).

Conclusion: one can identify a clear trend toward wide fixed layouts of the width 951 – 1000px.

1.5. Proportion between the content area and layout? (if fixed layout)

As you’ve seen above, each layout requires at least one sidebar which contains secondary content and provides navigation options to the users. However, what is the best way to provide users with a comfortable reading while clearly presenting all available navigation options? Or, put differently,
what proportion should one choose between the main content area and overall site layout? The less space the main content area takes, the more dominant the sidebar becomes and vice versa. Where does the balance lie?

The layout width on BoingBoing is 870px wide, the main content area is 550px wide. This results in a ratio 0.63.

According to our survey,

96% use at least a half of the whole layout to display the main content
(exceptions: CopyBlogger (0.48), SlashFilm (0.48),
54% used 50 – 60% of the layout for the main content
(Mashable, Lifehacker, Kottke, Blogoscoped, A List Apart, BoingBoing, DailyKos, TreeHugger, Scobleizer, Problogger, TUAW, bits.blogs.nytimes.com)
46% used 60 – 70% for the main content
(ars technica, TechCrunch, GigaOM, Dooce, Zenhabits, CNN Political Ticker, CrunchGear)
on average, 58% of the overall site layout is used to display the main content.

1.6. CSS-layout or table-based layout?

It is almost obvious that popular blogs, which are updated many times a day, should prefer CSS-solutions as these allow for better and easier maintenance and reduce load times. Consequently it is no wonder that

90% of the top 50 blogs used CSS-based layouts,
10% used tables or a combination of tables with CSS
(PerezHilton, Neatorama, CNN Political Ticker, Beppe Grillo, TreeHugger).

Furthermore, TreeHugger used the legendary onMouseOut-Javascript-event to imitate hover-effects — a design solution we have successfully managed to forget over the last years.

2. Typography

Content is king. This holds for “usual” web-sites as well as for blogs. Furthermore, it is designer’s job to make sure that the readers feel comfortable with reading or at least scanning the posts. And here is where typography comes into play. How can you achive best readability of your content? What font to use? Which font size to select? Our survey may provide you with some useful starting points for your design decisions.

2.1. Dark on light or light on dark?

As one would expect, 98% of the top blogs used dark text on a white background. Only PostSecret had a dark background color with light text displayed on it. However, this design decision is likely to be strongly related to the topic of the site.

Among top 50 blogs PostSecret is the only blog which uses dark background color with a light text on it.

2.2. How many characters per line?

To ensure best readability one needs to ensure comfortable reading. While some research results claim that an optimal line length is 52 – 68 characters per line (including punctuation marks and empty spaces), other studies show that even if the lines are getting longer it does not significantly affect usability. Since no rules of thumbs are provided, designers experiment with a variety of different line lengths.

To compute the max. number of characters per line we have used default setting of the browser as well as default typographic settings provided by the style sheets.

10% used 65-74 characters per line
(PostSecret, Beppegrillo, Perez Hilton, Scobleizer, Blogoscoped),
18% used 75-84 characters per line
(Dooce, Blogs.nytimes.com, Joystiq, CopyBlogger, TUAW, Slashfilm),
34% used 85-94 characters per line
(Lifehacker, Huffington Post, Kottke, ars Technica, Huffington Post, BoingBoing, Seth Godin, Treehugger, Problogger),
18% used 95-104 characters per line
(Mashable, ReadWriteWeb, Smashing Magazine, Google Blog, A List Apart, Search Engine Land),
16% used over 105 characters per line
(Engadget, TechCrunch, GigaOM, Wired, TMZ).

Based upon our findings we feel confident to suggest that the most usual (not necessarily most user-friendly) line length lies between 80 and 100 characters.

It is interesting to remark that not a single blog used justified text-alignment — 100% of the blogs used left text-alignment.

2.3. Primary typeface for body text?

Sans-serif fonts have become the de facto standard for body text on-screen, especially online. It has been suggested that this is because the small size of the font causes serif fonts to appear excessively cluttered on the screen. How is this suggestion reflected in the body text of popular blogs?

According to our survey,

34% ot top blogs use Verdana (sans-serif) for body text
(A List Apart, Kottke, TUAW, CopyBlogger, Dooce, ars technica, TechCrunch, Smashing Magazine),
24% used Lucida Grande (sans-serif, included with Mac OS X)
(Zenhabits, Mashable, Lifehacker, CrunchGear, Thinkprogress),
18% used Arial (sans-serif)
(ReadWriteWeb, Engadget, Google Blog, CNN Political Ticker),
14% used Georgia (serif)
(Scobleizer, GigaOM, Wired, BoingBoing, Huffington Post),
6% used Trebuchet MS (sans-serif)
(Andrew Sullivan, Seth Godin, Postsecret),
Helvetica Neue (ProBlogger) and Times New Roman (TPM) were used only once.

ars technica uses 85-94 characters per line and 12px Verdana for body text.

2.4. How large is the font size of body copy? (if em then convert and round)

The larger the font size is, the easier it is for readers to actually read the text presented on a given web-page. However, with a growing font size it becomes harder to read large blocks of text as one needs to scroll more frequently — and the eye needs to jump from one place to the other more often. So where does the optimum lie?

Here are the most common font sizes according to our research:

34% of the sites use 12px
(SearchEngineLand, TUAW, Mashable, ars technica, Engadget, Smashing, DoshDosh, TreeHugger),
30% use 13px
(Consumerist, CopyBlogger, Zenhabits, Valleywag, Lifehacker, Huffington Post, BoingBoing, Seth Godin, Google Blog),
14% use 14px
(TPM, GigaOM, Wired, ReadWriteWeb, Gigazine, ProBlogger),
12% use 11px
(A List Apart, Kottke, Neatorama, Dooce, TechCrunch, Dailykos),
4% use 15px
(Scobleizer),
the rest used 10px, 16px and 17px (each was used by exactly one site).

Zen Habits uses 13px Lucida Grande for its body text. Most blogs use 12px Verdana.

2.5. Primary typeface for headlines?

While sans-serifs are usually used for body text, it was often considered that designers tend to emphasize the headlines by using serif typefaces to present them. Does it hold true?

30% of the top blogs use Arial (sans-serif)
(CNN Political Ticker, Scobleizer, TPM, Crooksandlliars, Joystiq, Dooce, PerezHilton, ReadWriteWeb, Engadget, Google Blog, TreeHugger),
22% use Georgia (serif)
(A List Apart, Andrew Sullivan, Blogs.nytimes.com, GigaOM, Wired, Huffington Post, BoingBoing),
8% use Lucida Grande (sans-serif)
(Tuaw, ThinkProgress, Lifehacker, Crunchgear),
8% use Helvetica (sans-serif)
(Zenhabits, Mashable, ars technica, Smashing Magazine),
6% use Verdana (sans-serif)
(Blogoscoped, Neatorama, DailyKos),
6% use Trebuchet MS (sans-serif)
(Slashfilm, Postsecret, Seth Godin),
4% use Helvetica Neue (sans-serif)
(CopyBlogger, ProBlogger),

Besides, Calibri (SearchEngineLand), American Typewriter (Valleywag), Lucida Sans Unicode, Franklin Gothic Medium, Tahoma (TechCrunch) and no headline at all (Kottke) are used exactly once.

Darren Rowse uses Helvetica Neue for Problogger’s headlines. Most designers prefer Arial and Georgia (52%).

2.6. How large is the font size of headlines?

Finally, let’s take a look at font size of headlines. The larger the headline is, the more weight it is given to. However, if too many headlines compete for user’s attention the cognitive load increases and it is becoming harder to the users to actually consider the navigation options. In this context less sometimes can be more. The rule of thumb: if you have many posts presented on the start page, don’t forget to decrease the font-size of the header and highlight it visually, for instance with an eye-catching color.

24% use 20-22px
(BoingBoing, PerezHilton, Blogoscoped, Google Blog, TechCrunch, ReadWriteWeb),
22% use 23-25px
(CopyBlogger, ProBlogger, Lifehacker, Mashable),
22% use 17-19px
(Tuaw, Scobleizer, TreeHugger, A List Apart, Gizmodo),
16% use 14-16px
(YankoDesign, Dailykos, ars technica, Seth Godin),
6% use 26-29px
(Engadget, GigaOM, Wired, Dooce),
0% use 10-13px to highlight the headline.

Furthermore, 0% use 30-31px; Huffington Post uses 32px, Zenhabits 34px and Kottke has no headlines at all. Smashing Magazine is the definitive winner in the category “Biggest Headline Size”: our font size is 44px.

We conclude that it seems to work best to use headlines with the font size between 17 and 25px.

Bottom line

Let’s conclude the first part of the survey results with a brief overview of the main findings. Please keep in mind that the results of the survey should not be considered as guidelines for an effective blog design — this is a topic for another article.

large blogs require a multi-column layout solution (usually 3 columns suffice) (58%);
layouts are usually centered (94%),
layouts usually have a fixed width (px-based) (92%),
the width of the fixed layout varies between 951 and 1000px (56%),
58% of the overall site layout is used to display the main content,
CSS-layouts are used (90%),
the background is light, the body text is dark (98%),
the most usual (not necessarily most user-friendly) line length lies between 80 and 100 characters,
Verdana, Lucida Grande, Arial and Georgia are used for body text (90%),
the font size of body text varies between 12 and 14px (78%),
Arial and Georgia are used for headlines (52%),
headlines have the font size between 17 and 25px.

Please stay tuned and subscribe to our RSS-feed , we’ll present the second part of our findings next week.