Categories
Mozilla Web Development

Secrets In Websites

When a you browses the web, a fair amount of code is transfered from a server sitting in a cold lonely datacenter to your computer. It’s mostly serious business, but sometimes developers like to embed small jokes, and traditions.

A lot can be told by just a quick glance at a website’s design and it’s underlying data. From Infrastructure, design methodologies, CMS (content management system), among other things.

Here are some of amusing, interesting, and [in some cases] strange things I’ve seen on websites. Everything in this post is valid as of the time of the post. Keep in mind websites constantly change so anything seen here may be invalid as time goes on.

Yahoo! News

All pages served off of news.yahoo.com feature a comment towards the bottom that says:

<!– This page is brought to you by the letter W, the number 1 and fe57.news.sp1.yahoo.com –>

Obvious homage to Sesame Street’s “sponsor”.

Digg

Digg as served up a very elusive riddle in the comment:

<!– digg is up 959595–>

Some explanation can be found here. The best explanation is 9-5, 9-5, 9-5 or 3 days of work. Likely related to the $200 that it cost to develop Digg.

Apple

Apple to this day still serves up the copyright on its website as follows:

<p class="sosumi">Copyright &#169; 2007 Apple Inc. All rights reserved.</p>

The joke here is the class name for the paragraph (which is the copyright). Sosumi is known by Mac users as the System 7 sound. It’s name is actually a joke because of the Apple Corp. vs. Apple Computer case. It’s pronounced “So Sue Me”. A nice little tribute to Apple history, and poking fun at lawyers.

Amazon

Amazon has a peculular last few bits on it’s homepage:

<!– MEOW –>

No idea what to attribute that to other than someone who likes cats.

Slashdot

Tech site Slashdot has used http headers as a way of telling a nerdy joke:

X-Leela: He opened up relations with China. He doesn't want to hear about your ding-dong.

Leela, being a reference to Leela Turanga of Futurama. It’s random. Here’s another:

X-Bender: Try this, kids at home!

Bender was the robot on Futurama.

FAA

The FAA is running on:

X-Powered-By: Microsoft-IIS/6.0

Think they should run something that isn’t known for crashing quite as often? Same goes for the NTSB.

FBI/NSA

The FBI has a curious chunk of code:

<!– START OF Data Collection Server TAG –>
<script LANGUAGE="JavaScript">
</script>
<noscript>
<img BORDER="0" NAME="DCSIMG" WIDTH="1" HEIGHT="1" SRC="//209.235.0.153/njs.gif?dcsuri=/nojavascript&WT.js=No"/>
</noscript>
<!– END OF Data Collection Server TAG –>

“Data collection” eh? The NSA suprisingly has no such data collection. But perhaps that’s because they have no need? I would of thought by now information sharing between government agencies would have improved.

This is just a html bug just like virtually every other site on the web, just funny because of who it is…. so take off your tin foil hat

IBM

IBM invented a new version of HTML for itself. It’s been somewhat replaced at this point, but is still visible on some pages including IBM Research as of this post in early 2007.

The doctype reads:

< !DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd">

The differences are explained as:

  IBM XHTML 1.0 Transitional DTD
   This modified DTD module is based on the W3C XHTML 1.0 Transitional DTD
   and is identified by the SYSTEM identifier:
   "http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd"

   The following updates have been included to support 4.x browsers:
	- <noscript> tag is allowed to be placed between the <head></head> tags
	- Border, width and height attributes are allowed within the <input /> tag
	- <form> and <input /> tags are allowed to be placed between the <table>,<tr> or <td> tags
	- Marginheight, marginwidth, topmargin and leftmargin attributes are allowed within the <body> tag

Think this is completely meaningless and nobody even looks at doctypes? Well checkout what the Gecko (the rendering engine for Firefox) does:

// Special hack for IBM’s custom DOCTYPE.
if (!(resultFlags & PARSE_DTD_HAVE_INTERNAL_SUBSET) &&
    sysIDUCS2 == NS_LITERAL_STRING(
           "http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd")) {
  aParseMode = eDTDMode_quirks;
  aDocType = eHTML_Quirks;
}

Democratic Party / Republican National Committee

The Democratic party website uses XHTML 1.0 Transitional, CSS based layout, and runs Apache 2.x/PHP 5

The Republican National Committee website uses HTML 4.01 Transitional (though some xhtml-like self-closing <img/> and <input/> appear), table based layout, and runs Microsoft IIS 6.0/ASP.NET 1.1.4322. Source on the homepage indicates:

<meta content="Microsoft Visual Studio .NET 7.1" name="GENERATOR" />
<meta content="C#" name="CODE_LANGUAGE" />

Note: Self closing added by myself due to technical limitations here. They appear as “>” rather than “/>”

Perhaps all the above alludes to the idea of Democrats being more progressive, such as using open source php and Apache, xhtml (since xhtml was supposed to be easier to parse due to it’s relationship to xml… though that’s debatable), and the newer css based layout technique.

The GOP on the other hand stuck with more traditional html4 and table based layouts. Not sure what to make of the Microsoft use, but I guess some could link it to the United States decision to no longer seek breaking up the company during a Republican administration.

Ironically despite the contrasts here, the Democrats website is still showing 2006 as the copyright year, while Republicans updated to 2007.

51 replies on “Secrets In Websites”

yea yahoo was the best, slashdot one was known and they publicized it. Also check out robots.txt files ..there are some very interesting robots.txt files, including one site that has a robots.txt blog and others that serve out some hilarious stuff.

lol, you can’t quote the source code of someone else’s website because you’re a brainwashed XHTML-tard.

helpful hint: HTML 4.01 will look just as spiffy on your website, and won’t break the page when you miss a /

Apple to this day still serves up the copyright on it’s website as follows:

NO, IT’S ‘ITS WEBSITE’

LITERALLY YOU HAVE WRITTEN ‘.. ON IT IS WEBSITE..’

Alleagra: Fixed. Thanks for pointing that out.

Shii: When you work with a dozen or so CMS’s and learn they all have limitations you’d understand you have to deal with things sometimes, unless you want to retool the site for 2 lines.

My alma mater’s internal website used to be liberally sprinkled with <FNORD> </FNORD> tags, which I accidentally discovered once while teaching a HTML course (a reference, of course, to the classic Illuminatus! Trilogy). Now that they’ve redesigned it, those tags are gone, but I was please to see that there’s now a <h1 class=”fnord”> tag in their place.

Jonathan: Yes, I believe it is WebTrends, though I think a slightly older version of the code (not positive on that though). The way to tell is you’ll typically see the following in the <noscript> img url:

njs.gif?dcsuri=/nojavascript&WT.js=No

WT.js is the giveaway.

Secretos en los sitios web…

Cuando navegas por una web, gran cantidad de código es transferido entre tu equipo y el servidor que visitas. Se trata de algo muy serio, pero a veces los desarrolladores les gusta añadir pequeños chistes y tradiciones.
Descúbrelos.

just like Disney….hide an imbedded image or words or code.
ha ha. third grade hijinks. advanced I.T. personnel virgins.

Funny finds. xpoint.ru has a similar string in its web pages:

Which I won’t even try to translate, anybody not speaking Russian is unlikely to understand this even translated. The phrase is the answer to some internal fighting about the project’s direction several years ago.

Argh… You blog really added it as a comment. So the string is:

<!–Данный проект предпринимает попытку Ñ?озданиÑ? единÑ?твенно правильного и Ñ?ветлого будущего, в рамках одного, отдельно взÑ?того, узла вÑ?емирной Ñ?ети Интернет–>

Leela Turanga should be Turanga Leela, although she goes by the given name Leela, the family name Turanga comes first.

You make one terrible mistake on your website where you write:

“Bender was the robot on Futurama.”

This is absolutely insulting. Bender is the robot on Futurama. How dare you refer to him in the past tense? Bender lives!

That was an amusing collection. I didn’t know IBM had its own doctype.

What you didn’t comment on was that both political parties were using transitional doctypes. It seems to me that that is of more relevance than X/HTML choices. There seems little point in using XHTML unless you are serving it with the MIME type application/xhtml xml. (The Democratic Party is not using content-negotiation and is actually serving the page linked above as text/html to a capable UA (Firefox). So why use XHTML at all?) Actually, XHTML has no major advantage unless you want to imbed some MathML.

But there is an advantage in moving to a Strict doctype, because that enforces the separation of content and style.

I thought both parties had acceptable but hardly good websites. The Democrat website is the better of the two. It *is* actually using some tables for non-tabular content, but it looks like the Republican site, besides the tables, also has some spacer gifs. Moreover, what is worse, those spacer gifs don’t carry null alt attributes (alt=””). See, for example, error 12 below:

http://validator.w3.org/check?.....et=(detect automatically)&doctype=Inline&ss=1&outline=1&verbose=1

AFAIK, that would mean that a screen reader, having no alt text and no indication to ignore the image would fall back on reading out the filename of the image – “RNC_indx_rnc.gif”. That could get wearing after awhile for blind visitors.

I think the GOP needs to clean that site up a bit.

[…] Easter Eggs from Prominent Websites January 22nd, 2007 Every developer loves a good easter egg every now and then, and the only thingmore fun than hiding them is finding them. In “Secrets In Websites“, Robert Accenttura tracked down easter eggs hidden in the headers html source of 10 prominent websites. The eggs range from Yahoo! News’ homage to Sesame Street, to Slashdot embedding Futurama quotes in custom HTTP headers, to an unintentional easter egg produced by a server bug on the NSA’s website that produced a few bits about “top secret data collection”. […]

that Sesame Street homage is not so obvious to me at all

At least some (it’s been a long time) episodes include somewhere in them a character saying, “Today’s program was brought to you by the letter M.” (or some other letter, and possibly with slightly tweaked wording). The reference isn’t at all obscure if you’ve seen Sesame Street a few times.

” Delicious Delicacies for bots… humans just ignore ;-)” Hahahaha nice secret! Robert Accettura!

The IBM doctype check in the code cracked me up :).

Shii, you make an XML-based pretty printer first, then try to parse HTML to pretty print it. I can tell you which is less effort to make.

Nick: get over it and don’t write so much :).

~Grauw

The reason there is custom support for the IBM doctype in Gecko was to work around a problem where the presence of any DTD would put the browser in standards mode.

We needed the IBM DTD to cause quirks mode otherwise the images laid out incorrectly on the IBM homepage.

Ah history.

Shouldn’t web servers/optimized systems not transmit comments, end of line and so on in the name of bandwidth efficiency?

[…] no menos interesante serie de Accettura, Secrets In Websites, dividida hasta ahora en dos partes (I y II). En ella nos descubre aspectos poco conocidos del código que hay tras muchas de las […]

I implemented documentum solutions for several years.
As an example documentum (now EMC) provided scripts to help developers/integrators out with some complex stuff.

In one of the scripts I have found this comment (don’t remember the exact words).

“I don’t know what this line of code is about or what it does, but when you delete it the documentum server will crash completely.”

I love the futurama references. Am going to have to think up some easter eggs to have in websites from now on. Will add some fun to the process of developing websites.

I found the Amazon one by accident about a year back – I’ve just rechecked it, decided to google it, and came up with this page. It’s good to know I’m not the only one who reads source code when bored.

I’m omitting my name from this comment so it won’t be public, but you now (privately) have my email address, and you can google that to find out who I am if you really want to. (But please don’t make it public.)

I was idly googling around and just found this page. I can tell you exactly and authoritatively the story behind the Amazon “MEOW”, because I’m the one who put it there.

To be honest, it’s not all that interesting. I worked as a developer at Amazon some years back, and did a lot of work on their front-end framework (known internally as “Gurupa”). Shortly before I left the company, I inserted the MEOW comment at the end of one of the Gurupa page templates, just to see how long it would last. Well, it’s been more than five years (I left in 2005) and it’s still there. 🙂 It shows up not just on the home page, but also on most browse merchandising pages, such as those linked from the “shop all departments” menu in the upper left. We had just finished migrating the home page to Gurupa towards the end of my time at Amazon, and over the years, I’ve found the comment a convenient way to see that the home page is still running some of my code.

Unfortunately, there really isn’t any clever meaning to it. You guessed correctly in your blog post: I just like cats.

I found a few other references to the comment, including a discussion on StackOverflow where people were speculating on technical reasons for it: a check to see whether the entire html page has rendered, a marker related to reporting or search-and-replace, and so forth. Needless to say, none of that is correct, and I find it all very amusing.

Leave a Reply to Tim O'Brien Cancel reply

Your email address will not be published. Required fields are marked *