In 1998, I was a junior Program Manager on the Internet Explorer team at Microsoft. My first project was shipping Internet Explorer 4 Plus, a CD-ROM with Internet Explorer 4.0 and a whole lot of trialware junk that we sold for $49. Yep, gather around, kids – let me tell you about a time when browsers were sold in physical boxes at CompUSA. And a pack of gum cost a ha’penny, and Coke still had cocaine in it!
Anyway, I digress.
We had shipped Internet Explorer 4, and we were busy with Internet Explorer 5, which was chock full of awesomeness, like offline browsing using Channel Definition Format (remember when Pointcast was worth $450 million?), an all new Trident engine, and “Weblications,” a way-too-early-attempt to let devs build rich apps in the browser – though I guess that did lead to the invention of AJAX.
In those days, I was always at work until 10 PM, mainly because dinner was free and I had no life anyway.
And one of those nights, one of the senior devs Bharat Shyam asked me to come over to his office to look at something. I’m sure I was in the middle of writing some amazing yet useless spec/patent. So I walked over to his office and he was hunched over his 133 Mhz Pentium PC, one window full of Win32 C++ COM jibberish, and the other with his local build of IE5. He said, “check this out,” and in IE he added a favorite (aka bookmark, for you Chrome-heads). Amazingly enough, the favorite had a fancy icon to the left of it! Never before had anyone developed the “technology” to do such a thing. Marc Andressen hadn’t though of it, and that guy was rich, right?
Bharat said, “this is good, right? Check it in?” And I said, “yeah, sure, how does it work?” And he said that you just had to put a file named favicon.ico at the root of your IIS Server, and you were good to go! So I said, “sure that sounds good,” and I went back to my office.
The next day, my manager called me into his office. He said, “did you OK this checkin?” And I said, “yeah, sure.” And he scolded me because he said Bharat had taken advantage of me – he was looking for a junior PM to OK this late checkin, and I should have said no. I promised to do better.
But now I look back & realize that we did the right thing. Seriously, how risky was this feature?
And I still remember telling my friend Michael Radwin at Yahoo about favicon.ico. He was looking at Yapache logs for fun as he does, and he had noticed an unusual spike in HTTP requests for http://www.yahoo.com/favicon.ico. He said, what the hell is favicon.ico? And I explained it to him. He was so excited that he slammed a favicon.ico onto the server, which might have been one of the first official favicons in existence.
So there you go, the story of the invention of favicon.ico. For the kiddies.
Ray
Reblogged this on Dain Binder.
Another perfect case of enterprise bureaucracy (almost) getting in the way of innovation 🙂 Well done!
Pingback: Coder Read of The Day #20130903 | Coders Grid
The extra HTTP request has always seemed wasteful to me. The favicon should have been referenced by a “link” tag from the beginning.
Thank you for OKaying it back then.
Pingback: Favicon–What / Why / and How | Jimiz.net - Jim Becher on the web
You want what? Sure, go away so I can eat my free dinner!
“Seriously, how risky was this feature?”
Quite risky, and you were too young (and now, apparently too fond of it) to see it. Aside from the privacy issue (fixed by spamming the server with /favicon.ico requests), the fixed root-level URL was in no way GeoCities friendly. It should have been a “link” tag or something, not magic.
(And no. I am not happy with .well-known/ either…)
I think they also introduced in IE5.
I think they also introduced <link rel=”shortcut icon”> in IE5
Reminds me of the anecdote about how the blink tag came into existence.
woah… i wish favicon resolution was higher though
Best thing IE ever did.
Never thought that such small things can make such great stories. Who would have thought of this. In todays world, if something like this happens it would be a great news all over the world.
Someone on reddit found this about a security flaw in this feature:
http://web.cip.com.br/flaviovs/sec/favicon/index.html
http://technet.microsoft.com/en-us/security/bulletin/MS99-018
Pingback: Let's Get Rid of Default Favicons! And at the Same Time Tackle the Touch-Icon-Issue. » Metamonks
Thanks dude for sharing your true story, without faviocon bookmakrs were boring, with favicon you can quickly recognize which URK belong to which site.
With Firefox and Chrome growing and getting better and better, IE needs to be more innovative to play catch up game
Thanks
Javin
Pingback: Improve server response time | CL-UAT
Pingback: Improve server response time | XL-UAT
Pingback: 1p – Inventing Favicon.ico | Profit Goals
Pingback: 1 – Inventing Favicon.ico | blog.offeryour.com
“…all new Trident engine…”
What was the engine name before Trident? Or it was just the old Trident but new?
Reblogged this on Zach Melo.
You suck for doing that.
You, sir, have contributed to the welfare of humanity. Thank you.
Thank you for sharing this campfire story from the old days. 🙂
It was a good idea, but poor implementation. It should have been a meta tag, without dumb requests to see if the file is actually there.
Pingback: Inventing Favicon.ico
Pingback: Les liens de la semaine – Édition #126 | French Coding
Pingback: Founders Weekly #43 – This Week’s Best Startup & Growth Hacking Guides
Pingback: Изобретая favicon.ico | FNIT.RU
Epic story! If it wasn’t for you, Favicon Generator would not exist. So a huge thanks for leading me to make the app!
Pingback: inventing the favicon | Bas Wijdenes