Solving Problems like it's 1996 - Core Fonts for the Web

It's FONTS WEEK here at The Fortress! (Which is like Shark Week, just with fewer cages.)

The main theme of fonts week is that fonts are really complicated from a technology point of view. There's a whole rabbit-hole you can dive down to get fonts looking good. Not only do you need the letter glyphs themselves, you need a way to scale the characters up and down to various sizes, they need to look good on a screen and / or in print, and you need a way to handle all kinds of other micro-adjustments between characters.

For a long time, then, in the world of computers fonts were the domain of the printer itself, and your DOS-based computer (or early Apple, or UNIX machine, or what have you) would just tell the printer what font to use, and the printer would make it work - fueled by Adobe's excellent (but expensive) PostScript technology.

This all started to change in the early nineties when Apple, tired of paying out the nose for PostScript, came up with TrueType and licensed it to Microsoft - where it quickly became the naive way to handle fonts for both Windows and the Macintosh (Starting with versions 3.1 and System 7.)

Hand-waving away a lot of the technical details, this meant that fonts could now live in a font file on a computer, and the (then) new graphical displays could use the same files as the printers to show exactly what your document was going to look like when it was printed. At this point we got an explosion of "What You See is What You Get" word processors, with Word more or less leading the pack.

Suddenly the whole world goes font crazy, and we start knocking out fliers and newsletters that look like an explosion in a type foundry.

This all works great until the World Wide Web starts to go mainstream in 1995 or so. Because here's the catch - a computer can display pretty much any text in any font - as long as that font is installed on the computer in question. So, while the person writing a web page could specify any font they wanted for the text of their page, a person reading the page wouldn't see it like that unless they had the same font installed - and if they didn't, their browser would guess what font was the closest. And yes, that always looked terrible. (Sidebar: yes, this was one of the many reasons why all those geocities pages looked like they did.)

On top of all that, there was a legal problem - fonts are only licensed for use on one machine, technically, so if you used Helvetica, say, for your website, you couldn't just put your copy on the page to be downloaded. This was compounded by the fact that no one ever came up with a decent way to embed fonts in a website (thanks, largely, to the legal issues.)

So, it's 1996. The font thing on the web is clearly going to be a problem. Enter Microsoft, and their Core Fonts for the Web. The idea was simple - they took a batch of fonts they owned, and promised that those fonts would always be installed and available on all of their products, and they were willing to license these fonts for a very reasonable sum. Apple jumped at it, and the Core Fonts became the canonical fonts everyone used on the web.

It's a shame then, that they picked such sucky fonts.

The fonts in the set are:

• Andale Mono
• Arial
• Comic Sans MS
• Courier New
• Georgia
• Impact
• Times New Roman
• Trebuchet MS
• Verdana
• Webdings (Webdings)

(In case the machine you're on wont display the text right, text examples on the other side of the links.)

Verdana and Georgia were among Microsoft's first few attempts at making a font that looks good on the screen as opposed to the page, so those are good calls. Times New Roman had already become the standard word-processor font, so that's a good call too. Courier? Sure, why not. But the rest? In my mind, this is when Arial finally took over from Helvetica for good. Webdings? Why? No really, why would anyone actually use webdings? Impact and Andale? Fairly pointless as near as I can see. These are the two I always forget are in the set - and I don't think anyone used either until Impact became the font of choice for LOLCats. Trebuchet - dull and ugly. And my goodness - why Comic Sans? Of all the "fun" fonts to make standard across all computers on the planet, why did it have to be COMIC SANS?

So, we end up with the classic Microsoft solution to a problem. Don't get me wrong - I think the idea was fantastic, and it worked great - every machine on the planet, more or less, has these fonts installed, and they are what the web looks like now, for better or worse.

But what a pedestrian set of choices. Why two monospace fonts? Either Andale or Courier is a fine choice, but both? Arial, Trebuchet and Verdana have the same problem - having more than one basic sans serif font is probably a good idea, but those three are so similar as to make the choice pointless. Verdana, for example, was designed explicitly as a replacement for Arial for use on the screen, as was Georgia for Times New Roman. If we're going to install 10 fonts on everybody's computer, why not 10 really different fonts, and remove all the duplicates? Finally: did anyone, ever, at any time, use webdings for something that isn't utterly irritating? (Adding insult to injury on that front, while writing this post I discovered that Firefox straight-up won't use webdings, regardless as to whether you have it installed or not. I see I'm not the only one with a low opinion of the font.)

Moving forward to 2002, Microsoft finally pulled the plug on the idea, citing frequent copyright violations as their excuse. Apple still licenses them for OSX, however, and most of them are still included with Windows and Office, so their ubiquity looks likely to remain for a while.

Still. What a missed opportunity. Although it presents us with a fun mind game: if you could choose, which 10 fonts would you install on everybody's computer?