Moving the Font Matrix forward

In December, as part of 24 ways, I proposed that we web designers increase our font stacks with help of a font matrix. The font matrix is a list of all Western fonts installed with Windows and Mac operating systems and common software suites.

The implication is that there’s a high likelihood your readers will have more fonts installed than just the core web fonts, so why not specify these other fonts as well. But, as m’learned friend Stuart Langridge wrote (via email):

it would be approximately 975 times more cool (and I know this because I’ve measured it) if the font matrix was available in some form where those of us who don’t have all the fonts can still see what they look like.

And he’s entirely right; it would be great to see what all the fonts look like. There are a few resources which do show samples of most of the fonts in one place, and of course Google will happily lead you to resources for one font at a time.

But I think Stuart hasn’t set his sights high enough here. While it would be useful to see detailed samples of all the common fonts in one place – if I had the time I would put this together for one and all to use – what would be even better would be suggestions of font-family font stacks. Or even better, automatically generated font stacks, based on pre-determined associations between fonts. How would these associations be determined? Through the power of the collective, of course.

What I imagine (for I’m afraid this is all fantasy) is a website where folk can input a font, and then associate fonts they think would be suitable font-family back-ups. Some software magic could then amalgamate this community input and generate suitable font stacks, based on rules of including a generic font family, a Windows OS font, a Mac OS font, Linux OS fonts, optionally an Office font, and optionally a CS font.

So if that sounds like a challenge you’d be up for, I say go for it, and I’ll help where I can. Update Mulling this over in the shower (as you do) it occurred to me that, conceptually, what I’m proposing is a weighted tagging system where a given font is tagged with other fonts in order of preference. With the addition of some tags (fonts) being required from pre-defined pools.