drshilliams

Some good stuff already this year! No sure what's up with the formatting, fine in the preview. Oh well.

Ought—Room Inside the World

Post-punk

Top tracks

Desire

Disgraced in America

Hookworms—Microshift

Psychedelic/noise

Top tracks

Ullswater

Negative Space

A.A.L (Against All Logic)—2012–1017

Deep house/electronic

Top tracks

Now U Got Me Hooked

Rave on U

GoGoPenguin—A Humdrum Star

Jazz

Top tracks

Raven

Bardo

Ty Segall—Freedom's Goblin

Garage rock

Top tracks

Fanny Dog

Every 1's a Winner

Plus some other special mentions

Car Seat Headrest—Twin Fantasy

Fire!—The Hands

Shame—Songs of Praise

Superchunk—What a Time to Be Alive

@xrott: Of course there are ways of dealing with it, I'm saying that it's a major deciding factor in switching from a font-service to either self-hosting or web-safe alternatives.

I got the 228kb from logging into Typekit and creating a dummy project with 3 weights of the old typeface - if you want to add a second font to that kit for body copy you could end up looking at ~400kb. For large scale, high-performant websites it's a no go if you want a reasonable page speed.

@xrott:This page is currently 1.9MB with 111 requests - it's a pretty massive footprint for what is ultimately text content. 1.9MB is absolutely crazy. Assuming there were 3 weights to the previous typeface being used (that seems about right) that was 228kb which is pretty hefty. That's without a second typeface for body copy.

Regardless of whether the Paul Irish article is seven years old or not it's still massively relevant. Not everyone has access to fast internet, especially on mobile. I've noticed Giant Bomb has suffered from FOUT pretty badly recently. It's not just about the memory footprint of the font - it's about the response time with the Typekit servers. I'm not saying using webfonts are a bad idea (we use them everyday in work) but it needs to be implemented properly, especially at scale.

Although anyway the reason Rorie mentioned they've switched is due to pricing. Highest bracket on Typekit is $400 a month (not unreasonable for Giant Bomb to have that amount of traffic). Saving $4800 just for a font change probably makes sense.

This is presumably the whole point. Looks like body copy will only be using installed system fonts to decrease page load. If they were previously on Typekit either:

• The fonts are loaded asynchronously which results in flash of unstyled text http://www.paulirish.com/2009/fighting-the-font-face-fout/
• The fonts are blocking page load while the client waits to download them (not a good idea)

The typestack is a bit ridiculous at the moment, no real need to go with that amount of fallbacks, it's pretty unlikely most people will have any of those typefaces on their system.

The replacement for Jubilat seems a bit thinner so sizes may need to be tweaked in a couple of places. Macs should be fine however, San Francisco was specifically designed for digital . Windows will have to tough it out with Arial or Helvetica I assume (which is awful at small sizes).

Awesome, thanks for the advice guys. I'll definitely hold off for a while.

@turtlebird95 Already running Windows on an SSD plus Steam from another so no problems there :)

Yeah that was the problem, it was hard to tell whether the rest of the machine was OK or not - had some FPS drops on the latest Tomb Raider game so wasn't sure.

If there's not problems I'll just leave it, thanks!

Edit: Although would it be worth replacing the RAM at least? Looks like the board supports up to DDR3 2800 so I'm considering getting 16GB in there.

Bought a GTX970 last year but I'm looking to upgrade the rest of my machine. Currently running

• INTEL I5 3570K
• GIGABYTE GA-Z77X-D3H
• CRUCIAL BALLISTIX ELITE 8GB

Not sure if this is a bottleneck or not, so I'm not really sure whether I'd see much of a performance upgrade or not. Already running with a 600W power supply which I hope is still beefy enough if not though. It's hard to keep up with all the recommendations these days so any advice is appreciated!

Thanks

@hatking: Haha, in that case they'd need an evil corporate mascot!

Fair point, seems like it does improve legibility a little!