Google Fonts: more than a trillion served | Hacker News Reader

Google Fonts: more than a trillion served

google.com

80 pointsrobinhouston81 comments

I'm surprised nobody's pointed this out yet --

Windows Chrome, the #1 Google Fonts browser, pretty much sucks at rendering Google Fonts. Which means you probably don't want to use Google Fonts yet.

An outstanding issue [1] for the past few years has been that it can't antialias fonts correctly. The results are sometimes just so-so, and sometimes horrible, depending on what version of Windows you're running and what settings you have enabled. Small fixes have dropped but the fundamental issue remains.

When I first added Google Fonts to my site, I got word from several readers saying text was illegible. Thats when I discovered that tons of other designers had the same issue and it was Chrome's fault (search 'chrome windows fonts' and enjoy). I went back to Helvetica/Arial and every day this issue persists a little part of the designer in me dies.

When a company espouses a newfound interest in design [2] [3], but lets a major design issue on the largest platform for their products just sit there (visibly) unaddressed, you get a sense they don't have the right people in the right places to get stuff done yet.

Don't get me wrong, I love Google Fonts, it's a fantastic service for the design community. And Google has been building a lot of great design tools [4] that help push the web forward. Just.. close the loop on fonts already.

[1] https://code.google.com/p/chromium/issues/detail?id=137692 [2] http://www.fastcodesign.com/3016268/google-the-redesign [3] http://www.theverge.com/2013/1/24/3904134/google-redesign-ho... [4] https://www.google.com/webdesigner/

> When a company espouses a newfound interest in design, but lets a major design issue on the largest platform for their products just sit there (visibly) unaddressed, you get a sense they don't have the right people in the right places to get stuff done yet.

Same for Google Maps. I live in Romania and we have had high-rez satellite and maps of our country, including street view, but for 7+ years we haven't had GPS locations of individual house numbers.

Android being the most popular phone and Google Maps the most popular GPS/maps app in Romania, not to be able to locate a house on a long street makes it almost useless. GMaps simply points to the middle point of the street - if it is miles long, then you might be a long way from your actual destination.

Why can't a multi-billion dollar company get their act together and finish their work? Even small GPS companies sell products that can look up houses on the street, but not Google. Could it be that it is too expensive even for them, or that Romania is worthless for them to spend time to finish the job?

I raised the issue on reddit, send feedback directly to them and at one time I even met a Google Maps employee and described our frustration with the Romanian Maps to him directly, but nothing gets done. It's like we fell off the Google Map™ into "forgotten land". And it's not a minor issue, it's like the most central thing a GPS does.

I went back to Helvetica/Arial and every day this issue persists a little part of the designer in me dies.

On a similar note, caution is needed when using Helvetica in a web font stack. On Windows PCs, this could easily select a low quality screen font that was installed with a printer driver in preference to Arial. Designers who work on Apple gear, where specifying Helvetica tends to find a standard system font that renders reasonably well, sometimes don't realize this.

Thanks for the tip, that'd be bad! Is this still the case if you have Arial as a fallback? e.g. font-family:"Helvetica Neue",Helvetica,Arial,sans-serif;

Yes, because "Helvetica" is listed before "Arial", and some people have a bogus "Helvetica" installed.

If you just want a nice, legible sans-serif font, just put "font-family: sans-serif" into your CSS, and you should universally get a pretty good choice.

This. It's worth pointing out that the antialiasing issue only affects Windows PC's. But this is very much a font issue us developers have to hack around. -webkit-font-smoothing and setting the value to antialiased sometimes fixes this issue.

The weirdest part is that, Firefox on Windows renders those fonts fine.

Firefox is fine, yeah. Although, sometimes you do get a weird faux-bolding issue in Firefox for some fonts which cannot be fixed except for not using light text on a dark background and some custom fonts.

The only reason I came onto this article was to post this. Thank-you for saving me some time.

If you ever get lost in the tons of fonts offered, check out this curated list that I absolutely adore:

http://hellohappy.org/beautiful-web-type/

https://github.com/ubuwaits/beautiful-web-type

There are more like these. If I can find find them in my bookmarks, I'll paste them here.

Please look harder, I want to see!

yeah this was a big one for me too

Trillion visits logged.

Just remember, kids, that it's the only reason Google Fonts exist.

Yep, like everything, it's a trade off. In return you get a very robust platform and free fonts.

In return for what? Google's not profiting off of Google Fonts, because I'm sure that whatever data they get from it could be collected by their other services.

Logged visits to sites they would otherwise not get. Font request+referrer is useful information.

If you're concerned about your own privacy in this respect, you can install a plug-in for most browsers that will prevent a real Referer being sent with your requests. Google will still serve you the web font, they just won't know which page you were viewing at the time.

(I'm not making any comment on Google's general strategy or how many people would actually understand that they might wish to protect their privacy in this way, just pointing out that it can be done.)

My computer or more accurately the operating system came with dozens of fonts, maybe hundreds. Why can't you use one of those?

Because most of them suck?

And more than that, each operating system has a different set of fonts.

Imagine how ridiculous the web would look today if content creators had to restrict the images they used to some common subset of those preinstalled on clients' operating systems.

Would you condemn web typography to the use of "web safe fonts" forever?

Yes. I do indeed condemn creators to use a font which renders correctly and I also condemn them to use a size which is legible.

Isn't it the job of the user agent to have options to override the choices of a website? Otherwise, the subset of "fonts everybody is sure to have" is just too small.

Why not let everybody have websites look how they want them to look? I'm kinda sad this vision gets so little attention, that instead we all try to push our vision on our visitors. Even if the web turns app platform.. just like we have configurable desktop managers, browser could still let everybody customize how websites look - if only websites could stop with the "story telling" and "experiences", and concentrated on utility and information. We're too addicted too shiny for our own good already.

A trade off implies people understand what Google is doing, which they don't because it's secret.

Is it against the TOS to download the font and host it on your own server? It's very simple to do even though they don't give instructions.

No:

https://developers.google.com/fonts/faq#Download_Fonts

I am in the middle of a redesign and the new design uses Droid Serif. I wanted to avoid an extra HTTP request and bundle the WOFF font file in the CSS (using data URIs). The other option was to host the font on my server to avoid the DNS lookup for yet another domain. (CSS is in the critical path for page loads).

Both are easy to do but I abandoned both of those ideas after watching this interview http://www.youtube.com/watch?v=sqesm0euf9M

Google Fonts makes a lot of optimizations that are platform-specific. e.g. serve smaller files to Macs, larger files to Windows. It is not worth doing all this yourself.

Don't put anything more than 10KB into a CSS file as a data URI.

Where does this limit come from?

IE8 can handle up to 32KB[1], and more recent versions of all major browsers appear to have much higher limits. IE7 and below didn't support data: anyway.

[1] See http://caniuse.com/datauri, and confirmed by Microsoft's own technical documentation.

Oh I didn't mean that to be a browser limitation, but that it is generally a good idea performance wise.

Why would that be the case?

In terms of file size, a naive implementation of replacing reference to image files with data: URIs carries a theoretical overhead of 1/3. However in practice, as long as you're serving the relevant CSS file gzipped, you're likely to see less than 5% overhead on a single file. If you've got several similar image files converted to data: URIs within the same CSS file, you might even see a significant gain because the compression can remove more redundancy.

In terms of HTTP requests, generally fewer is better, so it's hard to go wrong here by using data: URIs instead of separate but relatively small image files.

The only major factor I can immediately think of that is clearly in favour of separate image files, unless you're really at the level where the decoding performance for data: URIs makes a significant difference, is that the image files can be cached independently, which may or may not be a win depending on how you manage your CSS and how often the various parts of it change relative to each other.

Don't forget that browsers will wait for the CSS file to download before rendering the page.

Far better to let "unnecessary" image files load later via separate HTTP requests than to delay everything just so that you can save one or two HTTP round-trips.

This is a big issue on mobile.

I'd advise comparing both methods on https://developers.google.com/speed/pagespeed/insights/

YMMV, but we did plenty of testing for one site I currently work on with a heavy mobile access focus, and it wasn't even close. We could have added hundreds of extra KB in a CSS file and still beaten even a single additional image file that needed a whole extra HTTP round trip for a visitor with a typical intermittent mobile connection (poor reception, on the move, etc.). The multiple levels of set-up required in that context can easily dominate the overall time to seeing a useful page.

Of course it's less of a dramatic comparison if you don't have all that set-up overhead to worry about, but then the time to download an extra 200K of CSS is negligible on basically any broadband or stable 3G or better connection today, so we don't tend to worry much about that sort of situation. In practice, our decision on how to transmit medium-sized images is often based on cache-related factors.

My least favorite thing about Google Fonts is that they look different on different browsers for each platform too. On my latest site design I had to put them on my server because they actually looked the worst in Chrome.

So you are claiming that reason of very existence of Google Fonts is to collect not-so-useful-and-easily-could-be-collected-by-other-services-of-Google data? Can we stop the whole Google bashing it is getting ridiculous.

They are an advertising company, which pretty much makes them data collection company. Its completely fair to question their motivations.

I personally think that Google does things like this because they can; they are not under money pressure (yet?) to have to really try to extract maximum value from every bit served or watt consumed, and can thus spend energy on initiatives like this that improve the web as a platform -- it's at best an indirect benefit, but fits their overall corporate strategy. And, it's cool.

I have my doubts about Google, but I don't have a problem giving them the benefit of the doubt here.

Make the web better for all forms of content so that people don't resort to proprietary tech?

sure, advertising. but the goal here is to make the web more user friendly and useful for people, through better looking design/readability etc. That helps us all, and if then in turn they can keep the ad platforms alive, good on them I say.

The motivations, as far as i know, are written on the about page:

"A web with web fonts is more beautiful, readable, accessible and open. " Google Fonts makes it quick and easy for everyone to use web fonts, including professional designers and developers. We believe that everyone should be able to bring quality typography to their web pages and applications."

If you have some actual evidence that the last part isn't true, and that it's really some big scheme to "log visits", what is it, exactly?

FWIW - my dealings with the team, which mostly involve reviewing occasional open source font contracts, makes me believe they really are just crazy about making web typography good. But i'm really interested in whatever sinister theories people have.

I mean, you know what they say about people who like "Signika Negative".

Your naivety is adoring.

Your hat is shiny.

If you want to skip the personal attacks and explain

1. What "trillions of visits logged" means, exactly. IE what you think is being logged past a request for a font.

2. What you think the data is actually used for

The web font loader is open source, so this should not be too hard to explain what it's sending that you think is valuable and why.

Or you know, you could just continue on with meaningless ad-hominen attacks, since well, you never responded to anything i asked.

You know you're arguing with Dale Gribble, right?

Yeah, well, i try to treat everyone the same, even if they seem, uh, off :)

Except the part where they allow you to download them and serve yourself (they're open source):

https://developers.google.com/fonts/faq#Download_Fonts

> You can download the fonts to use them for your mockups, in your documents or to host them on your own server.

Google Fonts exists because it's in Google's best interest if everything is online. Having freely available high-quality fonts is a [small] way to help move that along.

Update: No cookies are set by Google either, they're just doing analytics by the user agent sent. The actual font files have 1 year expires headers on them as well, so cookies would be of limited use.

Didn't realize this - thanks. :)

Also, without accessible web fonts, developers will have to use images containing rendered texts to get text pretty on their web pages, which is not so easy to index for a search engine.

Except for the fact that people don't download and serve them themselves.

> Except the part where they allow you to download them and serve yourself.

Yeah, you try that. It's a single format that you need to convert yourself to produce woffs, eots and other variations needed for the actual self-hosted deployment, and no matter how you convert, they will never come out looking the same as Google's native versions. This "go ahead, host it yourself" is disingenious and manipulative at best.

Here you go: http://www.fontsquirrel.com/tools/webfont-generator. Upload a font and get back a kit with everything you need.

Yeah, try it with Droid Sans, compare the results, report back.

@font-face generators are dime a dozen. The issue is that they generate kits of subpar quality for Google fonts.

Most of the time Font Squirrel has them made already, and you don't need to convert the already stripped down fonts from Google's severs...

Font Squirrel conversion botches up a lot of fonts, Droid Sans being the prime example.

Even if you download it elsewhere first (where elsewhere equals not Google)?

... All the fonts are open source, you can quite literally use Google's native versions. It's not exactly rocket science. You want Droid Sans?

http://themes.googleusercontent.com/static/fonts/droidsans/v...

... And you got these links from? What is the license for these .eot and .woff? Who says you can self-host them?

Rocket science or not, the bottom line is that these are not available from Google Fonts. You are expected to either use hosted version or muddle through the conversion process.

@eps

http://code.google.com/p/googlefontdirectory/

Only reason huh? nothing to do with prettifying the web and making developers happy, because that doesn't benefit Google in any way...

Not directly it doesn't.

If they cared about that they'd keep Reader open.

That's a a ridiculous stretch.

I don't see the connection there. Anyway it would appear that the universe has survived that ordeal.

Google Fonts, Google Public DNS, Google Analytics, Google+ like gadgets, Google Adsense, Google CDN for jQuery etc, Google services like Search and Gmail, Embedded Youtube videos, Google Chrome word correction sugesstions and safe sites checks, Firefox safe page checks, probably many things within Android backend.

Probably they know with 99.9999999% accuracy how many online web users are out there.

Considering that browsers have caches this isn't a huuuge issue.

This shows MSIE at ~18% split between IE10 and IE8 with a bit of an edge to the former. These are the most optimistic stats I've ever seen.

Still, a trillion pageviews is nothing to sneeze at. I guess for normal none-corporate browsing OldIE has really truly actually fallen below 10% and can finally be ignored?

I think these stats are more accurate than many others (I am looking at you Net Applications). However, Most of the popular large sites (Yahoo, Facebook, Amazon, e-bay and Chinese stuff etc.) do not use Google Fonts so IE stats are skewed.

For all intents and purposes those are outliers that can be ignored anyway unless you happen to be a Top25 website yourself.

Where do you see the IE version split? I only see the browser type analytics, not split by version.

There are two pie charts under "Views by platform": "Browsers" and "Browser Versions". Under versions, both "MSIE 10.0" and "MSIE 8.0" are listed.

Though assuming the two pie charts are based on the same data, I'm not sure how specific versions of Chrome and Mobile Safari have greater shares of the "Browser Versions" pie than all of their versions have in the combined "Browsers" chart.

A lot of this has to do with Twenty Twelve being the default theme for WordPress, which used Open Sans as its default font. Many people I've seen either use that theme or some derivative of it (myself included).

Interesting. It is a rare thing to see that Google shares information about the Browser and OS stats.

And that they broke out iOS (even iPhone/iPad separately), but not Android (I suspect it's in the oddly high Linux category).

They didn't break out android because it would contradict the usage numbers they would like people to believe.

Yes, i'm sure that's the reason ....</s>

It would actually be impressive if google was such a single cohesive whole

If Google is not a single cohesive whole, how can you speak for it?

Certain people have certain authority to talk about certain things.

I set policy around open source and open source licensing. I can speak officially to open source issues for Google (though at least here, i'm not unless I specifically say I am)

This does not mean it's a cohesive whole however.

In particular, a small team Of 3-4 people in the middle of a random non-android product area is not going to have the android team breathing down their neck asking them to "not break down numbers" or whatever the parent thinks happened here.

Since executives such as Eric Schmidt are happy to quote numbers about Android or Google+ in opaque and misleading ways, it seems entirely reasonable to believe that some other teams within Google may act to preserve the general message out of esprit-de-corps, whether or not they have android team members breathing down their necks.

Serif fonts are going out of business?

Only one in the top 10 is Droid Serif, still dropping.

The page we're on, the page linked, my browser and OS are all 100% sans-serif.

I wouldn't say going out of business, I don't think they will ever disappear, but definitely taking a back-seat role.

Reply on news.ycombinator.com