The Great Lie - Lorem Ipsum | Hacker News Reader

The Great Lie - Lorem Ipsum

ignoredbydinosaurs.com

83 pointsJgrubb40 comments

On a side note, what's the story with all these sites using blurry fonts lately? This article is completely illegible in the latest build of Chrome on Windows XP.

I see this on maybe 5% of the submissions here these days. Normally, it's a sans-serif font that's the culprit, and this is my first serif-blurry sighting. Here's the font-face:

    garamond-premier-pro-1,garamond-premier-pro-2,serif;

I've got photoshop installed, so chances are I actually have one of those local. But Chrome can't render it, so if you use it on your website it will look bad for everybody except you.

Now, since I'm in developer tools, I can change that to

    garamond,garamond-premier-pro-1,garamond-premier-pro-2,serif;

... and it will render in a readable form. So the question, I guess, is why would you specify a non-standard font without testing to ensure that it falls back to something that's at least readable for everybody else? Especially if you're a screen designer by trade, writing articles to show off your knowledge of screen design issues.

It's a [TypeKit](http://typekit.com/) font. Looks great on everything on a Mac, but thanks for the heads up.

edit: and I'm definitely not a screen designer, nor am I intending showing off knowledge of any sort of expertise. This was more of a "note to self". Cheers.

Just a heads up: HN doesn't use Markdown, so you can't link URLs like that. You can just type a URL and it'll become a link (shown truncated if its too long). Another popular convention is to use brackets as footnotes[1] and put the URL at the bottom of the comment.

[1] Like this.

I've seen the footnote thing plenty. I guess I thought they were auto-generated or something?? Leaving the error for all the world to see. Anyway, thanks...

On Windows, the only browser that consistently renders Typekit fonts legibly is Safari for Windows, due to its use of Apple's subpixel smoothing. I prefer the Apple style font rendering, which is one of the few compelling reasons to use Safari on Windows.

Windows and MacOS use different techniques when approximating a font to the pixel grid of the screen.

The Mac obeys the font, applying smoothing via anti-aliasing along the way, with no extra consideration given to the pixel grid.

Windows tries to snap things to the pixel grid. For most "medium" sized text this produces a sharper lighter look. For small text or (in your case) typefaces with light strokes to start with, if tends to make things look too light and a bit blurry instead.

These two pages give (respectively) an example of where the difference is clear to see and a description of the different processes: http://www.codinghorror.com/blog/2007/06/whats-wrong-with-ap... http://www.codinghorror.com/blog/2007/06/font-rendering-resp...

At higher resolutions Apple's way would definitely be better (you get a result closer to what the font designer indented). For low resolutions (like the 96dots-per-logial-inch Windows usually assumes on screen) the Windows way often produces much clearer text if you are not using font-smoothing in either case. Once you add font-smoothing techniques into the mix (they use slightly different sup-pixel mapping methods) it all gets a bit arbitrary, varying from typeface to typeface (and by size, the sub-pixel differences mean some colours are more affected in certain typefaces, and opinions vary from person to person too) as to which way around looks better.

The answer? Aside from just putting up with it (and testing your output on both types of font-renderer if you need/want to use typefaces away from the "standard" set) and arguing about which is technically the better mix of methods, the only real solution is probably to wait for nice high resolution screens that will make it a moot point. Unfortunately they are likely to be a long time coming, scaling the sort of DPI an iPhone4 boasts to 20"+ screens is likely to be much more expensive than the market will bare ATM.

lt;dr: MS and Apple render fonts differently. The difference is often not significant, but it is often enough that unless you use the de-facto "standard" set of typefaces you need to test on both systems to ensure your output isn't blurry or otherwise difficult to read on one or the other system.

So would using something natively installed on the machine, like Georgia for instance, solve this problem? Thanks for the lesson, BTW.

Watch out for bare vs. bear.

That sounds wrong to me. Pixel aliasing hints are part of the font design. Are you saying Apple doesn't honor them, or that this particular font doesn't have them? Both seem like killer bugs to me.

I'm guessing this particular font is unhinted assuming Cleartype, or was hinted with Cleartype in mind.

I've no idea what is actually possible or not; that's just a guess. But that page looks awful with Cleartype switched off, and kind of OK (or maybe Cleartype fans would think it looks good?) with it switched on.

Not that Apple's font renderer is not respecting those that are explicitly there, Apple is following the designer more closely. MS's renderer is snapping things to the grid on top of any such hints in an effort to make fonts look neater/sharper/both at low DPI (particularly when cleartype and other smoothing techniques are off), and that this can back-fire.

If you're in Firefox, you can disallow pages from choosing their own fonts. Preferences->Content, advanced in the font section, and uncheck the option.

Can we pretty please have this by default?

Your knowledge of web font usage is a bit out of date. @font-face allows use of custom fonts on the web, and works across all browsers[1].

As for your font rendering issues, either you need to turn ClearType on, or turn it off and on again and restart Chrome (it gets stuck sometimes). Or, you know, upgrade from XP.

[1] http://www.fontspring.com/blog/the-new-bulletproof-font-face...

Indeed. The medium defeats the message, when you're offering web design advice on an unreadable site.

First law of web site development:

The client's copy never arrives by the agreed deadline.

Second Law:

After the copy deadline, the client says they are a bit busy so could you just put something suitable in.

Upvoted for truth. In 5 years of professional web development I don't think I've ever had a client meet a copy deadline.

Third Law: After putting the copy in, the client changes their mind and sends you a Word doc without using outline mode so you can't see the changes and have to copy everything back into the site again from the start.

Saves as .txt and a text diff utility can be a real life-saver. It's not perfect, but with a reasonably-skilled user it should be a lot faster than redoing everything from scratch.

Actually, CSDiff allows you to diff Word documents (at least old versions, I've not yet tried it with .docx files) and it outputs a new one formatted set up with Track Changes on and highlighting the edits. Very useful.

That still sounds like a contract issue. It should be included in your contract that if the client does something like that, then they should pay extra. If that still doesn't alleviate the frustration, you aren't aren't charging them enough.

When it does arrive, instead of a headline + 100 words for the little box, it's a 5 page essay.

Or vice versa. I should've just dropped the first sentence of Lorem in for most of these pages.

Or you (directly or via the agreed project plan) tell them it could take seven days to sort everything after the copy is delivered. They deliver the copy very late (and not in the agreed format) and get upperty when they find not absolutely everything has been done by 0900 the next day. Because you should have been organised and ready to deal with the content as soon as it arrived...

My caveat to the second law: copy rates are 2x that of design. Now our team is more than happy to write for you.

Add that as a rider in the contract, and your problems will begin to disappear.

"The Great Lie" is a bit over the top, no?

This isn't a problem with Lorem Ipsum, but with planning and copywriting. You can't blame LI for the fact that the copy sucks.

In the context of Lorem Ipsum versus the revelations of the Pentagon Papers, maybe a little over the top, sure.

I think my main point was that it now occurs to me how flawed it is for an entire team to work on/refine/agree upon the design of a website and a key piece of company branding without putting apparently any thought up front into that most key piece of the website - the copy. So yes, I agree with you on all points.

I never start designing website before I get final version of all copy.

That's the rule.

No placeholders. No quick drafts either.

Web is all about type and words, I do pay a lot of attention how I present them, and I need to know what they are to do my job well.

I completely see the logic of what you're saying - how can you be expected to present words when you don't know what the words are - but does this hard line work in practice?

Do you demand every bit of copy or just the key bits? For instance if you were doing a travel site with 100 hotels would you want every hotel or just two or three as a template?

In the example you have given, yes, 2-3 examples would suffice. However, I usually design websites with carefully crafted text and design (that is, with a lot of content that won't change much and won't come from CMS), so in most cases, I ask for all copy.

We've tried sending design mockups to clients with the copy already inserted instead of Lorem Ipsum for this exact reason. It usually tends to lead to disappointment when the browser renders their copy slightly differently to Photoshop

This is another example of design being how something works not how something looks.

A website's copy is its function.

This can't be stressed enough.

Lorem Ipsum is great for sites where the copy is not the function. Say a message board. Its purpose is to provide unpredictable quantities of text. Using it for definite amounts, like marketing copy, is damn close to a misuse of the tool.

In fairness, I've actually had a good time on this little project. It's just a Wordpress site with 2 different templates, so it's not exactly complicated or high pressure, and I used it as an introduction to Compass (on which I'm still undecided).

I think I intended this post as a note to future clients on small projects, especially when they bring forth rather ambitious sitemaps and insist on drop-down menus (at the last minute (which are currently only serving to draw attention to pages "under construction")).

Thanks for reading.

Seems to me that you get paid either way. It's the client that loses if they don't provide the content they promised.

That's great, now you've got another crap looking website to add to your portfolio. Eventually you'll stop being paid. Just because the client's paying doesn't mean they've got a clue. It's our job to steer them towards great software, not take their money and produce mediocrity.

Actually with the exception of the copy, which may or may not get straightened out, it's a pretty decent little WP design. And like I said, I used the opportunity (which was basically a favor for a buddy) to take Compass for a spin, to fine tune my own little Sass library that I've been working on, and to get to know and work with a new designer. Building a cathedral was never the idea.

My main client is the one I'm obsessively steering toward great software.

Your past work is the only tangible evidence of your current ability. A good design with bad copy looks like a template, which may be OK if you design templates, but if you design websites, copy is extremely important.

Every day you wake up in the morning, work, and then you put your head on your pillow at night. You never get that day back, it’s gone forever and your life is one day shorter. You’ve either done good work or bad during that day. How many days of bad work can you tolerate before you decide to make a change in your life?

I think at some level people either decide that they don’t care about their work or they care enough that bad work is painful. It’s not for me to tell you or anybody else which way to go with this, but for me, bad work is painful and it’s up to me to work things out with clients such that I do good work and they are happy with the results.

And the alternative is to go to bed frustrated and hateful because a client didn't deliver on a promise? You can't change the client's past actions. You can't even improve them. The best you can do is go to bed knowing you did your best.

If a client does something such that I do bad work once in a while, fine. If one client in a while does things that make me do bad work on their project, that is regrettable but I’m not going to an early grave lamenting it. But if this state of affairs becomes normal, such that I am doing bad work nearly every day, for nearly every client, that is not fine with me personally.

Reply on news.ycombinator.com