Hacker News new | past | comments | ask | show | jobs | submit login
3px of padding makes all the difference in the world (garrytan.com)
240 points by jjb123 on May 28, 2013 | hide | past | favorite | 87 comments



Maybe the original design had the even padding but facebook A/B tested it and found that people are more likely to read the comments below if there is less padding on the bottom?

While I agree that is visually looks more balanced, it takes more than my opinion to decide that facebook made a design mistake.


This is a perfectly reasonable explanation that is often overlooked by designers. This also happens to be why I don't particularly love design anymore. What looks right and what converts the best are two different animals, and I don't like purposely creating ugly shit with my name on it.


This is often a result of designers failing to step outside their "designer window", if you will, and step into the users' shoes. It's possible to make something look right and translate that into conversions if you have enough of an outside perspective.

That said, it's pretty easy to get too involved in your own work to notice a bigger picture so breaks to step back are highly recommended.


When I was a/b testing UI designs for a stable of eCommerce sites, quite often the blatantly fugly design would convert much better. Especially for low-price-driven sites. My personal theory was that people figure the unpolished, crappy looking site is more bargain-driven (kinda like a dingy dollar store in real life).

Sticking a giant firetruck-red "CONVERT" button surrounded by flames and rainbows really does sell better. But boy does it look awful.


This is also evidenced by Amazon.com still looking really crowded and ugly. They test every last thing and if this is what they're using, it's because it works.


This same concept goes for those horrendously ugly, cookie-cutter, inline sales pages plastered with maroon Comic Sans headlines that have highlighter-yellow backgrounds. I've been there and done that, and the facts prevail. As a good/(as in one that creates aesthetically pleasing things) designer, you actually suck conversion wise.

I haven't done much research behind it, but I believe this has to do with why the flat UI is so pleasing. I can only imagine that it is easy for anyone to notice that it's beautiful regardless of technological/design experience/knowledge.


I am glad you pointed this out, your premise is a quintessential example of "UX is not UI."

It's not the visual or web designers primary job to A/B test or step into the users shoes on every detail, that's what a responsible and valuable UX person should be doing (note I didn't say UX designer).


The classic blog post on this: http://stopdesign.com/archive/2009/03/20/goodbye-google.html (aka "41 shades of blue").


Sometimes when things are too much in proportion we're left admiring the structure and not bother with the content. If something triggers our OCD, the padding in this case, it forces our attention to that spot. It's similar to how we're drawn to faces on a page.


Great comment. We have found this to be true in our own A/B testing. It sometimes makes me cringe, but you can't argue with data.


Per the CSS, the margin of the status text element is 5px on the top and bottom.

The margin on the "Like/Comment/Share" and Comment boxes is also 5px total. (Comment box is margin 1px, padding 4px)

Maybe it's symmetry of a different sort?


Was thinking the same thing but not for the comments but for the 'liking', 'sharing' and their money making 'promote'.


Then they should also A/B test the comments with big red fonts, because it's possible users will read it more.


He has a point that many interface designers will appreciate and others rankle at as nit-picking.

But I was struck by the irony of this particular observation being made via a blog that uses a body typeface with ever-so-slightly but definitely too-narrow tracking (letter spacing). Or maybe the kerning data is off. I would have it loosening by about 1/200em overall, with the most noticeable improvement becoming apparent on adjacent vertical strokes.

EDIT: the samples of this face, "Oxygen", that I see at the creator's site (Vernon Adams) don't suffer from the same issue so perhaps something has gone awry in the conversion to WOFF.


Yes, and his blog’s title uses an ASCII straight quote in lieu of an actual apostrophe.

Before that, Posterous’ default template had tiny, light grey, non-subpixel-antialiased text for body copy. It was excruciatingly hard to read. I never understood that; the point of a blog is to making reading and writing easy, but Posterous only ever cared about the “writing” side of that, it seemed. (This point is further illustrated by the obnoxious headers Posterous blogs had. By far the most visually noisy part of a Posterous post page was the call-to-action at the top: Sign up for Posterous, yourself!)

Actually, Garry, I’m pretty curious about this. What was the reasoning behind that small grey body text? I’m sure I am just being naïve here. Please enlighten me.


The real irony lies in the fact that Facebook has these "issues" and it doesn't stop them from raking in millions of users, but the perfect (well, almost) typography of Posthaven's site does not seem to be helping (I am aware of the unfair comparison).

Perfect typography does not dictate the success of your product.


Thanks for the feedback! I'll look into this. I'd been wondering about that myself.


To be fair, the number of users that read his blog vs. the number of users that use Facebook is a little bit different, and the amount of attention that should be devoted to each should vary accordingly.


"His blog" is the template for the entire Posthaven network.


I noticed the exact same thing. It was rather annoying. For example, in the word "like" (as in "And here's what it would look like fixed") the "li" had joined into a single shape.


I've never found a way to fine tune the letter-spacing using css, the smallest unit seems to be 0.07em which is usually too much. Would love to know how to do it if it's possible.


Usually you don't want to change letter spacing. It should be part of the font itself. By changing it via CSS letter-spacing, you alter the font's built-in spacing. This is particularly noticeable with letter combinations with very tight kerning, like 'AV' or 'WA' or 'li li' or 'ni in'. While most web fonts do not have the vast kerning tables that a professional typeset font will have, the built-in kerning is usually leagues above what you can accomplish with letter-spacing.


text-rendering: optimizeLegibility; may help in some cases


Is this safe to use now? If I recall correctly this crashed some browsers not too long ago, and could cause some extreme rendering glitches, such as text disappearing.


Uhm, didn't know about that it never happened to me


A reminder that the imbalance of 3px may be intentional:

It means that the post is closer to the comments are closer than to the user, and it pulls your eye more naturally to the comments than if they were perfectly equal.

Sometimes asymmetry is a useful design element, too.

Edit: I noticed that lukeholder has said the same thing.


For me, the current version is better. And in general, I don't get the obsession of designers to have text padded as much as possible. I like low padding and small fonts (unless it is put to extreme; but on good sites it never happens, whereas "overpadding" is more than common).

Perhaps it is about density of information.

Perhaps it is that I like to have a very different spacing between two lines of the same paragraphs and two lines with a paragraph break. (So my brain can effortlessly divide it into parts.)

Perhaps... it is just a matter of taste. But I guess I am not the only one.


I think it's less about padding-for-padding's-sake and more about maintaining symmetry in design. Notice that both the author's changes were fundamentally just restoring symmetry to elements within a box layout.


In this particular case the symmetry makes some sense. But I would rather cut the top (which is wasting a lot of space) than padded the bottom to match the top.


It must be a really quiet day for this to make the front page! I can't believe I just wasted time reading that.


And to think you wasted even that much more time making this comment! Your time is leaking everywhere!


The author is a YC partner, so it almost automatically gets up-votes for that. (He’s right, but so are you — it’s hardly newsworthy.)


One single timesuck in a reservoir of timesucks? HN is hardly the place to spend time if you actually have things to do.


Well, that makes the article better, eh ..?


A lot of people (have and will) scoff at the idea 3 whole pixels? And I agree it does seem trivial, and will those 3 pixels really change the way users interact and make users happier? Who knows. I know as a designer it does indeed look better when it's visually balanced, but all that aside, I am a huge fan of craftsmanship and attention to detail in ones craft. More so with crafts made by hand (wood, metal, glass) but also it translates over into design and programming.

As a designer I want to put love and attention into all my designs and I want each one to be better than the last as to always be improving. Those three pixels may not mean much to you, or most, but it's those types of things that when you pull out the magnifying glass or really admire a piece of art, that really matter. A lot of times I fail personally at making sure each detail is right, but I would hope that someone could one day look at my work and say he really paid attention to the details and he wasn't just rushing or not caring about the projects he did.

So I think it's less of wow is this really worth all the fuss over 3 pixels but more of a testament to how Facebook values the details in design. Sure Facebook has design that's decently pleasing to the eyes, but you have to wonder why this went unnoticed. If it was me coding the feed, I would've caught it and fixed but I wonder whoever was doing the coding actually stopped to think about it or tried to fix it and got shot down or whatever reason it could have been. It could've been an honest mistake, or a purposeful one, or just someone trying to meet a deadline.

To send a note out to everyone for a thought to think about, really try to put some sort of personal attention to the details when you're working. Whether it's programing or designing, you may not think opening up your code editor and finding the right line in the CSS to fix those 3 pixels is worth it, but someone one day might see that and have to wonder why and if you care about the quality of your work and the image your putting on your brand, I would take it personally to make sure it was perfect. I think this goes a long way with moving up and getting noticed. Anyone could skip it and it's not the end of the world, but finding people that care enough to put such attention and detail in their work to make sure everything is perfect are the kinds of people that are worth having on your team.


I've noticed Facebook has been doing a lot of AB testing on me lately. [Example: My Graph Search bar is now White-on-Blue instead of Blue-on-Blue. Also, the sidebar with birthday notifications used to display Friend Requests that told you who Friend Requested you; now it's gone.]

This spacing behavior is likely also a part of that. It just happened to me very, very recently, and there used to be sufficient padding in that UI ___location.


It's happening to me too. Every day something looks broken (read different). The Facebook logo was not centered before the graph search for a couple of days, the height of the chat box was changed and same goes for the padding. All I hope for them is to remove (or shift) the video chat button. I've had lots of accidental clicks on it.


If enough users have those accidental clicks, it might skew the a/b testing results :-)


I have noticed that most flame wars happen over small things; not simply bc they matter, but bc the opinions are so simple that everyone has one.


This is called 'bikeshedding': http://bikeshed.com/ . How do you all like that design?



I don't think it makes any difference.


I've spent about a minute on comparing both pictures to see "how much more crowded the text is in the unfixed version"©


Somebody should fix his OCD.

It's not a bug -- by any stretch of the imagination.

Unequal padding is a legitimate design decision as any other other.

But I see how it can annoy the hell out of compulsives.


Good designers are often perfectionists. Calling them "compuslives" is dismissive, and is also insulting to people who actually suffer from OCD.


>Good designers are often perfectionists.

Symmetry is not perfection. Plenty of unbeatable designs and works of art are non symmetric.

And the poster of TFA is not a "good designer" IIRC, merely some annoyed blogger. I was talking about HIS annoyance with an assymetry (that might very well be intended) and not about some designer sweating the small details.

>Calling them "compuslives" is dismissive, and is also insulting to people who actually suffer from OCD.

It might be dismissive, but it's also quite accurate. Could even be medically accurate for a lot of them.

(Seems you cannot say anything in this day and age without "insulting" someone. So be it. I have mild OCD too, and I could not care less about someone calling me "compulsive").


I guess you could call a YC partner "some annoyed blogger," since in this case he seems annoyed about a 3px difference. But, hey, guess what? His annoyance actually prompted an internal bug at Facebook. Somebody, at least, cared enough to consider the difference.

> Symmetry is not perfection.

Being a "perfectionist" is not the same as actually producing "perfection." It's about setting extremely high standards. I never said I agreed that the extra 3px was better, but I understand the importance of the decision, however minor.

I also understand that, in most cases, 3px won't actually make much of a difference. But hundreds of pixel-level decisions over the course of an entire project will make a difference in the final presentation. I'm not saying all of them are right or wrong, but unless you're going to A/B test every single debatable bit of padding or margin, deferring to a designer who cares about the presentation and aesthetics makes sense. It is often subjective, but so what? It's the designer's job to care about these things.

I still think that equating this to OCD is simply not accurate. As someone who does not suffer from the disorder, I guess I can't actually speak to whether or not it's insulting, so that was poor word choice. But I do know that living with OCD can be totally debilitating, and to say that it could be "medically accurate for a lot of them" (designers? perfectionists?) is just plain wrong. This is what OCD looks like:

http://www.youtube.com/watch?v=nrRqb5cvM5E http://www.youtube.com/watch?v=x4sadYeLHKU


Using the word 'perfectionist' is just as badly fitting for something as subjective as the decision about whether to center vertically or to have the padding on top twice as big as the padding on bottom.


<satire>

yeah, but over 1.8 billion page views.. thats 3px is a _lot_ of extra bandwidth.

Not to mention all those screen pixels gone to waste.. thats 3 times the screen width every time.. say 50x3x1280 px per page view.. jesus thats enough to power a small village in the third world.

Not to mention the Carbon footprint...

</satire>


Might I politely - as an Englishman - make the suggestion that if you feel the need to wrap it with "<satire>" tags, you're doing it wrong?


> Can you see how much more crowded the text is in the unfixed version, and how much more balanced it could be?

No. I had to open the two images in separate tabs and rapidly toggle to even see the difference.


I'm glad I cannot tell the difference otherwise my web browsing experience would be frustrating every day.


How can you call something completely subjective a "bug"?


if you believe beauty [ or even readability ] is a feature


why downvote ? I hilite a logical inference :

If readability is a feature, then lack of readability is a lack of that feature, which is a bug [ as measured against the goals of the product ]


Readability is not subjective. You answered a question with a non sequitur.


Readability is not subjective ?

So you have a mathematical formula for Readability, excellent.

That would be simpler than adjusting things by 3px until they just look spectacular and zing off the page.


You can measure how fast people read. It varies per person but it's not an opinion.

But this is not about readability. It's about what looks 'better', whether you prefer balance vs. having the comments closer. The real debate is about a subjective thing, and not about readability.


Can you see how much more crowded the text is in the unfixed version, and how much more balanced it could be?

Not really. No.


I'm interested to see if this gets changed or if it becomes an inside joke at Facebook.


That would be awesome if it became an inside joke!


Garry is wrong. What looks visually pleasing is not necessarily the right layout. Historically, typesetters only had paragraphs to set in doing a book layout. Now in the age of post DTP, layout can be more expressive. That is exactly what the lesser margin of 8px does. In this case, it means that you are liking the status message only 8px away, and not liking the author.


It's always tricky to adjust the padding/margin values because you need to take into account each case in which these 3 added pixels will have an impact. Maybe the style applied to this particular box is shared with other elements or maybe this box is used in other contexts where 3 pixels more will look odd. And I guess that for a website as dense as Facebook, it's easy to try to adjust something while breaking something else.

On a side note, whenever I set the margins of adjacent elements, I take some time to wonder if I'll put a margin-bottom on the 1st element or a margin-top on the 2nd one, or put both margin-top and bottom on the middle element and none on the 1st and 3rd. My rule of thumb ends up being the following one: use only one margin (top or bottom) on all adjacent elements, because some of them might disappear on some pages, so I need to provide some kind of flexibility in order to maintain a decent spacing for each and every scenario.


Facebook is probably built in a much mote modular way. I doubt fixing it would break anything else. Facebook engineers should be able to figure this out very quickly.


This is beginning to remind me of the colorists in Hollywood who color-time feature film. They will obsess day in and day out about the subtleties of a particular hue of some background out-of-focus trinket. Complete and utter minutiae.

I'm sorry, but as an average reader with decent visual tastes, I can safely say that those 3 pixels are not relevant to me.


I don't see how this is annoying and adding 3px is going to "make the difference in the world" Look at the authors blog. see how much space between the header and the "<Back to blog" and the content. Fix yours first. Its actually annoying than this


Padding is not always supposed to be "balanced". It makes we we for some sections to be closer or further to other sections. In this case, he extra padding probably looks very slightly better. But I don't think it's the way it is from lack of love.


This is the same obsessive minor attention to detail that turned Apple into a multi BILLION dollar company.

... and well deserved at that.

I say good on you Gary, its people like you who push the world forward, continually looking for ways to improve things no matter how minor.


Two things -

First: Where on Facebook is this style? In my current layout (which is the current "stable" layout - pre-graph-search), I don't see this anywhere. Not on my feed, not on my timeline. It looks like a timeline post (feed posts don't have a border around them like this does).

Second: What bugs me way more about the screenshots posted is that the left side of all the info doesn't line up. The large profile picture and the body text are (!) 3px more padded than the action bar, the "Like" list and the comment area. But! It should be noted that, again, I don't see this anywhere on FB, just the screens that Garry posted.


> Can you see how much more crowded the text is in the unfixed version, and how much more balanced it could be?

No way I would have seen that without it having been pointed out. I would have been looking at the actual content.


This guy is clearly a moron! That spacing needs 5 extra pixels, not 3!

Seriously though, great blog post. He is exactly right about the need for more spacing. I feel like he buries the lead... the last change he makes to the bottom comment section is a night and day difference. Not only does the last change produce an more significant improvement, it appears to make up at least the number of pixels lost in the extra padding. That should take away any possible objection someone might raise concerning a reduction in amount of content that fits on the screen.


WTF is "rageskitching"? I Have searched google far and wide for both "Rageskitch" and "Skitch" and as far as I can tell, it is hanging on to the bumper of a car while skating.


In a time of display resolution variances between 96 and 250+ ppi (slowly coming to big screen too), why are we still talking about fixed pixels instead of relative units?


The px unit in css is a bit misleading, the measurement doesn't always correspond to the size of the device pixels. From the spec[1]:

"For a CSS device, these dimensions are either anchored by relating the physical units to their physical measurements, or by relating the pixel unit to the reference pixel. For print media and similar high-resolution devices, the anchor unit should be one of the standard physical units (inches, centimeters, etc). For lower-resolution devices, and devices with unusual viewing distances, it is recommended instead that the anchor unit be the pixel unit. For such devices it is recommended that the pixel unit refer to the whole number of device pixels that best approximates the reference pixel."

[1] http://www.w3.org/TR/css3-values/#absolute-lengths


Seems fine to me. The content is all closely spaced and the header is set off with a larger spacing. This helps keep the header separate from the content, which in turn helps people skim past content they don't like. If they did this change, next he'd be demanding each piece of content have the same large spacing between it, they end up having a lot more spacing, showing less content above the fold, and having reduced engagement metrics.


If your really hate someone, teach them to recognize bad spacing and bad kerning: http://xkcd.com/1015/


I found this really annoying for some time now, too. I hope with someone like Garry being vocal about this issue Facebook will act and fix it.


I actually like the one more densely packed version. One of the core difference in design of Hacker News and other sites is just density of information. I prefer Hacker News densly packed layout a lot more than other sparse. So I guess it's all matter of opinion rather than absolute truth or one right way.


This is my observation: http://imgur.com/zfxvWzV


Isn't it possible that this is simply a bug? If you happen to have a ___location when you post a status, it goes under the comment with a padding of 8px above and below. If the ___location is missing then a default of 8px appear below the status text.

I don't think it makes a difference


He's almost right. 8px is not enough but it should be at least 13px, not 11px.


Facebook has since updated its design, and the padding is better balanced. My account got upgraded to the new design a month or so back. I assume they are rolling it out slowly.


Love the irony in this sentence: A site that fixes these problems is a site is being taken care of with the loving attention to detail of someone who can see the difference.


Even 1px matter ! With trained eyes you can see all the defaults on every websites... Annoying!


Pixel nazi


You said it before me !


vk.com seems to get both things right (judging paddings visually)




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: