I’ve already started blogging, with a post to introduce the new design, and another where I talk about my new WordPress setup. If you click through to my new blog, you’ll be able to enjoy a cleaner, more readable design, a faster-loading site, and an all-around nicer experience.

Also, I’ll be republishing some of this blog’s best content on SachaGreif.com, so it’ll be a chance to rediscover old favorites

I’m not sure what I’ll do with this site yet. I might just keep it as an archive, but if you have any cool ideas of what to do with it, I’d be glad to hear them!

First, let’s take a look at what we’re going to produce:

Our goal is to take a photo, and create a mosaic where every “pixel” is actually a die. The white-on-black die will be arranged according to their visual density to recreate the image, and the idea is to eventually use this new image as a guide to produce a real-life mosaic like in this video by Fujiya and Miyagi (however, this tutorial will only cover the Photoshop part!). We’ll also go through an optional (but easier to achieve) intermediate step where we encode the dice with colors instead of the actual die face.

UPDATE: for you lazy people out there, I’ve created a Photoshop action that automates the whole process with any photo! You can download it for free here along with the patterns.

To get started, let’s look at how the original script works. It converts an image to 6 shades of grey, divides it in squares, and then assign a die to each square according to the color’s density. For example, the lightest shade of grey is represented by a black die with six pips, while the one-pip side is used for the darkest shade.

So how would we do this with Photoshop? Well, let’s start by opening our photo (I’ll use the same photo as the author, in order to make it easier to compare results, I hope he doesn’t mind!).

1. Making the photo black & white

The first step will be turning our color photo into a black and white one. Fortunately, Photoshop has a menu item that does just that. In the Image menu, go to Adjustments > Black & White:

Let’s just use the default preset for now. Playing around with those value will have a big impact on the final result, so you can definitely come back to this step later if you want.

2. Reducing the number of shades of grey

Now that we have our black and white image, how do make sure we’re only using 6 shades of grey? Once again, Photoshop has the answer:

Still in the Image menu, go to Adjustments > Posterize.

And choose 6 levels, since a die has 6 “colors” (i.e. its faces).

3. Creating the tiles

We’re getting there! The next step is dividing the image into 16*16 pixels tiles. This time, let’s open the Filter menu and go to Pixelate > Mosaic.

And choose a 16px square size:

Looking good. We have our image divided into 16px tiles colored with 6 shades of grey.

4. Encoding the die faces with colors

We’ll also create our own system to make it easy to know which die should go where. However, instead of using die faces, we’ll simply use colors! Black will mean the 1-pip face, blue will mean 2 pips, and so on until white for the 6-pips face.

Go back to our trusty Image menu and pick Adjustments > Gradient Map:

Now here comes the fun part. Gradient Map simply maps a gradient to your document’s colors. Since we have 6 colors, let’s create a 6-color gradient. Click the gradient, and the Gradient Editor dialog should come up:

We want 6 colors, and we already have black and white so we only need four more. So click four times between the black and white color stops (those small things shaped like houses on the bottom row), and place the four color stops at 20%, 40%, 60%, and 80%.

Not much to look at yet, since our gradient maps to all black. Let’s fix this by picking 4 different colors (just double-click on the color stops to open the color picker). The goal here is to pick very distinct colors, since we’re going to assign a die face to each of them:

The problem now is that since this is a gradient, Photoshop has inserted a lot of extra shades between each color. We’ll fix this thanks to our good old friend, the GIF image format (also works with PNG8 by the way).

5. Removing the extra colors

In the File menu, choose Save for Web & Devices (don’t ask me what those “devices” are, I’ve never understood either).

Now choose the following settings: Format: GIF, Dither: 0%, and lastly remove all colors except the 6 we want (i.e. black, white, blue, green, red, yellow). To remove a color, just select it in the color palette and click the little trash can (you can shift-select multiple colors).

And there you go! Now all you need to do is write down on a piece of paper that black equals 1, red equals 2, etc. and you’re good to go.

Click to view full size

Now I was going to stop there, but then I realized it’s actually quite easy to accomplish the exact same effect as the original.

6. Reproducing the dice mosaic using patterns

All you need to do is create six 16 by 16 pixel patterns, one for each face of the die:

Then define them as patterns by going to the Edit menu and choosing (what else) Define Pattern.

Now once your 6 patterns are defined, go back to your 6-color image and choose the Magic Wand tool (the fourth one down in the tool palette in my version of Photoshop). Make sure to turn off “contiguous”, otherwise you won’t select all the pixels of the images, only the ones that are contiguous to the point you clicked.

Now select each of the 6 colors, and fill them with the corresponding pattern (paint bucket tool, and then choose “pattern”). Make sure “contiguous” is turned off here as well.

And here’s our end result! Pretty neat, don’t you think?

Click to view full size

Let me know if you use this technique to build an actual die portrait, and thanks again to Scott MacDonald for the inspiration!

Well, don’t worry. I’m not here to give you impractical advice or tell you something you already know. I know that if you had real content you wouldn’t be using Lorem Ipsum, and that in some cases it’s simply not possible to obtain it. But instead of Lorem Ipsum, Hipster Ipsum, or even Samuel L. Ipsum, I think you should use the first paragraph of a random Wikipedia article.

To explain why, let’s do a simple test. Please read the following paragraph carefully:

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in the answer is purple voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

Now what if I ask you what my favorite color is? Can you answer?

I’m pretty sure you can’t, despite the answer being included in that paragraph.

Why? Simply because as soon as your eye picks up on some incomprehensible latin words, you’ll glaze over the whole paragraph and stop reading. Believe me, there’s no way you can force your brain to actually read that gibberish.

Seeing vs Reading

This means that both you and the client will end up ignoring whole paragraphs, neglecting things like leading and line length just because you’re seeing the text, but not actually reading it.

On the other hand, something like a Wikipedia page or a news article has a real rhythm to it, and the familiar words will grab your attention the same way the final content will.

Why Latin?

By now, I figure most of the people using Lorem Ipsum don’t actually know the reason behind using Latin instead of plain english. It was simply to ensure that the fake filler text would stand out in magazine layouts and would not be mistaken for real content and left in when printing.

But do we still really need such precautions? I’m sure most people are smart enough to distinguish between fake and real content (provided you mention that the pages includes fake content, of course). And worst case scenario, even if some of the fake text makes it into the real website you can easily change it as soon as you notice, unlike a print magazine.

So give it a shot. Believe me, it will make it much easier to focus on typography and those long, boring text paragraphs will suddenly appear much more interesting.

Unsolicited redesigns are when a designer takes it upon themselves to redesign a site for free without having been asked by the company. I recently did a quick Quora redesign myself, and it prompted me to explore this topic.

Case Study #1: Zappos redesign by Metalab

Zappos redesign by Metalab

One of the main problems with unsolicited redesigns is that if you’re not careful, you’ll risk coming across as an arrogant know-it-all who thinks they can do a better job in a couple hours than a whole design team did in 6 months. This is what irked people about Metalab’s Zappos redesign. Calling the current website a “confusing mess” was probably not the most humble stance they could take.

But isn’t that arrogance justified? After all, their redesign did look much nicer than the original! Just look at all these pretty gradients!

However, the comparison is unfair. Unsolicited redesigns don’t have to deal with the same realities as the real deal.

Case Study #2: Facebook redesign by Jonathan Moreira

Facebook redesign by Jonathan Moreira

First of all, in most cases the redesign is only a static mockup. Can you really tell if the end result would be enjoyable to use? For example, Jonathan Moreira’s Facebook redesign looks awesome, but maybe that big black bar would get overwhelming in regular usage, or maybe the gradients and shadows would attract too much attention and make it harder to focus on the content.

Case Study #3: New York Times redesign by Andy Rutledge

New York Times redesign by Andy Rutledge

Second, in most cases redesigns can side-step business realities. For example, Andy Rutledge took the New York Times site as an example in his “News Redux” exercise (cached article). He achieves a clean and fresh look, but does so by doing away with a lot of ads that are probably necessary to the site’s survival, which would require changes to the company’s business model. Although in this case I agree with those changes, in the real world the design team probably doesn’t have the power to make them.

[Note] I initially wrote that Andy had ignored business realities, which isn’t true and is not what I meant. Sorry Andy!

And even if your version simply is better, maybe designers in the company are dealing with some other factors (politics, bureaucracy, etc.) that make it impossible for them to improve the design. Of course that shouldn’t stop you from volunteering a solution, but you should at least acknowledge that the problem is probably not lack of skills on the designer’s part, but simply different priorities inside the company.

Case Study #4: American Airlines redesign by Dustin Curtis

American Airlines redesign by Dustin Curtis

This is apparently what happened with Dustin Curtis’ redesign of American Airlines. Dustin acknowledged that the existing site’s main designer was pretty good, and the poor state of the site was due to other factors.

Case Study #5: Quora redesign by Sacha Greif

Quora redesign by Sacha Greif

So where does my own Quora redesign fit in?

Well first of all I tried to stay humble. I explained every design choice, and I let people vote on each of them. I never assumed that my version was objectively better, I just wanted to see what people thought of it.

And I also tried to keep close to the current’s site aesthetic and goals. There’s nothing wrong with being fancy and creative, but I feel it’s just too easy to achieve a meaningless “wow” factor when you’re taking liberties with a brand and going all over the place with colors and fonts.

I find it a much more interesting challenge to produce something that might conceivably be used for real. This is why I kept the same colors and fonts, and even kept the exact same width for the main content feed to make the hypothetical transition smoother.

So altogether, I think unsolicited redesigns can be a net positive, provided you keep a humble attitude. After all, there’s nothing wrong with throwing ideas around and debating different solutions to the same problem. As a community, we should encourage that kind of exchange and not worry so much about hurting the sensibilities of established designers (I’m sure they can take it!).

So let me know what you think about this. And I’d love to get your feedback on the Quora redesign, whether it’s good or not!

So if you’re looking to hire a good designer, here’s where I’d start looking:

1. Dribbble

In the battle of seducing great designers, Dribbble has won hands-down. Sure, other sites like Behance or Forrst (or even Deviantart) include amazing talent, but Dribbble is my personal favorite, especially for web projects.

Quick tip: Search by Location

Most people don’t know this, but you can search Dribbble designers by location. Just search for a city name, and then select “See full results” in the user column on the results page. For example, here’s a search for Vancouver designers.

2. Folyo

I’m biased on this one, since I’m the guy behind Folyo. But from the feedback I’ve had so far, people seem to really like the concept. Like other sites you can browse profiles for free, but you also have the possibility to pay a small $100 fee to send your project to all members.

Quick tip: Designers Map

Folyo recently added a designers map that lets you see all Folyo members around the world.

3. Zerply

Zerply is what LinkedIn would look like if it worked better, looked nicer, and didn’t send you notification emails every half hour. Although it’s not limited to designers, there’s a lot of very good ones on it already. You can search profiles using their tag system, for example here’s a search for London designers.

Quick tip: Endorsements

Zerply has a really nifty endorsement feature. Every profile lets you see how many people vouch for them for a variety of skills and tags. So endorsements can become a good proxy for talent and exeprience.

More resources:

• How to find the right designer (Attack of Design)
• How to pick a great designer (Folyo)
• How to write a good job description (Folyo)

• Mistake #1: Not having a portfolio
• Mistake #2: Using a cookie-cutter tagline to introduce yourself
• Mistake #3: Having tiny thumbnails
• Mistake #4: Only showing small screenshots of your work
• Mistake #5: Displaying inconsistent work
• Mistake #6: Not providing enough details about your work
• Mistake #7: Having an unfocused portfolio

In that guide, I advised to:

pick [a designer] whose current style is already a good fit for your needs.

I even quoted Jason Cohen:

I cannot stress this enough:designers don’t morph their style to match yours.

Here’s what Jin Yang tweeted in reply:

Is having a "style" such a bad thing?

So should you hire someone who “doesn’t have a style”? And as a designer, should you try to have one?

Well, I’m afraid I’ll have to disagree with Jin Yang on this one. I fully believe you should try to have your own style.

What “style” means

But before I explain why, let me explain what I mean by “style”.

For me, “style” is first and foremost visual style. The fonts, colors, shapes, or icons you use all contribute to it. So do small things like line width, light and reflections, gradients, or shadows.

Zappos redesign by Metalab

For example, Metalab are famous for their very crisp designs with lots of 1px highlights and vibrant colors, and their fondness for Hoefler & Frere-Jones fonts.

Furniture Icons by Tim Boelaars

Another great example of having a style is Tim Boelaars‘ work. He specializes in clean yet complex lines of uniform thickness, and his work is instantly recognizable.

Dana Tanamichi's amazing chalk work

But “style” does not always have to be purely visual. Maybe your “style” also includes the medium through which you express it, like Dana Tanamachi‘s chalk work.

What’s the point?

So why should you have a style? For me, saying you need a style is just another way of saying you need a focus. There’s no way a single person can be great at everything.

So we all have to choose if we want to be average at a lot of things, or great at a few. Personally, I know that I’d much rather pick the second option.

Having a style gives you an excuse to zero in on the thing you really want to do, and obsess over it until it’s perfect.

It also means that people will choose to want to work with you because they want to work with you.

Not because you’re the cheapest, or closest, or you were available, but because they specifically like what you do. Having a style is the best way to avoid being commoditized.

Hammers & nails

But what if your style is a bad fit for a project, and you do a bad job? Or even worse, don’t get hired at all because of it?

Well, you don’t see Dana Tanamichi being hired to design web apps (as far as I know…), and you don’t see me drawing on chalkboard either.

Is that really such a bad thing?

So don’t be afraid of having a style. People will say that when all you have is a hammer, everything looks like a nail. Well, I say there’s already enough pliers, screwdrivers, and drills out there. Keep hammering!

But I was soon confronted with another problem: now that I made it easy for people to find good designers, how do they pick one?

So here’s a short a guide I wrote to address this topic. Criteria include visuals, structure, experience, and a couple others. I also made an extra effort on the design, so go check it out.

Folyo is a newsletter of curated job offers, sent to great designers all over the world. It costs $100 to submit a job offer, and if it’s accepted it will then be posted to the site and sent to designers in the next newsletter.

Why Folyo?

Folyo was born out of two realizations: first, as a successful freelance designer, I didn’t know where to refer clients when I couldn’t accept their project. Second, I realized many companies had a hard time finding good designers. The old “ask around” method shows its limits when your ideal designer might very well be living halfway around the world.

So I decided to do something about these two related problems: I created a simple newsletter of job offers that I’d send to other reliable and experienced freelance designers.

After a couple months of maintaining this service, and getting great feedback from both designers and companies, I decided the concept was worth exploring further, and so I built Folyo.

Unlike a stand-alone newsletter, Folyo also gives clients a central place where they can manage designer’s replies. What’s more, having stats on who has replied means that I can also guarantee a minimum number of replies.

So go check it out and let me know what you think

But now, I’m starting to think the skeptics were right. Even though all the things that made the site’s success are still present, it seems like all the excitement has gone away. Here’s what I think are the main causes.

It’s not the next big thing anymore

The new kid on the block

For a long time, Quora was the next big thing, but that all came to a stop the day Google+ appeared. The early adopters crowd is not a faithful one, and they’ve already moved on to the hipper, younger, prettier site.

People like to ponder if Google+ will turn out to be a Facebook killer, but smaller social networks have probably more to fear than Facebook. After all, there’s only so much hours in a day that can be wasted idly browsing timelines, and it seems to me far more likely that people will get rid of Quora before they sacrifice Facebook.

It doesn’t solve a problem

Quora’s problem is that it has no problem. To solve, that is.

Facebook solves the problem of how to stay in contact with your 250 “friends”. Twitter solves… well, ok, Twitter didn’t solve any real problem, but has still grown to become extremely useful. But can Quora do the same? You might say that since Quora is a Q&A site after all, it solves the problem of answering your questions. But a closer look at the content shows that it’s actually a lot closer to a message board – just one where every topic title is phrased as a question.

And if you just want a practical answer to a practical question, there are other sites that will prove more effective.

Put it another way: would you be willing to pay a monthly subscription for Quora? I’m willing to bet the answer will be “no” for most people, but that it might be different for, say, Stack Overflow.

It’s no fun to use

Rounded corners and a gradient? On the same button?!

A lot of good things has been said about Quora’s great design, and people like to rave about how functional it is. And it’s true, Quora is very usable and well thought-out, and there are a lot of small details that contribute to the overall user experience.

But it’s no fun at all to use. The site is visually bland: there’s barely any color or images, and you won’t find any effect fancier than rounded corners. And there are no special interactions or animations to be found, nor any badges, points, karma, or other gaming mechanisms.

Social networks are the place people go when they have too much time on their hands. Thanks to apps, photos, and casual games, it’s very easy to lose yourself in Facebook. I’m not saying Quora needs to add Farmville, but right now using it feels like doing your taxes. In a very efficient and functional manner, to be sure, but still not much fun.

It’s not that interesting

And finally Quora’s biggest problem: it gets far less interesting after a few months of use. The site will serve you up the same topics and the same people day after day, and although it might suggest new people to follow, it won’t suggest new topics.

This is a classic consequence of designing a social network around “warrens” rather than “plazas”: after a while people become locked in their own little communities, and the lack of new blood leads to a slow decay, and eventually death, as all the interesting people just “evaporate”.

The results is that if like me you joined Quora for its great startup-related content, your timeline is probably filled with questions like “How many employees does Airbnb have?”, which frankly, I don’t care a damn about. And since the startup community is still the most active on Quora by far, other topics will more often than not be drowned in the tech flood.

What should Quora do?

That’s a good question. If only there was a Q&A site where you could ask it!

But seriously, I think if Quora wants to survive it has to move away from it’s Q&A roots, and embrace its de-facto status as a kind of tech world Reddit. Quora has a great community, but it needs to be re-energized. Maybe a redesign is the way to go?

It also desperately needs to break down some of those internal walls, and make people venture out of their warrens. I might not care one bit about gardening, but if some question about the best way to kill slugs got 234 answers, then show it to me and make me care. I’m sick of  reading about how many employees AirBnB has (by the way, it’s 130).