Anchors & Curators

Curated content on the internet is all too much like the employee-picks shelf at the local bookstore (or video store). A selection, maybe even an interestingly thematic one, but just a collection of existing items. And presented with no other information to help you decide, but what you would find if simply searching on your own. Consider instead what people with Curator on their business card actually do. They collect, investigate, verify, and organize. They then build exhibits of some items from the collection, in a manner that tells a story. They add their own content. Maps, brochures, labels. Often, it all works together to make the point, to reveal deeper truths about the collection, and to encourage the viewer to explore further on their own. Compare these two paintings, though 20 years apart, the similarities help point out how the artist has grown in the intervening time. Or, consider the classic format of the national nightly news. Behind the scenes, the anchor is the Head of News. He sets the tone for the whole department, and makes key decisions about what and how it will be presented. Then as the on-air anchor he gives an intro to the story, frames not just the basic facts but the reason we care. If there's a graphic over his shoulder, it may well be a map, so you know where this is taking place. And then he hands off to a reporter who was on site, or at least pretends to be based on footage retrieved otherwise. And maybe, if it's a multi-faceted story, then you go to another reporter who gives another point of view; after the on-the-spot report, the reaction in Washington.
All this is all in relation to a number of posts* and discussions I've read and had with people about the point of online media. And it's coming up since I am being paid during the day to actually develop an IA for a very large product catalog, and integrate existing (magazine and other) content and personal details into one experience. I have seen at least a little of most of the available experiences. I don't subscribe to them because I am un-thrilled with the consumable media available. I still read paper magazines, because even cut down to save on costs, They do a much better job than pretty much any digital version. Web and tablet versions seem to have lost the key point of sidebars or related stories, and just use paper paradigms instead of developing their own; or learning from education, museum information design and broadcast radio/TV media even.
If this seems like a subtle change, it's not. Go back to the examples, and look at how much internet content is served. I call most of that a subset of "portal theory." An article is reduced to a smaller version, on a category page. It can be reduced to a smaller-yet version, even a title-only for a higher level category, for cross-linking or for the (portal) home page. Instead, I am starting to think that I want to break that whole model, and use the intelligence of humans, not just to tag and categorize and group things, but to differently re-order, to choose what is presented and not just sort, and to not just crop but actually rewrite content to present it most relevantly to the context. No, I don't have diagrams, or mockups for you. Yet. Maybe later. The only impediment, as I see it, is people. Good, smart, dedicated people who can write have to be found, persuaded and paid. And too much focus is on technology solutions, and paying for software. We have the software and interaction nailed enough. Now it's time to bring people back into the job of presenting information. Give me a news aggregator with the voice of an anchor, and I'll listen. But it could also be a great opportunity. To give a new type of voice -- or a reason to hope for the future -- to writers, and to the whole profession of journalism. If not this, then something like it simply must happen. Not to preserve a dying business, not just to make money in a new market, but to keep the public informed so we can all make decisions about the way we live.
* Now that I look for the links, way too many seem to be centered around Khoi Vinh. And he's a good voice in design and interactive publishing. But I swear others are talking about it also. And some rather interesting ones I can't quote.

** When searching for the links I ran across this, which I swear I didn't read before. It also has good points about curated computing. Very different ones, but the over use (or I say, mis-use) of the term "curated" is covered pretty well in there, if circuituously.

No shopping carts

We did some shopping today, mostly for christmas. At the entrance to Toys "R" Us there's this sign that seems to say "No shopping carts." Being obviously untrue, I approached. Its like something out of a Tufte book. The repeated, easily identifiable item -- the shopping cart -- is by far the most bold item. The "no" slash is always thru the cart, and hardly touches the case that is bad about it. Much lighter, so you cannot see it at any distance, are the variable items, the child. And sometimes the parent, but bisecting the white circle, I didn't even notice her till I got home and opened the picture. I also like how the labels are impossibly tiny (the bottom of the large sign is above head height) impossible to scan, and way too wordy. And how some have consequences, like "this may cause the cart to tip over." But what will happen if I let my child push the cart? I don't know! Okay, National Treasure sucks, so I went ahead and did it they way I think it should have been. Note I kept most of their drawing style even. The cart is gray, as its iconic, and repeated often. The heavy prohibition slashed circle was removed in favor of the simple X. Let me put it over the prohibited part, implying a specific location or activity, instead of implying everything about the situation is wrong. The parent figure is only present when needed to show she's not paying attention; otherwise it might have been lost in the clutter. And, the titles were changed to be shorter, and the "Do not" part is on another line to make the instruction component scanable. Though now that I look at them, the capitalization is nonsensical; it scans well, but doesn't make sense. Hmm... There were two additional images, and a title block above, but I didn't get them in the photo. So, its still incomplete, sure. I believe one graphic was how to use it right. In this case, I like green circles to say "do it this way."

Why don't we use italics online, anyway?

At the risk of publishing something that relates directly to my work, Sprint has recently changed their branding. The type standards, currently print-based, say everything is in italics (well, oblique, but that's another post entirely).

My kneejerk reaction (to no avail in the first iteration) is "no italics online!" But then I had to prove it. And this was way harder than I thought it would be. It was sort of interesting trying to investigate such a deep-seated heuristic.

As it turns out, I never found the text of any study that proves it. There's reference to one, and that's good enough for me. Instead, I am relying on the "everyone says it" method, as well as proving it thru demonstrations of the grid effect. Oh, and one disclaimer. I didn't find all these links myself. The whole list is a straight up ripoff of those emailed to me by Fred Pilgrim, off to the south about 3 yards.

Overall, its just about web-resolution. Type traditionally is smoothly drawn or pressed; pen strokes or lead type create lines going any direction. Any irregularity from beat-up type, or paper texture was inconsequential at normal sizes (usually) and regardless had no direction, so italics with their angles suffered no discernability issues. In reproducable print, from the dawn of photo-lithography to today, the screen (whether angled, straight or stochiastic, square, round or oval) is of such high resolution that all shapes appear essentially straight. Except for very straight lines very close to (but not quite /at/) the screen angle, everything looks smooth and sharp.

Digital displays are generally around 80 ppi. Some are higher, some even nudging over 100, but generally not. If resolutions exceeding 200 ppi become commonplace, this will begin to become moot, but this is possibly decades away due to low-cost and high-longevity of current display devices if nothing else. Anyway, unlike print, the low resolution of type on the digital display means that angled items, like the vertical in italic type are rendered as a series of steps. Cleartype, Smoothtype and other technologies try to mitigate this with anti-aliasing, but its imperfect; the sharpness cannot be preserved with such technologies. Its just a limit of physics.

There are additional issues with readability (comprehension, fatigue) of italics or oblique faces in print, but they are not quite as severe as the online component so I am ignoring it for now.

Anyway, here's what a bunch of others say about this. Interesting quotes about italics are here directly, but follow the link for context and more useful info. The IE box model bug is a great one I had forgotten about. Seen a similar one with Verdana (I think) having a space/letterform issue with their Ws.

http://www.grc.nasa.gov/WWW/usability/textfontcss.html

Research has shown that people can read faster if words are written using mixed case rather than all upper case. Blocks of bold and italics text are more difficult to read. Black is easier to read than colored fonts.

http://www.jimbyrne.co.uk/show.php?contentid=53 Jim Byrne is the founder of the Guild of Accessible Web Designers and has been a Web accessibility specialist since 1996.

Avoid using italics for small text sizes: the problems of screen display of outline fonts has not entirely disappeared. Italized fonts look particularly bad at small sizes - as italics do not easy to render using a square pixel grid. If you must use italics, avoid using them for large blocks of text.

http://www.bindu-design.co.nz/pcnz.html

Italics don't work on current low resolution screens: Staying with the same page we see that the chief executive's message, is written in italics. Research has established that reading from computer screens is about 25 percent slower than reading from paper. Even users who don't know about this human-factor research, usually say they do not enjoy reading from a computer screen. Text in italics should be avoided on web sites, as italics are even harder to read on current low resolution (below 80 dpi) screens. I predict that the screen readability problems of today will be solved by the year 2008. By that time we will see screens of 300 dpi in common use.

http://www.smartisans.com/articles/web_writing.aspx

While the Stanford-Poynter study focused on reading news on the Web, the results were similar to various studies conducted by Jakob Nielson, Jarod Spool, and others in the mid-1990's... ...But don't use italics for emphasis because italics are difficult to read on a computer monitor. There are just not enough pixels to render italics clearly.

http://www.456bereastreet.com/archive/200501/ie_and_italics/ Interesting info about Internet Explorer & Italics

Ever seen IE/Win ruin a layout just because a few words are in italics? I have. Just the other day I was pulling my hair out over this bug...

Ha! CC numbers can be linked to issuer!

Back when I started with this whole web design thing I was able to find some scripts to take advantage of some of the check encoding built into credit card numbers. I think we even had a little checksum so we could tell if it was a validly formatted number, but most of all there was a system to the numbering, so we could tell issuer networks apart.

We had to provide this tag for whether the card is Visa, MC, AMEX or whatever to the processing bank, but we could determine that based on looking at the number. I knew this to be true from earlier in my life, so went looking and it was great.

Now, a few years after I get to Sprint I am told this no longer works. Though no one can show me proof, and I don't believe it fully, I'm told that due to the large number of cards issued, that scheme doesn't work so we'll need to ask the customer which card they are using. Its always lame to provide an extra step, but everyone else seems to have done it, so who am I to argue.

Well, its demonstrably not true:


I just recently bought something thru this store with a PayPal checkout scheme. On the left, you see they accept the usual assortment of cards. On the right, after entering my valid card number (that is not it!) it shows which one you have. No user input required. Neat.

I suppose its possible they are using Ajax to actually pre-process the card number, but for a bunch of reasons, I doubt that. I think its parsing the CC number itself on the client side.



Some other time I'll rant about how amazingly poorly the CVV2 code is implemented from a design, usability, comprehension and mostly security point of view. Note that mine was auto-filled by the browser. That's secure.

Misusing Mood Boards

It seems there is no design tool that cannot be misused. I don't recall calling them them "mood boards" in school, but I've used things like this over time. see hundreds of other examples Instead of describing this myself, I'll quote someone else:

...You create a bunch of logo ideas in Illustrator. You print them out on your fancy high-resolution color printer and then show the client. Although the print-outs are just rough examples of what a design could be, the inherent precision of your computer and printer will always make your work seem “finished.” No matter how rough, the client can’t help but think, “ok, what I see is what I’ll get.” So, instead of the broader concepts, the client will zero-in on the details. During early explorations, you can short circuit time-wasting debate about blue versus purple or Times New Roman versus Comic Sans by using mood boards... ...So what do they look like? Just think back to kindergarten. Like your first art projects, mood boards are collages that center around an idea, concept, or feeling. You can create them on the computer, but it’s more fun to literally to gather and paste photos, magazine cut-outs, and objects on a board... Read the whole article

So, I'm getting presented this nascent new... thing here at my day job, and the agency which will remain nameless says they came up with the final message style by running a half dozen mood boards in front of some focus groups. I want to see the mood boards, and finally realize I am seeing them. Obviously I cannot show them, so go to a magazine, pick any shiny ad, and tear it out. That's it. Not clustered with others, not cut apart, just one concept ad. Huh? That's not a moodboard. In fact, its not a valid test of any sort, even if the focus group is well-run. And frankly, the high concepts aren't that reflected in each of the comps. So? Why do I care? Well, aside from being bad process, its a clear way to get nothing out of it. At this point we might as well just have let the designers at the agency pick their favorite.

Violating Communications Paradigms is Not Just for the Internet

Take the much misused button. On the web its really for submitting data. At Sprint we even made standards to this effect some time back. We had to because it was so often misused as a "more important" link. Its big and (at the time) red, so it must be more important than a simple colored, underlined word. The current Sprint home page has a button for "all phones" that does nothing more than the links above it, except be big and yellow. Other prominent examples abound.

I've always given particular grief to interactive systems due to their immaturity about such issues. Newspapers don't mess up this sort of thing. Road signs, regardless of country, are often another good one. In the US orange, for example, means construction.



And nothing else. Reserving it for construction means that any driver can pay special attention to such signs and know there is an exception message contained therein. If you've never been here, there's a temporary, probably dangerous situation. If you are used to this place, disregard all you know and start looking at the signs again because something probably changed (like, the road is closed). This is all well reinforced by using the same orange for pylons, with white stripes for barricades, etc. I even saw a team of guys paving on the way to work this morning, everyone one of them wearing construction-orange T-shirts.

Which makes this annoying:



This is near my house, but I saw scads of them throughout extreme northern California. Something the highway department thinks is just "really important" at the time gets a sign. Often, its just not that important. Like, 6% grade ahead. Which is perfectly well served with a yellow sign everywhere else in the world. Yellow is a fine color for warning signs.

What I think of this is pretty much what I think of over-using buttons. If there is a problem that legitimately is solved by using an orange sign, you need to solve it some other way instead. Aside from misleading everyone, eventually the value of the orange sign will be diluted, and comprehension or urgency failures will occur. In traffic signage, this could have much more dire consequences than any website.