Thursday, December 21, 2006

Yahoo! Sign-in Seal - what exactly does this accomplish?

I'm no security expert here, but I like to think things through. Yahoo! has a feature which is, it seems, a few months old. But it's new to me. It allows you to create a custom image - here I've made one that says "yahoo sucks!" - and every time you access a Yahoo! login screen, you see the image and know it's legit.

It took me, as a semi-regular Yahoo! user, a few months to even notice the little plug for this feature on the Yahoo! login page. I don't often look around at the various graphics on login pages, since they're usually cluttered with ads and help for users without accounts. Now that I've set a witty statement as my sign-in seal, I may notice it more. But if it didn't show up, or an error displayed where the sign-in seal used to be saying "sign-in seal temporarily unavailable", would I suspect a phishing scam? Would most users?

Yahoo! uses some interesting methods to store the reference to the image. I was at first concerned that simply deleting cookies or the browser cache would result in a broken seal, but Yahoo! took precautions.

But Yahoo! doesn't seem certain about the reliability of their security feature:

What if I don't see my sign-in seal?

You could be on a fraudulent site, but there might be other reasons why you can't see it. For example, someone else using your computer may have deleted or changed your seal, your cookies or files on your computer may have been deleted, or you're using a partner or international Yahoo! site (like BT Yahoo! or Yahoo! India). To be safe, look for these other clues to make sure you're on a genuine Yahoo! sign-in screen.

Yahoo! is too quick to list other explanations to the missing sign-in seal. If a user has taken the time to look up Yahoo!'s help information about their missing seal, the company could at least provide them with a link to some more information on phishing and how to avoid being scammed. This brief explanation contains no links, nothing to do other than "look for these other clues" and then log in and hope for the best.

How much phishing has this sign-in seal prevented? Probably not much, if any at all. Quite telling is the little pop-up message (shown above) asking if that's your sign-in seal. Phishing sites won't bother to put that little pop-up message in, and that's the only time when users will need to see that message.

Sunday, December 17, 2006

Canandaigua, NY paper covers Wegmans protest

Great article in the Messenger Post:

Katie Barber shivered in the stiff wind, but her sign, which read “Wegmans, Please Go Cage-Free,” didn’t tremble. The CA junior and about five of her fellow students protested conditions at Wegmans’ Wolcott egg farm.

“Their egg farm is very cruel,” the 17-year-old said. “Their chickens aren’t being cared for properly.”

Barber said Wegmans’ chickens live out their lives in cages with floors no larger than a common office file folder. Once they’re no longer capable of giving eggs, they’re slaughtered and sold.

Wednesday, November 29, 2006

Learn What American Accent You Have

What American accent do you have?
Your Result: The Inland North

You may think you speak "Standard English straight out of the dictionary" but when you step away from the Great Lakes you get asked annoying questions like "Are you from Wisconsin?" or "Are you from Chicago?" Chances are you call carbonated drinks "pop."

The Northeast
The Midland
The South
North Central
The West
What American accent do you have?
Take More Quizzes

I'm usually not one to take online quizzes, but regional accents interest me. It's amazing how accurate this quiz is in my case.

Friday, November 17, 2006

Malkin: A Student Baits the Taser?

Is Michelle Malkin serious? She has entered into the realm of satire. If an individual refuses to show identification, does she really think shooting them repeatedly with a taser is the appropriate response? At least we can agree that tasering them after they've been handcuffed is excessive, right? Please?

From what I've read, even the police agree that this man made no threats of violence and was nonviolently resisting after they tasered him. What sort of scary world would Michelle Malkin prefer to live in? Doesn't tasering people who refuse to show ID make the world less safe?

Then a police officer named Joel writes to Michelle, and it only gets more bizarre. He claims that the problem here was California-style political correctness. I guess that in liberal P.C. society, we're now politely tasering people who resist showing ID, who knew? According to Joel, it's better to "hit a person with your nightclub one time as hard as you can in the leg" instead of resorting to the gentle taser approach.

Tasers are not always non-lethal, and become much more dangerous when they're used repeatedly on someone. Tasering a person for a few seconds may be defensible in situations where they appear to pose a threat - tasering a nonviolent protester multiple times while they are physically restrained is indefensible.

Thursday, November 16, 2006

The Woman of Super Mario World

Via ffwd, Serious Games Source has an article by Gonzalo Frasca about Super Princess Peach
and the gender stereotypes that Nintendo reinforces in the game. It's less of a manifesto and more of an attempt to take the game more seriously, which appears to be mission of the site.

The most interesting part of the article is the part where Frasca compares the Japanese version of the game's advertisement to the American version. The contrasts between our two cultures are quite stark.

But this was strange to read:

In Mario’s universe, Princess Peach - formerly known as Princess Toadstool in Western countries until 1996 - has always played the damsel in distress role. Until now, all she did was wait until Mario showed up and rescued her.

My favorite playable character in Super Mario Brothers 2 was always Princess Toadstool. When she jumped, she seemed to defy gravity. I preferred her everywhere but the ice world. Of course, who knows if she was really capable of all that - we find out at the end of the game that it was all in Mario's head.

So I don't know how Frasca could have forgotten about Super Mario 2. Back then, the Princess was a much less emotional character, probably due more to 8-bit sprite limitations than anything else.

Taking video games seriously is a good idea, I'm sure, because video games play a major role in society. But I think you'd want to discuss the effects games actually have on attitudes and habits, rather than limiting the discussion to subject matter. And incidentally, any truly serious discussion about gender in Mario World would include Birdo. I remember learning, as a child, about Transgenderism through the Super Mario 2 manual.

Thoughts on the Wegmans Boycott

Many friends of mine quite naturally boycott Wegmans out of personal reasons. They can't stomach regularly handing hundreds of dollars of cash to a company "whose influence," as one corrections officer put it, "is the reason you went to jail."

A boycott of all Wegmans products is more of a symbolic move that an action intended to bring the company to its knees. I can completely understand why people - after seeing what Wegmans does to animals when they think nobody's watching - would want to withdraw their economic support from the operation.

Compassionate Consumers isn't calling for a boycott, but we support people who do. Whether individuals or families decide to buy cage-free eggs, boycott all animal products, or refuse to shop at Wegmans, they're making an ethical choice that should be admired.

Wegmans has a "Nature's Marketplace" because they make more money off of those products than they do conventional products, not out of their love for vegetarians. If cage-free eggs are considered "Nature's Marketplace" products, then the Wegmans Cruelty campaign has increased the number of natural shoppers more than it could ever take away.

That's one of the biggest problems with the campaign - cage free eggs only cost 5 to 10 cents more per carton to produce. Companies that sell exclusively cage-free eggs often charge little or nothing more for those eggs than companies that sell battery cage eggs. If Wegmans simply did the right thing and stopped selling battery cage eggs, they would lose out on their market for "upscale" cage-free eggs. The Wegmans Cruelty campaign currently serves as free advertising for these pricey eggs.

Anyone who's been to the egg section of their Wegmans will probably be amazed at the number of cage-free eggs available. This represents chickens living much better lives than they would at Wegmans Egg Farm. But this also probably represents more money for Wegmans and less incentive to change.

It's hard to know what's right, so we encourage any positive step that people take.

Wednesday, November 15, 2006

The Brave New World of Software Icons

Via Daring Fireball (which uses a nice, simple icon in its design), Craig Hockenberry has a post up at Iconfactory comparing raster (pixel-based) versus vector (shape-based) images in the context of the needs of software developers. He makes the argument that raster is the way to go for a few reasons, but the argument that caught my eye was that vector images are so much larger than rasters, which brings my own understanding of the world completely into question.

Size: Today's photorealistic icons require a lot of vectors. More than you may realize. Unless you're dealing with simple line art, effects such as gradients, shadows, and highlights result in enormous files. As an example, compare this 512×512 pixel PNG file of the CandyBar icon with a PDF file containing the same image. The PNG file is about 100 KB while its PDF counterpart is a whopping 3 MB. Consider a five icon toolbar with PNG files versus a toolbar with PDF files—500 KB versus 15 MB. Your ISP will love you and your PDF icons!

Which may be true for icons like Hockenberry's example Candybar icon. The icon is so full of blends and dropshadows that it looks more like a stylized 3D rendering than a 2D illustration. Don't blame Iconfactory, though - that's what the vast majority of icons in modern operating systems and software look like. Hockenberry quite usefully links to the PNG and PDF files of the same image, so it's time to download a vector illustration that's about the size of a three-minute MP3.

Looking into the icon PDF

Call me old fashioned, but 3 MB does not sound like the proper size of a vector illustration, no matter how photorealistic. Take a second to imagine using this icon in a downloadable PDF of your instruction manual, on every page, for however many pages. You'd run into some problems fast.

I opened up the PDF in Acrobat to see what was going on. First, under Document Properties we learn that the file was made with Adobe Illustrator CS3, remarkable not only in that the software hasn't been released yet, but also because in my experience, various overhead in Illustrator can result in much larger files.

As you may know, a PDF file can contain either raster or vector elements, or a combination of both. Let's see what this file is made of. There's an underutilized but useful tool in Acrobat under Advanced -> PDF Optimizer -> Audit Space Usage (and I notice that Christopher Lloyd, the blogger and not the actor, has done the same thing, but for some reason I don't quite understand to a different file which he saved out). When I ran the audit on the Candybar file, here were the results:

At least 89.47% of this PDF is comprised of raster graphics. Vector graphics are listed primarily under the "content streams" category, but also show up in "shading information" and some of the other categories depending on their complexity. But here we have a file that is arguably far more raster than vector. Some of the raster objects may have been complex vector elements in Illustrator, but most of these rasters are probably raster effects that Illustrator renders on screen and then to the PDF when it's saved out.

To learn a little more, I opened the PDF up in Illustrator CS2 to see its underlying structure. Here I have the Links pallet open to demonstrate the number of images in the document:

I counted over 100 images in the PDF, from the large drop shadow behind the candy bar, to tiny round purple blobs, to graphics that can't be seen because they're behind the candy wrapper. Apparently, Illustrator is even ignoring some of the images - Stephen Deken found 281 rasterized images using command line tools that are out of my league.

Are all these images really necessary? With just a couple minutes of work I was able to remove the non-essential images and come up with a 248 KB PDF that looks almost (but not quite) as good as the original and opens much faster in Preview than the 3 MB version.

Now 248 KB is still pretty big, and with some redesign we could get it to be a lot smaller. You could squeeze a simple flat color illustration, or one with basic blends, down to 10 KB, and those 10 kilobytes will look great at any resolution and scale factor. But a modern icon designer would probably find and icon like that completely unlickable.

Modern icon design - more raster than vector to begin with

Just look at the dock in Mac OS X, and you see a world of glassy drop shadowed lickable icons. These have become common on the Mac over the past few years, and with Microsoft finally catching up with a new Windows interface the world will only get more lickable. These icons are so complex that it may very well be impossible to create vector-based resolution-independent versions that work well with modern computer hardware and internet connections.

But this isn't the fault of the inherent shortcomings of vector graphics. This is due in part to the limitations of the PDF format, which for PDF interpreters in current use requires that elements like drop shadows and complex blends and gradient meshes become raster graphics instead of vector data. When we introduce the problems of raster graphics into our ultra-high-resolution vector graphics, that's when we get 3 MB icons.

The other problems with vectors in resolution independence that Hockenberry mentions - speed and appearance (when they're scaled down) - are mitigated by reducing our reliance on raster graphics in icon design. Speed will be much less of an issue with simpler icons, and icons with less bells and whistles are naturally clearer at smaller sizes.

Hockenberry links to a post by Sven-S. Porst which discusses icon resolution. Porst has a point that pixel-by-pixel editing can bring out a lot of detail that scaled-down vector graphics can't, but I wonder how many modern icon designers really give the ultra-small version of their icon much thought. I don't see much evidence of that on OS X. Porst has hope that icons will end up much simpler if vector graphics are part of the process in icon design, and I second that. Unfortunately, thanks to all of Illustrator's features, vector graphics are very rarely just vector graphics.

And wait - we're forgetting how commonplace vectors are

There's one part of computer interface graphics that have just about completely transitioned over to vectors. Remember the days of bitmapped fonts? Long gone, replaced by scalable vector fonts. The same data is used whether you view your font on screen or on a sheet of paper. Deken, in his post, makes a point about how efficiently fonts are rendered quite well:

I suspect that just before you read this, your computer rasterized every letter in this essay using a vector-based font.

The transition to vector wasn't without its bumps. At small sizes, raster fonts looked much better than vector, so technologies like font hinting stepped in to improve readability of small vector fonts. And with sub-pixel rendering on LCD screens, new uses have been discovered for vectors that the original inventors and designers never thought of.

And thank goodness for the restrictions in place for vector font graphics. I can only begin to imagine the ugly multi-colored drop shadowed fonts we'd be designing if it wasn't for the single-color vector limitations that have long been in place.

So when you're hovering over your OS X raster dock icons and you see the name of your program pop up with a nice drop shadow behind it, know that at least part of what you're seeing is rasterized vector data.

Some current situations where vector computer interface graphics would rock

- Users zoom in on interface elements for accessibility reasons; these would scale up nicely if they were vectors
- Screenshots with live vector data would be useful in print media, where screenshots often look pixelated
- Screencasts and videos that show details in interfaces can benefit from vector images

These situations are limited, but I already am quite happy that at least the text part of computer graphics are vector-based. When I'm making videos that discuss web sites I find myself using an application like Adobe Acrobat and its web capture feature rather than a web browser to render the text of a web site because the vector text in video looks way clearer than a raster screenshot.

Some obscure uses are probably not enough for software developers to use vector icon graphics as opposed to raster, but I imagine as time goes by new needs for scalable computer interface graphics will arise. I don't think it will matter, for most purposes, whether these needs are filled with vector images or higher resolution raster images. Like raster vs. vector in print media, the choice will largely be according to the nature of the image or text.

And in conclusion

Icon designers once lived in a world of restrictions. Limited colors and limited sizes resulted in clunky bitmapped images that seemed like more of an afterthought. These icons, though, were at least descriptive when you had a large set of them on screen, and their file sizes were nice and small.

Modern hardware and operating systems have resulted in fancier icons, but at the cost of usability. Our computer screens have only gotten a couple inches bigger, on average, over the past couple decades. But when scaled down to small sizes, many modern icons look less like icons and more like colorful blurs with drop shadows on them. Some tricks - like the OS X dock's "Magnification" feature - help users figure out what they're clicking on, but it's hard to come up with a reason why users are happier with your shiny blue icon than they were with clearer icons from 5 or 10 years ago.

Some forms of design still have restrictions that may help the designer in the long run. Logo design is often constrained by the need to use less colors to save money on print runs. This may result in a cleaner looking, more professional logo that stands up well over time. Computer icon designers don't have the same kind of restrictions, and that may not be a good thing.

Support for vector graphics in computer user interfaces is a really neat idea, and we can imagine a whole slew of limited situations in the present and future where one could find them useful. In fact, we're already finding new reasons to love vector-based fonts.

But icons aren't ready to be displayed as vectors, in part because often by their very nature they are composed of raster images, in part because higher resolution raster images work just fine in most cases, in part because there's really no file format ready to meet our needs, and in part because a few more years will have to pass before icon designers decide that there's a bitter taste to all these lickable photorealistic icons.

Tuesday, November 14, 2006

Getting Started

I'm 27 now. It's time for me to stop keeping my opinions and concerns to myself. So I'm keeping this online journal about topics that interest me, some of which I feel should be important part of modern discourse but which are often swept to the side by our bizarre media, and others that I just like to think and write about. I also hope it serves to convey my thoughts of what's happening in my own life, most notably in the case of the People of New York State vs. Adam Durand.

I have a strange writing style. Perhaps I can hone it here. I don't write on a regular enough basis, I feel, to work on structure or develop concise arguments, so I'm interested in seeing what happens when I practice.