I think 80 is overkill. I'd say anything above 60 is secure enough for most passwords. One of my main password (consisting of only lowercase letters) has an entropy of 78.7 bits, which is good enough to safely guard sensitive information like financial records, but then again, it's 21 characters long :)
Matters not. You still have no idea which one of my main passwords, how many they are nor where its used. Also, this particular password is non-critical which is why I used it as an example, but you're welcome to try and hack my accounts ;)
If you are adding to the 8 char password you aren't creating unique passwords. They share the initial 8 chars with each other. Still, this method is much more secure then using password.ly.
This merely gives the user a false sense of security. What if a hacker get hold of your master password? I rather advocate the use of different high-entropy passwords divided in security tiers. A junk password for places you don't care about or fully trust, a generic password to use on trusted services and secure passwords for crucial services.
People would have to gain access to my machine, plus figure out my master password to compromise me. But I gain unique, random, 24-32 character long passwords for all my log ins. I think the benefit greatly outweighs the risk.
Not related to the article. Seems Bret doesn't care much about us Windows users. It's more or less impossible to scroll with your scroll-wheel with that custom scroller. Personally I think he's trying too hard with worrydream.com. I feel it's hard to navigate, with bad responsiveness and clunky, even on Chrome. I believe the bad UX comes from the skeuomorphic design. I've never seen such design work good for a software UI.
That aside, Bret Victor is a genius and I love most of is work. He's a great inspiration to me, and should be to anyone involved in any form of software development.
This is more of a show of expression that creating a usable site.
Again this reminds me of the Flash days, where people were doing things because they could rather than because they should. CSS text rotation is a great example of this.
Still I think sites like this are a great playground, and show what can be done.
As others have said most flash sites of this kind did the exact same hack and provided an HTML fallback that used JS to replace the fallback div with the correct Object/embed for the browser.
Back in the days , making a flash site meant making an accessible fallback as well, and i always designed a pure html fallback for each full flash website i made ,Heavy javascript sites are not accessible by magic.
The irony is this site would have been smoother using flash. On my mack , this webiste is pretty slow and hard to use.
I'm not sure if you're having the same experience, then. Scrolling is almost impossible for me. The only way is if I manually drag the indicator bar, and even then it's clunky.
I'm just going to assume he designed it for a different system that doesn't have these problems. On my end (Opera/Win7), it really does not show "what can be done" :)
What I mean is there's a difference between coding a site that has so much eyecandy the designer wrecks usability (which happened in the old Flash days) and coding a showcase site that looks fine (but flashy and perhaps not entirely intuitive) on some systems, but it a real pain on others. The latter didn't happen nearly as much for Flash because the plugin's the same everywhere, but I think it's more excusable than the former for a showcase website like this (even though it's too bad I can't enjoy the demo--at all).
I'm on Ubuntu and scrolling with the mousewheel feels like I'm scrolling through tar. A lot of these sites with fancy scrolling hooks seem to do the same, really annoying.
I think it's fine to hijack the default scrollbars so long as all you change is their appearance. If you can't do it without changing the way scrolling via the wheel feels in any way, you need to trash it and use the default scroll bars.
Maybe it's designed for people with touchscreens? On Windows, in Chrome, the scrollwheel sucks, but if you click the page and move the mouse up or down to scroll (dragging), it's much more responsive.
> I believe the bad UX comes from the skeuomorphic design
I can't see a single thing that has its function derived from a real-life object. Maybe I didn't dig deep enough this time, but why not just call it bloat? (Which I think it is - even though it's really butter smooth on Safari)
How about... The diapositives, motion blur, scratched wooden surface background and shadowed borders (simulated lighting)? Those are the main design elements of the site, and are all skeuomorphic.
Nothing of this has or had anything to do with functionality on either side of the screen. The motion blur is not even tied to motion on the site (which may or may not be a good thing).
Most "3D hacker interfaces" in Hollywood movies aren't skeuomorphic, yet they blink and rotate and blur and zoom all the time. Same for many Asian TV channels. In what way is worrydream.com bad that doesn't apply to these elements, and why is it caused by textures from real life? Aren't they just all confusing because they are doing too much at once?
I think he was referring to Scroll combined with blurred preview of item your hovering over. I could be wrong but it was designed to give a feel of driving in your car through a street full of houses or something.
I thought the point was to use very low-res JPEG files and load the full versions later. The motion blur is still visible when the screenshots are not moving too.
