SafePasswd.com Update 11/19

I updated SafePasswd.com tonight. Updates include better generation of memorable passwords, easier to handle length selector, and better quality bar. Most of the changes were to algorithms, rather than major features. Hopefully the quality of passwords generated is now slightly improved. It’s not always about the big things, sometimes it’s the smaller refinements that need to be done.

20,000 Passwords Analyzed

An interesting perspective on 20,000 Passwords. As noted in the comments, the data collection skews the results a bit, since most people who fall for phishing scams aren’t knowledgeable enough to know a good password form a bad password.

But it’s possible to generate a safe password with ease even if your not a technically inclined ;-).

SafePasswd.com Update

So it’s been about 10 days since I launched SafePasswd.com, and so far the response has been extremely positive. I made a few small changes to the system to improve the quality of “memorable” passwords, and I have now made the default length of passwords a minimum of 10, up from 8 (it’s actually is random between 10 and 14). A few slight UI fixes were also made.

Overall, very good first week. Thanks to those who gave feedback.

The Need For Browser Testing

Ok, I’ve done a fair amount of work over the years in browser compatibility. From web development work, to writing the reporter tool. I’m well acquainted with the stress of testing your beautiful site against a dozen different browsers/versions/platforms. I just recently did so with SafePasswd.com which I launched the other day. IE5,5.5,6, Firefox 1.0, 1.5, and soon 3.0 (2.0 is pretty similar to 1.5, so likely not much trouble), Safari, Opera! It’s a drag.

What drives me nuts is I can’t keep parallel versions of IE on 1 system to test against. I want to test against IE 5.5, 6, and 7 (RC1 as of this post). This hack lets you have IE 7 as a standalone, but the broken stuff is rather critical.

So what does a web developer do?

  • Keep several computers lying around with different browser versions? This seems costly?
  • VMWare (or Parallels) with different configurations? This too is rather expensive, as Windows licenses aren’t just given out like Linux.
  • Install/uninstall each and every time? That’s excessively time consuming to test between just IE6 and IE 7.

I know there are some services out there that will give you screenshots of your page, but that doesn’t work for things like JavaScript functionality testing, and debugging. So those are effectively worthless for most purposes.

So what is the recommended approach to testing between IE versions? I haven’t been able to find any recommendation from Microsoft on the topic (if anyone knows of one, please point it out). Perhaps it’s a topic for the guys over at the IE Blog?

The same question goes for Safari? How can I a Mac OS X Tiger guy, test how my apps ran with whatever version of Safari someone with Mac OS X 10.3.6 would have? Or 10.2 (though to be honest, I have a 10.2 machine around)? Perhaps it’s a topic for their blog as well?

For the record, all Firefox releases can be found here, Opera can be found here.

So what’s the “best practice”? So far it seems the jury is out on IE and Safari. Firefox and Opera are a pain, but easily done. So what do you do?

Introducing SafePasswd.com

SafePasswd

I don’t think I’ll ever get tired of tinkering. Way back when, I wrote a script to help me generate cool random passwords. I thought people wouldn’t mind one that didn’t suck… so recently I got to work on that.

The site is still in beta, and quite a few things aren’t quite done yet, and some things are still being tested out. There will be advanced options to further customize password creation, as well as some API’s for those who want to quickly plug in automated password generation into their “Web 2.0″ applications. Those will be coming in the near future.

So check it out, and let me know what you think. It’s designed to be simple and helpful. It’s not Google and it’s not Digg. Just trying to make online life a little simpler.