Strong passwords are often hard to remember. That seems to be because the very patterns which allow us to form memories are also the things which hackers can get to grips with most easily.
If you use any simple algorithm to generate a password, for example, then that is a source of insecurity.
Using a slightly more complicated algorithm, which generates a unique password to each site, seems like a better idea.
URLs are, by definition, unique to websites, so today’s invention relies on that.
Say I want to login to https://iotd.patrickandrews.com/wp-admin. One way would be to always use a password consisting of the 2nd, 3rd, 4th, 10th, characters ie otdc…Logging into http://www.bbc.co.uk/news/education would use password bccw, etc.
This removes any need to memorise anything other than a single string of digits. One’s browser could easily remember the algorithm and when it loses the information, as mine often does, it would be easy to reproduce manually, as required.
I’d take it one step further, because I have such a poor memory. Instead of 2,3,4,10 etc, I might choose a sequence to generate my string of digits, such as ‘primes minus 2’ ie 3-2=1st, 5-2=3rd, 7-2=5th, 11-2=9th, etc or even some variant on the logistic map over integers (which is already used to generate pseudorandom bit sequences).