I wrote a plugin for dealing with logins a few years ago at a company
I worked for. One thing I had was the ability of an admin account to
login as another user so you could see that users views and manage
their accounts. Perhaps I should have used a standard rails plugin,
but I wasn’t sure how to accomplish the same thing that way and would
have had to figure out the inns and outs of the plugin. I had to write
my own code, but if I needed to make changes to do various odd things
in the future I was somewhat more familiar with the plugin. It also
got me to get a feel for how to write a plugin so it seemed like a
good thing for me in a way, yet I still suspect some people may think
it’s not such a good thing to write your own authentication (although
alot of rails books have examples of sessions and so on with login
etc)
I need to revisit this however as I am not sure if I mention it in a
job interview if it would be considered a plus that I wrote a plugin
or a negative that I didn’t use one of the standard ones.
good thing for me in a way, yet I still suspect some people may think
it’s not such a good thing to write your own authentication (although
alot of rails books have examples of sessions and so on with login
etc)
I need to revisit this however as I am not sure if I mention it in a
job interview if it would be considered a plus that I wrote a plugin
or a negative that I didn’t use one of the standard ones.
As may be obvious from my tag line, I am there too. I see a lot of
positions that require knowledge of Devise (and CanCan). So, I suggest
you use Devise and fork it where you need additional features (and
send Pull Requests if the feature could be generally useful). I think
that is the strongest argument: use what is there and continue to
build on top. Not the easiest solution, but surely the strongest one.