@ericjperry - the PR to better support array parameter values started causing some tests failures. I dove in a bit & saw you were checking if they dialect is actually postgres & if so doing something different, otherwise taking the old code-path so as to not change how the other dialects work when creating parameter values. Not sure 100% where the error is, but something is causing some tests to fail.

I definitely like the direction you went in, but it might be good to instead of saying if(this._myClass == Postgres) actually break out the parameter converter into each dialect and override the postgres one? or something like that - basically all the dialects are calling the postgres dialect to get parameter values, which seems weird.

However you decide to approach it I'll be fine with merging it if all the tests pass. I don't know why the travis tests passed on your PR because they definitely break when I run them locally...do you have time to take a look & see if you can get the tests to pass?

Again, I really appreciate the effort - there was some failure of automated process that didn't catch this until I merged into master, so I gotta pop it back off for now. Until then I'm gonna back your changes out with this PR.

Interesting...I'll have to look into that. The reason I was doing the _myClass check is because all of the dialects inherit the Postgres class (which was not my doing) and I only want this code to run for Postgres. I can't override it for Postgres specifically, because it is already in the Postgres dialect file. I will double check and see if there's a cleaner way to take care of this and/or fix the tests.

I feel you on the not my doing thing - I haven't written any of the dialects except the postgres one. Such is the glory and trial of open source!