Speaking of which, if you're not going through the garden it means you're going through the front door. There are a few good reasons for tracking with compression. Savings in time, the difference between the sound of the hardware versus plugins, and especially the idea that if you record agressively and print your effects, you can tie the mix engineer's hands. If you know exactly how you want things to sound while tracking and work this way, you're painting things into a corner with a much smaller set of variables. So if you're going to track and mix, you're going to tie your own hands. I'm not sure this is a good way to learn compression. If you can hear exactly what you're doing and have the confidence that it's exactly what you want, that's fine. If not it's easy to screw it up and then it can't be undone.
If you're trying to learn compression I would track dry and run plugins. Start with the stock plugins in your DAW and then move on to some freebies. Play with them. Abuse them. Smash the vocals to oblivion. Make them distort. Make them pump. As you learn how these things sound your ears will become more sensitive to subtle (& probably more suitable) amounts of compression.
The plugin workflow isn't quite the same as with hardware but the end results can be very similar. Sometimes people will run 2 or 3 instances of a plugin in series, maybe with an EQ or 2 tucked in before or after. This can help to get the desired crush without a bunch of unwanted artifacts. Another popular thing to do is run a layer of volume automation before hitting the compressor. Being more conservative about it also allows you to run gain plugins or whatever you need to audition the compression and flip back and forth between "on" and "bypass", which has to be done at the same percieved volume level. Depending on the track and the goal, sometimes the best compression is none.