RAMI said:
Yes, on individual tracks, not usually on the master bus.
Not usually, if you don't know what you're doing and you're unsure. If you're unsure, rather don't.
My opinions on this master bus EQ/Comp/FX hootnanny...
1. Compression on the master bus is a completely common technique that has been used for years. However, it's not just a matter of slapping anything on there. Usually a dedicated, purpose built BUS COMPRESSOR is used, like the SSL G384, or a Smart C2, or a Crane Song STC-8, or the API 2500, etc (many of the great bus compressors, however, are variations on the SSL design). Of course, those are very expensive and they're outboard. Pros use them to get a certain sound that you've heard on a million albums and that's what their purpose is. Also, they will strap them on at THE BEGINNING of the mix and MIX INTO them. In other words, they'll start with a setting they know they like, go ahead and mix, and then push the mix into the threshold a bit to get the desired gain reduction and effect. Timing is then adjusted (attack and release). Most guys that I know use this technique will only push it to about 2dB gain reduction or so because that's generally where you get a bit of glue out of the comp. This is the common approach but certainly not the only one. A little more, a little less,...it all depends on the program material.
Now, plugins. Do plugins do the same thing? Probably not. They don't have the same non-linearities (even a lot of the good emus). However, I myself use a bus compressor plugin when I need to and I treat it like it's analogue. I confess. I like the results. I also only use it lightly and play around with the timing in order to get the right amount of squish with the right amount of push. Bootsy's Density MkIII is a killer plugin for this and I highly recommend it.
2. EQ is also a completely common technique. It's been used for yonks. This is the whole reason why the Pultec EQP1a and MEQ5 were invented and why they're called PROGRAM EQUALIZERS. However, obviously they impart a bit of mojo that many engineers have found pleasing. They sound good just plugged in without any changes. EQ is a little bit of a different beast because it's usually to correct a spectral issue. GENERALLY it's a good idea to address the top and low end in a mix instead of fiddling too much with the mids. As mentioned before, the midrange should have particular attention during the mix phase (and also the recording phase if possible). However, saying that, if a mix is too dull, turn up the high shelf a bit. If it's too bassy, turn down the low shelf. It's really that simple. If there's too much 250 overall, turn it down a bit. Simple. Just know that you're not going to fix huge spectral issues by applying a global process. Not going to happen. Mix bus processing is usually the icing on the cake and used to finish off an otherwise GOOD MIX.
3. FX can be applied creatively. Ever heard those 80's mixes where whole bridges were flanged? That's creativity and there's no rules for that. It's got fit the song, though, of course, so think deep. If it doesn't need it, don't use it. Simple.
4. Overall, blanket processing is not a good idea if there are deeper problems at the track level. Check those first. Also, anything done as a matter of course in audio is also a bad idea.
Hope that helps. My 2c.
Cheers
) can make for hell of a big diff in how all this is going to pan out. As well, likely make for some fairly significant differences in how you perceive and react to this mix' as it progresses. 
