What happened to job inter-play? Was it any good?

[Shurrikhan]

Reading a lot of the replies, all I can say is welcome to how schizophrenic this community can be. They keep raging against the current blandness we're getting, but immediately revert into saying that everything was bad or frustrating back then as well.

Either way, what we had had some kind of balance between collective and individual. It's never been a fully team synergistic game at all, let's not kid ourselves. And the amount of scripted things was still staggering, even if a lot less so since the whole had a lot more moving parts that could go wrong beyond just fight mechanics.

That's the thing though: most of the past attempts at inter-role interaction pointed out as such... weren't actually interactive across roles except in terms of capacity (i.e., no more than Shroud of Saints or Second Wind did from the start).

In the vast majority of cases, it was either effectively just an in-role CD for which you could more directly and impactfully shunt loss onto someone else (by starving them of resource or crippling their and thereby wasting party output to support your negligence in a way redundant with and less interesting than, say, your ability to greed mechanics), or they simply meant that you now had a group mitigation skill available as DPS (to use of schedule because it's not single-target and is therefore noncontextual/nonreactive anyways). They were uninteresting forms of "free" resource around which the game was tuned anyways (ultimately meaning that you're simply punished for not using them as near to on CD as can be leveraged, with most having no barrier to such thoughtless use for effective play).

________________

What was your favorite interplay interaction that existed back then (and may or may not still exist today)?

If I had to pick a single example: Clemency while healers were jailed (A3S, I think it was). ...Or DPS-kiting melee mobs in dungeons while speed-leveling.

More generally, though: Early on, CC, Mana Wall¹, Paladin absorbs/heals (even as small as they were), and tank flat (as compared to %DR) self-sustain if that would ever be considered cross-role. Such allowed DPS to bait, kite, and place enemies without disrupting the tank's gather, to prevent loss of positionals due to unfortunately-timed special attacks moving the tank at just the wrong moment OR to take pressure off the tank in emergencies, emergency recovery tools, etc. Later, Rescue² and maybe Mana Shift³.

¹ BLM's Mana Wall, before its removal, had the hilarious impact of being able either to fully immune a physical AoE or skillshot type boss attack... or to split a split-damage tankbuster with a tank. It was a short-CD, 100%-mitigation-against-a-single-hit personal defensive... with effectively extendable purpose.
² Fun for dungeon speed-running. Would have been more significant in raid-fights, too, had the game not by then already replaced a majority of displacing mechanics with a "Opt out of mechanic" button-press in the forms of Arm's Length and Surecast.
³ Mana Shift was nice in that it didn't produce additional resource out of thin air (except, perhaps, as casted by a BLM when it wouldn't delay their return to Astral Fire) and therefore wasn't tuned around, but could do a lot to alleviate emergencies. As it was single-target, it was also much more responsive.

Honorable Mentions:
If Casters had run out of MP only from using specifically cross-role skills (rez, AoE Phoenix heals, Titan tanking, etc.) and Mage's Ballad had been oGCD, I'd also have considered it as belonging in that list. Not Army's Paeon, though Mage's allowing for greater use of expensive burst tools would have been fine so long as non-casters also were to then use MP, though that would have then partly rendered Foe Requiem redundant even if it had affected all damage. Foe Requiem, if it had been oGCD, though, would at least have engaged enough with party output dynamics that, until this excessive fixation on precisely 2-minute cycles, it would have been a tool that actually feels supportive to use where Battle Voice and Radiant Finale, imo, do not (since it's up to everyone else to use their shit when you pop it, with you simply holding maybe once or twice per fight based on boss jumps). Similarly, Palisade could have been decent... if there were actually just the right number of emergencies throughout content that it could be relied on while also requiring it to be reactive; too many, and you find alternate solutions that render it redundant, while too few make it feel pointless.

Finally, enmity skills on non-tanks would have had at least a bit going for them... so long as their having any impact were no longer dependent first on enmity-multiplying tank stances (i.e., those Enmity stances being replaced with passive bonuses and non-tanks' Diversion skill instead giving that modifier briefly to said non-tanks) and tanks now-disgustingly-high passive eHP bonuses (which are mostly just a change to make non-tanks go squish even faster) weren't so high. As it stood, though, their use cases were limited to mechanics that would otherwise require very precise tank swaps on things that nonetheless wouldn't clobber non-tanks any harder than tanks, which would be beyond awkward to shoe-horn into any significant portion of fights. And even if we had all those things, we'd have been better off with just slightly more sophisticated enmity modification through positioning and slightly greater control over our moment-to-moment throughput (and therefore Enmity) dynamics through slightly increased bankability... rather than requiring a whole discrete selection of Enmity-increased skills. Especially so long as Enmity remains solely a way of determining who gets the auto-attacks and certain special attacks.

While the way CC diminishing returns work needed a rehaul and a few (sparing and situational but impactful) extra tools would have been appreciated, those did far more than "Esuna but better because oGCD" or "hit this to save your tank, assuming equal gear and skill, ~30s of damage loss (so long as everyone else with this ability also presses it)" or "use on CD so your casters are allowed a normal resource economy" ever did. Of those auxiliary "inter-role" skills, all but maybe 10% or 20%, depending on how broadly one defines them, were uninteractive, one-dimensional uninteresting tools. In short, bloat. They should have been trimmed down to those few that were capable of more and then started bending design philosophy slightly around those.