Machinist is almost perfect but damage sucks

[Shurrikhan]

It also had the most complex or demanding rotations on the spot, it had meaningful support that wasn't just about pressing a button every 2 minutes.

It's never had the most complex or demanding of rotations relative to other jobs in the same time period. HW Monk, Dragoon, BLM, and Ninja, etc., had wildly more kit complexity against striking dummies than Bard or MCH had at that time, with the latter's saving grace for complexity largely being those mobility constraints, clunky and ill-thought-out WM and GB were. Similarly, even StB-launch Warrior arguably had more complexity going for it rotationally than StB Bard. (MCH had complexity, admittedly, but given the sorry state of it, it's difficult to tell how much of that was accidental so I'm not sure I can rightly use it as a benchmark.)

The rest, though, I'd already agreed on in the post before (re-quoted below). Was this meant to be in disagreement, or just confirming common ground, or...? Sorry, just a bit confused by the inclusion here.

It had support, then support and casting, then support and contextual actions available through rotation with unique forms of complexity...

I'd also argue that positioning within the design of ARR/HW and the current twitchy bullshit we're being served today has literally nothing in common.

Even as far back as HW still had a more in common than not in terms of movement requirements relative to mobility available. When Accuracy was still a thing, you were incentivized to stay behind the enemy to free up a large amount of Accuracy to be replaced with Crit/Det (or, at slightly higher accuracy, to maintain RoD when unable to rear attack), which kept you on your toes a bit more. Similarly, while they were further "move slightly, wait, move slightly, wait", less frequent but longer movement requirements from knockbacks, safezones, proximity AoEs, etc., still meant that a saved Straighter Shot (and Faint as a fallback) or readied MCH combo action were pretty damn relevant. You created the "twitches" yourself if you had cast times or positionals, but you still just as often pushed the limits of how much you could move without damage loss.

While I mentioned that "support" was something Rangers had unique to them, that was probably an overstatement if speaking in terms of gameplay rather than mere capacity. Our Ranger support, as it was then, mostly amounted to indirect gameplay sabotage for others ("Bring a Ranger to be allowed to actually use your toolkit for more than 4 minutes at a time"), akin to Lucid Dreaming now or Invigorate before but done as a compositional check (no Ranger, no sustainable play, which was frankly a shit design). Its benefit was mostly just in that Bard had to remember to sync Ballad/Paeon's exit to just before a DoT refresh (and not to actually try to use as soon or often as desirable for the support itself, since it came at 1.2 GCDs of cost just to start and therefore was a loss to start any earlier than absolutely necessary if you'd have to stop it early), while MCH saw no real gameplay variance from it (since the controls were oGCD and their costs were independent of the MCH's own actions, merely costing it its second auto-attack).

Even when we had the extra control of no CDs, it was still basically pre-pull Foes until nearly oom, then Army's once absolutely necessary and until nearly oom, with faint reserve for Ballad only if it looked like it might turn necessary, basically exactly as done via the Stormblood CDs but with a bit less MP regen typically produced (as it could previously be traded for extra magic damage). Battle Voice was more interesting then, definitely, but otherwise... it was more a matter of a better glimpse and at the compelling idea lost behind the implementation than actual difference in practice.

Accordingly, I have to wonder what, more precisely, you'd want to see changed from the current model that would newly unlock a Ranger that would appeal to you. Having played them from 1.x onward, I can't say the actual execution at any point gives us much to go off of.

I feel like we'd agree on the vibe of what's generally a better direction for Ranger job design, certainly. But to me, it'd probably(?) take almost as much revision to undermechanics (i.e., game fundamentals encounter mechanics could then manipulate) as to encounter design, with most things short of that feeling pretty similar between HW, so I'm curious what the... milestones, so to speak, in getting to a more worthwhile context would be for you.

I'd be all for the cues being more spread out while still being just as hard to reach, as they were back in earlier days, or even having them be a bit softer or by other means more forgiving (say, linearly-diminished proximity AoEs, etc., rather than clearly indicated bimodal AoEs slicing up the arena ad nauseum), but making those feel compelling would likely require some deal of constraint to play around.

Anyways, to circle back to the initial disagreement, imo, if not positionals (a sector of the enemy sighted, which you'd then have X seconds to follow up on with a spender from that angle) or linear AoEs to manually line up (cleaving an additional dummy portion of the enemy via skillshot), etc., then, yeah, cast-times would be a damn good tool by which to add interest especially because of the extent of dodging typically required (not just in the last couple expansions, but even --speaking relative to available mobility-- as far back as HW).

[Obviously, compelling branching choices go a long way, too, but even those would amount to contextual difficulty just like constraints to positioning (positionals, cast-times, etc.), as they'd otherwise just be a more convoluted but still-rote rotation.]