Moving Through Multiple Provinces

[licker]

Thank you Kristoffer.

I knew this already, but its nice to hear it again from time to time.

I'm not disappointed or surprised by your answer, its what you've said all along.

I don't see the real issues with bugs or other problems as some have been suggesting if this is implemented intelligently, however, I really don't understand it in terms of if you don't want to use it you won't have to, and if your forces get stuck you will find them anyway by hitting 'n', commanders cannot be issued an illegal movement command as it is, how this changes that I don't know, or see.

Think about it, set a move order to a distant province, and each turn the commander will issue a *legal* (obviously) movement order to get closer to that province. There might be some fudging on that, but already there is a kind of path finding for non flying move 3 units, this is just expanding that to more than move 3.

Each new turn the move order is checked against the saved destination province and if it fails (due to the province being enemy, or no longer having a clear path) then the unit default to defend like basically every other failed command defaults to. So at the end of your turn if you hit 'n' relilgiously, as imagine most serious people do, you find all your stray commanders without orders and can adjust as needed.

I don't know how this results in lots of bugs, or even anything remotely unintended. Unless its implemented badly, but well, I assume there is at least some QC process you all have...

[Gandalf Parker]

I do know enough about coding to agree that..
A) it would not be fun to code and insert at this point
B) it would have bugs which would take awhile to iron out
C) its not impossible but probably more trouble than its worth

Maybe in the early birth of Dom3 when it was first considered. But now that the devs have moved on to a new project this would definetly not fit into a fun and simple patch thing which might get tossed in during a break from their real jobs and other projects.

[Kristoffer O]

> I knew this already, but its nice to hear it again from time to time.

> I'm not disappointed or surprised by your answer, its what you've said all along.

> I don't see the real issues with bugs or other problems as some have been suggesting if this is implemented intelligently, however, I really don't understand it in terms of if you don't want to use it you won't have to, and if your forces get stuck you will find them anyway by hitting 'n', commanders cannot be issued an illegal movement command as it is, how this changes that I don't know, or see.

> Think about it, set a move order to a distant province, and each turn the commander will issue a *legal* (obviously) movement order to get closer to that province. There might be some fudging on that, but already there is a kind of path finding for non flying move 3 units, this is just expanding that to more than move 3.

> Each new turn the move order is checked against the saved destination province and if it fails (due to the province being enemy, or no longer having a clear path) then the unit default to defend like basically every other failed command defaults to. So at the end of your turn if you hit 'n' relilgiously, as imagine most serious people do, you find all your stray commanders without orders and can adjust as needed.

> I don't know how this results in lots of bugs, or even anything remotely unintended. Unless its implemented badly, but well, I assume there is at least some QC process you all have...

It is more like residual orders I'm afraid of. If orders are not cleared properly there is a chance that orders get transfered from one dying commander to another etc. THis could result in strange movement on AI units or other players orders. Not sure how likely it is, but this is one type of bug I can imagine right now.

Other possible bugs might include strange movement behaviour of stealthy troops when waypoints switch ownership.

Other not movement related bugs might also be introduced. Since move is just another order the mechanic would involve an order after order mechanic. It is possible that the AI decided to use the system or that commanders got strange orders turns from now. I'm not sure if this is possible, but I guess it might happen.

QC is by far more boring than actual coding. THe less QC we do the better we feel. If we do too little QC we will release a buggy patch and players will feel as if we use them as testers. We will also be forced to release new bug-fix-patches with little meaningful content. If we test a lot we will feel sick.

[lch]

Originally Posted by licker

I must say the bulk of you completely miss the point.

It does not have to be pathfinding afterall. You can set up your movement however you like, and the file can simply store the movement command to the next province# in your chain. Is this massive bloat? Seems doubtful to me, unless you intentionally make huge move chains, but even there you could cap the length of the chain at 5 or some other arbitrary number.

[...]

Seriously, I don't see the resistance to this, until a dev says its too hard to code. I won't pretend to know what it would take to add it, and I don't really think any of you should pretend that either.

I didn't miss your point, my argument was "it's going to be too much work for too little gain, and buggish" from the start. You're right, a chain of provinces to move to, set by the player by hand, would be the easiest approach and would eliminate a number of problems by making the player responsible for the route that he chose. But as I said, this is still decidedly non-trivial. Consider the following:

First, you need to be aware of your armies potential when moving. Because as you are aware, units may have Forest survival and similar skills, hence moving faster or slower through specific terrain. No problem you say, just walk as far as you can each turn. Give them 1 map move orders for the entire chain and then try to travel as far as possible along that chain as you can every turn. Very well, but what if that chain is getting broken by enemies taking a province that was once yours in the meantime? If it was the end point in a 2 map move or 3 map move movement, then you'll have to travel to the province before that, and then take the province back. You'll probably want to address that enemy army first and move somewhere where you're better defended. Things get tricky. All in all, vfb already summed up nicely a couple of points that I have against a feature like this.

Regarding site searching: It is my impression that the game so far just isn't aware about what provinces are capitals. Thus capitals are being sitesearched just like every other province. It would of course be problematic to include the info what all the capitals are in the turn files, because people could then extract that information and abuse it. But it might be a good idea to add a new flag (as in a bit in the province bitmask) for capitals, so that the mages can skip this province even if the site searching order was given during the client's turn preparation.

[Edi]

I'd also like to add that even fairly innocuous changes can result in big headaches. In patch 3.20, move order validation was improved. This means that sneaking out of sieges and force marching were eliminated.

What is not generally known is that it introduced a MAJOR bug that would have had everyone on the forum screaming bloody murder and it would have ruined or at least severely hindered all ongoing MP games on hosting. And the implemented change was FAR less than adding a waypoint system that allows sequential movement from one turn to another. Fortunately that one was caught and squashed, which is also the reason for so many versions in such a short order.

So just take Kristoffer's word for it when he says it's a no go. If he can imagine the kind of consequences he enumerated right off the bat, I don't want to know what kind of exotics it might introduce into the game. We still have the immortality weirdness that has not been hunted down, the F9 and D9 bless bugs plus some other stuff. Like Kristoffer, I prefer the current system even if it's not the smoothest possible, because it works without major problems.

[licker]

Fair enough, I will continue to request it for Dom4 then.

I am in the QC field myself, though not limited to software QC, which is it's own beast.

QC is only boring if you get it all right the first time

No problem you say, just walk as far as you can each turn. Give them 1 map move orders for the entire chain and then try to travel as far as possible along that chain as you can every turn. Very well, but what if that chain is getting broken by enemies taking a province that was once yours in the meantime?

Asked and answered... seriously with the tinyist amount of intelligence built into the system none of the so far proposed issues (other than the potential bugs, which are of a different nature) are even possible.

If the chain is broken the move order is canceled, and the commander is defaulted to defend. This works for monthly rituals, if there are not enough gems (or targets) the commander is reset to defend.

[Gandalf Parker]

To add to KO's reference to passed orders..
For many things the game does not keep track of units by nation. It tracks them by ID number. The commander you create is "in line" with other peoples commanders and the AIs and the Independents. This has popped up before where a commanders was taken out by spell or event but their actions were passed down to the unit next in queue. It was mostly caught and fixed during beta-group testing but it did hold things up sometimes. It is not simple since much of the game is not modular.

Also, when discussing present waypoint code, keep in mind that this game was created on an Atari long ago and the base code is mostly converted code. Those of you with coding backgrounds saying that waypoints are easy might realize what a bear it is to insert new code routines into ancient spaghetti and cross-platform converted code. I have no idea how much of the game is that way but since we are talking about some of the first portions of the game to be written Im thinking this might be part of the problem. Building on such pre-packaged algorithms was waypoints is much easier than inserting them later. Johan is an excellent programmer but seeing this game develop from infancy to now I suspect that answers of "its not that simple" is probably realistic and cannot be easily compared to other programming projects.

I am hopeful that the new project might have waypoint code in it along with some of the other desires we had that we didnt get because it would involve rewriting the game from scratch.

[lch]

Originally Posted by licker

Asked and answered... seriously with the tinyist amount of intelligence built into the system none of the so far proposed issues (other than the potential bugs, which are of a different nature) are even possible.

If the chain is broken the move order is canceled, and the commander is defaulted to defend. This works for monthly rituals, if there are not enough gems (or targets) the commander is reset to defend.

Is the whole move order canceled as soon as one of the provinces in the path are no longer passable without trouble? Because if that's the case, then it might be hardly usable in some games, as you'll end up reacting to how the map looks around the battlefront all the time. If you only stop until it directly poses a problem, not in some turns later, you'll have the unwanted effect that your reinforcement armies may run into the hands of your powerful enemy without you knowing. I always hated Wesnoth for that. You set a unit to auto-move somewhere. On the beginning of your turn, new enemies are spotted in front of the Shroud of Darkness. But your lone units still happily end their movement one hex in front of the enemy, as long as that enemy doesn't block their path they're trying to move through, even if that gives an abysmal defense bonus. And in Wesnoth it is extremely important to have a good defense modifier, as is hitting first.

As I said, the feature request would definitely be "nice to have". But given the problems, and that in development feature requests are considered last concerning implementation, I can hardly see it appear in the future.

[licker]

I would break the move command as soon as any province from the chain is missing.

clearly you still need to be aware of the situation, and not blindly leave your armies on some automarch which no longer makes sense due to other conditions, but that in no way makes haveing the option to set an automarch a bad idea, just one you don't want to use blindly.

Then again what do you do blindly in Dom3?

Probably not much, so you may decide to never use it, others may decide it helps them cut down on the micro, others may use it, and get screwed by it by not reacting to changes, but honestly, those in the latter group are screwed anyway if they aren't paying attention.

Might as well complain about forgetting to set some scripts to mages or units because you didn't realize they were going to get attacked.

There's always lots of little things you should do when you have a big nation to manage, makeing some of those little things less tedious is a good thing, getting rid of those little things would be ideal, but probably not in the realm of play style for dom3.

I'm channeling MoO3 there, but that's a different game design concept entirely. Though it could work in a dom type setting/game. Think Majesty or RotK if you prefer those, but basically one leader setting policy, but not actually dirtying their hands with the nitty gritty.

Sure the nitty gritty appeals to some people, and that's fine, I'm talking about completely different games, though I see the advantages in empolying some of those game mechanics to all TBSs which suffer from mid to late game micro management hell.

[SlipperyJim]

Originally Posted by licker

There's always lots of little things you should do when you have a big nation to manage, makeing some of those little things less tedious is a good thing, getting rid of those little things would be ideal, but probably not in the realm of play style for dom3.

I'm channeling MoO3 there, but that's a different game design concept entirely. Though it could work in a dom type setting/game. Think Majesty or RotK if you prefer those, but basically one leader setting policy, but not actually dirtying their hands with the nitty gritty.

Sure the nitty gritty appeals to some people, and that's fine, I'm talking about completely different games, though I see the advantages in empolying some of those game mechanics to all TBSs which suffer from mid to late game micro management hell.

Ever played the original Dominions? Dominions: Priests, Prophets, and Pretenders ... aka Dom:PPP?

No re-usable commander scripts. No automated site searching. No monthly spell-casting commands. No automated pooling of gems and blood slaves.

I feel like one of those cranky old men who's always telling the kids, "When I was your age, we had to WALK to school! Uphill! Both ways! In nine feet of snow!"

Anyway, the point I'm trying to make is this: Dominions 3 has seriously reduced the amount of micromanagement that we used to have. Can they go farther? Sure, there's always going to be room for improvement. But they've come a long way.