GB vs. WL

General Question/Answer/Announcement about NSD. We are a small independent game development team and we value our community. If you ask, we'll answer.
User avatar
norb
Reactions:
Posts: 3778
Joined: Mon Nov 26, 2007 9:59 am
Location: Central Florida
Contact:

GB vs. WL

Post by norb »

GB was the first MP game of ours, followed up by WL. A lot of work went into WL, hopefully most of it good. I read some comments and I realize that I'm not sure which game people are referring to.

So please give me a hand on this thread:

What does WL do that you wish GB had
What does WL do better
What does WL do worse

So what in your opinion did we improve and what did we break. I was just running both games on my computer and I get an early d3d error running GB that I don't get with WL that shows up in the log.

I am wondering if it's worth trying to combine the two so that we just have one thing to work on and can improve it for both games.
DarkRob
Reactions:
Posts: 352
Joined: Mon Aug 15, 2016 5:56 am

Re: GB vs. WL

Post by DarkRob »

In my opinion Waterloo was superior to the civil war games in almost every way. I wasnt actually that excited 8 years ago when I first heard the next Scourge of War game was going to be on Waterloo, as I have always been more of an ACW guy.

But Waterloo won me over with its more complex combat, the interplay between different types of arms, and the vast open battlefields of Europe.

You mention possibly combining the two games and that does make me think, there are a number of Waterloo features that if retroactively applied to the civil war games would make them better, at least to me.

The way of placing units feels much better to me in Waterloo. Placing units with right click and being able to hold the button down while moving the mouse to adjust facing before releasing is much better than having to left click and then manually having to adjust facing by hitting the wheel left/right buttons multiple times like in Gettysburg.
If this system were unified to the Waterloo system for both games, it would do alot for my muscle memory when it comes to going back and forth between the series.

There is a quirk of the Waterloo interface, which, while being an accident, and I do understand its not supposed to be this way, it nevertheless makes the game better. In Waterloo, certain brigade level commands still function from the brigade commander, even if all his units are under TC. Among them are double quick, retreat, charge, advance while engaged, and fall back while engaged. This gives the player alot more control over his units once they become engaged in combat.
In the civil war games, no functions of the brigade commander work while his units are TC,d
Again, I understand that the civil war games are how its supposed to be, and its Waterloo thats the accident. But accident or not, its better. Couple this quirk with multiple levels of TC like TC on/off all subordinates and it makes everything so much more quicker and more efficient when it comes to the ever important dance of jumping back and forth between AI control and TC.
If the civil war games could retroactively work this way, I might be more inclined to go back to them more often, maybe even more so as given the more spread out nature of the forces, hills and dense forests obscuring everything, more control would be even more useful in those games.
Im probably pipedreaming this one though and I understand that. But a guy can dream now cant he?

I dont know how far you want to take this, whether we're just talking remaster, or whether you want to improve the games as well. But Il ramble on anyway and you can take what you like and discard the rest.

Waterloo is a great game, but even so, there are some things that could use some, shall we say, tweaking.

1) There are a number of truly busted scenarios in the base waterloo game that could be fixed. and I think easily fixed with some minor scripting rework and objective placement/value adjustments. Most of the really broken ones are broken in silly ways that could probably be fixed in a few minutes. I wont list them all here, if youre interested, PM me and Il tell you which ones and why theyre broken.

2) The AI still mostly fights like its a civil war AI, with some napoleonics tacked on. The biggest area it could use some help in is how it uses combined arms. Its extremely vulnerable to having combined arms tactics used on it, and lacks the ability to return in kind.
Three key things
A) It leave its artillery behind on the attack, which always means I have the guns and it doesnt.
B) It needs to better support infantry attacks with cavalry. Sending infantry to attack one part of the line, but sending cavalry to attack a different part of the line is not cavalry supporting infantry. Instead they both end up getting destroyed because they both need each other. Infantry has to protect the cavalry, while cavalry controls the enemy formations allowing their infantry to gain the advantage.
C) The AI needs to up its skirmisher game. It doesnt use them nearly enough. If Im shooting one of its line units to pieces using just skirmishers, it should at the least know to kick out a few of its own to screen that unit. Most of the time it doesnt.

That brings me to skirmishers in general. Perhaps no type of unit in the game is as overpowered in so many useful, ultilitarian ways, as skirmishers.
Theres nothing wrong with the way they work on a by unit level. They generally do better in a firefight against an enemy line unit and thats fine, there has to be a reason to use them in the first place.
The problem is when you start to multiply them and multiply them. Swarm the field with them, and screen everything with them that they start to become ridiculous.
I just did a speedrun of the full battle of waterloo as the allies, and Im not kidding, I fought the whole thing with pretty much only skirmishers, theyre that good. You can see it for yourself if you ever feel like checking it out.
https://www.youtube.com/watch?v=GeEdf-oUfVc&t=2174s

Skirmishers just need to be reigned in somehow. Right now they have no limits. I can kick out as many as I want from any unit I want, from as many units as I want, until the unit literally runs out of men. I can send them as far away from their parent unit as I want, and do anything i want with them.

Thats what they really need, some kind of limitation. Maybe a line unit can only kick out so many skirmishers, or perhaps only certain types of units, or a certain number of units per brigade. Or maybe even only units of a high enough troop quality can kick out skirmishers, say level 5: seasoned and above. Just something that makes them not so limitless. Right now they are an unlimited and constantly recyclable source of manpower.

Well, Im sure I could think of more, but this is enough for now. Also none of this is really meant as heavy criticism. Scourge of War Waterloo is my favorite wargame ever made. By and large, its a masterpiece.

One other thing. If you decide to maybe revisit the toolbars for both games, let Reb do them. Hes the master.
Puntalancero
Reactions:
Posts: 2
Joined: Sat Sep 03, 2022 4:21 pm

Re: GB vs. WL

Post by Puntalancero »

In my comment in the other thread I was referring to the WL. In any case, the maxim to follow should be to optimize what is there. Fluidity and reliability. And focus on the MP, which is probably why there are, seven years later, active players and many others interested in SoW.

When you have a fully playable game, step it up and release a DLC with new gameplay options and new stuff. No one in this family would blame you.
User avatar
norb
Reactions:
Posts: 3778
Joined: Mon Nov 26, 2007 9:59 am
Location: Central Florida
Contact:

Re: GB vs. WL

Post by norb »

Thank you. I appreciate the write up. I believe it would be better to just have the one engine rather than two. I would be able to add more features and it would get tested more. This is perfect as my biggest concern would be that there is someway that WL works or plays that people do NOT want in GB.
voltigeur
Reactions:
Posts: 244
Joined: Tue Dec 14, 2010 5:51 am

Re: GB vs. WL

Post by voltigeur »

Great writeup by DR.

If one engine is decided upon, here's some thoughts:

1. Split out the AI dll's for separate ACW and NAP use (allow fine tuning as the tactics are quite different between periods, don't lose the different 'flavours')
2. Settle on a single toolbar system to keep things simple and uniform (although obviously there will be some changes between the two eras).
FWIW, I always have the context menu in WL disabled - I prefer to use Reb's grog toolbar.
3. on that point, I agree with DR - Get Reb to do the toolbars, he's a genius!
4. Also agree with DR regarding skirmishers - they need range and number limiting.
5. Multiplayer is, as we know, where the game shines. Given that Raknet has long since stopped development, and the peer system known to be unstable, is it time to look at a dedicated server option? I am sure there would be plenty of enthusiasts out there willing to host servers, me included.
6. Consider removing the WL campaign feature. I found it somewhat under developed and did not play it more that a couple of times. Put precious time into other more popular items.

cheers Volt
Biondo
Reactions:
Posts: 111
Joined: Wed Apr 13, 2016 9:26 pm

Re: GB vs. WL

Post by Biondo »

norb wrote: Mon Sep 05, 2022 10:35 pm GB was the first MP game of ours, followed up by WL. A lot of work went into WL, hopefully most of it good. I read some comments and I realize that I'm not sure which game people are referring to.

So please give me a hand on this thread:

What does WL do that you wish GB had
What does WL do better
What does WL do worse

So what in your opinion did we improve and what did we break. I was just running both games on my computer and I get an early d3d error running GB that I don't get with WL that shows up in the log.

I am wondering if it's worth trying to combine the two so that we just have one thing to work on and can improve it for both games.

What does WL do that you wish GB had

- As Rob said, the way you place unit is far better in WL, especially when you play with couriers and in GB you have to wait that the courier reach the unit to be able to wheel that unit in the direction you want

- possibility to save and resume a MP game

- the campaign feature, I remember all the problems which affected the development of the campaign but I liked the idea very much; so if there's time to develop it, it would be good to have also in GB

- I think the modding possibilities are another point in WL favour; we had map tools and the possibility to mod the AI

- possibility to occupy buildings although I'm not sure how many times this happens in the ACW


What does WL do worse

A couple of minors things:

- it's missing the Move to this map point feature that we had in GB. Actually we lived without it for so many years but it would be good to have it back

- about the positioning of the units, as I said before WL is better but I feel that the arrow we had in GB was more visible than the green pointer we had in WL at least when you play in HITS. It's not a matter of color (green, yellow, whatever) but it's that the pointer is a 2D image drawn parallel to the ground. Probably having it tilted of 30 degrees it would help when you play from the ground and doesn't bother the player when the camera is high in the sky



Ok, those are my 2 cent about the difference between WL and GB. I also have some other stuff to suggest but when the time comes
r59
Reactions:
Posts: 101
Joined: Fri Apr 14, 2017 8:17 pm

Re: GB vs. WL

Post by r59 »

And that pretty much covers it all, I'd say! :D
Many great points here, fellas. I agree with all of them. Congrats.
Good luck to Norb if he's really persuaded merging the two codebases is the best place to restart from!

One secondary order / technical thing that I could add up that WL had implemented and GB didn't, is multithreaded LOS.
We must be honest, it really didn't paid off in terms of speed gains.
It also carried some bugs and race conditions I'm not going to list in details here.
But just the first issue coming to mind now: there were chances that within the same game tick, two units A and B might (running their LOS computations on different threads) end up seeing another enemy unit C's visibility state variable with inconsistent values (true or false). This was happening according to the execution timing of each thread, not based on any gameplay condition...
And from there the two could have taken different paths down the code like pushing or not C in the enemy list or target list, coming to different AI decisions, etc.
That clearly broke MP determinism and as we know it had to be disabled.
But altought it really doesn't catch the eye for a number of reasons, it actually makes LOS results in some cases randomly asymmetrical and impredictable in SP too.
I think it'd wise at this stage to cut it out altogether and not only for MP.
As it's my understing Norb still considers LOS the main source of concern about the game performances (I partly agree), when he might eventually try to improve/polish it - if he feels appropriate, he could do in a "safer" single-threaded context.
Focusing on code and data, avoiding over-complications.
There's little to gain to keep it in that form and at this stage imho.

Good luck again.
Martin James
Reactions:
Posts: 569
Joined: Mon Apr 11, 2011 9:23 pm

Re: GB vs. WL

Post by Martin James »

I think you are asking good questions, Norb. Here are some comments that have been gathered from a small group of players and modders that I'm in touch with, on one aspect. That is whether you should combine GB and Waterloo, so the NSD team can work on just one engine in future.

That’s a very rational desire and would, hopefully, help stream-line the update process and modding going forward. However, it also generates concerns on the modding front.

GB spawned a very active modding community, with a large number of mods (and modders), several covering other time periods (eg Malburian, Sikh Wars, Crimean War etc).

To our knowledge Waterloo has significantly fewer mods, albeit substantial and impressive ones such as the KS mod and Les Cent Jours. If you still feel modding is an important factor in the longevity of the game, it would be an idea to explore the reasons for that difference. The following are some which occurred to us:

• The requirements of MyGui mean that OOB and scenario.csv construction in Waterloo has an extra hoop to go through, due to the need to create a Names file. And the latter needs to be adjusted every time an OOB is tweaked

• There are differences between GB and Waterloo logistics and OOB-type file formats, which means GB mods need work to be ported over.

• Many sprites from existing GB mods do not work in Waterloo, because of the way the latter reads the sprite sheets. Losing all of these great sprites would be a shame given the many hours of work involved in producing them.

• There are differences in map requirements between the 2 games. Thankfully, this can be worked around, and many GB maps have already been converted to Waterloo. However, that conversion again takes extra effort. Ongoing compatibility in the combined engine would be helpful.

• The inability of the current Waterloo engine to represent features of some other wars (eg lack of dismounted cavalry, presence of squares and skirmishers, which are not appropriate for all wars/periods). It would be very helpful if these attributes could be ported across/switched off as necessary to allow different periods to be represented in the combined engine.

• The concern that the modder would need to learn programming skills, both to make full use of Waterloo’s capabilities, and also adjust aspects of the AI which would not be appropriate the particular mod

Please note we like Waterloo. Several useful new features come with it. And most of the above can be overcome or worked-around. But cumulatively that requires a considerable effort. Rightly or wrongly, that has proved a barrier for the modding community as a whole, so far.

Several once keen modders have dropped away. Some of us are still active. As mentioned, I am part of a small informal group which is still producing mods. We did do one Waterloo mod several years ago, but since then have stuck to GB for the reasons listed above.

If all work is to be centralised on the Waterloo engine going forward, then it would be helpful if the revised engine was sufficiently adaptable to incorporate the mods - particularly the sprites and maps - produced so far, and for modding to be easier more generally. In summary, it would be helpful to have:
1. better modding tools
2. some way of ensuring all those years of GB modding work are not lost.
3. the ability to reflect other periods than Napoleonic

Of course, we appreciate what might be done depends on your plans for the engine going forward and how ambitious those are. Whatever you guys decide to do however, we wish you well. Good luck NSD!

Martin J
User avatar
norb
Reactions:
Posts: 3778
Joined: Mon Nov 26, 2007 9:59 am
Location: Central Florida
Contact:

Re: GB vs. WL

Post by norb »

Great tips, thank you. Although we probably won't be backward compatible, we will write the tools for everyone to use. Heck, if we do this we have a ton of scenarios to port over myself. And being the laziest coder out there, I'll write the tools.
Martin James
Reactions:
Posts: 569
Joined: Mon Apr 11, 2011 9:23 pm

Re: GB vs. WL

Post by Martin James »

You're welcome Norb. Thanks re the tools. Being the laziest modder out there, I will use them ;)

Martin
Post Reply