Is CMBS worth buying?

[Battlefront.com]

3 minutes ago, Peter Panzer said:

Hey Steve:

Granted it may be early, but have you guys started sketching out possible timeframes and settings for CMx3?  Perhaps a past hypothetical (Fulda Gap '84), future fictitious action (Korea 2025) or will we return to WWII (Barbarossa '41)?

Personally, I consider Battle for Normandy to be a real masterpiece at this point with Black Sea showing an exciting degree of promise.  I'm curious if you can share a thought on at least the chronological direction the next iteration of CM may take.

By the way, congrats on Final Blitzkrieg entering the homestretch - being able to play out scenarios in the Hurtgenwald is very enticing...

Definitely too early to say what the first game will be.  We're also weighing the options of keeping CMx2 going in parallel or freezing it as we did CMx1.  There are significant arguments in favor and in opposition to each of those choices.  We'll make the decision later on when we see how CMx3 turns out.

Steve

[Michael Emrys]

On March 28, 2016 at 8:39 PM, Battlefront.com said:

Having said that, the problem CM faces is that it is really 4 games in one

1.  Turn based "strategy" (think old top down hex type games)

2.  Turn based "tactical" (think tabletop miniatures)

3.  RealTime "strategy" (think Command and Conquer)

4.  RealTime "tactical" (think First Person Shooter)

I have always seen CM as a kind of electronic tabletop miniatures game. Not only the look of it and the scale, but the kind of players it attracts. Consider: for many miniatures players, half the fun lies in painting their armies. CM has its rabid modders. CM gives you all the views over the battlefield that the tabletop provides without the stiff neck and no need to have a measuring stick to check ranges and whether or not an LOS is valid or not. There are other similarities. So CM is very much a Type 2 (from your list) game. I don't see it as being very much a Type 1 at all. For one thing, the scale is unlike the vast majority of Type 1 games I ever played, although somewhat like, say, Panzerblitz or SL or its offspring ASL and a few others. I've never played much Type 3 or any Type 4, so I won't even attempt a comparison there. But the resemblance to tabletop gaming is what has struck me from the beginning.

Michael

[BletchleyGeek]

23 minutes ago, Battlefront.com said:

Sorry to hear that it didn't end well.  Kinda part of the point I'm making

I'm sorry too.

24 minutes ago, Battlefront.com said:

Yesh, that's too flawed to even bother touching.  Let's just say that I've been doing this long enough to know the result of such an attempt would not be worth the money invested in it.

Well, and I have pulled off similar "stunts" in the past without too much trouble. YMMV of course, and it is interesting to hear about it.

 

30 minutes ago, Battlefront.com said:

You're missing the point.  When you jump from Tactical to Operational there is a huge shift in concept as well as task responsibilities.  When you shift from Operational to Strategic there's another big change.  ARMA didn't have to mess with that.  They stayed firmly within Tactical.

Basically, your argument in the real world amounts to saying that you can run a basic training course in a military and crank out Privates and Colonels equally well because the same principles that the Private needs to do his job are inherently scalable by snapping fingers.  Or perhaps you are saying that you can put all recruits through higher education courses and that's going to make then excellent riflemen because that knowledge can be scaled down.  This is absolutely not true.

There isn't such a "huge shift".

Let me clarify one thing: when I refer to the command level I see them defined implicitly by what is the smallest "entity" or "organisation" to which you issue commands. That, in CMx2 can be as small as whatever task-oriented grouping a squad is broken into to as big as a regiment (I have never tried to put more than 3 Bn's on a map, probably there's some people who have tried to play the game with a whole division...). The player plays the role of the decision maker, he who decides what to do when rather than the "implementer". In CMx2 both roles are quite mixed in together. Rarely a player takes command of single riflemen (see the exception below) but more often takes command of a Bn-ish force. As you go higher in the hierarchy, the tools offered to the players to implement the job of what would be the S-3/XO or S-2 become pretty much inexistent, as proven time and again by some excellent AARs. 

Going back to the privates vs. colonels example you put forward. CMx2 isn't about directing privates but you can take command of a truck driver and have him try and go Audie Murphy on some enemies, but that's certainly a side effect of the design. Also a side effect of the design is the possibility to do a meaningful S-3/S-2 job.  What I am trying to say is that amongst the different levels of command indeed there is a change in the "repertoire" and "scope" of such tasks: the temporal, spatial and material scope does indeed grow and the methods to tackle certain situations need to change and some others are inherently new. That's what I referred to when I said "new heuristics and algorithms may be needed". I used those two works because we're talking about pieces of software, that when autonomous, or semi-autonomous, are made up of some suitably arranged mixture of both heuristics and algorithms.

In CMx2 there's plenty of autonomy for the individual pieces - within the scope of the commands given to them and modified by their morale status - but no autonomy whatsoever for any entity beyond the rifleman (well, there's another exception, the assault command available for certain squads).

Yet the basic elements making up the description of the "tasks" those routines have to deal with remain: the impact of communications (i.e. can I see what is the platoon commander signalling? can I receive confirmation from the Pentagon to launch my Tomahawks on top of Kim's palace? can I tell the commander of the 3rd tank in my platoon to drive around that KV-1 and engage it from the rear?), handling the possible impact  of varying degrees of training, experience, fatigue and motivation of subordinates (who are also somewhat autonomous entities), working out a plan that is detailed enough to guide subordinates but flexible enough so that it allows room for handling contingencies, and so on. 

Colour me reductionist, but armies can totally and usefully interpreted as "distributed programs" running on people's minds designed to harness complexity. Just look at how the history of military organisations has evolved, from the early Bronze Age armies consisting of one or two charismatic leaders and a what essentially is a mob, to 2016 US Army, going through Mauritius of Nassau book of regulations for his infantry. I am not saying there's a clear arrow of progress - the charismatic leader + mob combo model has been shown to be able to produce upsets for much more recent ways of doing things - but if you look at all those ways of organising people and material for war, there's quite a few common patterns and concerns. 

I didn't say anywhere "by snapping fingers", either. I may sound a bit too subversive, but it essentially follows from my personal interpretation following the study of Robert S. Leonhard works. Is he wrong, perhaps, but he makes a quite compelling case.

46 minutes ago, Battlefront.com said:

The UI has to be kept in synch with the player's needs at that moment and, as I stated in detail above, those moments are not neatly separated as with ARMA 3.  They also sit in the development queue along with non-UI related tasks that people have requested that we do because we don't have a team of programmers.  When Charles is coding a UI feature he's not coding a game play feature and vice versa.

Well, the answer to that is that you perhaps need to find a group of skilled, talented and trusted people you can bring into the process on an on-call fashion. I am not saying it is easy to organise or manage - the traditional monthly salary paycheck and employer/employee relations are "easy" because we've doing it for about 150 years now, and there's a lot of common knowledge about practices what works and what not.

49 minutes ago, Battlefront.com said:

We've had debates about command delays many, many, many, many times.  There's so many different schools of thought on this subject.  And again, there's extremely different PoV from the WeGo guys vs. the RT guys vs. the Command guys vs. the micro control guys.  There is no agreement, so any design we pick is going to be a poor choice for quite a few people.

What I was proposing was a "to each their own" solution, which of course, implies coming up with something that nobody has come up with before.

I let you to your web programming, and I go back to the 10th reading of Where the Wild Things Are today.

[Kinophile]

6 hours ago, BletchleyGeek said:

Well, the answer to that is that you perhaps need to find a group of skilled, talented and trusted people you can bring into the process on an on-call fashion. 

This. 

Steve, have you guys seriously and properly looked at how an open source, BFC managed, community engaged effort would look? 

With the hard core  community you have a superb resource available in terms of  manpower and dedication -  and for free! 

Of every 'proper' mod I've worked on as part of a team, the strongest are those with a serious and mature core with a strong technical background - ie. Every CM player ever. 

There are platforms available for organizing such efforts.  A clear and simple constitution is critical,  as is a defined,  clear objective. 

A superb example would be the FreeSpace community, who have essentially rebuilt the original space sim and given it a lease of life (stronger processing, vastly improved graphics, AMAZING editor) that would have been inconceivable when it first came out. 

The trick, naturally, is to balance and fend off over enthusiastic or blinkered members,  but that's essentially personnel management. 

I don't mean to sound preachy,  this post is also for those who have not done, or are not much aware of modding as a development process. 

But if you guys are strapped for time and money,  then please,  please reach out. Or even have a mechanism for you to put out a specific task or objective and see what people can offer as an answer. This would retain command and control in your end and allow us to participate in a strictly mission - oriented capacity. Financial renumeration would be specifically prohibited but public acclaim guaranteed.

You could put up the to-do list,  with detailed criteria, and see who bites. You could then allocate a task based on the contender's examples, Resumé and proposed solution 

It then becomes a matter of review and direction rather than you guys slogging through ALL the code, graphics, testing, analysis etc and retains your control of the direction of the game. 

You have a very dedicated and willing pool of people to draw on - use us! 

Edited March 29, 2016 by kinophile

[Placebo]

2 hours ago, kinophile said:

Steve, have you guys seriously and properly looked at how an open source, BFC managed, community engaged effort would look? 

With the hard core  community you have a superb resource available in terms of  manpower and dedication -  and for free! 

The phrase "as easy as herding cats" springs to mind  

I certainly agree with you modders have done amazing things in transforming games and devoted huge amounts of time and effort to a project for free, but to plug that into a business model is a different thing.

[Kinophile]

5 minutes ago, Placebo said:

The phrase "as easy as herding cats" springs to mind  

I certainly agree with you modders have done amazing things in transforming games and devoted huge amounts of time and effort to a project for free, but to plug that into a business model is a different thing.

True. Then maybe the second option,  if a semi-public to do list of available tasks would work.  No management required,  just review when/if product is uploaded. 

[Battlefront.com]

11 hours ago, BletchleyGeek said:

Well, and I have pulled off similar "stunts" in the past without too much trouble. YMMV of course, and it is interesting to hear about it.

The primary assumption you're making is that there is One UI To Rule Them All solution that is just begging to be coded without much more than a few day's design work.  Having worked with CM now for pretty close to 20 years I'm not confident that such a UI exists.  A better one than we currently have?  Sure, I believe that's possible since I'm working on exactly that.  But it's going to take me months of elapsed time to develop it.  There's a big difference in terms of what comes out of a long period of design that offers enough time for reflection and revision vs. narrow and ridged time.  I can definitely come up with a decent UI redo in 1-2 weeks, but it's risky.

11 hours ago, BletchleyGeek said:

There isn't such a "huge shift".

Which is why we don't see eye to eye at a very fundamental level.  When I'm talking about UI I'm talking about in the context of Combat Mission.  You're talking about it in the context of a fantasy game (i.e. one that doesn't exist anywhere on Earth except in your imagination).  We might as well be speaking different languages because the two have absolutely nothing to do with each other.

11 hours ago, BletchleyGeek said:

Let me clarify one thing: when I refer to the command level I see them defined implicitly by what is the smallest "entity" or "organisation" to which you issue commands. That, in CMx2 can be as small as whatever task-oriented grouping a squad is broken into to as big as a regiment (I have never tried to put more than 3 Bn's on a map, probably there's some people who have tried to play the game with a whole division...). The player plays the role of the decision maker, he who decides what to do when rather than the "implementer". In CMx2 both roles are quite mixed in together. Rarely a player takes command of single riflemen (see the exception below) but more often takes command of a Bn-ish force. As you go higher in the hierarchy, the tools offered to the players to implement the job of what would be the S-3/XO or S-2 become pretty much inexistent, as proven time and again by some excellent AARs. 

Yes, CM requires the player to "wear many hats".  That's the essence of what I've been saying all along, so obviously no disagreement here.

11 hours ago, BletchleyGeek said:

Going back to the privates vs. colonels example you put forward. CMx2 isn't about directing privates but you can take command of a truck driver and have him try and go Audie Murphy on some enemies, but that's certainly a side effect of the design. Also a side effect of the design is the possibility to do a meaningful S-3/S-2 job.  What I am trying to say is that amongst the different levels of command indeed there is a change in the "repertoire" and "scope" of such tasks: the temporal, spatial and material scope does indeed grow and the methods to tackle certain situations need to change and some others are inherently new. That's what I referred to when I said "new heuristics and algorithms may be needed". I used those two works because we're talking about pieces of software, that when autonomous, or semi-autonomous, are made up of some suitably arranged mixture of both heuristics and algorithms.

You are making an argument for a Command level game.  It's been made before, since about 1999 in fact, so for sure you are in good company thinking this is a good way to go.  However, ever time this gets discussed the people that want a Command level game get stomped on by the majority.  They want a "table top" experience such as Emrys described, so what you are suggesting is something they would not want to play.  Which poses a bit of a problem for us if that's our majority customer AND there's not enough Command level guys to make up the difference.

But that's just the philosophy of it.  There's also the engineering challenge...

11 hours ago, BletchleyGeek said:

In CMx2 there's plenty of autonomy for the individual pieces - within the scope of the commands given to them and modified by their morale status - but no autonomy whatsoever for any entity beyond the rifleman (well, there's another exception, the assault command available for certain squads).

Yup, because that's not what people want (first problem) and it's not what we can deliver (second problem) due to the massive amounts of near perfect AI that would be needed to pull it off.  Given everything else we have to do aside from AI, it simply is not practical to do.  Not even "open source" as that's not viable in and of itself.

11 hours ago, BletchleyGeek said:

I didn't say anywhere "by snapping fingers", either. I may sound a bit too subversive, but it essentially follows from my personal interpretation following the study of Robert S. Leonhard works. Is he wrong, perhaps, but he makes a quite compelling case.

There's a difference between theory and reality.  In theory I'd love to have such a game.  In reality I know we'll never be the ones to make it.  We're not even going to try.  It's simply too difficult and too uncertain.  In my view it's akin to corporate suicide.  No thanks.

11 hours ago, BletchleyGeek said:

Well, the answer to that is that you perhaps need to find a group of skilled, talented and trusted people you can bring into the process on an on-call fashion. I am not saying it is easy to organise or manage - the traditional monthly salary paycheck and employer/employee relations are "easy" because we've doing it for about 150 years now, and there's a lot of common knowledge about practices what works and what not.

It's not viable.  As someone who has been in this business for most of his adult life, I find very few things I can think of that I'm more certain of than that.  It's akin to saying that we can have world peace if everybody just treated each other with respect.  Nice theory and I welcome you to have at trying to make it reality, but until then I'm happy to pay my tax Dollars to fund police and military protection for my person and property.

11 hours ago, BletchleyGeek said:

What I was proposing was a "to each their own" solution, which of course, implies coming up with something that nobody has come up with before.

UI designs support the game mechanics, not the other way around.  Proposing a heuristic UI for CM without making the game itself heuristic is simply a disaster concept.  Since CM is not heuristic, nor is it possible to become that, you've not really challenged any of the points I raised about the compromises necessary in UI design to make CM (the one that exists) work reasonably well.

Proposing a new, ground up CM game that has almost nothing to do with the previous CM game, then having a totally new UI concept that supports it, is inherently OK as long as the game mechanism is viable from an engineering and economic standpoint.  I doubt it is, which is why you won't be seeing us tackling that sort of game.  Not with CMx3, not with CMx4, not ever.

11 hours ago, BletchleyGeek said:

I let you to your web programming, and I go back to the 10th reading of Where the Wild Things Are today.

At least you're not rereading Where's Waldo

Steve

[Battlefront.com]

3 hours ago, Placebo said:

The phrase "as easy as herding cats" springs to mind  

I certainly agree with you modders have done amazing things in transforming games and devoted huge amounts of time and effort to a project for free, but to plug that into a business model is a different thing.

Well put.  Even fairly modest opening up of the engine to modding poses major technical challenges which require significant engineering costs to us.  Costs which not only won't be compensated for but will take away from sales we would have ordinarily had access to.  We've had this discussion more times than I can count, but it always comes down to the same thing... we are in a niche market and not a very lucrative one.  Any business model change that requires more work and more risk while offering less chance of reward is not something we're interested in.

3 hours ago, kinophile said:

True. Then maybe the second option,  if a semi-public to do list of available tasks would work.  No management required,  just review when/if product is uploaded. 

Management and support are always required since our name is on it. 

There's an old sentiment that basically states that the more people you have working, the more chaotic things are, the more need for management, the less productive each person is per hour worked, and the less accountable any one person is to any one particular problem/solution.  It's not the sort of mess we want to get involved with.

Steve

[Kinophile]

To be a bit clearer, I'm thinking more of "spreading the workload", e.g. farm out the more mundane, time intensive tasks, rather than opening up the engine for outsiders to tinker with and inevitably, break.

But that's true about some form of management required, and the dangers with open-source diluting the focus, discipline and attention to detail that's gotten BFC to where it is now.

Edited March 29, 2016 by kinophile

[A Canadian Cat]

16 minutes ago, Battlefront.com said:

The primary assumption you're making is that there is One UI To Rule Them All solution that is just begging to be coded without much more than a few day's design work.  Having worked with CM now for pretty close to 20 years I'm not confident that such a UI exists.  A better one than we currently have?  Sure, I believe that's possible since I'm working on exactly that.  But it's going to take me months of elapsed time to develop it. 

Not to mention that the way to successfully go about any UI design really comes down to being iterative. Clearly some up front work (what Steve is doing now) to define the goals and the basic concept and constructs is needed.  But anything that is designed will need refinement as and after it gets implemented.  No design has come off the drawing board and into real life as a "go away and code it project".  OK no successful design . I have been doing UI design and programming for years and when it works best there is a lot of thought up front to define the plan and then lots of trial and error to get the details right. The designer(s) and the developer(s) along with tester(s) work together to manage that process.  I have never seen it work where some designer dreamed it, some developer built it and some tester verified it.  OK again I have never seen that be successful.  Sadly I see that kind of junk happen frequently.

I think, and I hope, Steve gets that he sure sounds like he does to me. <ptich> Personally I am looking forward to getting my hands dirty as a tester in his process too.  Usually I'm either the designer or the developer so it should be a fun change.</pitch>

[AttorneyAtWar]

I am not going to dive deep into this discussion, but maybe an interesting idea would be to show off the UI concept Steve (once its done obviously) and see what players think on the forums. Maybe you have already done this with the beta testers, but I figured that might be a good way to take the temperature so to speak of how people feel about it.

[kraze]

15 hours ago, Battlefront.com said:

The first person shooter portion of ARMA 3 is very established and extremely simplistic.  Move around, shoot, select weapons, hide, etc.  Most of it is achieved through a couple of keys that cycle through options.  Movement is achieved through long established keystrokes for footwork, mouse for aimpoint.  I could design this UI in my sleep because it's been a standard since the mid 1990s. 

 

Just don't tell ArmA players the same about its scrollmenu UI

 

Saying that UI is not a problem with CM. Once you get used to it (which is pretty quick, just not as quick as in more casual RTS games) - UI is not a problem at all.

 

Main problem with CMBS compared to CMSF at least for me is how "iterative" the improvements are. It's basically the same game but with features++. They are more than welcome of course and it's a great game but there was no innovation or at least new game elements in 8 years to make a real difference. The structure of campaigns is the same (except in CMBS thus far it's short non-branching campaigns), mission structure is still the same: basic text briefing->gameplay->dry post-mission stat screen. Thankfully we now have triggers but from what I understand you need separate dedicated AI controlled units to make them work - so still no real reactivity on AI part to ever changing conditions on the battlefield.

Saying all that I do enjoy CMBS immensely, the gameplay mechanics are second to none. But the series need a real step forward. For example even adding some basic, if not slightly dynamic strategic map with you being able to pick your own fights - even as a separate, additional gamemode - would make a real difference. It's what makes Close Combat players name CC games with dynamic map the best in the series. It's what makes Graviteam Tactics so replayable, unpredictable and enjoyable. And those are "hardcore" strategy games, right up there with CM. But even more "casual" games like Dawn of War (Dark Crusade and Soulstorm expacks), Total War and even Star Wars: Empire at War did that and it extended their staying power on HDD tenfold.

Edited March 29, 2016 by kraze

[Kinophile]

8 minutes ago, Raptorx7 said:

I am not going to dive deep into this discussion, but maybe an interesting idea would be to show off the UI concept Steve (once its done obviously) and see what players think on the forums. Maybe you have already done this with the beta testers, but I figured that might be a good way to take the temperature so to speak of how people feel about it.

Dunno, that could be a bit out of context, without the actual gameplay to test it. It would be just a Oh I like These Colors, I dont Like These Shapes, etc.

[AttorneyAtWar]

11 minutes ago, kinophile said:

Dunno, that could be a bit out of context, without the actual gameplay to test it. It would be just a Oh I like These Colors, I dont Like These Shapes, etc.

That would be true unless this UI is going to or could be adapted to CMx2, than we have plenty of experience to test from.

Edited March 29, 2016 by Raptorx7

[womble]

For me, probably the most important thing about the UI is that it is customisable. BFC have made some good progress along these lines, from "hotkey.txt" via the options menu, and the various different camera modes. Moving along this path so that more of the interface element parameters can be defined by the player would be a welcome further iteration. I would like to be able to define the "untypable" keystrokes, (like <DEL> for "Clear Target", and any of the <[modifier]+[alphanumeric]> combos. Being able to define what mousekey presses do would be good, and the colours and opacity of interface elements like deployment zones and map labels. And a command pane that scales to a defined size, whatever the screen resolution...

[Battlefront.com]

2 hours ago, kinophile said:

To be a bit clearer, I'm thinking more of "spreading the workload", e.g. farm out the more mundane, time intensive tasks, rather than opening up the engine for outsiders to tinker with and inevitably, break.

The problem is effort in requires money out for it to be worth doing.  The new Battle Pack concept is farmed out and early results are very good, so there will be more of that.  However, this was largely viable because it required very little management or engineering effort on our part compared to most other possibilities.  We have done partnerships for larger things, such as CM Afghanistan and CM Fortress Italy.  A lot of work but worth the effort.  Meaning, I won't rule out something like that, but the conditions have to be very special to make them good ideas.

2 hours ago, IanL said:

Not to mention that the way to successfully go about any UI design really comes down to being iterative. Clearly some up front work (what Steve is doing now) to define the goals and the basic concept and constructs is needed.  But anything that is designed will need refinement as and after it gets implemented.  No design has come off the drawing board and into real life as a "go away and code it project".  OK no successful design

Yup, and we learned that with both CMBO and CMSF, the first releases with brand new UI designs.  We spent almost an extra year reworking CMBO for CMBB and about that much time reworking CMSF to be CMBN.  Neither incorporated every change we envisioned.  The difference is that CMx2 continues to be improved with the stuff we didn't get into CMSF or CMBN.

2 hours ago, Raptorx7 said:

I am not going to dive deep into this discussion, but maybe an interesting idea would be to show off the UI concept Steve (once its done obviously) and see what players think on the forums. Maybe you have already done this with the beta testers, but I figured that might be a good way to take the temperature so to speak of how people feel about it.

Testers will certainly get to kick things around before we implement.  That's standard operating procedure for us.  But releasing the concepts to the masses tends to not work well.  It opens up the floodgates too much to people with too little understanding of how to design UIs.

2 hours ago, kraze said:

Main problem with CMBS compared to CMSF at least for me is how "iterative" the improvements are. It's basically the same game but with features++. They are more than welcome of course and it's a great game but there was no innovation or at least new game elements in 8 years to make a real difference.

I agree with the first part, I do not agree with the second part.  The biggest one I can think of right away is the Quick Battle system that is in CMSF and CMA, but not anything since.  The primary reason we redesigned it is people wanted different control (UI) over the units and the old system simply wasn't designed for that control.  It was a massive redesign effort and I think everybody who likes QBs would agree it's a major "difference".  We made lots of smaller changes which have made a "difference" to some players, such as the popup hierarchical menu and the different mouse control options.  But yes, in the end CM is still a 4x4 truck and not a Lamborghini or a hovercraft.

Quote

The structure of campaigns is the same (except in CMBS thus far it's short non-branching campaigns), mission structure is still the same: basic text briefing->gameplay->dry post-mission stat screen.

Change in one area comes at the expense of change in another area.  People don't complain about the campaign screens therefore we haven't focused our attention on them.  We could change them, but then something else would not be done.

Quote

Thankfully we now have triggers but from what I understand you need separate dedicated AI controlled units to make them work - so still no real reactivity on AI part to ever changing conditions on the battlefield.

The Triggers allows the AI Plan to be more flexible and context sensitive.  It does have the ability to be reactive, though not dynamically.  It is an area we intend to improve in the future.  AI is generally an "expensive" thing to invest in so it means lots of small and medium sized feature improvement requests won't be possible.

Quote

Saying all that I do enjoy CMBS immensely, the gameplay mechanics are second to none. But the series need a real step forward. For example even adding some basic, if not slightly dynamic strategic map with you being able to pick your own fights - even as a separate, additional gamemode - would make a real difference. It's what makes Close Combat players name CC games with dynamic map the best in the series. It's what makes Graviteam Tactics so replayable, unpredictable and enjoyable. And those are "hardcore" strategy games, right up there with CM. But even more "casual" games like Dawn of War (Dark Crusade and Soulstorm expacks), Total War and even Star Wars: Empire at War did that and it extended their staying power on HDD tenfold.

There are arguments for and against a "dynamic campaign" model.  We've been having such debates since CMBO days.  Players vastly, vastly, VASTLY underestimate the effort needed to pull off an even crappy implementation.  We'd probably have to stop everything else we do for the majority of a year just to get this one feature implemented.  My sense from all the discussions we've had in the past is we'd be skinned alive if we made that decision

Steve

[MikeyD]

This thread has gone weird.

[BletchleyGeek]

3 hours ago, Battlefront.com said:

The primary assumption you're making is that there is One UI To Rule Them All solution that is just begging to be coded without much more than a few day's design work.  Having worked with CM now for pretty close to 20 years I'm not confident that such a UI exists.  A better one than we currently have?  Sure, I believe that's possible since I'm working on exactly that.  But it's going to take me months of elapsed time to develop it.  There's a big difference in terms of what comes out of a long period of design that offers enough time for reflection and revision vs. narrow and ridged time.  I can definitely come up with a decent UI redo in 1-2 weeks, but it's risky.

Hold on a second. Maybe it wasn't clear what I was talking about - adding a "schematic map view" to the existing UI. You don't seem to completely understand what I mean by a "prototype" - a fully coded, integrated yet still in flux first iteration on the design. I understand that your way of doing things is more like writing up a proposal with some illustrations, open up a consultation process, etc. etc. What I believe in is to keep the written part short and sweet and present a fully-interactive, working and integrated prototype/draft etc. that not only serves to illustrate but also to ground the discussion into something "tangible". Then the kind of feedback you get changes: you get less "rants" about what the perfect UI would be and more "hey, that combo of font and layout looks a bit cramped". You already learnt some PHP, you can definitely learn some C++ and Charles looks to me as quite a sensei to learn from, Steve

3 hours ago, Battlefront.com said:

You're talking about it in the context of a fantasy game (i.e. one that doesn't exist anywhere on Earth except in your imagination). 

Well, there's a private GitHub repository with some work in progress... I like to test my ideas by doing stuff

 

3 hours ago, Battlefront.com said:

It's simply too difficult and too uncertain.  In my view it's akin to corporate suicide.  No thanks.

Quote

Fair enough, Steve. You're right that it is difficult and the outcome is not clear. But I wasn't really asking you guys to deliver that, but rather to introduce some elements to improve the "quality of life" for players that enjoy the "bigger is better" extreme.

3 hours ago, Battlefront.com said:

It's not viable.  As someone who has been in this business for most of his adult life, I find very few things I can think of that I'm more certain of than that.  It's akin to saying that we can have world peace if everybody just treated each other with respect.  Nice theory and I welcome you to have at trying to make it reality, but until then I'm happy to pay my tax Dollars to fund police and military protection for my person and property.

Quote

You mean you may not see it viable. I know literally dozens of people who do freelancing programming and design work, and have done so for almost 20 years. I have done it to get some cash on the side. YMMV indeed, and of course there are risks associated to it. But if you want to avoid those risks, you're my guest to go and seek some capital to operate in a more traditional way.

3 hours ago, Battlefront.com said:

UI designs support the game mechanics, not the other way around.  Proposing a heuristic UI for CM without making the game itself heuristic is simply a disaster concept.  Since CM is not heuristic, nor is it possible to become that, you've not really challenged any of the points I raised about the compromises necessary in UI design to make CM (the one that exists) work reasonably well.

Well, that "CM is not heuristic" may be a bit too much strong of a statement, if for heuristics you mean "rule of thumb, approximation".

In the TacAI you guys are using dozens of heuristics - say for instance the implementation of the Assault command for squads. How does your AI routine come up with the bounds each element of the squad needs to do? Is it somehow generating several candidate bound patterns and simulating possible outcomes resulting of each decision? Or is it rather something like "let x be range to assault order location, if x less than 200 meters then set bound to x / 2, otherwise set bound to x /3"?.

Also in many of your ballistic models there has been at some point a "captain's call" deciding where to put a stop to the faithful modelling of physics, enacting a rule of thumb, good enough approximation to avoid ridiculous amounts of computation.

Even your rendering engine is a massive collection of "heuristics", because what we see in our screens isn't a combo of pure raytracing and radiosity, but an approximation via triangles, textures and color blending.

Thanks for the answers, even if I see that we disagree in several fundamental respects, I appreciate we can have a conversation.

[BletchleyGeek]

By the way and going back on topic

Yes, buy CMBS, it's worth your money.

[A Canadian Cat]

1 hour ago, MikeyD said:

This thread has gone weird.

No way this has been very interesting.

[A Canadian Cat]

1 hour ago, Battlefront.com said:

Yup, and we learned that with both CMBO and CMSF, the first releases with brand new UI designs.  We spent almost an extra year reworking CMBO for CMBB and about that much time reworking CMSF to be CMBN.  Neither incorporated every change we envisioned.  The difference is that CMx2 continues to be improved with the stuff we didn't get into CMSF or CMBN.

 

 

 

 

 

Yeah that too. I was talking about iterating improvements and refinements during development...

1 hour ago, Battlefront.com said:

Testers will certainly get to kick things around before we implement.  That's standard operating procedure for us.

Yeah this. I am sure your work on the UI changed and improved during CMSF development.

 

1 hour ago, Battlefront.com said:

 But releasing the concepts to the masses tends to not work well.  It opens up the floodgates too much to people with too little understanding of how to design UIs.

Probably true, with the possible exception of a person or two on this thread.

[BletchleyGeek]

11 minutes ago, IanL said:

 

Quote

 But releasing the concepts to the masses tends to not work well.  It opens up the floodgates too much to people with too little understanding of how to design UIs.

Probably true, with the possible exception of a person or two on this thread.

Certainly not to "the masses", but yes to existing users who volunteer to give their opinion in a civil manner (i.e. you know who they are) or go through some sort of online survey (allowing for anonymous feedback which usually brings about surprisingly candid asessment and insight). Google Forms is awesomely functional and handy for the latter. Of course, the living prototype thing is and needs to be limited to a group of trusted (read here, people who can keep things secret) persons. The problem with that is to have in that "inner circle" - your beta testers / volunteers - a plurality of views and a culture that encourages robust yet civil discussion. But,  as one makes decisions, it is all but inevitable to "disappoint" someone, and those persons may and will eventually drop off and disconnect from the development process. How to keep a clear direction but still transmit to everyone that they haven't "fallen off the wagon" is something for which I sadly don't have an idea how can be solved.

 

Edited March 29, 2016 by BletchleyGeek

[Battlefront.com]

5 hours ago, BletchleyGeek said:

Hold on a second. Maybe it wasn't clear what I was talking about - adding a "schematic map view" to the existing UI. You don't seem to completely understand what I mean by a "prototype"

Ah!  Yes, a prototype of that nature is indeed a totally different beast.  But my point remains the same   We can not afford the expense of that elaborate of a "mockup" (which is the terminology we tend to use, though prototype is perfectly fine as well).  We don't have the time to do it ourselves and paying a contractor to do it (which involves our time as well) isn't likely going to work out in the end either.  Obviously if someone wants to spend 4-6 weeks of their free time making something for free, that's a different thing. Just not one that we can build a business plan on.

Quote

Well, there's a private GitHub repository with some work in progress... I like to test my ideas by doing stuff

As I said, you go straight to the head of the class GitHub is pretty slick.

Quote

Fair enough, Steve. You're right that it is difficult and the outcome is not clear. But I wasn't really asking you guys to deliver that, but rather to introduce some elements to improve the "quality of life" for players that enjoy the "bigger is better" extreme.

I'm not opposed to that at all.  In fact, one of the primary features in v4.0 is aimed squarely at the "bigger picture" type players.  But it is an iterative change and not a fundamental one.  A fundamental change is simply not viable.

Quote

You mean you may not see it viable. I know literally dozens of people who do freelancing programming and design work, and have done so for almost 20 years. I have done it to get some cash on the side. YMMV indeed, and of course there are risks associated to it. But if you want to avoid those risks, you're my guest to go and seek some capital to operate in a more traditional way.

We do outside contracting work on a regular basis.  In fact, that's how Phil got his start with us (we hired him to do the Mac port of CMx2).  However, he was brought in to pick up the ball dropped by the previous contractor tasked with the same thing.  Worked out well for us in the end as the first contractor, a friend of ours, refused to take any money because he left us high and dry.  This despite having done a significant amount of work already.  MikeyD, whom some of you might recognize from this very post, was brought on as a freelancer something like 10 years ago.  Chris was also brought on as a contract worker.  Our UI artwork is done by a freelancer.  We do contract work with probably a half dozen others.

The message here is that we're not shy when it comes to outsourcing work we know we need to do and we know will be paid back in the end.  We're shy about contracting out work that has a chance of putting us out of business or simply chewing up our margins.

Quote

Well, that "CM is not heuristic" may be a bit too much strong of a statement, if for heuristics you mean "rule of thumb, approximation".

Sure, some subsystems are.  But that's not what you were talking about.  You were talking about scaling the game system depending on level of play so you could have a command level game experience with tactical level details automatically working themselves out.  The game is fundamentally not designed to do that.  Which is why the UI is not designed to do that either.

Quote

Also in many of your ballistic models there has been at some point a "captain's call" deciding where to put a stop to the faithful modelling of physics, enacting a rule of thumb, good enough approximation to avoid ridiculous amounts of computation.

Actually, this is not the way CM's game guts are coded.  Either the feature has a rule of thumb built into its standard method for handling a specific circumstance, or it does it long hand.  There's no "well, it's been a busy few cycles so we'll just guess this time around". 

Quote

Thanks for the answers, even if I see that we disagree in several fundamental respects, I appreciate we can have a conversation.

Likewise.  Some people think that just because I take up a contrary position that I'm trying to shut down discussion.  The contrary is true.  The only way to arrive at a meaningful end point is for two or more sides to argue (in the useful sense of the term) until one side clearly carries the day or a definite impasse is reached.  Pandering to customers by saying "oh yes, you're right" or "that's something to think about, thanks" is about the most insulting thing I can think of doing to someone taking the time to express an opinion.

Steve

[Battlefront.com]

4 hours ago, BletchleyGeek said:

 But,  as one makes decisions, it is all but inevitable to "disappoint" someone, and those persons may and will eventually drop off and disconnect from the development process. How to keep a clear direction but still transmit to everyone that they haven't "fallen off the wagon" is something for which I sadly don't have an idea how can be solved.

This is why we engage in conversations/debates with our customers, but do not invite them to be a part of the design process in any official way.  If you ask people for their opinions they too often mistake it as being asked to have ownership in the outcome.  While most people here would probably be just fine, a certain percentage would muck up the system.

That is why we have a handful of carefully selected people as testers. In the past 17 years I can only think of one person was kicked off the time because he mistook his role as tester to be one of designer, stockholder, member of the board, etc.  That speaks to the quality of people we've chosen for our testers as well as our ability to sniff out quality.  There's lots more to choose from too, but too many people in the kitchen isn't a good idea.

Steve

[BletchleyGeek]

33 minutes ago, Battlefront.com said:

We do outside contracting work on a regular basis.  In fact, that's how Phil got his start with us (we hired him to do the Mac port of CMx2).  However, he was brought in to pick up the ball dropped by the previous contractor tasked with the same thing.  Worked out well for us in the end as the first contractor, a friend of ours, refused to take any money because he left us high and dry.  This despite having done a significant amount of work already.  MikeyD, whom some of you might recognize from this very post, was brought on as a freelancer something like 10 years ago.  Chris was also brought on as a contract worker.  Our UI artwork is done by a freelancer.  We do contract work with probably a half dozen others.

That's quite a story, Steve. Thanks for sharing.

33 minutes ago, Battlefront.com said:

I'm not opposed to that at all.  In fact, one of the primary features in v4.0 is aimed squarely at the "bigger picture" type players.  But it is an iterative change and not a fundamental one.  A fundamental change is simply not viable.

That is understood. And it does sounds really great. Another area where you can iterate and get a significant quality of life improvement would be the editor (allowing to switch overlays without having to close down the app and shuffling the "special map overlay.bmp" file for instance).

 

33 minutes ago, Battlefront.com said:

But that's not what you were talking about.  You were talking about scaling the game system depending on level of play so you could have a command level game experience with tactical level details automatically working themselves out.  The game is fundamentally not designed to do that.  Which is why the UI is not designed to do that either.

That wasn't my intention - I meant heuristics/algorithms as the particular way the "rule of thumb"/"long hand" approach is implemented at every level and for every task and behaviour that applies. The kind of approach you understood I was referring to does work for the "dynamic campaigns" we used to get on Falcon, Eurofighter or Operation Flashpoint (this last one I am not entirely sure), where pretty much everything happening where the player wasn't physically present was "faked" (i.e. simulated with a more or less crude model). The other setting where I have seen something like that is in military simulations (see "federated systems" where you can pretty much integrate a human empire at any level of the simulation, to do the "mechanical turk" and "implement" some aspect of the simulation, e.g. when it comes to determine whether or not a certain outcome is achieved or throwing nasty events for the participants to have fun with). Both are indeed very different from what you're guys doing, since one can pretty much go and check things on the ground pretty much anywhere at any time (within the scope of the scenario) and you don't get to see these men to put away the toys 

27 minutes ago, Battlefront.com said:

That is why we have a handful of carefully selected people as testers. In the past 17 years I can only think of one person was kicked off the time because he mistook his role as tester to be one of designer, stockholder, member of the board, etc. 

1 case like that in 17 years is a pretty good record. Yet users can be even more dangerous as they can totally destroy your credibility, and here I am thinking of what happened to Harpoon makers.

Let me wrap up the discussion and reconvene at some point in the future.

Edited March 30, 2016 by BletchleyGeek