Will there ever be a "release" version of C2C?

Hello C2C Devs,

A couple of years ago I had a computer that was almost 3 years old (built it in 2008, with gaming-level parts, though I never buy the extra-premium just-released/top-of-the-line parts). In either 2008 or 2009 (cannot remember), I went back and played a full game of Civ IV w/expansions (largest map, max players, etc...). I did play on a high speed, but all the way until the last turn, I experienced no noticeable turn lag.

Now I have the following PC specs:
i5-2500k 3.3ghz (Sandy Bridge, which is barely slower than the latest Ivy Bridge, as Ivy Bridge was the first Intel die shrink in a while to not actually increase general/gaming performance much...and unless there is something specific with c2c, the i7s offer little additional gaming performance compared to the i5s)
8GB RAM/Memory
160GB Intel 320 SSD (with Civ4/c2c installed on it)
AMD Radeon HD 6850 w/2GB Memory

I tried playing a full game of C2C recently, with all the advanced features enabled (except the ones that are known to be heavily unbalanced/problematic, not yet truly integrated into c2c). I used a Gigantic PerfectMongoose 3.1 map, the default number of civs for Gigantic (15), and Marathon game speed. Despite thoroughly enjoying the general gameplay of c2c (I restarted many times after playing multiple games for several hours, as I discovered new strategies / tactics that I knew would have helped me get a better start), in my final and by far the longest game, I had to permanently quit shortly after reaching AD. Why? I was already having around 1.5 minute turn lag, and it was only slowly but surely getting worse.

A more powerful computer (and really, still very close to top-of-the-line in CPU power, gaming wise) was having 1.5minute turn lag on c2c while a slower computer was having zero turn lag on default Civ IV Beyond the Sword.

Then there is the major balance issues I keep reading about everywhere, though I cannot speak of those myself since I have not played long or regularly.

Then there is the stability / crash problems (I have had plenty of crashes not even caused by MAFs...if I didn't have a SSD, I probably would have quit long ago due to that alone).

Then there is the lack of automation options for the "animal herding" part of the game all the way through to the Classical age (or later...never have got past the Classical part). It is fun at first, but then it gets old micromanaging all the animals captured again and again.

Then there is the current state of War AI. The most powerful civilization in the game, with a 50% higher score than any other and a 100% higher score than my own, was barely fighting me after he declared war on me, despite his cities surrounding my own and me actively attacking him.

This is "version 27" of a mod that appears to have begun over four years ago, and the core of which is based on mods begun even earlier. Yet its "stable versions" are still of the same quality level of "early beta" mods, with SVN versions being alpha-like. Yes, I know the base Civ IV code is an unoptimized mess and you guys have drastically optimized it. Yes, I know that you have added so much that the optimizations' gains in performance have been nullified by that continuously growing content. Yes, I know you guys are hardcore Civ players that are willing to deal with a lot to play with these advanced features. Yes, I know you guys are volunteers using your own unpaid time to provide a free product. Yes, I know you guys openly go with the "more is better" philosophy.

However, you have taken the "more is better" just too far, if one of your goals is still to to provide an enjoyable experience to not just the most extreme hardcore civ players. It seems like most of the devs even admit that they rarely play past the Classical age anymore due to the various issues with the mod. I can very easily see how some of you might enjoy working on the mod as a hobby much more than you do actually playing it, and there is nothing wrong with that if that is the case. But then you should be honest that there is no longer a goal for a playable mod, and that this is more of an experimental modding hobby group oriented towards fellow modders, rather than a mod oriented towards fans/players.

Quantity is great as long as all it does is add content and/or additional complexity to the game. I know many games/mods go with the "less is more" to reduce complexity and appeal to much more "simple" gamers. But when the quantity becomes so great as to significantly reduce the playability of the game (such as massive balance, lag, and stability issues), the line has been crossed on how much quantity is good. I am not saying that a great, playable C2C mod with all the advanced features (both present and future) is not possible. But I am saying that, if the goal of the mod is still to provide an enjoyable experience to common Civ players/fans, the devs should slam the brakes on new content/features and figure out how the heck to get all the stuff you have now to work properly, THEN start adding more. In some cases it might even be necessary to wait for the next release of CPUs to start adding new super-advanced features to a release version that is playable performance-wise. Perhaps divide c2c into three different release types instead of just two: Release (reasonably stability, balance, and performance with gaming-level PCs), Beta (the "stable versions" you have now), SVN.

Because all I see now is "yes we realize X, Y, and Z are major problems, but we need to fix A first, but before we fix A we need to finish developing/adding B, but before we finish developing/adding B, we need to fix C, but before we fix C we need to finish developing/adding D, but before we finish developing/adding D, we need to fix E, but..." infinitely. There are big problems that need to be fixed, and yet there is still an infinite amount of new content in the pipeline that you guys add before fixing the current big problems, and that infinite new content comes with a supply of infinite new problems.

Sorry for the rant, but from reading your posts it just feels like you guys do not even truly realize the situation, and as I really like the premise of the mod (and actually tried it for a short time many months ago before I got busy in RL), it is frustrating that I see no playable version in sight, with the current "more is better, at any cost" philosophy. I am not saying you owe fans anything, but if your goal is to keep existing fans and bring in new ones...

I know you may have recognized this but throughout the adding of additional content from the beginning of our work on C2C (coming from RoM/AND) we have improved on all of the issues you mentioned regarding processing, with the one exception that it has become a bit more RAM demanding.

In watching the profiler run through turns, it appears to me our biggest cause for turn delay is not at all the amount of content we have so much as it is the enormous amount of detailed analysis the AI undergoes to reach conclusions about what to research and what to build. All that analysis does make for a much better ai so is the tradeoff there? Hard to say for now since the AI still lacks a lot in the military department despite its massive quantum leap in the handling of civil matters.

Load time will always be a matter of the content volume, but over the last 4 versions or so we've seen HUGE improvements in that department.

We have a few modders here with the skills to so vastly streamline for performance and they are both on a bit of a hiatus for the moment so the rest of us continue with our own tasks, goals, and responsibilities as needed. Will there be a final release? Sure... in a few years perhaps. But there IS still a lot more to achieve to realize the 'vision' and as we go, we keep repairing bugs here and there, I feel at a rate that exceeds the emergence of new ones (with the exception of OOS errors at the moment unfortunately.)

We all do feel a lot of responsibility for the fans. And I'm sure long term fans would be able to note a lot of improvement has taken place in parallel to the new content! I'm quite certain that will continue for some time to come until we have a much more polished mod overall. Improving modder skills (I'm pointing at myself here!) will also contribute to that factor throughout the process.

Apart from FFH, and probably vanilla RFC since that was included with BtS, the most popular mods are RoM and C2C. The critics, however constructive, need to stop kidding themselves (that they speak for some nebulous majority).

Welcome to the wonderful world of hobby-made products where we have a very finite number of people working on this for months at a time and would really love to take a break from A to work on B. Look at it this way, you love ham, but must eat ham for breakfast, lunch and dinner. Very quickly you'll get sick of ham and want turkey instead. If you were to get paid for eating ham, sure you could muscle it down until you're served turkey, but here we have no one nor (in the spirit of free mods - and staying lawsuit-free) want anyone to pay us for eating ham.

And, believe it or not, even when people DO get paid for creating stuff, this pattern you highlighted is still prevalent - just look at Bethesda.

It is my belief that the C2C team is amazing for the work they do during their hobby time without getting paid. They truly are a spectacle to behold as it is modding groups like this that go on to create indie companies and games that put the big companies on their toes that allows all other games you play to be as high-quality as they are. As it is now, Indie games dominate the gaming industry at an astonishing rate of 2 to 1.

SuperFreeza said:

.... It seems like most of the devs even admit that they rarely play past the Classical age anymore due to the various issues with the mod....

Click to expand...

You are a bit behind the times; now it is the Renaissance we don't play past often. The reason is that we think we have most of the problems ironed out in the prehistoric to classical and are now working on the game play in the middle ages and Renaissance to early industrial while still adding content to the modern and later eras and tidying up the early periods. Yes there are still some big things to add which will cross all time periods, extending the diplomacy for instance.

SuperFreeza said:

Then there is the lack of automation options for the "animal herding" part of the game all the way through to the Classical age (or later...never have got past the Classical part). It is fun at first, but then it gets old micromanaging all the animals captured again and again.

Click to expand...

There is AI for this, it is "just" not available for the player. This is also complex, I could probably code something now for horse/camel/elephant to go to the nearest city without any horse/camel/elephant herds or in vicinity resource or the associated national wonder. Wolves I would probably go with build Prehistoric Wolf myth, then in the recon building city build Dog Breeder and Kennel and wolf cage, else build Dog Breeder in any city. At least i could if I could figure out what it is that I am doing wrong with the Mission and Outcome system. So I can come up with a number of buttons which would reduce the micro management but not by much.

I tried playing a full game of C2C recently, with all the advanced features enabled (except the ones that are known to be heavily unbalanced/problematic, not yet truly integrated into c2c). I used a Gigantic PerfectMongoose 3.1 map, the default number of civs for Gigantic (15), and Marathon game speed. Despite thoroughly enjoying the general gameplay of c2c (I restarted many times after playing multiple games for several hours, as I discovered new strategies / tactics that I knew would have helped me get a better start), in my final and by far the longest game, I had to permanently quit shortly after reaching AD. Why? I was already having around 1.5 minute turn lag, and it was only slowly but surely getting worse.

Click to expand...

Gigantic PM map with 15 AI, Marathon, 1.5 minute EoTs at AD, not bad. Were all 15 AI still alive?

Were you using REV and had more AI being added to the game thru revolutions? Did you have Barb World or Barb Civ Options "On". How many AI were in the game still at AD, that's the biggest question that needs answered.

Still 1.5 minutes on a Gigantic PM map with 15 AI is not bad for a 32 bit game engine that is close to 9 yrs old.

Why do you think it's unplayable? It is a turn based game after all not RTS of FPS. Much more data to be stored and processed each turn and only using 1 core at that.

Instead of trying the upper limits why not test it on an avg setting? Huge Map (PM is fine), 10 AI, NO Rev, No Barb Civ, Epic game speed, Monarch difficulty level with Standard Leaders (so you won't be using the New Leader Trait system which is a work in progress) and see how it goes. Would be a more realistic test.

Game enough to give it another go at these settings?

JosEPh

Welcome to CFC!

I can see both sides of the coin (I'm not affiliated with C2C, by the way). I agree with you, SuperFreeza, that C2C could use some more polish. I started a game of it (I think version 2.5, so slightly old) early this fall, and it was tons of fun until around 2000 BC, when it crashed for no apparent reason. I lost 4 turns where a lot had happened, so I didn't feel like re-playing them right away. And then I got into other stuff before I returned to it, and, well, I haven't played C2C since then. In part because I figure there's a good chance that it'd just crash again if I picked up that same save game, or if I started a new game, it might also crash eventually. I have played vanilla Civ4 since then, which I have confidence won't crash (at least in single player). So, the crashes in particular are definitely decreasing C2C's appeal with at least some Civ fans.

On the other hand, I do development (both as a hobby and as a job), so I'm definitely familiar with certain tasks being much more appealing to work on than others. And in terms of performance and balance, that's stuff where you have to be careful when changing too, so you don't accidentally make things worse. But in general, the enjoyable development will be prioritized if it reasonably can be.

Nevertheless, as a developer, both professionally and for hobby projects, crashes are something that have to be high priority in almost any workflow. That's not to say no crashes will slip through the cracks, but they should be fixed relatively quickly. I don't know what the C2C project's status in this regard is - but if I were to consider funding a modding team to develop a full game, I'd certainly want to see crashes and other serious issues be a top priority, well above balance and above performance as well.

If we're going by analogies, my spur-of-the-moment one would be that you're the household repairman, and crashes are the equivalent of leaky roofs. You might prefer replacing the electrical wiring, replacing the floors, or putting up new wallpaper, but you really ought to stop the leaks before working on things under the roof (and sadly, just placing a tarp over the roof doesn't count in this case).

CTD crashes are actually not ALL that hard to fix in the big picture and are yes, the top priority. It is VERY likely that whatever crash you experienced has been repaired and if not, its only because you never sent us a savegame that displays the issue! lol

Note to all: Send in saves that lead to repeatable crashes whenever you encounter one!!!

SuperFreeza said:

Because all I see now is "yes we realize X, Y, and Z are major problems, but we need to fix A first, but before we fix A we need to finish developing/adding B, but before we finish developing/adding B, we need to fix C, but before we fix C we need to finish developing/adding D, but before we finish developing/adding D, we need to fix E, but..." infinitely. There are big problems that need to be fixed, and yet there is still an infinite amount of new content in the pipeline that you guys add before fixing the current big problems, and that infinite new content comes with a supply of infinite new problems.

Click to expand...

All I can say is that fixing X, Y and Z is usually only temporary solution since doing A, B and C usually changes the balance of X, Y and Z. And there have been times were we HAVE fixed X, Y and/or Z but it has been just unbalanced again by further experimentation and balancing. Such as the Civics or the tech tree.

The game is more like a living ecosystem where each modder has their own niche. When one adds something it can effects all the other modders and of course the player too. Creating a balanced game is like making a balanced ecosystem. Sure one can break it down into a simple food chain, but it thrives best when its a complex food web. Where each part is connected to the other and works in harmony.

And much like nature through trial and error we weed out what works and what doesn't work to eventually select out the best stuff. Sure we have have our "vestigial parts" or "junk DNA". But I think we come out stronger and more robust.

The main problem with your comparison is the map size. You picked a map that has, I think, something over twice as many plots as the biggest regular BtS map has. Try comparing C2C to BtS on the same size map. C2C will still be noticeably slower, but not as much slower.

It also has vastly more "stuff" in it, so it takes a lot longer for the AI to determine the value of each thing that it can build at that time, evaluate civics, pick a tech to research, etc. C2C adds more stuff to that game than BtS came with in the first place. A lot more. With way over twice the buildings and such, it should not surprise you that it takes longer to figure out what to do.

So try a more normal size map and see how the turn time is. Like perhaps "Large", to begin with.

Currently playing a large map [Terra] right now... Just started the Industrial era, about 20 seconds and then it gets to combat outside of my turn.
10 AI players, started with less than that, but REV (not BARB civ though) on.
My actual turn takes longer, because I don't mind micro-ing my 37 cities.
Pride of Nations takes FAR longer to process a turn, I tend to do other chores around the house...
[3302/14305 turns, 959 BC - Eternity speed]

God-Emperor said:

So try a more normal size map and see how the turn time is. Like perhaps "Large", to begin with.

Click to expand...

I agree here, Large is always "fun" to play, and i make it to the TransHuman Era each time withOUT a CTD, now. Course i am more of a builder type player, so i keep fighting to a very very low minimum. I just love the way it plays out this way, but be vary wary of this approach, in the Modern Era you'll probably be over run, if your not careful.

Acularius said:

[3302/14305 turns, 959 BC - Eternity speed]

Click to expand...

No way, I can barely play with marathon speed, because i am very very impatient.

Quintillus said:

Welcome to CFC!

Click to expand...

I agree here completely and also for trying this Great Mod, C2C

JosEPh_II said:

Gigantic PM map with 15 AI, Marathon, 1.5 minute EoTs at AD, not bad. Were all 15 AI still alive?

Were you using REV and had more AI being added to the game thru revolutions? Did you have Barb World or Barb Civ Options "On". How many AI were in the game still at AD, that's the biggest question that needs answered.

Still 1.5 minutes on a Gigantic PM map with 15 AI is not bad for a 32 bit game engine that is close to 9 yrs old.

Why do you think it's unplayable? It is a turn based game after all not RTS of FPS. Much more data to be stored and processed each turn and only using 1 core at that.

Instead of trying the upper limits why not test it on an avg setting? Huge Map (PM is fine), 10 AI, NO Rev, No Barb Civ, Epic game speed, Monarch difficulty level with Standard Leaders (so you won't be using the New Leader Trait system which is a work in progress) and see how it goes. Would be a more realistic test.

Game enough to give it another go at these settings?

JosEPh

Click to expand...

1.5 mins at that point is pretty good. Now the real thing that will speed up turns is not reducing the AI processing (which will make the AI worse), but to multi-thread it. The Firaxis devs kindly included a mutex and other threading things, so we can safely multi-thread parts of the DLL which do not call the main engine (or more specifically where we can control the lifetime of an object exactly).

ls612 said:

1.5 mins at that point is pretty good. Now the real thing that will speed up turns is not reducing the AI processing (which will make the AI worse), but to multi-thread it. The Firaxis devs kindly included a mutex and other threading things, so we can safely multi-thread parts of the DLL which do not call the main engine (or more specifically where we can control the lifetime of an object exactly).

Click to expand...

They included threading things? Where?

AIAndy said:

They included threading things? Where?

Click to expand...

The mutex, recursive_mutex, and read_write_mutex object files to name some. I could swear that back in september you said that this was good news for efforts to multi-thread C2C, but I certainly could be remembering poorly.

ls612 said:

The mutex, recursive_mutex, and read_write_mutex object files to name some. I could swear that back in september you said that this was good news for efforts to multi-thread C2C, but I certainly could be remembering poorly.

Click to expand...

I added those. The good news back then was that the runtime used by the exe (and therefore also the DLL) is the multithreaded version so there is no problem with memory allocation from threads.

AIAndy said:

I added those. The good news back then was that the runtime used by the exe (and therefore also the DLL) is the multithreaded version so there is no problem with memory allocation from threads.

Click to expand...

Ah, my bad. It still means what I said above though, that we can multi-thread anything where we can completely control an object's lifetime (IE, most of the AI), right?

ls612 said:

Ah, my bad. It still means what I said above though, that we can multi-thread anything where we can completely control an object's lifetime (IE, most of the AI), right?

Click to expand...

Yeah, we can multi-thread most things (the limits are where calls to the exe are involved).

Thunderbrd said:

{QUOTE=AIAndy;12171841}Yeah, we can multi-thread most%2

Click to expand...

That was the edited version?