Dennis Shirk confirms DLL is on its way

[PawelS]

I don't think we need to worry about people that abandon the game after a year or two, mods are made mainly for dedicated fans who stay with it for longer. You know, Civ3 is still being modded And mods aren't made for money, so it's not that important how many people download them, but of course a solid user base is important to test the mod, provide feedback and ideas, or even join the development team.

An expansion is possible, as 2K Greg mentioned that "there are some much bigger plans we hope to announce in the coming months".

 

[orlanth]

It's not a question of whether WE will still be here. To me, it's a question of whether anyone else will still be around to download our mods. The traffic level in this forum is way down from what it was only a few months ago, so if it's another six months before we get the DLL and a year before mods really make use of it, it's just hard to see there being a huge demand at that point.

If you build it they will come Most of the really significant Civ4 modding only got going after quite a while and still found a big and appreciative audience. I think a lot of people are just not interested in DLC and minor "mods" that add a few civs and units to the existing game, but would be enthusiastic about total conversion mods that change the setting, remake mechanics, and effectively make it a new game. FWIW I started playing your AoM mods and definitely looking forward to them being finished, as well as the other total conversion mods in early stages of development on this board.

 

[Spatzimaus]

Most of the really significant Civ4 modding only got going after quite a while and still found a big and appreciative audience.

The problem with using Civ4 as a precedent is that it was an expansion-based game. Demand would slack off, then they'd release an expansion and suddenly the folks that had shelved the game came back all at once. Many of those folks (myself included) wouldn't keep track of which mods were progressing in the interim, but during that brief resurgence of activity I'd download pretty much anything that looked promising. If the expansions were good (like BTS) then I might stay actively playing the game for quite a while, but if they weren't then that period wouldn't last long. So if you were making a mod, it was only really a question of whether it was ready by the time the expansion came out.

We're not getting that with Civ5. DLCs just don't have the same sort of effect, and there haven't been any expansions announced. That's why I mention forum activity; without any reason to expect a sudden uptick in traffic, I just don't see the trend spontaneously reversing itself. And the Civ4 DLL came out about six months after release IIRC, meaning "good" mods were appearing within the first year; we're already over a year for Civ5, with a minimum of six more months until DLL-altering mods are out... that's a pretty long time.

Now who knows. Maybe there ARE a ton of people who'll jump from Civ4 to Civ5 once we get the good mods out. I'd love to see that happen; it's just that I don't think you can really use what happened with Civ4 as a good indicator of how Civ5 will progress.

 

[bitula]

I dont understand why does it take so long to handover that source. Is the dll source currently compiled into the main executable? But even if so, why would it take to separate it into a dll so long? Can someone explain? I believe there are some "political" reasons (like DLC buisness), not technical ones. I'm one needing the dll to rewrite the whole game, not just make it "better", it cant be made better for me, but could be a good base for a complete change. So adding new lula functions will not make me return to the game, and I think Im not alone with this.

 

[Pouakai]

I presume what they are working on is the support for DLL modifying mods.

 

[bitula]

I presume what they are working on is the support for DLL modifying mods.

Doesnt this mean that different mods just replace the dll? Well, aditionally probably swap them in-process (so as to avoid a restart of the game). This cant take many months to implement, can it? Btw, just publishing the dll source would be enough for a start, the extra cream could go later, dont we think so? The point is to start a larger community project as soon as possible.

 

[The_J]

In progress code is always a mess.
You have to clean that stuff up, comment it, bring structure in it, and take care of possibly not own code, which might be licensed from somewhere else.
And that definitely takes up a big bunch of time.

-> and to add there: Everyone who's crying "give us the source now" has never worked with a bad documented project. You don't want to see code, which's not meant to be seen by anyeone else.

 

[bitula]

In progress code is always a mess.
You have to clean that stuff up, comment it, bring structure in it, and take care of possibly not own code, which might be licensed from somewhere else.
And that definitely takes up a big bunch of time.

-> and to add there: Everyone who's crying "give us the source now" has never worked with a bad documented project. You don't want to see code, which's not meant to be seen by anyeone else.

Did the core dll exist from start or it is being created separetly? In the latter case some work needs to be done to assemble the code, but still not 14 months, or how much time passed since release? Btw, if the plan to release a core dll source existed from start of the project why wouldn't they implement it so from the very start? Just to create some extra work? Regarding documentation:
1) Commenting the code doesnt take 14 months.
2) Its not that important. A good C++ programmer can easily debug it and understand it, a bad one won't be able to understand it even with good comments.
Regarding structure:
In case the dll is being created after the release and they insist to have an ultra-nice structure, that may take long time. Thats why it would be nice to know wether the dll currently exists or not (probably can be looked into in the game folder). Another possibility is that the code is ill-structured and needs to be restructered, well, hopefully this is not the case...

 

[zarakand]

It's most likely a combination of resources and DLC sales for the delay.

 

[The_J]

Did the core dll exist from start or it is being created separetly?

o_O check your game folder, if there's a gamecore dll, or something like that.

In the latter case some work needs to be done to assemble the code, but still not 14 months, or how much time passed since release?

You're aware of the fact that development is still in progress, and that the code is probably changed every week, yes?

Btw, if the plan to release a core dll source existed from start of the project why wouldn't they implement it so from the very start?

You might guess so, but software architecture is always a mess.

2) Its not that important. A good C++ programmer can easily debug it and understand it, a bad one won't be able to understand it even with good comments.

Oh...sure.
I worked once on a project, which was probably 100 fold smaller than Civ.
There was no documentation, no instructions, no comments (nearly), and nobody who had a clue of the project.
It took us with 2 people 3 weeks of full worktime (so 8 hous per day) to hop through some of the basic functions, and another 2 weeks to find out why the stuff doesn't compile.
If you want a sourcecode like that, you'll see the first mods which utilize a custom dll in maybe 2 years, if even that.

 

[deanej]

And I thought civ4's DLL was badly commented.

 

[The_J]

Haven't seen many comments in there, but i also haven't looked deep into it.
But from what i've seen, at least function and variable names are often telling you enough. Which is already something, which is not always given. And you have no idea how it is for Civ5.

 

[bitula]

o_O check your game folder, if there's a gamecore dll, or something like that.

Uuuh, ok, but I have uninstalled it since long. So you say it doesnt exist yet? Very strange...

You're aware of the fact that development is still in progress, and that the code is probably changed every week, yes?

I know, but one can merge the diffs, shouldnt be a problem for a programmer, right? Btw, DLC-s add a bare minimum if at all anyithing to the core source and patches are quite minor nowadays. Not sure, when was the last patch?

Oh...sure.
I worked once on a project, which was probably 100 fold smaller than Civ.
There was no documentation, no instructions, no comments (nearly), and nobody who had a clue of the project.
It took us with 2 people 3 weeks of full worktime (so 8 hous per day) to hop through some of the basic functions, and another 2 weeks to find out why the stuff doesn't compile.
If you want a sourcecode like that, you'll see the first mods which utilize a custom dll in maybe 2 years, if even that.

Well it took me 2 months to understand (about 80%) a code with zero comments larger than CIV alone in my current company. BTW AFAIR CIV4 dll source was quite small, right? 1 week was more then enough to understand it to a level where you could actively change/enhance it. Im really confused that you say Civ5 codebase is huge. It should be comparably small IMO. Anyway we shouldnt bother with the larger part: the graphical engine code, just the game logic and engine api interface for graphics which would be in the coredll (and LUA). Well, but if you think that the whole project and the source code is a huge mess, thats a huge problem in itself, I dont even want to think of that, I want to pretend that firaxis experts create well structured and well designed code, and that they employ proficient programmers, not newbies who learn c++ by exprimenting in the source code, lol, but ok, I did see messy codes, sure, restructuring such codes frequently take more time then rewriting everyting from scratch.

 

[Valkrionn]

Here are some potential stumbling blocks. No idea if any apply or not, of course.

1) The code likely needs to be cleaned up and organized.
2) Using Steam exclusively (ie, as drm) likely mandates a different approach than was used for civ4 (distributed over Steam, but not subject to Steam drm). Could be a large stumbling block, here.
3) Quite likely that whoever is working on the DLL source also has to work on other aspects of the game.
4) They could have come up with something they want to implement that would make life easier for us when we do get the source.

Could easily go on, here. Point being, we have no idea what the cause of the delay is.

 

[deanej]

Why would steam affect releasing the DLL? Even with civ4, we didn't get the complete source (ex: no graphics engine stuff or DRM). Is there some steam agreement that says "thou shalt not distribute any part of the source with your game"? As far as I'm concerned, they can leave the DRM stuff in the exe.

 

[Valkrionn]

I didn't say it would prohibit it; All I meant was that it would necessitate a different loading architecture due to Steam's DRM requirements than was used in Civ4, meaning there is actual programming work to be done on their end to make it work.

In other words, the game likely can't load modded DLLs in the same way it did for civ4, and has to use another method, requiring development time. That's all.

 

[bitula]

I didn't say it would prohibit it; All I meant was that it would necessitate a different loading architecture due to Steam's DRM requirements than was used in Civ4, meaning there is actual programming work to be done on their end to make it work.

In other words, the game likely can't load modded DLLs in the same way it did for civ4, and has to use another method, requiring development time. That's all.

There is only one way to load a dll under windows: with windows api within the process, so I dont see what could Steam change about this. Some administrative structure needs to be implemented for loading these dlls in a generic manner, well let me estimate the effort: 1 week? Btw, maybe it is ready since long, last time Dennis wrote something like that they are testing the hot deploy, which I think is the dll load. But, I agree we have no idea why is the delay. If its aint a top secret Dennis could probably reveal the cause. At least wether it is technical or "strategical". I still believe it is related to the DLC buisness and that firaxis does not want to concurr with community programmers until it has income.

 

[bvanevery]

It's not a question of whether WE will still be here. To me, it's a question of whether anyone else will still be around to download our mods. The traffic level in this forum is way down from what it was only a few months ago, so if it's another six months before we get the DLL and a year before mods really make use of it, it's just hard to see there being a huge demand at that point.

Do you have an inherent objection to jumping ship to an open source project, such as Freeciv? There are tradeoffs, but at least you have control over your own coding destiny.

 

[Valkrionn]

There is only one way to load a dll under windows: with windows api within the process, so I dont see what could Steam change about this. Some administrative structure needs to be implemented for loading these dlls in a generic manner, well let me estimate the effort: 1 week? Btw, maybe it is ready since long, last time Dennis wrote something like that they are testing the hot deploy, which I think is the dll load. But, I agree we have no idea why is the delay. If its aint a top secret Dennis could probably reveal the cause. At least wether it is technical or "strategical". I still believe it is related to the DLC buisness and that firaxis does not want to concurr with community programmers until it has income.

Steam could enforce it's own methods for calling the same api. requiring a rewrite. Or it could be stricter about how the DLL is loaded (rather, WHEN and WHERE it is loaded). Or any number of similar things. Point is that it would be a delay; How long of a delay is a different matter. I never said it would be a long one, simply that it would exist, but you ignore one of the other points I made. Namely, the guy working on this would also be working on multiple other projects for the game. It's not a large group of people working on it.

The primary point behind my post is simply, we don't know exactly what the delay is, we don't know if it's some sinister plot to gain profit at the expense of a good modding community or simply not enough coders, not enough time. I dislike the "sinister plot" hypothesis, so I wanted to give a few examples of perfectly valid reasons for a delay.


All that said, it has been FAR too long of a delay.

 

[bitula]

Steam could enforce it's own methods for calling the same api.

What does steam has to do with a game's dll? IMO its being loaded from the exe. But anyway if you say like 1% of work resource is being allocated to the dll project, then ofcourse the delay is comprehensible. What are others doing? Expansion pack? Or they are on an other game project? I agree on that we donot know anything, but since it seems to be a secret I vote for the "profit plot" and not technical issues. Otherwise why couldnt they explain what is going on and what are the difficulties they face atm. 14 months have passed since release, this is abnormally long, dont you think so?