Impressions of a newcomer

posted on June 28th, 2012, 9:41 am

Hello! This is my first message on this forum, and I would like to greet everyone. It is somewhat lengthy, but it provides some feedback from an avarage human like me who walked by and came in..

A bit about myself: when I was a teenager, I was a big StarTrek fan and watched all the movies and played all the games, including the original Armada II... later on I favored more the harder sci-fi genres.. nevertheless I still love playing some StarTrek games (SFB included) from time to time. Just about a month ago or so I came across this STAII mod and it seemed very inviting - I liked all - the website, the screenshots, the graphics, the online database, the news section and the project progress section. So, after downloading and playing a little, I have decided to share some of the impressions I've got.

First of all - I was very impressed with the obvious amount of work put in this project. I remember the original STAII and now there is so much done, that the sheer splendour all around the project makes me happy and compels to play more. The gameplay mechanics, as described, promises much strategy and planning and fun...

However, in the real world, when I tried playing the game, I was losing a match after match. When I went to the website for answers, and dug deeper into the documentation, discovering what it's all about, I was overwhelmed by the complexity of the in-game relations and interactions. After reading the database again and again and playing several games with different factions and watching some replays from the video section, I finally started to understand something. It turned out that there is a lot of hidden peculiarities just about any aspect of the game. Since the game is mostly about fleet combination and ship battles, a big deal of the spice comes directly into those aspects. Soon I realized that the descriptions that are given when hovering above a ship's icon do not give even 50% of all the info that I should be aware of in order to succeed and it is impossible to win the game by knowing just that.

The system implies the following unobvious (but extensively documented textually) discriminations:
- ship sizes
- ship roles
- ship (not separate weapon!) firing ranges
- weapon damage for different roles and sizes
- ship targeting priorities based on ship roles
- passives
- mastering micro and specials
- did i miss something?

After spending some time with all that stuff and trying to play, I have realized that I can't memorize all this since the application of the rules to the ships seems arbitrary (aside from the build availability order). Any associations from other StarTrek games are either wrong or nonexistent. All in all, playing and finding a good strategy and tactics feels (for a noob like me) like mastering some hidden arcane knowledge which resembles alchemy the most. And if I have learned something about feds vs klingons, then I am totally lost if the next time I face romulans... The gameplay system is very unforgiving to mistakes, what makes learning even harder. I have studied the community and I guess that the old-time gamers were facing a steady growth of the gameplay system, which is very different from when all this just happens to you out of the blue.

From the mathematical point of view, the game system is very profound and admirable. Despite the inherit buginess of the engine, it manages to deliver very nice gameplay experience (once you get into it). And I think that it is some kind of disconnection between the visual and the logical parts of the game which brings some chaos into my perception. I guess I'm not one of the kind. Helping this could make the game more popular.

I really like the game and am continuing to learn, but I would like to ask making some more friendly tutorial system. What about making concise infographics charts with all information displayed in a highly condensed and perceptable view? (maybe online? interactive HTML5?)

Also, in my point of view, Fleet Operations, as a separate game, and the team has outgrown the Armada II engine and is restricted by its limits in their growth. Don't you have any plans on migrating to another engine, or, better, creating an own one? From what I see, you are showing a prodigious creativity, marvelous ability to work and great wisdom in creating the variegated and balanced system. I would definitely like for the project to evolve and clear more bars.

posted on June 28th, 2012, 10:25 am

In the next patch a lot of these things you mention are being changed . the devs have said they realise that the interactions are sometimes not obvious enough, and too many things have to be memorised. so they're changing it and making things simpler.

size is being removed in favour of a new system of profiles which will make things simpler. things will be simpler and tooltips will be more valuable.

targeting priorities are a big thing that are being changed. defensive profile ships will be shot at first, so you can more effectively add tanks to a fleet.

About the game engine, i don't think they have any plans to move engines. they've never expressed desire to make that huge leap before, and i think they are happy with what the a2 engine can do with a little 'persuasion'.

welcome to the forum

posted on June 28th, 2012, 12:09 pm

Good to hear about the upcoming patch, I have seen some rumors, but never seen an official announcement or any approximate date.. don't you know when?

And what's for the engine - yes, I understand that it is not that easy.. I just wish sometimes that could happen.. I am myself a software developer and I have experience from developing an engine for RTS from ground up (except for the low-level graphics) using OpenGL.. currently I am developing interactive software and some games for the employee, using JMonkeyEngine and Scala. I would say that moving to own engine is not that hard too. This is a real possibility one is interested. I could even take part in that, if that would happen sometime... From my experience I can say that today it is not necessary to use C++ for that, where a small team can spend years before any significant result. Today it all can be accomplished in a fairly lesser period..

posted on June 28th, 2012, 12:22 pm

Heyho and thanks for your detailed feedback!

You are indeed not the only one who found the game mechanics to be less beginner friendly. That's one of the key points we are currently working on for the next major version. Besides a completely revamped gameplay (see the last news for more details) we are also working on a better presentation so everything is obvious without having to dug deep into the guide. The guide should be for strategies, not to tell you what your ships do

The Armada 2 engine is not the most bug-free out there, but it is still quite capable. Nevertheless we are constantly working on improving things, including the renderer. The next major version will come with some performance and reliability improvements, too.

I take the freedom to move this post into the FO General section

posted on June 28th, 2012, 1:00 pm

Welcome noncom. You speak out what many people stated about Fleet Operations. It really has become hard to understand for new players. I play FO since Beta2 and until today I did not fully understand it for the reasons your posted. And honestly, reading through the guide and applying the gained knowledge is just too time consuming for me. It's like learning, learning, learning the game but totally fail to finally enjoy the game since you never stop learning. On the other side I'm happy that this is a game that withstands the tendency of the main stream culture to make things too simple and arcade like. But as Myles stated already there will be several changes with the next version that seek a balance between a too great complexity and interaction and a too simple system.

I think the current state of the game is the result of the process of implemented ideas. But now with the set of available features it is possible to rearrange them to a more logical and easier to understand unity. What was stated in the last news post provides a confident feeling about that issue.

The engine itself was a matter of discussion years ago. The devs had a look in quite some engines to possibly migrate but couldn't find one that served their needs. So they sticked to this one, completing the code and extending it continuously.

Again, welcome in our community.

posted on June 28th, 2012, 1:38 pm

Ok, good to know that! And I really wanted to add that as Jan said, I really like that the game is not arcade and involves some deep thinking processes. To no extent would I like to see it turning into an arcade RTS ))) The changes you are foretelling are promising and I hope that the system structure optimisation would go well, becoming easier to reason about but not stupid-simple like in many RTSs! Since despite its obscurity, it is one of the things that attracts the most. I would be glad to try out that new version! No approximate date is known though.

As for the engine - so looks like you have the source code. That is good because that would let you fix all the things eventually. A simple game assets and scripting mod could not hope for that which would be sad.

posted on June 28th, 2012, 2:13 pm

Good to hear about the upcoming patch, I have seen some rumors, but never seen an official announcement or any approximate date.. don't you know when?

Everything gets put in the News Section

The really big update was announced in a news post here and little things have been announced since.

posted on June 28th, 2012, 3:56 pm

As for the engine - so looks like you have the source code. That is good because that would let you fix all the things eventually. A simple game assets and scripting mod could not hope for that which would be sad.

Actually that's just it, they don't have the source. They're doing all of their engine modifications with hooks into the original, bypassing or suplimenting functionality. In point of fact, if they did have the source code development would be much faster and they could do a good deal more towards improving and altering the engine. I for one would love to see the ancient DX 8 graphics core replaced with a nice shiny DX 11 one. Unfortunatly the code (if it still exists) is property of MadDoc and Activision (or whatever they're called these days) and seeing as parts of Armada and or Armada II code are (supposedly) in Star Trek: Legacy as well, we probably won't be seeing it for a while. That said there may come a day when the engine has been tweeked and prodded to its limit; it will be interesting to see what happens then.

From my experience I can say that today it is not necessary to use C++ for that, where a small team can spend years before any significant result. Today it all can be accomplished in a fairly lesser period..

While it's true that there are a good number of engine bases out there which allow you to quickly scaffold your game up to a working point, these still require a good deal of work and customization to get the game logic in place. Furthermore your logic often ends up in the form of scripts with the engine becomming more of a sandbox. Thus your logic usually gets worse performance and that can be a game killer. Finally you're also often stuck with the graphics features the game offers and you're stuck with having to use the file formats the engine supports (this mainly applies to models). Ultimately one may be looking at the same amount of work as if you wrote your own engine. I should like to refer you to the Star Trek: Excalibur project. They're over there writing their own engine, have been for some time, and will for a while yet, but it promises to be lightyears beyond any known Trek game and will probably put most pro studios to shame.

posted on June 28th, 2012, 8:16 pm

The really big update was announced in a news post here and little things have been announced since.

Oh, I have read it long ago, I just did not realize that this is the new system and I had no big experience with the game then, so it just went in and went out. Thanks for pointing me there again, I will pay more attention.

That said there may come a day when the engine has been tweeked and prodded to its limit; it will be interesting to see what happens then.

Well, actually the limitation is already here. It does not show itself entirely yet, but I can clearly sense its presense. It stays out of view for many because of the habits developed over the course of time of living with the engine. Just a question of getting used. Accustomed to feel and live in terms of the present limitations, thoughts and imagination does not even try to step over the boundary. In other words, there are vast new possibilities which are unthinkable in terms of present means. And what stroke me very hard as one of the first impressions - and why I have included it in the "first impressions" post, - is the contrast between the great potential, dignity, talent and dedication of the team and their working environment. And the constant community demand for modifications is too, partly inspired by that hidden need. Oh well, I can't say the current implementation is bad, on contrary it is outstandingly perfect for its tooling, more beautiful than the tooling itself. And there are two ways from such a situation - dare and launch an entirely new stem or keep improving on what is there. I think that both ways have their benifits, but the daring way is inevitable sooner or later if the thing wants to survive.. that's just life and.. this project is alive. However, I see that currently the train of thought is to leave it as it is, improving the concept. In this case the game will still keep getting better and what is more important, the team will keep gaining experience which could later be, if they finally decide to switch, of a great use.

these still require a good deal of work and customization to get the game logic in place. Furthermore your logic often ends up in the form of scripts with the engine becomming more of a sandbox. Thus your logic usually gets worse performance and that can be a game killer

Yes, writing all requires elaboration. But it is fun and full of possibilities. I do not see big problems in logic being scripted or programmed (in case of opensource). Actually I think, there is no need to make the game as vastly moddable as Armada is, because you have the source, everyone has the source, so what's the point? Modular architecture would allow for easy customization, especially if it was written in languages like Java or C#. Scripting languages like Lua can be used for some designer-oriented things.

Finally you're also often stuck with the graphics features the game offers and you're stuck with having to use the file formats the engine supports (this mainly applies to models)

I do not quite understand what you mean here. There is a bunch of opensource graphical engines which already support many advanced graphic features. You can do shaders, GPU, CUDA, all the stuff. If it is not there, you can add. OpenGL or DirectX (I prefer OpenGL, because DirectX is a developers hell). And while its true that engines have their inherent asset loaders (how else?), you can write your own if you want. However, most game engines already have quite well implemented loaders of some popular formats. Currently I do not have these problems, working for production, so I am talking from experience, which might differ from yours.

I should like to refer you to the Star Trek: Excalibur project. They're over there writing their own engine, have been for some time, and will for a while yet, but it promises to be lightyears beyond any known Trek game and will probably put most pro studios to shame.

I am watching that project too and I totally admire it. They put a great deal of effort. BTW, looks like it has a similar relation with Bridge Commander as Fleet Operations has with Armada II (http://www.playhard.ru/Article.aspx?id=3699). They say that BC engine was exhausted so they are stepping to the new ground. I think same is true for Fleet Operations - either it steps over with this wonderful team, or with some another(can be?), or it never steps over. Quite simple here.. Excalibur is great, however, the graphics engine is not something unseen before. It is just that an engine of this quality has never been used in StarTrek games, that's true. There are FPS engines that graphically far surpass what an indie team could do. But besides the excellent graphics they also got control over how the universe itself is created and works. That is the real win. And honestly, I think that for an RTS, it is not necessary to have this kind of graphics as Excalibur has. Even the current graphics looks pretty nice thanks to the effort of the team. Surely it can be improved but there is no need to make it look better than Excalibur. As the practice shows, no one ever has time to examine the graphics engine lighting or shading details in RTS games. In my opinion, the biggest win for Fleet Operations would be the total control of the game world and its laws and what's possible there. If I were team, I would probably at least explore what's there, since, as far as I understand, much time has passed since the last attempt. The option to stay is always there )

Here, see, you got me as a techie ))) talking about all that programming stuff... I just want to say that all I talk about is a kind of a thought experiment, I am not claiming that I am right or know what to do or how life works or stuff... it's just a contemplation about programming games and that is my profession so it kinda gets me )))

posted on June 28th, 2012, 9:53 pm

Yes, writing all requires elaboration. But it is fun and full of possibilities. I do not see big problems in logic being scripted or programmed (in case of opensource). Actually I think, there is no need to make the game as vastly moddable as Armada is, because you have the source, everyone has the source, so what's the point? Modular architecture would allow for easy customization, especially if it was written in languages like Java or C#. Scripting languages like Lua can be used for some designer-oriented things.

Scripting works well for smaller systems or places where individual unit specifications are needed, but where core logic is conscerned you realy don't want it scripted for the sake of performance. That's not to say there's anything wrong with modularity; modularity is a fundamental principle of all programming, but scripting has its place and it's not core logic. As for moddability, that aspect has been the life blood of A2 (or all games for that matter; console games are missing out); FO is case in point. To not have the game as moddable as it has been would be its death.

I do not quite understand what you mean here. There is a bunch of opensource graphical engines which already support many advanced graphic features. You can do shaders, GPU, CUDA, all the stuff. If it is not there, you can add. OpenGL or DirectX (I prefer OpenGL, because DirectX is a developers hell). And while its true that engines have their inherent asset loaders (how else?), you can write your own if you want. However, most game engines already have quite well implemented loaders of some popular formats. Currently I do not have these problems, working for production, so I am talking from experience, which might differ from yours.

I was thinking in particular about model assets. A2 uses the fairly obscure .sod. While I believe there is a format specification floating around (I believe I even have a copy somewhere) it's not exactly a standard. And to port all models to a new format would be a fairly significant undertaking as would be writing a custom loader. Also, I find the DX 10+ API to be perfectly sensible - far better than DX9 - and VS 2012 is going to have DX debugging support so graphics development may improve. That said, working with the graphics card is never easy.

I am watching that project too and I totally admire it. They put a great deal of effort. BTW, looks like it has a similar relation with Bridge Commander as Fleet Operations has with Armada II (http://www.playhard.ru/Article.aspx?id=3699). They say that BC engine was exhausted so they are stepping to the new ground. I think same is true for Fleet Operations - either it steps over with this wonderful team, or with some another(can be?), or it never steps over. Quite simple here.. Excalibur is great, however, the graphics engine is not something unseen before. It is just that an engine of this quality has never been used in StarTrek games, that's true. There are FPS engines that graphically far surpass what an indie team could do. But besides the excellent graphics they also got control over how the universe itself is created and works. That is the real win. And honestly, I think that for an RTS, it is not necessary to have this kind of graphics as Excalibur has. Even the current graphics looks pretty nice thanks to the effort of the team. Surely it can be improved but there is no need to make it look better than Excalibur. As the practice shows, no one ever has time to examine the graphics engine lighting or shading details in RTS games. In my opinion, the biggest win for Fleet Operations would be the total control of the game world and its laws and what's possible there. If I were team, I would probably at least explore what's there, since, as far as I understand, much time has passed since the last attempt. The option to stay is always there )

I was speaking more in the sense that for being such a small group they have managed to produce something of professional quality and could therefore put a pro studio to shame. Furthermore, I suspect that the current DX9 graphics engine they have will not be what makes it into the final game. And yes, clearly the engine level of Excalibur is wholy unnecessary for an RTS (you would break the computer if you tried to get that level of detail into an RTS)

posted on June 28th, 2012, 10:27 pm

May I ask what kind of projects you are working on, noncom?

posted on June 29th, 2012, 1:54 am

Well, the most basic thing I could say is against AI, are you playing AI or people? People I don't really know. And the two are very different games. but for AI...

Which faction are you, and what faction is the AI? Federation is the easiest AI probably, Borg don't tech up fast, the small yard dominion AI is pretty easy, but if you let it get to large yard and are not prepared yourself it comes back with a vengeance, Klingon is difficult, but if you can handle their initial number offensive of lower ship quality you can come back in mid/late game, I find Romulan to be tough.

Than decide what faction you play best as and learn to beat the AI. I love Klingon faction vs AI Federation. Playing against two/three merci feds is loads of fun on maps with large number of moons and decent size (blue amber).

Consider AI strength, begin on normal ... pretty easy. Then hard. Eventually merciless.

Consider the map. Is incredibly small? 6400ish, or medium (-)15000(+), or larger 25000(+).

Right there, with AI faction, your faction, AI strength, and map size there are a lot of variables before even getting into any ship builds specifically.

In the beginning I found small maps particularly difficult if not impossible to beat AI, now I rather enjoy them.

Establishing economy (miners, mining stations, and good number of di and tri moons is critical). Doing this, and you'll find yourself working out what ships are good economical choices for biggest bang for your buck.

Just some general inputs.

posted on June 29th, 2012, 9:30 am

Scripting works well for smaller systems or places where individual unit specifications are needed, but where core logic is conscerned you realy don't want it scripted for the sake of performance. That's not to say there's anything wrong with modularity; modularity is a fundamental principle of all programming, but scripting has its place and it's not core logic. As for moddability, that aspect has been the life blood of A2 (or all games for that matter; console games are missing out); FO is case in point. To not have the game as moddable as it has been would be its death.

Ok, I see where you are getting at and agree with you. Making the game moddable is not only a good architecture pattern and fun in its own, it also turns it into a kind of a sandbox which is beloved by many for a good reason. It should not be too hard, even at the scale of an RTS and in its base it depends on the APIs used inside the game. And the quality of APIs is a direct result of good architecture and programmers work. Whatever goes on, the community must be happy.

I was thinking in particular about model assets. A2 uses the fairly obscure .sod. While I believe there is a format specification floating around (I believe I even have a copy somewhere) it's not exactly a standard. And to port all models to a new format would be a fairly significant undertaking as would be writing a custom loader

Well, all that looks like just a one plain significant undertaking, no?

Also, I find the DX 10+ API to be perfectly sensible - far better than DX9 - and VS 2012 is going to have DX debugging support so graphics development may improve. That said, working with the graphics card is never easy.

Yes, they have improved much in the latest versions of DX, actually they have changed the API almost entirely... so it became more adequate from my point of view. But for my deeply personal religious reasons, I still prefer not to use MSVS and stay away from DX. Working with GPU is not easy, but if it had once been exotics, novadays, it is a pretty common thing, and tools evolve.. I am not a GPU pro though.

Furthermore, I suspect that the current DX9 graphics engine they have will not be what makes it into the final game

I can't wait to see the working product. It promises to be very interesting and technically advanced, much to learn there. And I also would like to revisit a BC-like gamelpay. With all that procedurally-generated universe....

May I ask what kind of projects you are working on, noncom?

Currently I am working for a multimedia integration company. Now I am mainly developing normal input and multitouch interactive 2D and 3D software, augmented reality, optical recognition, kinect-based software... Some projects are more like simple games, some are plain multimedia, some are like educational or informational applications. That kind of stuff.

Just some general inputs.

Haha))) I was on contrary, playing only on small maps, did not think that the map size affects difficulty that much. I can easily beat normal always, so I plating only with hard. We have also played 2x2 with my friend against randoms and were always getting destroyed in the first 30 minutes Yeah, I think we should stick with race choices and practice much there, we'll do that!

Doh, the post is getting long (

posted on June 29th, 2012, 2:59 pm

Yeah, when I first started my style suited larger maps only.

I wanted to play a game where I got to my battleships, and so did the AI, then we duked it out. But that only worked at my level in the beginning on larger maps.

Now i can manage against a few of the harders AIs on smaller maps, I just changed build order from moving up the tech tree to going for stronger higher crew ships. Focus on capturing early enemy destroyers, and building larger number of smaller ships. And it makes it feel better when I do get a few battleships out on the board.

On larger maps, style changes. I get huge economy, completely ignore building weak vessels, and go right in to mid ship cruisers and large battleships. By the time enemy starts invading, my economy is built, and although I have fewer ships cause I tech up, ai is predictable and starts with higher number of weaker ships. So with some turrets it all works out.

Not to mention how the nebulas can be super helpful or harmful. i.E. Borg. green nebula is useless for borg, they have no shields to repair. Where as nebulas that disable shields don't deter borg at all, they have all hull. So if you get a fight with borg where you shields are disabled you are almost screwed. Terrain is pretty big.

posted on June 29th, 2012, 5:28 pm

Welcome! I suck at online matches too. If I see that your on Tunngle I will do a 1v1 with you if you would like.