Comments - Expack

Well, now everyone has full access to the game's source code: Frictionalgames.com

Bet this will help you quite a bit.

Seems odd that the ocean map as shown in this update has no horizon emulation (e.g. terrain and objects don't 'curve' out of view). Right now, it just looks like everything past the horizon disappears in a very unnatural way. (Or, in technical terms, it's obvious anything off the water map's top border is not being drawn anymore.)

Just from observing games which do have horizon emulation, such as Mickey Mania (see Youtube.com and Youtube.com ), I have a few suggestions:

* Put the ocean map on a custom bit of geometry which bends at the end. With proper UV mapping and some padding of the actual ocean map to account for the very edges of it, the horizon emulation would happen naturally.
* Use a shader to progressively 'roll' the map as it gets closer and closer to the top (obviously starting at a point near the visible edge of the map).

Hope this helps!

I know this'll be far-off, but might it be possible to extend the animation to characters on flat screens, like Father Elijah in Dead Money, the scientists in Old World Blues, or the Yes Man in the base game? It seems like lip-syncing, or a semblance of it, would be a logical extension of an animated and interactive (e.g. the Pip-Boy games) Pip-Boy interface.

What benefits do you forsee PhysX Flex providing for Excalibur? Yes, what is effectively a unified physics model is pretty cool and makes complex physical interactions (i.e. a cloth laying over a wooden bridge with wide slats being overburdened and breaking from the weight of collected rain) easier to do, but I'm interested to hear what sort of usage ideas have been tossed around.

I think you answered your own question: EA doesn't have a small team like Rebellion - which likely has the nimbleness a larger company/dev team like DICE lacks.

Based on the distinct lack of content, I suspect this game was built with a mixture of the mindset of Battlefield 1943 with limitations based on meeting requirements from the licencor (LucasArts/Disney) - that is, a side game to the main Battlefield experience, but using similar design elements from the main series with concept-appropriate modifications.

I also see one of two reasons why there are no space battles: 1. The audience the game will be marketed towards won't be interested in pure space battles. 2. The full version of the previously-mentioned IGN quote is merely PR speak for this:

"The last time we tried anything close to pure space combat (which is more like pure air battles from a design standpoint) in Battlefield 2142 - Titan mode - we ran into some MAJOR issues. The netcode caused odd behavior with the Titans and whatever was on them, they were the hardest thing to prevent exploiting of flaws (exploits)...in short, it was a pain, and we got a lot of flak for that. When we looked at doing space battles in Battlefront, we realized we still don't have the kind of experience to do space right, so instead of incur the wrath of Star Wars fans (and possibly our licencor), we're focusing on what we do best: planetary battles. It's not what everyone wants, but we think it's worse to make a game where half of it is utter garbage than making a game where everything just works - even if it's at the expense of cutting some core features." -John Doe, fictional character

Have you considered allowing the engine to procedurally create ships using the ship designs the players built? That way, the kinds of ships which are used for custom ships isn't based on what models already exist.

Why is the DirectSound software driver still being forced? Not enough people owning compatible soundcards? Unified experience? Something I'm missing?...

Ah! I was thinking that might be the reason - which is QUITE understandable - but I wanted to know the real reason first so I didn't make myself look like a fool. :)

What inspired G.R.I.T's unique look? To me, it looks like the Fallout universe finally met the 1980's, what with the more modern-looking metals and materials meeting all the grime, gunk, and intentionally-retro styling.

Can you give us an idea of what your art toolchain looks like? I imagine you try and utilize as many time-saving tools and techniques as will make your life easier while ensuring the level of quality you want.

I must say that the way things are colored makes them stand out VERY well - in fact, I'd be willing to wager that the coloring would hold up if you reduced the resolution to, oh, 320x240 and dithered everything with 256 colors or - if you're feeling especially retro - 64 colors! (I'm kidding about the wager part, BTW. I made the tweaks I mentioned to G.R.I.T. screenshots from IndieDB in Photoshop, hence I meant what I said about the coloring standing out.) In fact, I'd suggest you separate your DOF from pretty much every other game's DOF by desaturating what isn't in focus (see Edvardtoth.com ). In this way, the player's eyes will be further drawn towards the important things.

Also, you wouldn't happen to be looking for a voice actor who can learn tricks other than acting, would you?

Why not both on UE4?

I'm surprised that no mention was made in the changelog that hardware audio and EAX/EFX support was removed from the game in favor of software DirectSound support. Not only are sounds now too loud like they were in the original release, there's none of the nice echo and occlusion effects provided by EAX/EFX.

Thanks for pointing that out - I hadn't thought to look in archived news for some odd reason....

What does procedural generation DO in the case of this mod? It's not very clear in the announcement.

Dang...I just love the X-Ray Engine's volumetric god rays! They're so much more realistic and atmospheric than the screen-space god rays most games use - and they're the only kind which actually stays on-screen when you're not staring at the sun.

Regarding the strange reverb: I found that only happened in the original release with Generic Software as the sound device. If I changed it to my Creative SoundBlaster X-Fi sound card, which is capable of EAX5 and true hardware-accelerated sound via OpenAL, the reverb issues went away. Switching to my actual sound card also caused some sounds to correctly decrease and increase in distance, but I suspect that may be due to forcing elevation-based and close-range sound processing to be enabled though my sound card's control panel.

Where is Strelok's stash? I've looked all over the Research Facility underground, and I haven't found anything out-of-place nor have I found any way to access the ventilation system (the code on Strelok's book cover suggests his stash is somewhere in there).

The game crashes to the desktop when opening a safe in-game.

Very nice indeed - quite the radical difference. However, I'd be curious to see a comparison between the R2 and R3 renderers to see how much of a difference there is.

Looks very cool and detailed! However, the lack of shadows makes the scene look very odd and unnatural. Any plans to improve lighting later on so scenes like this can accept lighting and are self-shading?

Actually, the devs are planning on adding voxel water - possibly with physics (how complex has not be revealed yet). (See Board.pathea.net )

I absolutely LOVED how Earth 2150 combined three different varients on the RTS formula in one game! For example, the LC were the easiest to play as since their economy, base construction, and power grid was so simple, yet they lacked the more powerful weapons of the UCS and the ED; in contrast, the ED had the most convoluted economy and power grid, yet had access to the best weapons. It would be really cool to be able to do that here - for example, sacrifice better weapons for an easily-protected economy and power grid, sacrifice your early-game firepower for greater mid-to-endgame firepower, etc.

Just a question regarding behaviors: how advanced can we get with the behaviors? For example, if my minion has Kamikaze behavior, can I do things like have it ram a minion based on other behaviors? Along those lines, can I also have my minion drop certain (or all) weapons to lighten itself so it can increase the speed at which it rams with an enemy minion? Could I have certain weapons used by my minion have higher reload and/or repair priority over other weapons (i.e. have an anti-air minion's all-purpose miniguns reload first so it can defend itself against air and land target while its anti-air missles are repaired)?....

Good to hear. :)

I think one of the ways which the problem could be alleviated short-term is including a deathmatch mode where between 2-8 players (I'm not sure how well 9+ players would balance out), each on their own 'team', battles to see who remains standing. I think this may end up being a popular mode since many people who play CTF matches tend to treat it like a team deathmatch mode anyways - until, of course, the enemy reminds them what mode they're actually playing by starting to capture their base. In this way, you could have a mode which actually allows for small player numbers and much easier match-balancing until a larger player base is established.

I really appreciate how you and your fellow jammer team are using a Scrum-like development method one might use at a game jam - it potentially allows for faster bugfixing, adaptation to unintended issues, and iteration on gameplay modes and mechanics. It also has the potential for a smaller, more loyal fanbase who vocalize their thoughts, opinions, and feedback to you more easily since they feel more confident their ideas will be heard and acted upon. (I'm not saying this isn't the case now; I'm just saying I believe your development method will make this fact more evident to players than might be typical since you and your jammer team can act upon ideas deemed "good" quicker.) In fact, it's for this reason that I struggle to recommend Steam Greenlight for Robocraft since I imagine it'll be quite hard to jam with thousands of players simultaneously....

Also, beyond the issue I brought up above, the game thus far is very well balanced - even with subscriptions and purchasing parts using real money - because of how progress is tied to actually playing the game through TP and tier-based CPU points.

Thanks to you and your jammer team for making such an awesome game, and I hope to be able to continue to jam with you to make Robocraft as good as it can be!

If you're a low-tier player with weak weapons and you go up against a team of heavily-armed and multi-layered, heavily-armored robots - this combo seems preferred by high-tier players - they'll tear through your bot's armor easily and quickly destroy your pilot. Furthermore, this requires low-level players to focus fire to actually do damage to high-tier opponents - which is problematic since said high-tier players focus their fire on the weakest robots first. Since, in my experience, this tends to be the norm, low-tier players tend to die quicker than mid or high-tier players against other high-tier players regardless of skill, preventing them from earning much-needed RP and TP.

For someone unwilling to pay, the above situation may prove disconcerting to the point of leaving. However, for someone willing to pay, they can merely purchase a subscription - which, at $22.50-worth of GC for a three-month subscription (assuming a one-day membership requires only $0.25-worth of GC), is quite cheap - and get 50% more RP and TP for the subscription's duration. Thus, even when they end up dying quickly, they'll still get a nice amount of TP and RP. As they start rising the ranks, gain access to better parts - most importantly armor - and gain CPU points from rising the tiers, they'll start surviving longer and actually start improving their skills, meaning they'll make more kills. This can be further accelerated by purchasing and using GC to buy parts without needing to wait for enough RP. All this, in my opinion, leads to non-paying player being out-classed and out-gunned by paying players and being stuck as cannon-fodder for high-tier players until they acquire enough RP and TP to survive and compete.

Initially, I would agree for two reasons: tech points, which are needed to access anything more than Level 1 tech, cannot be bought, only accelerated in terms of the player's post-match acquisition of them via premium subscriptions, and CPU points, which limits how much stuff you can use to build a robot (on a per-robot basis) and is only increased once a higher tier is reached; combined, this means players willing to pay will still have to wait like everyone else to actually build their dream robots.

However, with the introduction of armor, I disagree. Long story short, I think armor has unintentionally tipped Robocraft towards pay-to-win thanks to reducing the amount of RP and TP non high-tier players - especially low-tier players - can earn without resorting to the 50% boost gained via a premium subscription due to the high-tier players decking their robots in multi-layer heavy armor and high-to-max tier weapons. However, if I may, I'd like to elaborate.

In terms of flyers, I would agree - those things to tend to be physically fragile. However, in terms of hovering robots being weaker than tanks, I disagree due to practical experience - the better higher-tier hover robots I've seen not only have thick layers of heavy armor encompassing the entire vehicle - including their hover blades - they also tend to have, at minimum, 8 of the highest-tier weapons available. As such, even though they have 6 degrees of vulnerability, they're just as hard to take down as tanks.

Correct me if I'm wrong, but couldn't players simply surround their hover blades with armored blocks, making it nigh-impossible to shoot them down that way?

Why not, say, FXAA or SMAA (Sub-pixel Morphological Anti-aliasing)? Both work as good, low-processing-cost AA via a post-process, either integrated into the game or as a injected effect via the NVIDIA Control Panel (although this only does FXAA) the AMD-approved RadeonPro application, depending on your graphics card, or a third-party injector. That way, you can still have anti-aliasing without needing to implement MSAA yet or worry about having a deferred renderer.

However, since both FXAA and SMAA do tend to blur the image - especially FXAA, which, when not tweaked or integrated directly into the game engine, can blur text as well - I'd definitely say NOT to jump into it as the defacto AA technique for in-game screenshots without doing some careful comparisons.

A game about insanity...interesting. Not many games like that - and certainly not this beautiful-looking. Looking forward to seeing the released product, whether it be free or commercial!

Thanks for the video link! Looks really cool.

Yes, I totally agree with your desire to avoid specifically making a "MIRV" part - after all, as you implied, that would defeat the purpose of players making their own.

Also, although I don't know if this will even affect you, the third-party physics engine which Unity3D uses by default, PhysX, tends to have serious issues with collision detection and can be performance-hungry since it's meant to be run on NVIDIA GPUs. Thankfully, there are other alternatives, such as Bullet, which is also free and open-source. For a short, succinct comparison of PhysX and Bullet, I'll refer you to this comparison video made by the developers of Rawbots, a Unity3D-based game where you can build and program your own robots using a wide variety of parts: Youtube.com

EDIT: Corrected some faulty facts on my part and clarified certain parts.