Symmetrical planets are getting there...

Exactly, all you have to do is record the order that you place the features in (with what ever editor you end up using) and you have your seed.

You don't have to generate and compare it to every possible planet because seed X will always generate planet X.

I am starting to get that creepy feeling, that there is a different understanding of what a seed actually is.

Can you explain a bit what a seed is for you?
Say, something like a single number (like a 6/8/32/etc digit number), or maybe a record of all positions and types of all features placed on the planet?

criz.

Does not having HR to work on anymore give you more time for things like this? Not trying to be a jerk, or anything, I'm just curious where devs are spending their time, currently, as I'm sure thousands of other players are too. It doesn't matter to me if the things mentioned get finished, but I'm curious where the efforts are spent. It's that reason that jables posted this thread in the first place, right?

There weren't any engineers assigned to HR; so there wasn't any impact on PA.

jables, please have them do something about Orbital to ground invasion balance! Currently, it's almost impossible to crack a planet in a longer multiplayer game, when there are no game enders (like smashable planets, or Annihilasers) left in play. Even if one player is more skilled and has a better economy, the planet turtling player can just sit on a planet and fill every available space with nothing but Anti-Nukes and Umbrellas (which also makes it impossible for an invading force to build a teleporter on the ground). Both defenses are currently cheaper and much more effective than their offensive counterparts, which makes it unbearably tedious to play the game at that point. This tends to make the game more of a test of wills to see who will get tired of playing first, or who can build enough units to make the game crash. I've been involved in many of these types of games to date (with the most recent being just last night), and it's not much fun for anyone involved. This, in my mind, is one of the worst aspects of this game currently.

Obviously a simple balance fix would be to adjust the build cost of the defensive structures, which is probably needed, but I'd be more interested in increasing the HP of an Anchor (which can currently be killed relatively easily by both orbital fighters and umbrellas), or doing both cost adjustments and HP adjustments for the units mentioned. An even more interesting option would be to introduce another unit type, something orbital with a lot of HP. It could be something hugely expensive, like a harbinger type unit in space, or maybe even just an orbital teleporter? One of the pains of trying to crack a planet is the time it takes to bring reinforcements to the enemy planet, which also eliminates the element of surprise as the enemy can see your forces coming in space. An orbital teleporter would eliminate that problem, and also enable the leveraging of tech that's already in the game. An orbital teleporter unit that would allow you to drop air units into the atmosphere below would be kinda of cool too, and provide another interesting solution to the orbital invasion problem.

I know that has been said, but isn't that Varrak prominently displayed (in yellow) in the photo below from the HR KS page? He's also in the HR KS video talking about being interested to get to work on the game. So, if what you are saying were true, then what was he doing there? Not to mention the other engineers in the photo.

He works on the engine, given that the games were planned to use the same engine...

Do you know what a seed is? It is simply a number, that initialises a pseudo random number generator. Most of these generators have an absolutely huge period, that means that they repeat after lets say 2^12391 iterations. Now, what you are trying to say, is that it will be easy to find a seed, such that the resulting sequence matches your predetermined sequence. Lets do some math:

Let's assume you have only 10000 numbers in our sequence.
We also need to assume the size of the PRNG output space, so lets assume 32 bit output.

Probability of getting the seed right = (correct numbers in space / size of space) ^ number of matches required)
(10000/2^32)^10000 = 10^-56329ish

That is a really small number. And if you could try 1 billion seeds every second (testing each of the 10000 numbers in the sequence) it will only take you about 10^56312 years at worse. Not bad at all.

Watch the video again, it was definitely implied that there would be more engineering effort to get everything to work properly in HR. In any case, I didn't mean to touch off an HR discussion. I was just trying to get some extra info about dev work on PA.

sure there was more engineering involved once HR development work would have started for real, so once the ks is successful. That never happened.

It'll probably be even worse when you finally finish at the end of time and realize the sequence you searched for does not even exist in the whole possible output of the prng.

Which was exactly my point.

I think we have been confusing planet seeds with random seeds...

Would I be correct in saying that it works the following way:
- Seed the pseudo random number generator.
- Have the pseudo random number generator generate a planet number/seed
-Take the planet seed and generate a planet.



I'm talking about reversing the planet seed, not the random number. If X number will always give Y planet, then we should be able to go backwards and take Y planet and get X number.

The planet seed is the random number seed. That is, the random number generator is seeded with the planet seed number.
However, even if it wasn't, it's still not feasible to reverse it, assuming you start with a planet that has a seed that matches.

So in order for planets to come out the same for every client, the random number generator has to use the same formula for each client.

If we know the formula and the result. How can we not be able to calculate what seed was used?

Because there's no way to automatically reverse it - the only way is to pick a seed, generate the planet off it, then compare that planet to the original planet to see if it matches. If not, pick the next seed and try again, and so on. Not impossible, but not feasible, since it would take years for a single correct match .

Edit: It's like the opposite of a hashing function; a hashing function takes some input and generates a small string that represents it (but doesn't encode it). You can tell if the input matches a hash, but you can't re-create the input from the hash. Planet generation is the opposite; you can generate the output from the seed, but can't get the seed from the output. In both cases (hashes & seeds), it's possible in theory to brute force it by trying every combination; realistically though, it's not a feasible solution.

Edit Edit: This is assuming you are just given a generated planet. If you can monitor and store information about the planet generation process for the particular planet, then it becomes more feasible (especially since you are guaranteed that the resulting planet will match a seed, whereas an arbitrary planet will almost certainly not match one). But in that case, you already have the seed anyway

Lore is wholly unimportant and I actually wished it had never existed in this game. The resources spent on it (however large or small the spending was) came from our pockets.

When I play a video game I do it because I want to game. When I want a story I go and read a novel. Storytelling is not the strength of the medium (video games) and that's why good writers are never drawn to video games and there are no good examples of storytelling in video games, if you think there are then you are not reading enough.

you seem to have no idea about gamestorytelling ... games dont need to tell stories like movies or novels ... the biggest strengs in gaming is that you can be part of it and have story be told that way or by your actions ... movies or novels won't ever be able to tell a story like a videogame can... it does't matter what you think about storytelling in videogames ... fact is you can do it in far more creative ways novels or movies ever would allow you to ...

If you think a game should tell a story the way a novel does then you're not playing enough.

There is so many stories in games or even a game it would scare away most novelwriters who work alone. Guild Wars 2 -> 3 books, Halo Universe -> 6+ books (not counting reproductions, Starcraft Universe .. well lets say easily 20+ Books.

Now They are novels and there is no gameplay in reading novels. But everything discribed in those book come back in to the games. They are all related. You are correct that good writers are not drawn into videogames. They attract the best.

Also if you think that game development only consists of dev's artists, management and PR stuf you are wrong. (though this is just an assumption I make out of your post). In Guild Wars 2 for example they hired a economist who worked in Wall Street to keep track of the ingame economy. Game Developement attracts people from all disciplines. And writers are IMO the most prominent one. Nearly every game has lore of some sort and for good reason.

Also no good storytelling? True The Walking Dead or The Wolf Amongst Us. They have their flaws, But's it's still good storytelling.

I think the problem is when people conflate things like story and storytelling.