Jump to content

Sench

Pathfinder
  • Content Count

    7
  • Joined

  • Last visited

Community Reputation

1 Neutral

About Sench

  • Rank
    Pathfinder
  1. While technically possible, it wouldn't work to a satisfactory degree. Each grid cell is a separate server instance. If you put an island "on the border" between two cells, it would only show up in one of them, and a large part of it would be inaccessible beyond the cell border. You could circumvent that by placing the island on each side, but such an island wouldn't serve much utility since it would technically be two separate island of identical shape. Its intended two parts would exist in separate server instances. Issue number three is how border travel goes. You leave a cell at the very edge, but you are deposited some distance away from it, so you'd have to offset the islands to compensate for that. Also ships traveling across water cells spawn at water level, but with islands, there's a risk of spawning in underground.
  2. Except "you have to have X types of Y" to make good gear is punishing gameplay, whereas "you can do X with multiple types of Y" is rewarding gameplay.
  3. While I can appreciate the small touch of having different kinds of wood come from different trees etc, at the moment it's a very minor feature that has no impact on gameplay but makes inventory management significantly more annoying. Every type of wood, stone, flint, oil, hide etc is identical in gameplay terms, with the exception of not stacking. So here's what can be done with it: 1. Simple solution, remove them altogether, keeping only the base variant of each resource type. Trust me, nobody's going to complain about that tiny bit of immersion in this kind of game (or very few will, but many others will appreciate the convenience). 2. Interesting solution, make non-base materials provide added qualities to crafted items. For example, basic wood in a pike does nothing special. Strongwood increases durability, lightwood makes the item weigh less etc. Also lightwood itself could weigh less than basic wood. Of course, this would require some kind of hook into reading which specific type of resource is being used and in what proportion to the total of that class, as well as one to impart bonuses on crafted items, and added parameters defining what advantages each material can provide in each item. Knowing how the resource system is set up, I suspect that's impossible, but that would be very cool and it would more than justify gathering different types of resources and the annoyance of inventory management.
  4. Better yet, when a boat is over capabity (crew or cargo), make it unable to move, like players, instead of taking damage. It's still unusable but you've just solved a lot of griefing problems.
  5. After setting up a local game server, I noticed two rather important things. One, islands are largely reused in this game. For example, we have island shape idenitified as "Cay_A" in the internal files. It has 4 types: equatorial (_EE), equatorial variant b (_EE_B), tropical (_TR), western tropical (_WR). Each variant appears in the game world twice, except tropical which does so three times. However, they all have the exact same shape, and the climate differences between the four are negligible, so effectively you have the same island appearing 9 times. This is the rule, not the exception. Each unique island shape has several climate variants, some with sub-variants, and almost all of these are used in the game world twice, with few used once and some used more. Of course, an island like "Mnt_C" has desert, tropical and tundra types, so those are more varied despite the same shape. On the other hand, this is how many times it appears in the game: 3x _CH, 3x _CH_B, 2x _CL, 2x _CL_B, 2x _ET, 2x _ET_B, 2x _ET_C, 2x _ET_D, 3x _TR, 4x _WF, 2x _WR, 2x _WR_B, 1x _WR_C, 6x _WU. The two-letter code signifies the biome, and the added _B or _C signifies variants, like an obelisk being in a different place. So we get 6 completely identical tundra islands, 8 almost identical east tropical islands, which are likely very similar to west tropical and "generic" tropical as well, etc. Again, this is the rule, not the exception. Identical, near-identical or very similar islands appear in the game world many times over. Worse yet, some "event" islands (where power stones are found on the official map) aren't uniquely shaped, either. "Mnt_R" has 2 "event" variants (desert and arid biomes), and that's with a total of 8 event islands. And then it has 6x _WU, 7x _WU_B, 7x _WU_C, 7x _WU_D variants. To reiterate: 27 islands with the same shape and biome, split into four groups of identical islands. At this point I should mention that the islands in each starter region are identical, most are used in two regions with a similar biome, and starter islands share their shape and biomes with some non-starter ones (so their sub-variant is having the NPC port). That's several islands with 8 identical copies in the world each. The other important thing about the game world is that about 95% of its surface area is open sea, where practically nothing happens. Now, in a game about pirates, you should have more sea than land. But the ratio doesn't need to be quite so dire, especially since so little happens at sea (at least in the present state of the game). That was the prelude, now for the suggestion. I understand some of the reasoning behind a map this big, and all of the reasons behind reusing islands. But is it necessary? The game world is redundant. It takes upward of 10 minutes to travel between two islands in the same region, upward of 25 to get to another region. The main quest (and character progression being tied to it) forces you to travel all over, spending many hours on uneventful sailing, and while visiting different corners of the world, you will see some new and exciting things... but mostly the exact same islands over and over again. Why have a massive world of 225 regions when you need to visit at most 7 to see all the unique stuff? You need to separate players? Separate servers will achieve that more efficiently, taking away their ability to get to someone else rather than putting it behind hours of waiting. You need to keep players together? Smaller maps will once again achieve that more efficiently, with players taking less time going to do something and more actually doing it. Most of the world is irrelevant to any single player. A player cares about: their current region, their home base region, possibly regions around their home base (on PVP servers). That's 10 regions out of 225. Set a server to 25 (5x5 grid) and suddenly you have 9 servers with more stuff happening than you had in that 1 oversized server, without any part of the world feeling out of reach. It shouldn't take hours of waiting (current state of sailing) just to see a different environment. And let's be honest, nobody was taking "40000 concurrent players" seriously anyway. TL;DR: 1. Reduce map size severely - I'd suggest a 5x5 grid instead of the 15x15 as it is now. 2. Increase island density somewhat - sailing between islands takes too long. Reduce island visibility range if performance is an issue. 3. Only use each island shape once, at least for islands where base building is intended (ice caps are probably okay to rehash), or at least make the variants very distinct (massively different biomes and decorations) and still don't use one shape more than 3 times (with a smaller world it shouldn't be a problem).
  6. Sorry if this has been answered before, but is there a way to make your 1x1 grid not have the level 8 limitation?
  7. After spending a couple of hours yesterday trying to set up a local server for myself and some friends (as I did for ARK and pretty much every game that let me), I realized the current server system is absolutely terrible. For those who do not know, the game world is separated into cells, each one is a separate server thread. I understand that it was necessary to split them up due to the massive size of the map and large player base. What I don't understand is why it was necessary for each such thread to be an independent server instance in terms of loading game assets. Every single server thread loads all base game assets, even ones that aren't used in that particular instance. With a single cell you have some overhead, but it's negligible. A 2x2 world has your server running 3 redundant copies of every base game asset, and 2x2 is too small to fit even the majority of unique island types. A 3x3, which shouldn't be too claustrophobic, has 8 redundancies. But if you want to make a server with the main quest line working well, you'd need a 4x4 or larger, with 15+ redundancies. I understand that the current system is much easier in the coding department, but quite frankly, it's terrible. There should be a single master thread that holds base game assets, and the cells should be instance-threads which would only process things specific to their cells. Things already work that way from a gameplay perspective, whereas such a setup would provide a massive save on system resources overall and potentially improve performance.
×
×
  • Create New...