Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - Player7

Pages: [1]
1
Feature Requests / Re: Multiplayer
« on: March 30, 2021, 06:03:31 PM »
Fair enough, yeah nothing Unity does is speedy.

well I've bought various Mirror related assets before so maybe I'll just go ahead with it. I've got Pirates for VP now aswel.. can't say I'll enjoy Mirror never supporting the faster domain play reload stuff, as that was the appeal with MLAPI as Unity is an absolute slug to get into iterating with code changes these days.


Anyway I had a look through Pirates and well it's a bit heavy just to run it, like having to go through the menu's etc as the way it is setup is 2 scenes etc..

I'm just wanting something simple like the Mirror examples of a single scene setup.. something where the main scene voxel demo scenes work even just a simple flycam controller where each connecting player can just place/break stuff and inventory (btw I think you mentioned inventory isn't server side? could that be changed it seems like a easy exploit that doesn't sound good for anyone to be building a game on top of these days)... so the very basics of multiplayer that is setup to work with VP without all the main menu/options/character selections/terrain gen etc.

Because it just feels like looking at Pirates, I'm gonna end up spending more time trying to dismantle so much of it because it's not what is needed to build on top of and well it doesn't help Mirror does make things take longer to startup, and then because I'm doing that all future updates will never just have a very basic scene that is maintained.

So my request is can we not have in the Pirates VP package a very simple example of Multiplayer VP setup  (and infact maybe this simpler example scene(s) could be what in the future you just add MLAPI examples for, I'm not sure everyone really wants a full game template and honestly it's a bit buggy in some ways so its not the kind of fully polished game template where I'd want to build on from as I'd already started my own character controller and other bits of the game I'm working on, so I just want basics of MP setup and what is needed to get that working)...

Much like how the original Mirror for VP started with the video of setting things up in Mirror and the code for getting VP changes synced.. that direction was kinda ideal keeps things simple.. I'm sure the full on template will be useful to have for codesnippets and for progressing VP with a working multiplayer solution but yeah simple multiplayer scene setup.. or even just video tutorial for that continuing on the old mirror+vp video.. but with the code that was done for that in in this full template? (or some pointers to getting it setup that way.. quick guide?

2
Feature Requests / Multiplayer
« on: March 05, 2021, 03:52:58 PM »
On the roadmap for voxelplay ... could MLAPI  be added.. it is the new official multiplayer framework to replace unet. I know pirates template has Mirror support, but Mirror has some downsides like it will probably never work with unity's domain play reload feature, which is huge time saver to unity's increasingly sluggish build play times.

I'd buy the pirates template if MLAPI was on the roadmap for that even, for now I've held off as I think it's just going to become to deeply embedded with tying everything into mirror.

And well think some just wanted something to get started with all the core base multiplayer stuff working with voxelplay, the more you're having to rip out things that are not generically useful in code/prefab etc the better I find. I get its aimed at maybe a different audience but I'm just wanting basics of multiplayer support and enough examples to get started, not an entire game built really again that also comes into price range paying more for a template of things I don't need just for multiplayer support in a framework that I don't want to be using for the reason of it doesn't support domain play reload and then my entire build iteration times will go through the roof (which only gets worse with the more you code you have in your project) 😑

3
Announcements & Showcase / Re: Voxel Play Community Poll
« on: June 26, 2020, 06:25:34 PM »
I guess at much higher view distances an LOD system would be overal beneficial,  I was just thinking at view distances of like 4-8 chunks that maybe the chunk recalculation to polygon reduction wouldn't be that worth it, ie having a chunk that is already generated to be regenerated just for a low/high LOD.


4
Extensions & Resource Packs / Re: [Extension] Terrain More Voxels 1.1
« on: June 23, 2020, 01:36:33 PM »
can't see why this doesn't come built in? if you don't use and fill in the additional list then it just goes by the default?

5
Announcements & Showcase / Re: Voxel Play Community Poll
« on: June 22, 2020, 09:08:12 PM »
"LOD is very good for optimisation" ... is it? in the case of voxelplay I don't think it would be, the polygon count reduction really worth the cpu time taken to reduce all distant chunks..  seems like extra calculations on something that has already been done when the chunk is loaded into view area, with larger chunk view distances sure the polygon count ads up....  but with voxel density VP has its not exactly breaking the GPU here.

...and you're looking at a video where all the voxels are of the same material type.. consider many different voxel material types the LOD stuff wouldn't look that great...and I still think the reduction in polygon vs the time spent redrawing existing chunk meshes again is a bit of waste.


LOD works great for smooth terrain or higher smaller voxel density where you would already have a lot more polygon being used per unit and then the poly count reduction would be a pay off as visually you'd see less LOD swap... the fact that video shows obvious LOD swap just proves the point of it being an almost pointless and premature optimisation.. modern GPU's can handle those kind of poly counts.

happy to be proven on this, but I just don't see an LOD feature being all that beneficial compared to just reducing actual chunk view distance .... then again if this could double/tripple chunk view then maybe worth it?


"Also for networking, if you want to add networking then i think it's best to make it as an Addon kind not directly integrated,"

My thinking on it would be that ... or similar to how this now dead voxel c# lib (its no where near VP in features) did it ....https://github.com/richardbiely/Voxelmetric/tree/alpha_3/VoxeUnity/Assets/Voxelmetric/Code/VM 

6
Announcements & Showcase / Re: Voxel Play Community Poll
« on: June 21, 2020, 09:36:40 PM »
I'd like to see basic library networking support added (at least it would make it a bit easier to support a different one with all the important and relevant voxelplay specific stuff added...

(Integration with third-party assets) For a specific network library I would nominate LiteLibnet (free to use and its the only c# lib that actually works with unity faster playmode reload, its also seems to be a good library... mirror doesn't because they've got to much static usage and way too much weird reflection stuff that breaks the faster playmode from working along with a number of other unity network libs so all no goes for me)  and being able to hit play and test things is hugely important.. voxelplay alone is already quiet a heavy framework to playtest and develop on top of..without the faster unity playmode reload it was taking me almost 1min to play test a change.

So other than networking example.

I'd like to see the examples and input handling moved to the newer input handling system package.

Inventory/Console etc these things i have my own or third party and consider them less relevant, though I went with my own character controller based on Kinematic controller + Animancer for the animations handling.. and well frankly what a mess doing character controllers is, but the tying in both those systems into using Voxelplays own framework in order to utilise the existing fps controller classes and methods for like character walking on different surfaces.. etc its a messy maybe I just hate doing character controllers

Tooling...

"Other things I could think of I guess would be the constructor tool, I personally found it too clunky to work with. I bought Qubicle and just import from there, which is much easier. Of course I realise that making it comparable to Qubicle would be a lot of work and probably an asset of its own."

Agree, so I was working a worldedit plugin just never finished it, but tooling is more important to me than algorithmic world generation stuff.. rather have artist tools (brushes, pinch/pull/smooth/block paint..radius adjustments basically worldedit/voxelsniper operations supported out the box) the cuboid/sphere selections, with save/load schematic support ( there was never any mc to voxelplay importer but would be useful) to create and adjust world terrain or models from large to small... good tooling but again is so many areas to want improvements in.


The voxel based lighting is another area i find not perfect, but then maybe future realtime raytraced lighting is something that URP might support? and then the voxeplay unique voxel lighting solution would be less relavent? i dunno


optimisation, I'm not sure how much more faster voxelplay could get.. is there more room for further reducing average ms of voxelplay when the player isn't moving across chunk or even when they are.. increasing the view distance is goal.. I dont see LOD as something worth doing in voxelplay because it would just looked weird, and VP doesn't do smooth terrain

8
Support / Smooth lighting
« on: December 13, 2018, 11:08:13 PM »
old image

recent-ish beta with the new smooth lighting.. but we seem to have lost the lighting getting darker further it is away from a light source.... on hq demo scene

9
Announcements & Showcase / Re: Playable demos (updated for v2.1)!
« on: November 23, 2018, 09:06:38 PM »
is the door working in the demo version?


Pages: [1]