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 - wiggi

Pages: [1] 2 3 ... 9
1
Development / Re: Next Hardfork
« on: June 16, 2017, 07:26:13 PM »

1. Life Span of Hunters = ~60 to 80 Blocks


The number should be lower, not much higher than 25, to make it impossible to keep the status quo (that new spawned hunters are always outnumbered on the better harvest areas) by steadily spawning 1, 2 or even 3 hunters per block, which would be easy for the dominators.


2
Development / Re: Huntercoin alternative gameplay
« on: June 14, 2017, 09:39:26 PM »
Great idea! I wanted to test, but it shows an error. What can be wrong?


I could reproduce this error by starting "huntercoin-qt.exe" directly instead of the shortcuts "huntercoin-qt.exe - 1" and "huntercoin-qt.exe - 2 - scrypt".

It would then use the normal Windows data folder for Huntercoin "C:\User\...\AppData\Roaming\Huntercoin". The error happens if there is already a blockchain (mainnet, probably) that is not compatible with this testnet-in-a-box.


3
Development / Huntercoin alternative gameplay
« on: June 08, 2017, 07:06:52 PM »
I think the "human mining" capability should not be proportional to the time invested (or even steeper than that), so in this prototype, game rules have been changed in order to make the difference between players who can only play for a few minutes per day, and players who can invest unlimited time, as small as possible.

As side effects:

- it becomes impossible to use bots (they would need to think strategically)
- it would also be much more difficult for POW miners to gain an advantage in the game (by
  not processing transactions and discarding their own blocks selectively)
- and the number of transaction generated by players is very small

Features:

- Character levels, and leveled chars are near immortal.
- Character equipment, weapons and items that determine combat abilities.
- Smart monsters.
- Safe zones with teleporters connecting them.
  NPC merchants to buy items from.
  (coins earned by these NPC can later be re-used for a fully decentralised and automated bounty system)
- Low risk/high reward for players. Cheating death is cheap, but
  monsters can drop potentially thousands of coins if killed.
- Universal network-wide AI to micromanage all actions on the battlefield.
- Coins have a purpose in-game, and players have an incentive to accumulate them.
- New player, after they managed to survive a few days, can command the most powerful monsters
  (they have a "longest time not bother me, first serve" policy)
- Combat is resolved only determined by character position and abilities, not by transactions in the block immediately before the combat happens


Testnet-in-a-box with source:
hunttest_20170607.zip, 24.2 MB
https://mega.nz/#!CA5iRSCQ!omJpbIAt5kow59IXhpy9wpL-cJOxE2o6qcEgzK8J_IQ


Some pics from older version (layout is now a bit different)

Starting new hunter:


Center with NPCs:



It takes some time for the character  (i.e. 50 blocks on this testnet, 25 blocks would give a 50% chance) to memorize the coordinates, this is to balance the disadvantage of sending your move before knowing other player's moves. In latest version hunters can wait inside the safe zone.




The game's nature changes depending of whether 1 round is 200 blocks (like in the testnet) or 2000 blocks. Normal players wouldn't be able to do 1 move every 200 minutes, day and night, so while this kills the bots, domination by human player groups is still possible.

However, the average player would have no problems to input 1 move per 30h forever if they can choose the exact time freely.


4
Technical / Support / v1.4 blockchain (archived)
« on: February 09, 2017, 07:46:40 PM »
v1.4 blockchain as of February 11, 2017:

Code: [Select]
blk0001.dat
md5: fdd10b8114aff98f255058e4e0170eac
https://mega.nz/#!fZwjnJDI!YMt8we93bO1QMlUh-qFg1Ed0tdSpeJg9Um7OtQr5ihk

blk0002.dat
md5: f93a05dfa057ada05d3623e1d7a2571c
https://mega.nz/#!eBgzjIDb!1w5TAAtl51DGzWOyVKbWDGPYw47sF1ozuVplnGXSZOY

blk0003.dat
md5: a1f824f2af3e178f7de818f6f480e9b1
https://mega.nz/#!jV4TEC5Y!bYGOzYmEKV7kBEEbAcl9ySZfD2S1ebM8_xgRe5iBOWw
https://mega.nz/#!WNpTxIqR!A-MKWEC2oyYqKMl5zAvSnZVWF00e1z3EJjcTPqAlzVo

blk0004.dat
md5: bd4416796ba0271f2cbfa0562b2ee40c
https://mega.nz/#!2MYyAaTY!sfyU-nmWdBV-hqVhpojBN6mX2OsruEzeQns9CvVlnxQ
https://mega.nz/#!uMpAjYja!zbeTbh1l1gA_HdS67Dq7atzy-ybI31KBFSmfOBbaB0Q

blk0005.dat
md5: 27b4f91716ffacabcf29dd7792136b0c
https://mega.nz/#!DcgkBRhQ!XAKtgKM2sbZTRvxNYFu7i_gx8D4fF141Nc5goRhIywQ
https://mega.nz/#!GdAUnALR!sFBo0-IOfPOOquohJnkPL3pZQdZk9VDoIaZbDSbX-wQ

blk0006.dat
md5: 992e0e81c1ed134da48dee84b3bbde2c
https://mega.nz/#!LF5hjLhZ!eLolCrl8nRCAkWBiG7RHJXYEp6PEcjUKL2FGB-epLvg

blkindex.dat
md5: 7468a2f9bffbe59ff71a6c2dbe473cce
https://mega.nz/#!OQJRybZY!djFmO6lDUzQFdkr4QVByYGoblMGaThMd4HTj1Rs49dc
https://mega.nz/#!zEg1GA5R!3L1Pc5T77JJFFnCFAWrVbHt22GbivRO1rihYJdxFCbE
https://mega.nz/#!SVRzxZJC!5BNSKtOkUfUmfKW8ca6Qkunb0N7gLVBf58EtR5CZsqo

game.dat, md5: 7af4be9b2a092f09081a553a4e185994
game_sv4.dat, md5: 80b5951614a396498932329664e27750
nameindexfull.dat, md5: 1af8434a9399052fe08db1fddd44903a
utxo.dat, md5: 63fa216adac2e5986282e0490da26884
https://mega.nz/#!HVgBULzB!XSco_H5ZmSXidu-Q6v36Xto82SOenHd04jEgOrWts8w

5
Development / Re: [ANN] Huntercoin Core - ready for first testing
« on: October 22, 2016, 04:21:40 PM »
Is there any guide for creating and moving hunters by RPC, or is it just "use the source, Luke"?

It would be the latter, but in some pics if the forum you see the client showing the transactions as plain text.

Example: http://i.imgur.com/5UfrEOR.jpg

This is also the format that you would use if moving hunters by RPC. The console accepts the same format, only need escape characters, \" instead of "


6
Development / Re: Gameplay change (hard fork) wish list
« on: October 21, 2016, 08:08:25 PM »
Definitions to visualize things for client developers:

Compile with the "const unsigned char Game::SpawnMap[MAP_HEIGHT][MAP_WIDTH]" from extras folder (just printed this, not tested if it looks right)

edit: "extras" folder removed. The printout of the SpawnMap data can be found here (used for visualization):
https://github.com/wiggi/huntercoin/blob/betterQt-timesave2/src/gamemap.cpp

A tile is player spawn tile if
Code: [Select]
(SpawnMap[y][x] == 2)
A hunter is protected (i.e. invulnerable) if
Code: [Select]
((characterstate.stay_in_spawn_area < 6) || (characterstate.stay_in_spawn_area > 8))
A hunter is spectator if
Code: [Select]
(characterstate.stay_in_spawn_area > 8)
A tile is bank spawn tile if
Code: [Select]
(SpawnMap[y][x] == 1) but there's no need to visualize this.


7
Development / Re: Gameplay change (hard fork) wish list
« on: October 21, 2016, 08:04:47 PM »
New and (unless bugs found) final testnet in a box

Copy the "hunttest" folder as "C:\hunttest"

huntercoin-qt.exe is huntercore

3 nodes, node 2 will connect to node 1, node 3 will also connect to node 1, node 2 is not connected to 3

The shortcuts determine as which node any of the executables is started. (don't start the exes directly)

If huntercore was started once, then one of the 2 other executables may crash if the same node already did run huntercore
(because wallets are not downward compatible)

Note that if huntercore is started for the first time, it must first "download" the blockchain from the other nodes (1h or so)


hunttest-timesave-box3b.zip, 29.5 MB
https://mega.nz/#!HYEnSLyS!32irzYYVznFMK_o3JXWpS7NitFlrfWN_5SDdHqi-DMc

To run, it also requires blk0001.dat, blkindex.dat, game.dat, nameindexfull.dat, and utxo.dat from testnet blockchain
(copy to testnet folder of \hunt1, \hunt2 and \hunt3)
Snapshot of the testnet blockchain at height 331025 for reference:
hunttest_timesave4_blockchain.zip, 322.3 MB
https://mega.nz/#!CNcCRAgY!84D2j1ynXRZGebkkFQ3Mcg5zyP2Di93sgmbscSmddjI

edit: "extras" folder removed. The printout of the SpawnMap data can be found here (used for visualization):
https://github.com/wiggi/huntercoin/blob/betterQt-timesave2/src/gamemap.cpp


features same as
http://forum.huntercoin.org/index.php/topic,22265.msg29486.html#msg29486

plus "ghosting with phasing in"
repeated every 500 blocks so that every "timezone" has one event per day
1 ghosting event will accumulate about 700 coins that are released for the hunters at once when it ends

schedule:
at block 300 out of full 500, every 4th coin spawn is ghosted (game continues normally but coins already accumulate)
at block 450 out of full 500, 3 out of 4 coin spawns are ghosted (2 out of 4 would look ugly graphically)
at block 480 out of full 500, full ghosting
at block 500, return to normal

1:4 ghosting



fork height is set to 331500, first ghosting will begin at 331800 and end at 332000

The number behind hunter names is (stay_in_spawn_area) the most important variable for debugging.
It will remain at 0 during the last 10 blocks before ghosting ends, for newly spawned hunters.
(so they have more time to consider strategy before the new war begins :)


8
Development / Re: Gameplay change (hard fork) wish list
« on: October 21, 2016, 03:21:37 PM »
if i remember the fee change happened after the first disaster after the specified height (not sure tho) anyway you could apply the same logic: after the first disaster with height > xxx apply changes, this way there is no risk that hunters with different value coexists.

After disaster the old poisoned hunters still live for 30 or 50 chronons, so combat between 10 HUC hunters and 200 HUC hunters was possible imo, but may have not actually occurred.

With this fork's system in place, I think it will be technically easy in future forks to change hunter cost in a way to make pvp combat between "different" hunters impossible.
e.g.
chronon T+0: all hunter on map made spectator
        T+5: hunter price change goes into effect
        T+10: all hunters on map are banked and logged out (as if on bank tile)
        T+15: pvp combat resumes normally when 5-block-spawn-invulnerability of the first new hunters is finished

Then hunter cost could be different for every day of the week, going from very cheap to very expensive (for the sake of experimentation, in case we can't decide on something else)


9
Development / Re: Gameplay change (hard fork) wish list
« on: October 21, 2016, 03:14:26 PM »

wiggi, it sounds like you would prefer a higher cost game with fewer players. did i read correct you would prefer ~50 individuals on the board at one time? is that not orders of magnitude too low for a "successful" global decentralized game? when you have time, i'd like to hear your thoughts on this one?

We don't really know how the game will look like and play like post-fork, but 50-100 individuals on the board at one time translate into 1000-5000 more or less active players depending on how long they play on average. Not "global decentralized game" but an very optimistic goal for next year. If this is reached, and "too much competition" or "insane hunter price" becomes the limiting factor, then we must be ready with a total map layout overhaul (with more and more diverse harvest areas) and a clever way to determine hunter (refundable) cost. These are tasks for next year.


10
Development / Re: Gameplay change (hard fork) wish list
« on: October 18, 2016, 04:04:59 PM »

I think we should release changes on the new huntercoincore.
Yep, release on the old Huntercoin and on Huntercore at the same time.
I'll make the (hopefully) final testnet in a box with both Huntercoin-qt and Huntercore-qt.


Btw, can someone remember, when the "lifesteal" fork began, did old hunters that had cost only 10 HUC co-exist on the map with the new 200 HUC hunters. (What happened if you life-drained them etc)


11
Development / Re: Gameplay change (hard fork) wish list
« on: October 16, 2016, 12:54:51 PM »
To expand on option C. I was thinking of a longer time window to determine fees - like days. The purpose being to regulate the cost with median map density over days, not short time scales. generally, all players on the board at any given time would cost a similar amount. However, even on shorter time scales, it could add an additional strategic element to the game.

I would prefer option C too, but in practical terms, we don't know what the formula should be. Reacting to sudden popularity when it causes increase in both hunter population and HUC:fiat price (which the formula doesn't know) with a spike in hunter cost would perhaps not be good.

The current gamestate also has no capacity for additional statistical numbers, to add them would cause delay with new rounds of testing and would remove the fallback option for polo and the miners to just put a "new old" daemon in and leave the existing blockchain in place. (We don't know if Huntercore daemon will choke on polo's probably massive wallet which was created by the old daemon)

I personally would prefer a game where 50 human players play a relatively short session, and have 1 or at most 2 hunters each that they monitor personally at all time because they are expensive. (*) If around this time next year Huntercoin is really popular, then hunters can become cheaper, but together with other changes like a map redesign with more harvest areas to accommodate for more players.

(*) and more about psychological tension than clicking alot of things, the "monitor personally" would also level the playing field for mobile/light clients that won't have automatic movement and destruct functions

12
Development / Re: Gameplay change (hard fork) wish list
« on: October 14, 2016, 03:26:34 PM »
New testnet in a box:

hunttest-timesave-box2.zip, 14.8 MB
https://mega.nz/#!rYEhUYiY!__xU043U6WiZvW-3jWkQZ202vIFzLfmnrRf--TXCKT0

Snapshot of the testnet blockchain at height 331025 for reference:
hunttest_timesave4_blockchain.zip, 322.3 MB
https://mega.nz/#!CNcCRAgY!84D2j1ynXRZGebkkFQ3Mcg5zyP2Di93sgmbscSmddjI

copy blk0001.dat, blkindex.dat, game.dat, nameindexfull.dat, and utxo.dat from testnet blockchain to testnet folder of hunt1 and hunt2


It's now 2 executables (based on betterQt, and based on official v1.3.01) and 4 links (choose 2 to run)
It looks like all graphical glitches (with the ghosting) are gone

source:
https://github.com/wiggi/huntercoin/tree/betterQt-timesave2
https://github.com/wiggi/huntercoin/tree/master-timesave2

Source is cleaned up so that all changes for the fork are in 1 commit.
(it's 24 changes, and 24 * "FORK_TIMESAVE", easy to find and to read)


fork hits at 331100 (on testnet, mainnet height not set yet)

what happens post-fork:

reduction of possible player spawn tiles (as in OP)

player spawn tiles work like banks, but faster (2 blocks until logout)

player spawn tiles visualized in old-Ot style as semi opaque rectangles

reduction of possible bank tiles (as in OP)

all banks reset at once when fork hits

logout on bank tile in 1 block (i.e. you can't be on a bank tile and not logout)

new hunter cost 1 HUC (down from 5)

destruct cost 1 HUC (down from 20)

hunter invulnerability after spawn for 5 blocks (ends if waypoints set, or stand still to become spectator)

spectators are invulnerable and also can't move, refund normally 15 blocks after becoming one.

ghosting to break up attempts of domination (100 blocks after every full 1000 blocks, 50 blocks after every full 500)

when ghosting is in effect, coins spawn normally, but can't get picked up and accumulate on map

ghosting is visualized with semi-opaque coins


Note that important state changes (hunter invulnerable, hunter is spectator, coins ghosted) can directly be calculated from
2 gamestate variables, either characterstate.stay_in_spawn_area or gamestate.nHeight, so visualization should be easy for all clients that work by fetching the gamestate from daemon.


13
Development / Re: Gameplay change (hard fork) wish list
« on: October 12, 2016, 09:17:45 PM »

1 - about blocks ghosted
I don't think that is a good idea because it limits at some extents the game for casual players who just want to go in, play some minutes and go out and didn't help about dominator issue, because my direct experience with them says that they are always active even for pvp (at least one of them). I'm always a beliver that the main anti-dominator will be having lots of players to face, so the more the map is populated, the less the dominators can ...dominate

this mode rises even some technical aspect to consider (but not that importants): the GUI should reflect that coins aren't pickable (so remove them from the map or change the aspect) and it should be clear to users, so game_state (thinking to my client extensions or other possible ones) should include that block is in "block ghosted" mode (but I think that at a gameplay level, that features doesn't fit well)
But i'm more concerned on the feature itself that technical aspects


Because of encapsulation issues, I did implementation for GUI by copying the formula that calculates the state from block height. So:
https://github.com/wiggi/huntercoin/blob/master-timesave/src/qt/gamemapview.cpp#L97
All clients can do this too.


Perhaps we can have both, not annoy players too much and still have a big wave of new hunters, by making the ghosting partial.
Example: for 150 blocks, every 4th coin spawn is ghosted (game continues normally but coins already accumulate)
    then for 30 blocks, 3 out of 4 coin spawns are ghosted (2 out of 4 would look ugly graphically)
    then for 20 blocks, full ghosting
repeated every 500 blocks so that every "timezone" has one event per day

can be done by just changing the formula. It's not ghosted if...
Code: [Select]
if ( (((x % 2) + (y % 2) == 0) && (nHeight % 500 < 480)) ||
     (((x % 2) + (y % 2) <= 1) && (nHeight % 500 < 450)) ||
     (nHeight % 500 < 350) )

Would have to check for each hunter if they're allowed to loot at the current chronon and coordinates, but still fast and easy.


14
Development / Re: Gameplay change (hard fork) wish list
« on: October 12, 2016, 03:18:45 PM »
testnet in a box: it's always for Windows. Traditionally put it in C:\hunttest then just use the 2 links. It will also run while the windows machine (or VM) is offline.

waiting for domob: would want him to take a look before the fork is actually deployed for Huntercoin mainnet. Everyone else should take a look too.

20% of blocks ghosted: it would be 15%, one 100 blocks break and one 50 blocks break per 1000. We can vote the numbers, or even kick this out in case everyone is confident that it is not needed to un-dominate the map.

Hunter cost: I think, even with expensive hunters, every harvest area will get at least 1 hunter (i.e. 80 hunters on map)
When it comes to fighting, hunter cost determines both player fear, and player greed. ;)
We could vote hunter cost.

But unless hunter cost is *very* low, I expect population around 100, so the game would cater to >= 100 actual players at the same time. If overcrowded, it gets kinda nasty (maybe in a fun way), with both risk and reward skyrocketing.

Other proposed features: I wouldn't want to add anything else because of complexity.
For viewing pleasure, there's now a v1.3 old-Qt based version with exactly 1 "FORK_TIMESAVE" per code change. (Github branch master-timesave) Count is 24, this is more than enough already.
I'll make another testnet in a box, and of cause port it to Huntercore (without the graphics part) when needed.

chronoDollars: is fully functional, but should first have lots of use and testing under heavy load conditions before it can become part of the Huntercoin core protocol

15
Development / Re: Gameplay change (hard fork) wish list
« on: October 11, 2016, 03:28:21 PM »
Brief recap: it looks like we've settled on tile reduction as the solution to our gameplay problems... it looks like he's successfully made it so that Hunters always spawn near coins (spawn tile reduction), but the bank tiles haven't been reduced, yet.
You didn't actually look at it, right?

Player spawn tile reduction is working.
Bank spawn tile reduction is working.
Additionally you can logout on the player spawn tiles, faster than currently on banks
Fee reduction to 1 HUC is working.
Hunters who stand there for longer than 5 chronons will irrevocably become spectator.

I like the ghost coins idea, but I think the frequency is too high.
Yes, could be reduced a bit. perhaps:
 for 100 blocks after every full 1000 blocks, and
 for 50 blocks after every full 500 blocks?
If nerfed too much, it might be not enough to serve its purpose.


I'm curious about spectator lifetime, though, I didn't realize they'd have a lifetime. Why don't they just stay there until banked like other Hunters?
Snailbrain's request due to fear of too much spam (You read it on skype. Need more sleep ;)

Concerning Hunter cost, I'm undecided. Too low has its dangers, too high has its dangers.


Also, we must not forget to adjust the fees and Hunter cost, especially with recent price increases. We also really need to revisit the feasibility of tapping the gamefund with something like random coin drops, because it's getting too big. I predicted a while ago that it could scare investors, who view it as a ticking timebomb waiting to hit exchanges.
Yes. Perhaps *not confiscating* hunter loot for the gamefund (in case of disaster) would be a one-liner (in HandleKilledLoot?) and could easily go into this fork. If we've managed to un-dominate the game there will be time to think of intelligent gameplay ideas to tap the gamefund.

Or perhaps it's better to not thinking about gamefund at all because this is what creates the irrational fear of it being dumped (which is just as impossible as with the other 27000000 coins that are not mined. They don't exist. The gamefund coins don't exist, too)


Pages: [1] 2 3 ... 9