Page 1 of 2

MonsterHunt Versions & History

Posted: Tue Dec 26, 2017 1:38 am
by SC]-[WARTZ_{HoF}
MonsterHunt v400
(4.6 MiB) Downloaded 151 times
MonsterHunt v450
(4.68 MiB) Downloaded 128 times
MonsterHunt v500
(6.18 MiB) Downloaded 144 times
MonsterHunt v503
(6.14 MiB) Downloaded 145 times

Release 1 (Internal)

- Initial release

Release 2 (Internal)

- Fixed lots of tiny bugs
- Added subtle little things you probably won't notice
- Removed 'Deaths' listing from the scoreboard
- Fixed some spelling errors :-)
- Greatly expanded the ReadMe file

Release 3 (Internal)

- Added new game type - Monster Arena!
- Included three Monster Arena maps, MA-Winged, MA-Queens and MA-Revenge.
- Included new Monster Hunt map, MH-Canyon.
- Created icons for MonsterEnd triggers to make it less confusing when making MH and MA maps.
- Fixed "*Player* is now on" message that was displayed when a player joined a game, now reads "*Player* has joined the hunt!"
- Fixed some death messages eg. "Shrimp was slimed. by a Slith". Note the full stop after "was slimed." all those have been fixed.
- Improved bot intelligence, they should now be able to complete the map.
- Added support for Unreal 1 weapons, giving you a potential total of 21 weapons.
- May have fixed bug where the monsters weren't showing up on some people's systems. (experimental)
- Added a small Monster Hunt logo under the health and armour readings on the HUD.
- Changed the difficulty descriptions in the bot config window to suit Monster Hunt.
- Added support for up to 32 players, I don't reccomend you use that many though, but the option is there if you want it.
- Added Monster Hunt server browser for finding online games.
- Removed ammo from maps when using the Chainsaw Arena mutator.
- Fixed lots of bugs in MH-Forbidden.
- Created this HTML help file, looks a lot nicer than NotePad don't you think?
- Thought up little slogan, "Don't think, shoot!!" (you may recoginse this from the best PlayStation game ever made, Metal Gear Solid!).

Release 4 (First public release)

- Added maps: MH-Trials, MA-RocketArena and MA-Crimson
- Added death messages when a monster is killed
- Added server browser for Monster Arena
- Server browsers actually work now :-)
- Fixed "Hunt Successful!" message even if time ran out, will now read "Time up, hunt failed!" when time runs out
- Fixed problem where monsters wouldn't work on some people's systems for real this time
- Fixed problems with Monster Arena map lists
- Created 3 new monster classes: LavaTitan, LavaSlith and RockTentacle
- Added the Monster Hunt credits to the Mod menu
- Revised all maps and fixed lots of bugs
- New background images for the Rules, Settings and Bots config screens when starting a game.
- Created the MonsterEvent class, allows messages to be displayed. Can be used to inform players of progress or things that have happned.
- Weapons and ammo respawn faster in Monster Arena
- Monster Hunt now uses it's own .ini file "MonsterHunt.ini" for storing it's information

Release 5

- Fixed Unreal weapons not working in multiplayer, also added decals to their projectiles - Thanks to UsAaR for help.
- Added a scoring system so you score points depending on which type of monster you kill. Scores are as follows: Titan, Queen and WarLord = 5 points. Giant GasBag and Giant Manta = 4 points. SkaarjWarrior, MercenaryElete and Brute = 3 points. Krall, Mercenary, SkaarjTrooper, Slith and GasBag = 2 points. Everything else = 1 point. You also loose 5 points for killing innocent creatures and yourself. 10 extra points are awarded for killing a boss.
- Fixed bug where monsters would attack spectators before a round begins.
- Fixed bug where where if no time limit was set, but you finished the level, it would say "Time up, hunt failed!".

Release 6

- Bots can now follow waypoints (if supported by the map) and complete complex objectives and end the level by themselves.
- Bots now fight monsters and monsters fight bots.
- Monster difficulty is influenced by bot difficulty settings, eg. if you set the bot skill to Novice, monsters will be easy to beat, set the bot skill to Godlike and monsters are much harder to beat.
- A life limit can be set, when players (and bots) run out of lives, they are sent into spectator mode. When all players and bots are dead, the game ends.
- Tournament mode can no longer be set, it will always be off in multiplayer games and always on in single player.
- Fixed "You are the winner" announcer message from playing when you actually lost (ran out of time or team has no lives left).
- Team skins are now optional (set when starting a game), no need to use red all the time.
- 5 Points are lost for being killed when there is no life limit, this should help prevent "kamakazi" style gameplay.
- Revised all Hunt maps and fixed a lot of bugs in them.
- Added new map, HM-NaliVillage][.
- Lots of other minor bugs have been fixed up along the way.

Re: MonsterHunt Versions & History

Posted: Tue Dec 26, 2017 7:35 am
by medor

Re: MonsterHunt Versions & History

Posted: Fri Dec 29, 2017 6:34 am
by Nelsona
I'm not going to say that by using 504 you will not have any crash, but let's say crashes are reduced. At this moment the only thing almost removing all crashes (perhaps still random occurrences are still happening) is XC_Engine which has a lot of things rewritten and add-ons, offering a lot of options.
So far what was helpful to get rid of old borks brought back in stage is NavAdder which I wrote and which is supporting updates more time from now on by simply adding files without editing anything. I think this way makes me to feel better than overloading controllers with a lot of "fixes". If one of fixes is not good in other map you'll have things ruined. Path modules can be changed and even new Player-Side stuff can be triggered without to do anything else than a file copy operation. I'm not sure if exist something more simple for a server.

The heavy part is writing patching files which probably only a coder with a minimal experience can do - this is indeed an important problem so far.

Another possible versions which I was doing and talking about at probably are not gonna see the public light because I'm not interested about more "coders" having skill in Copy-Paste unable to manage the stuff and even messing up codes - I could see even some of my codes rewritten in a dumber format and then I'm not gonna participate at more MH trash supply.

Re: MonsterHunt Versions & History

Posted: Mon Jan 01, 2018 2:44 pm
by Cancer{HoF}
wonder if the BAH MH mod somewhere out there

Re: MonsterHunt Versions & History

Posted: Sun Apr 15, 2018 11:32 am
by medor
medor wrote:
Tue Dec 26, 2017 7:35 am
Missing the excellent no crash 504 ... rch_mode=f
not for mappers

and add some patch
MHHelp_MH504_30_05_2015 ... ... ...
working with NewNet with custom weapons UN1.DisableWeapons

Working with RX family UN1.RXPosError

may be some bugs to found ?

Sometime we can't open the mapvote ( two different test build mapvote) is it due to the ini ? MH2_NS_V2.ini

Re: MonsterHunt Versions & History

Posted: Thu Apr 19, 2018 4:39 am
by Nelsona
Entertaining, but as I recall so far I was running that MonsterHunt before XC age exactly as supposed. When I'm speaking about "as supposed" for sure I won't mention "mods" and all that.

NewNet in MH !!?? Can I ask a question ? Why do we do crap all the time ?
In server a player X can claim a kill, after 180-200 ms another one Y claims the same kill because of REAL PING - often occurrence serial kills in MH. Both of these players were killing and scoring for the same Monster ? WTF is this ? A new fantasy a la 2018 ? So if a fake "mod" has bugs, MonsterHunt is guilty for that, right ? Did anyone figure these things ?
This "original" MH will not be in my attention anyway at this moment, I have XC_MonsterHunt which is my current MH game-serving support and I have some MH2 type un-shared yet, supporting teaming for more than 1600 monsters without recursive crashes (noobs cannot use factories because they do have a poor brain). In that MH2 I will look at stuff trying to discard iterations at game-start as much as possible. So far as I recall game starts and might run in maps with 3200+ creatures - if we are interested to ruin our own servers. XC_Monsterhunt probably is supporting 5000+ creatures but it might not be the smoother ever thingy.
Anyway for preventing future crashes I might drop some love for these loaded maps doing injections and getting a double number of creatures but NOT ALL of them loaded in Map. That's trash making not mapping.
As for NewNet fantasies I'm not even curious to look there. Explanations: Sometimes I start playing using a lagged 2G connection through a tethering wireless device having 180+ ms ping. Why ? Because I like ON-Line feeling, it's the part of UT's charm. Learning to mitigate ping in MH is a hobby for me. NewNet at this point has no fun and does FAKE effects and that's why I won't move a finger accordingly.

MH architecture related to MH2 types will suffer probably a work-around - not sure when I'll drop an eye. This way in attaching stuff for making a map to work I think belongs to the past. A new and perhaps more properly coded MH is based on EXPANDING MonsterHunt and by using this way Actors embedded in map for MonsterHunt will work in that expanded MH. All the time these MH2 games were child of TeamGamePlus injecting their actors and making mapping creativity more limited - This is already noticed by me as the wrong way, and then doesn't worth too much development at this point.

More MH techs and details based on my observations from all these years are doable in a separate section with some explanations, some people might not understand "a single word" from there (as they claimed) as long as they do not even have a clue how do a game-type works, differences between a game-type and a mutator. Even written with a poor grammar, these might help CODERS not random readers or supposed mappers/modders.

Re: MonsterHunt Versions & History

Posted: Sat Apr 21, 2018 9:59 am
by medor
The original goal was not to make newnet compatible for the MH.
But just I realized that it works currently testing with custom weapons and newnet UN1.DisableWeapons.(BT game test under request Dizzy)

To answer your question what for.
If just the movements can be more pleasant for play on a distant server .... why not.

Ho and we forget the question
Sometime we can't open the mapvote ( two different test build mapvote) is it due to the ini ? MH2_NS_V2.ini

Re: MonsterHunt Versions & History

Posted: Sun Apr 22, 2018 5:52 am
by Nelsona
I think I answered, in any of my MH versions I could open any MapVoteLA type (those functional after that explosion of sudden "MapVotes"). If mod would crash Votes or such why would I release it ?
I had a server in that time using for months MH2_NS_V2 (incomplete for me at this time) and I did not have issues with MapVote but more issues in some maps. You can modify INI settings as you like, see if it do helps. But heavy timers Are a Pain, perhaps I should drop out a MH2 based on state code letting people to see another way of doing a regenerator with less timer load, I'll think about that. Or I really have to launch some stuff which I did not write yet - it's just a project/idea which I have in my mind toward regeneration type mods and also doing some "control". Those should be smoother as oil taking in account that they will not go intensive with iterations. As always, my focus has orientation to game-play - monster is just A Monster not a stupid lamb, and fancy colors are far from my needs.
medor wrote:
Sat Apr 21, 2018 9:59 am
If just the movements can be more pleasant for play on a distant server .... why not.
I don't get any duplicate kill and/or strange effect coming from these "NewNet" things. In 2018 I intend to grow quality not to fallback in the past. As you can see I'm trying to improve MH play with this XCGE and NavAdder rather than duplicating maps/mods in so called fixed versions.

Re: MonsterHunt Versions & History

Posted: Thu Nov 14, 2019 5:42 pm
by Nelsona
What could I say about MonsterHunt after this time of silence? I think tests, servers differently configured with more or less similar mods, now they show that what I did for MonsterHunt is somehow normal and in acceptable conditions. What map makers do is preferable to stay less discussed since some of them have no clue what they are doing - that's about as much to say at the moment.

Re: MonsterHunt Versions & History

Posted: Mon Feb 10, 2020 5:18 pm
by Nelsona
Bumping Monsterhunt thread a bit.
Previous version of MH2 which I did has an internal deal toward performance but I don't think that is the best ever solution for making things aggressive but smoother in the same time.
Counting Monsters happens often if server has a tick-rate closer to setup and they are counted more rare when tick-rate is falling for preventing an extra-load. I used in that time some UScript things working somehow but I'm not sure how accurate are these. I think there is another method, a native one which I did not know in that time having probably a better accuracy.
Perhaps I need to think at a MH controller using as much as possible natives and linked lists depending on number of Pawns which are in map at time. All these counters involving replication I think have to be aligned toward performance making more space for processing Pawns instead of trying to keep a straight line for updating HUD when map is heavily loaded. I would like to rewrite these things but it might take other 1-2 years, heck... I'm thinking...

As a project for spoofing visuals, monster detecting players in "SightRadius" range should "see" them, decreasing this range if tick-rate goes at 75% of that "proposed" by admin value. I believe this will generate dumb monsters in servers with super duper high tick-rate but which in reality won't even touch 35-45 real tick-rate as under-resourced as they are in MH games - yes, take this in account because I know what I saw in some of now days MH "servers", and then, my attempt to align game at machine's performance capability will only make a mess in certain cases because "button admins" think they know what they do... it will take other ages of working...

I believe I will want to introduce a class into controller activated by a benchmark boolean configurable value stuff at UScript Level but calling timers and native things from Engine and Core as much as possible, even XC_Engine where is available. This will have a purpose to print 5 instances of TickRate configured Vs Real TickRate and reporting what machine is capable in purpose to have a smoother game as possible and controller dynamically adjusting replication stuff. What replication are we talking about ? That HUD data called Monsters and Hunters which have iterations that can be reduced if game has no resources at a moment for them (big explosions, chunks, high load generally). In other hand, we might want a controller working permanently like a clock but using a number of pawns/second adjusted according to tick-rate deviation. Low tick-rate = Less pawns/second - maybe Min 32. High tick-rate = More pawns/second - Max. 1024, instead of instant counting 700 - 800 creatures and making tick spikes.

If you understand what I mean and you are willing to participate in stage we can start to discuss classes and various old problems not solved.

For Bots, perhaps there is not much to do in original Server environment but an extra Control disabling somehow MonsterWaypoints which are not target and... accelerating Bot according to native Hunting code because here Bot is hunting after all. As another common task which might take place is a delayed general pathing tweak also in points/second- Here I believe that we don't need PrunedPaths[x] any, we cannot operate all navigation network in the same time because I can predict a crash based on recursive iterations 1000 points × 16 × whatever stuff... × other native operations = Trash-Box. Probably in those "Invasion" craps we can remove uni-directional paths going out of SpawnPoint actors with WalkJump reachFlags these being probably and more than sure unreachable - you might imagine how much we need to iterate here and that's why this has to be a background state not a single function turning controller in a kamikaze habit, like some MapVoteLA whatever lager and crusher. Here it's the other chapter when Paths list of a navigation point is fully loaded, I don't think we are okay with paths list full filled when NavigationPoint linked to other 8 nearby points might be more than enough and reducing a lot of processing. Probably I would like to reconfigure NavigationPointList in such a case, excepting borked ones from stage and leaving them out of network, this should be doable even in servers without XC_Engine. Known crusher maps (with navigation borked), I don't think will have a sudden recovery... lol. Here I might discuss that loop in spot between two points, editing NavigationPoint if a nearby one has the same distance declared... and another set of iterations would be required.
Definitely these changes must be done around a few seconds because PostbeginPlay() it's not an answer for all these Tasks.

At this moment I'm thinking to wait if UT469 won't have some native fixes before INITGAME and boosted by engine directly, I'm very curious, and then I think I want to wait results before doing some useless work in advance.

For subject duplicated actors I think I cannot do anything in run-time as long as it's place to network issues based on replication and... having chances to crash client or server or... BOTH, here it's under discussion duplicated Movers... but perhaps worth crushing machines in private for some testing... or it's waste of time. Huh ?

As for a maybe crazy idea about controller might be... a delayed initialization. What sort of thing ? We do have a bool with value False located in PrebeginPlay and all those things. Until extra tweaks are operated, nothing is initialized and then bool goes to True and functions called for execution post tweaking :!: .