Alpha 6.1.3

User avatar
GFX47
Dev
Dev
Posts: 2914

Alpha 6.1.3

Post#1 » 23 Dec 2016, 20:17

Release candidate ready for testing!

Windows 64-bit > https://developer.cloud.unity3d.com/share/Zy4Je3jzRb/
Windows 32-bit > https://developer.cloud.unity3d.com/share/WyGNWEnGRW/
Android > https://developer.cloud.unity3d.com/share/bk-Re43fRW/
Linux > https://developer.cloud.unity3d.com/share/Wkf0-EnzCb/
Mac > https://developer.cloud.unity3d.com/share/b1fYZ42MR-/

Changelog > http://roadmap.gladiabots.com


======================================================================================
UPGRADE NOTE:
Upgrading is irreversible, no downgrade possible.
======================================================================================

User avatar
NullPointer
Autonomous Entity
Autonomous Entity
Posts: 539

Re: Alpha 6.1.3

Post#2 » 23 Dec 2016, 20:22

The changes in the nodes helped a little bit, but I still think the Tag filter should appear before the distance filter, sounds more important to me.

User avatar
GFX47
Dev
Dev
Posts: 2914

Re: Alpha 6.1.3

Post#3 » 23 Dec 2016, 20:24

It will be for another update, sorry.

User avatar
NullPointer
Autonomous Entity
Autonomous Entity
Posts: 539

Re: Alpha 6.1.3

Post#4 » 23 Dec 2016, 20:27

No worries, thanks for the new release!

User avatar
HBomb
Algorithm
Algorithm
Posts: 93

Re: Alpha 6.1.3

Post#5 » 23 Dec 2016, 22:48

I only know of this because my game just complained that it was too out of date for multi-player on patch 6.1.2.

Now I'm curious... Do the people with the Android store version get this notification? Or is it just those with test builds that aren't the latest?

User avatar
GFX47
Dev
Dev
Posts: 2914

Re: Alpha 6.1.3

Post#6 » 23 Dec 2016, 23:58

Just the 6.1.x versions.

User avatar
GFX47
Dev
Dev
Posts: 2914

Re: Alpha 6.1.3

Post#7 » 24 Dec 2016, 00:18

OK guys, I think it's time to go live with 6.1.
I'll take a one week break to spend some time with my family and enjoy the end of the year.
I'll continue to check the forum and try to find some time to answer.
No dev before next year though ;)

pier4r
Skynet
Skynet
Posts: 3390

Re: Alpha 6.1.3

Post#8 » 24 Dec 2016, 00:28

I was expecting this after 5.3.1, sometimes you go too fast (setting expectations too high as well). Once a nice release is done, where many enjoy it, just take your time for the next one. Or at least this would be my approach.

It is also a test the long run test (some weeks), to see if with optimizations people find out something not so good or the emergent grouping in the multiplayer suggests other type of changes.
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

User avatar
HBomb
Algorithm
Algorithm
Posts: 93

Re: Alpha 6.1.3

Post#9 » 24 Dec 2016, 00:41

GFX47 wrote:OK guys, I think it's time to go live with 6.1.
I'll take a one week break to spend some time with my family and enjoy the end of the year.
I'll continue to check the forum and try to find some time to answer.
No dev before next year though ;)


6.1.3, right? I miss people...

Also, enjoy the break. You've done plenty recently and many need time to understand tags and subroutines before any other major changes. :)

User avatar
GFX47
Dev
Dev
Posts: 2914

Re: Alpha 6.1.3

Post#10 » 24 Dec 2016, 00:44

Yep, 6.1.3. I said 6.1 because previous public version was 6(.0) ;)

User avatar
Chunkymonkey
Automaton
Automaton
Posts: 139

Re: Alpha 6.1.3

Post#11 » 24 Dec 2016, 00:55

GFX47 wrote:OK guys, I think it's time to go live with 6.1.
I'll take a one week break to spend some time with my family and enjoy the end of the year.
I'll continue to check the forum and try to find some time to answer.
No dev before next year though ;)


Oks thanks. Merry Christmas and Happy New Year!

User avatar
Ritter Runkel
Neural Network
Neural Network
Posts: 498

Re: Alpha 6.1.3

Post#12 » 24 Dec 2016, 05:40

Have some nice days of and thanks allot. Merry Xmas

User avatar
Kanishka
Skynet
Skynet
Posts: 1421
Contact:

Re: Alpha 6.1.3

Post#13 » 24 Dec 2016, 13:52

I will play again when the next update comes. Till then, I plan to take a rest.
Fixes break an AI more than bugs do. :ugeek:

Gladiabots Off-Topic Chat


My Stats: Kanishka_RN3;Kanishka_MiPad

User avatar
NullPointer
Autonomous Entity
Autonomous Entity
Posts: 539

Re: Alpha 6.1.3

Post#14 » 24 Dec 2016, 16:17

My impressions of tagging so far:

- Not so easy to use effectively.
- Can beat my best AIs that don't use tagging or Currently Targeted.
- Not as good as Currently Targeted, even with faster ticks.
- My AIs became considerably bigger in order to reproduce the Currently Targeted behaviour.
- Not all Currently Targeted behaviour was worth simulating with tagging.
- Much easier to implement complex strategies.
- Can be used to simulate a stateful AI.

The last one may be a game changer for me, but I didn't have the chance to play with this idea yet.

All in all, it was a good improvement to the game, but I wonder how new players feel about this. Maybe tagging shouldn't be available from League 1.

User avatar
HBomb
Algorithm
Algorithm
Posts: 93

Re: Alpha 6.1.3

Post#15 » 24 Dec 2016, 17:35

NullPointer wrote:- Can be used to simulate a stateful AI


Can confirm. 4/5 of my team tags are basically states and are doing the world of good.

This level of persistence is a definite game changer for macro management/strategies.

MGBlitz81
Automaton
Automaton
Posts: 135

Re: Alpha 6.1.3

Post#16 » 24 Dec 2016, 18:06

I say it again... bring back currently targeted! It would work well with the tagging system, they go hand in hand. I don't understand the point of taking it away. It's not that it was a quick fix or a short cut, it's a feature that any AI should be able to do... LOCK ON and STAY LOCKED! Games with hardly any AI at all at least makes the enemies lock on to you and vice versa. Besides, the tagging system is NOT USER FRIENDLY. It's ok for people who dig this kind of stuff, but hey... you are going to get a lot of middle school and high school kids who don't obsess about programming and Boolean logic downloading your game. They can't use the tagging feature, they probably can barely figure out the currently targeted node.

Imagine playing any mech game where once you lock on you lose it because another enemy came a little bit closer. Or watch it lock and then switch back and forth as enemies are dancing in your field of view. This shouldn't be a slowed, multi node task. It's as simple as saying attack short range or flee from nearest machine gun bot. If i'm shooting an enemy, it's my priority and I should be able to choose to continue to attack it under all situations without having to tag or untag stuff.

Including currently targeted with tags makes tagging take a back seat like it should. You can create great AI using currently targeted in intelligent ways, but if you utilize the tagging system with it then your great AI gets even better. Keep in mind GFX47, you are making a GAME, not a programming language.

Ok, rant is over. Merry Christmas!

User avatar
NullPointer
Autonomous Entity
Autonomous Entity
Posts: 539

Re: Alpha 6.1.3

Post#17 » 24 Dec 2016, 18:43

I agree with Blitz, the game would be better with tagging and Currently Target co-existing. We could use tagging to focus on the strategy rather than using it to poorly simulate the missing Currently Targeted.

Tagging makes the game better on its own, but is not enough to make Alpha 6 better than 5.3 if CT is removed.

ntk
Algorithm
Algorithm
Posts: 75

Re: Alpha 6.1.3

Post#18 » 24 Dec 2016, 20:03

I'm tired of develop AI for maps i hate... people in free time like do nice things. It become a work... i wait a game on seven wonder / set your priority two days and not played yet.

I wait private/gost matchs

I think the game is near the point where it can exploid or deploid

MGBlitz81
Automaton
Automaton
Posts: 135

Re: Alpha 6.1.3

Post#19 » 24 Dec 2016, 20:28

No one is playing because tagging isn't a good alternative for currently targeted and everyone's AI was messed up because of it.

Just hold out for a while, it will be fixed.

mcompany
Autonomous Entity
Autonomous Entity
Posts: 872

Re: Alpha 6.1.3

Post#20 » 24 Dec 2016, 21:06

My opinions/thoughts of the new tagging system:
1. Currently targeted doesn't seem worth replacing with tags (except for snipers... I'll get to that). Unless the tick system is ridiculously fast, as long as we can't set a target the same tick as we start aiming, it won't be better setting up a currently targeted over just simply attacking first and asking questions later. It eventually becomes a waste of time and makes for AIs that are laggier than the opponent (that said, I'm still not totally against tagging being forced to the tick system to punish over use, it's just that currently, normal use is punished as well)
2. The game rewards not switching targets (with 1000 ms of aiming for even assaults and shotguns, and a whopping 300 ms for sniper aiming). Currently targeted allowed for that to be easily achievable, and was a blessing for snipers who are basically useless without any method to keep them focused. However tagging does the opposite in many cases. Tagging punishes you for trying to keep track of your target (now, I will say that 4 ticks per second makes the punishment not too harsh, but that's basically what it still is). Thus, the only real you should keep track of a target is because snipers have no choice but to have some method of doing this.
3.Tagging is easier to explain than currently targeted, I'll give gfx that. But explaining the uses is SO MUCH HARDER. Thus, if you were to think of points as money, trying to find good uses of tagging just makes "the rich get richer". Not very new player friendly imo,unless that player likes real complex games (although that's what gfx could be going for, idk).
4. I tested in 6.1.2 on the clone apk before 6.1.3 went public, and I can say, setting states on your bots is useful, but can they still are slower than not setting the states in the first place. Setting a state before the battle begins?: Alright, fine , the enemy is lagging as well to set their first attack target. Need to set a state to make sure you checked another state?: You already lost. Changing a state mid-battle?: Fine, but you better make sure you can quickly get back to fighting. Are they useful? Sure, but they come at a cost and I'm not sure that price is worth it (but someone else might prove me wrong). The only maps I saw improvement by using states over not using states was almost always because either the maps started the enemy out of range, or because the lag actually helped for better positioning when attacking (and even then, I was just able to test assaults only vs bots that were still made by me, so it might of even then not been optimal).
5. 6.1 is Actually 6.0, 6.0 is actually 5.4. Just think of how different it was to get used to bot classes. Just give it some time before saying whether people is dropping the game or not. (That said, I do think tags still need some more, because their form is the the best with these current bots)

All in all, I don't like the tagging system as much as currently targeted and I think it is much more limited (despite the fact fact that it was supposed to have more options), but at the same time, this change is so major, that I'm not sure if it is even possible to say much on it.

Return to “Release candidates”

Who is online

Users browsing this forum: No registered users and 1 guest