Help needed testing 3.2.x Campaign games!

Discuss the future of Warzone 2100 with us.
Forgon
Code contributor
Code contributor
Posts: 298
Joined: 07 Dec 2016, 22:23

Re: Help needed testing 3.2.x Campaign games!

Post by Forgon »

Bethrezen wrote:[...] although it seems I'm going to have to reupload the screen shot again because they have been deleted DO'H !!!! stupid image hosts what i will do is add the link to the provided vid as that better demonstrates the issue than a screen shot and written explanation. [...]
Why don't you just upload your screenshots to the forums?
Each post can contain three attachments.
Bethrezen
Regular
Regular
Posts: 661
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen »

Why don't you just upload your screenshots to the forums?
Each post can contain three attachments.
In short I need to be able to upload more then 3 if you look at the outstanding general issues thread I'm already up to 10 screen shots plus 1 save, and now that i have the ability to capture video to then I may well be attaching short clips to demonstrate issues since often its easier to show incorrect behaviour than it is to try and explain/describe it using screenshots.

If I could upload unlimited attachments then would just upload them to the forum because that would save some hassle but unfortunately I can't so I just have to use external hosts for this stuff it's an annoying inconvenience but c'est la vie.
Forgon
Code contributor
Code contributor
Posts: 298
Joined: 07 Dec 2016, 22:23

Re: Help needed testing 3.2.x Campaign games!

Post by Forgon »

@Bethrezen: Concerning your "outstanding general issues" thread:
Bethrezen wrote: [...] In short I need to be able to upload more then 3 if you look at the outstanding general issues thread I'm already up to 10 screen shots plus 1 save [...]
Why don't you just split your bug list into several posts?
Bethrezen wrote: [...] and now that i have the ability to capture video to then I may well be attaching short clips to demonstrate issues since often its easier to show incorrect behaviour than it is to try and explain/describe it using screenshots.
For simple bugs like overlapping GUI elements I still prefer screenshots. More savegames would be nice to easier reproduce some problems, like bug 11.

You can also remove bugs 9 and 10 from your list, since I provided patches for both of them.
pastdue
Warzone 2100 Team Member
Warzone 2100 Team Member
Posts: 339
Joined: 13 Aug 2017, 17:44

Re: Help needed testing 3.2.x Campaign games!

Post by pastdue »

Berserk Cyborg wrote:
pastdue wrote:Could you please test a build of the work-in-progress reduce_qt_dependency_1 branch, and let me know whether it fixes the savegame issues for you? (If not, steps to reproduce the bug with that branch so I can fix it?)
Using Qt 5.9.5 and compiled with GCC 7.3. Mostly everything appears to be working correctly on saveload (no libcampaign crashes/errors)! :wheee:

I did receive two warnings while loading a save on Beta 1. It did not impact anything from my testing of the mission afterwards.

Code: Select all

warning |09:18:21: [toInt:149] Failed to convert string 'me' to int because of error: stoi (**Further warnings of this type are suppressed.)
warning |09:18:21: [toInt:149] Failed to convert string 'scriptName' to int because of error: stoi (**Further warnings of this type are suppressed.)
Those warnings are harmless in this case, and I'll tweak the code to avoid outputting them. (i.e. No behavior change, working as designed, but a new warning logged.)

EDIT: And this (not-so-useful warning) is now fixed in the branch.
Berserk Cyborg wrote:Edit:
I tested a skirmish and got the above warnings on saveload. Should an AI script be attached via the debug menu then these messages are shown on saveload:

Code: Select all

info    |09:54:25: [findEngineForPlayer:790] Script context for player 0 and script name semperfi not found
info    |09:54:25: [findEngineForPlayer:790] Assert in Warzone: qtscript.cpp:790 (false), last script event: '<none>'
This bug seems to affect master as well (does it not reproduce for you on the master branch?). Good catch - the steps you outlined reproduce it consistently (for me) on any branch - will work on this separately.

EDIT: Looks like this second bug has to do with loading player / AI scripts, autogame, and properly handling the case where the script debugger is used to attach a script to a player, and then the game is saved + loaded. I think it's fair to consider this out-of-scope for the "reduce qt" branch (as it's a pre-existing bug in code that hasn't been touched by that branch), and I'll file it on the to-do list for later. Might want per's input on this.
Last edited by pastdue on 06 May 2018, 21:48, edited 2 times in total.
User avatar
Berserk Cyborg
Code contributor
Code contributor
Posts: 938
Joined: 26 Sep 2016, 19:56

Re: Help needed testing 3.2.x Campaign games!

Post by Berserk Cyborg »

I attempted to fix the player not being able to see enemy transporters in f11a3bd2b10c9dde08c163cb0ff0a2a309333449.
Bethrezen
Regular
Regular
Posts: 661
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen »

Why don't you just split your bug list into several posts?
its not really something I'd though about but in general keeping everything in one place makes editing easier since i only have to modify 1 post.
For simple bugs like overlapping GUI elements I still prefer screenshots. More savegames would be nice to easier reproduce some problems, like bug 11.
Usually I'll just use what I think is most appropriate, unless asked for something specific.

With regard to the specific case of issue 11 a save is not really necessary since this can happen on any level, now while I can't be sure about this i think the problem is related to the angle of the camera when the camera is at just the right angle it throws out the hover zone for the object, and you will often have to pan, tilt, or rotate the camera or hover to the left, right above or bellow the object till you can find the hover zone.

why precisely the camera angle throws out the hover zone for objects i don't know, because it shouldn't and it doesn't in v1.10

question is there some way to make the hover zone for objects visible ? that way it might make it easier to determine what is actually going on with this one at the very least it would allow me to test my theory that this is related to the camera angle.
You can also remove bugs 9 and 10 from your list, since I provided patches for both of them.
have those patches actually been implemented, i don't know and can't test it because for the time being master wont load on xp due to a compatibility issue and pastdue is still working on a fix with his reduce_qt_dependency_1 branch which should make cross compiling for xp somewhat easier

or at least i assume that's what he meant when he posted
pastdue wrote:The Qt change was an unintentional side-effect of using a more modern version of MXE to cross-compile (which we needed for non-Qt reasons).

While there is a workaround for MXE, this should also be resolved once I get automated builds working with MSVC, which doesn't have the same complexities making XP-compatible builds with a specific (XP-compatible) version of Qt.
Forgon
Code contributor
Code contributor
Posts: 298
Joined: 07 Dec 2016, 22:23

Re: Help needed testing 3.2.x Campaign games!

Post by Forgon »

Bethrezen wrote:[...] have those patches actually been implemented, i don't know [...]
Yes.
Bethrezen wrote:[...] and can't test it because for the time being master wont load on xp due to a compatibility issue and pastdue is still working on a fix with his reduce_qt_dependency_1 branch which should make cross compiling for xp somewhat easier [...]
You could test it by running Linux from a CD or virtual machine.
Our Linux Compile Guide contains the instructions neccesary to build the master version.
User avatar
alfred007
Regular
Regular
Posts: 619
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Post by alfred007 »

My try to replace firePause with reloadTime failed. As I found out later the ROF is calculated in stats.cpp with firePause. And after I removed firePause the Collective was firing without a pause. So I reverted my changes. I think we should let it as it is because it would be too much work for a just cosmetic change without any significant influence on gameplay. But I made the announced changes in weapons.json and started testing it from beta 01 on.

@ Berserk Cyborg
Transporter is again visible in beta 01 with commit f11a3bd. I also used your libcampaign mod and no errors or info messages occurred in the log files. Logs are attached.

@ Bethrezen
I want to ask you to test my mod with the new weapon effect ALL ROUNDER for cannons and help me with balancing if you can use it with the latest master that is working for you. The changes in weapons.json are

Light Cannon: damage up to 40
Medium Cannon: damage up to 60
Heavy Cannon: damage up to 90
HVC: Damage up to 70
Assault Cannon: damage up to 36

Needle Gun: firePause up to 45
Rail Gun: firePause down to 50
Gauss Cannon: firePause down to 60

Assault Gun: firePause up to 5

Lancer: damage down to 120

Bombard: damage up to 90
Pepperpot: damage down to 30

Hellstorm Howitzer: numRounds up to 3, firePause down to 5, newly defined reloadTime 150
Ground Shaker: damage up to 300

Ripple Rockets: numRounds up to 8, damage up to 125

Angel Missile: numRounds up to 6, damage down to 150, reloadTime up to 450
Archangel Missile: numRounds up to 6, damage up to 300, reloadTime up to 600

In research.json I set the value for damage upgrade for lancers up to 25 because I noticed that lancers were in disadvantage compared to cannons with a value of 20 due to the decreased damage of the lancers. And it makes no sense to get now lancers no longer useful. I hope that we can manage, that lancers and cannons are both (nearly) equally useful in beta campaign until the Assault Gun is researched. I used python hover heavy cannon tanks in my tests of beta 01 and even if they are compared to lancers in disadvantage against tanks they are in favour against cyborgs and buildings.

The changes in structurmodifier.json and weaponmodifier.json are self-explanatory.

The libcampaign mod of Berserk Cyborg is also included; if you don't want to use it, delete it.

Thanks in advance. :)
Attachments
logs beta 1.zip
(3.43 KiB) Downloaded 130 times
Updated-Campaign.wz
(53 KiB) Downloaded 129 times
Bethrezen
Regular
Regular
Posts: 661
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen »

@ Bethrezen
I want to ask you to test my mod with the new weapon effect ALL ROUNDER for cannons and help me with balancing if you can use it with the latest master that is working for you. The changes in weapons.json are.
Sure I can help you out if you want to attempt to do some balance work, I'll just continue to use the last master that worked for me till the incompatibility issue is sorted out it makes bug testing difficult because I'm not seeing the effect of bug fixes, but balancing is something I can do as that can simply be put in to a mod.

What I would suggest though is perhaps a slight change of method, I know it's tempting to go striate for the problem weapons but I feel like that approach is perhaps a little haphazard.

So it might be an idea to start from the very beginning and deal with 1 weapon at a time so that it is balanced in context with the level, take for example the machine-gun very first weapon you get, now due to the fact that you get 3 damage upgrades for that back to back it very quickly becomes way to strong against scav's yet flamers are never used because they are way to weak, so not only does it require a little rebalancing but since the MGs damage problem is caused by getting 3 damage upgrades back to back that also necessities switching the upgrades round a bit.

To go back to some of the discussions we had with Hironaru about this.

Here is the current order you get tech on alpha 01
At the first enemy held oil well you get:
Hardened MG bullets, APDSB MG bullets, APDSB MG bullets mk2

At the first base and the second oil well you get: Flamer turret

At the Second base you get: The Machinegun guard tower

At the final base you get the following tech: Engineering, Tank trap, Mobile Repair tech
So what is wrong with this picture? Several things actually.

1.) You get the machine-gun guard tower before the engineering upgrade that makes no sense.

2.) You get 3 damage upgrades for the machine-gun back to back that is way too much and makes the machine-gun to powerful to quickly.

3.) Scav's out range guard towers which makes there use completely pointless.

4.) The repair turret is in the wrong place that should be at the second base with the flamer turret.

So here is what I would propose.

1.) At the first enemy held oil well you get:

Artefact 1 = Engineering > MG towers and Tank traps

2.) At the first base and the second oil well you get:

Artefact 2 = Flame turret + Repair turret

3.) Now this is where I would diverge from the original game and instead of giving the player 3 damage upgrades for the machine-gun I would do

MG upgrade 1 - Armour piercing bullets - increases damage
MG upgrade 2 - Improved weapon parts - increase accuracy and range
MG upgrade 3 - Improved cooling - Increases rate of fire

And then I would divide them over the last 2 bases.

I would also apply the changes that Hironaru made to the flamer so that it is roughly equal in power with the machine-gun so that using either is a viable choice, id also adjust flamers damage so that the damage is done by the burn instead of on impact like it is with machine-guns, I'd also apply the changes he made to the MG towers so that scav's no longer out range them.

from there we can then move on to alpha 02, and make any necessary changes.
User avatar
alfred007
Regular
Regular
Posts: 619
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Post by alfred007 »

I agree that there are a lot of things to do for a complete rebalancing. The project that Hironaru started is very interesting and will be of course very helpful for a total rebalancing of the game. But it will need a looooooot of time. In a first step, I just wanted to fix some issues that are relatively fast to fix like the overpowered lancers, useless heavy cannons and the overpowered assault gun. I think this are things we can fix before the next release. And after the next release, we can then concentrate our time to Hironaru's BOMB project.

For now, I wanted to give the player the viable alternative between lancers and heavy cannons. And maybe, if we have the time for it before the next release, to give the player for gamma mission the alternative between using a combat group with only assault cannons or using a mixed combat group with tank killers and assault guns instead of using only assault guns nearly till the end of the game.
Bethrezen
Regular
Regular
Posts: 661
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen »

I get where you are coming from the problem however is not just that lancers and the like are to strong when compared to other weapons its also that the other weapon are simply not viable choices now right now.

I highlighted cannons as a sort of test case since cannons are typically used as an alternative to lancers so they are pretty analogues to each other and should therefore exhibit similar perform but the same issues I highlighted with cannons also exist for basically every other weapon as well.

So even if you don’t touch the artefacts or turrets you are still going to need rebalance all weapons if they are to become viable choices and if you are going to do all weapons then the best way to go about that is to start at alpha 01 and then do them one weapon at a time, because right now the only weapons worth using for pretty much the entire game are Lancers, Machine Guns and there variants because pretty much every other weapon is useless for similar reasons as cannons, and this is the point I was trying to make.

Now artillery and vtol weapons are a slightly different matter but they can probably be left until later.

I mean just to put this into context for you lets take the mini rocker pod and compare that to its closest equivalent the heavy machinegun and for a third point of comparison lets add Lancer in to the mix then compare all 3

Currently these are the stats the mini rocket pod has just before taking off for alpha 12

Range - 7.5
Damage - 38
Rate of fire - 150
Weight - 200

And the heavy machinegun has these stats

Range - 7.5
Damage - 36
Rate of fire - 107
Weight - 600

And lancers have these stats

Range - 9
Damage - 256
Rate of fire - 10
Weight - 250

So if we do Damage X Rate of Fire for all 3 weapons we get the following

Mini Rocker Pod
38 x 150 = 5700 damage per minute

Heavy Machinegun
36 x 107 = 3852 damage per minute

Lancer
256 x 10 = 2560 damage per minute

So we can see from this that in theory the lancer is the weakest of the 3 weapon however the numbers don’t tell the full story.

And if you have ever used mini rocket pods then you will know that Mini Rocket Pods doesn't come anywhere remotely close to doing 5700 damage per minute and in fact it performs significantly worse then both lancers and Heavy Machinegun due in part to its poor accuracy, so if mini rocket pods are ever going to become a viable choice they need to have the accuracy issue fixed which will in turn likely necessitate adjusting its damage and / or rate of fire

we see similar issues when we compare Machineguns and Flamers again they should have roughly similar performance but they don't the Machineguns are significantly better and you see this again and again and again.

so you can see why I'm saying that really you need to go back to alpha 1 and balance 1 weapon at a time in context with the level and other weapons that are available on that level so that they all come out with roughly similar performance so that they are all viable choices.

Now don’t get me wrong this is a big and time consuming job so much so that i can't see this being ready for next release even if you limit your self to just a small sub section of the available weapons say tank mounted direct fire weapons for example

the other problem with this is that if you balance weapons say heavy cannons for example for fighting the collective then don't you run the risk of making them to strong again the New Paradigm so again its better to start from the start and then if they prove to weak against later opponents we can just buff the upgrades slightly

a little something for you to think about
User avatar
Berserk Cyborg
Code contributor
Code contributor
Posts: 938
Joined: 26 Sep 2016, 19:56

Re: Help needed testing 3.2.x Campaign games!

Post by Berserk Cyborg »

Qt reduction branch is in for review: https://github.com/Warzone2100/warzone2100/pull/170.
--
Bethrezen wrote:I would also apply the changes that Hironaru made to the flamer so that it is roughly equal in power with the machine-gun so that using either is a viable choice, id also adjust flamers damage so that the damage is done by the burn instead of on impact like it is with machine-guns, I'd also apply the changes he made to the MG towers so that scav's no longer out range them.

from there we can then move on to alpha 02, and make any necessary changes.
Here is something to try. I brought back Hironaru's efforts into this mod and used your research order and included custom modifiers based off of alfred007's work. I'll merge in alfred007's weapon stat changes into the mod next.

There are two MG upgrades: first increases damage and range, and the other does the same (with lesser increases) and a bullet speed increase. More upgrades can be unlocked when we get the twin MG researched,
if necessary. Flamer impact damage is lower than the overall periodical damage. I say its useful now.

The easiest way to remove the OPness of lancer and assault gun is to make the anti-tank and anti-personnel very weak against hard structures and somewhat for bunkers. Anti-personnel is way too strong against tracks and arguably half-tracks. Much of which the mod's modifiers already achieve.

This is only for Alpha 1, so don't use this any farther (for saves anyway).
updated-campaign.wz
Bethrezen
Regular
Regular
Posts: 661
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen »

Ok so just gave that mod a go and it seems to be totally broken not sure if its just coz I'm running it on master warzone2100-master-20180204-051309-ab17b9b or what but I'm getting errors

Here are the logs

Code: Select all

--- Starting log [C:\Documents and Settings\<User Name>\Desktop\Warzone 2100_portable-master\Warzone 2100 master\logs\WZlog-0512_005631.txt]---
error   |12:56:41: [js_enumArea:3774] Label scavBase1Cleanup not found
error   |12:56:41: [callFunction:211] 0 : camSetEnemyBases(bases = [object Object]) at script/campaign/libcampaign.js:1025
error   |12:56:41: [callFunction:211] 1 : eventStartLevel() at script/campaign/cam1a.js:173
error   |12:56:41: [callFunction:211] 2 : <global>() at -1
info    |12:56:41: [callFunction:214] Uncaught exception calling function "eventStartLevel" at line 1025: ReferenceError: labels.contains(label) failed in js_enumArea at line 3774
info    |12:56:41: [callFunction:214] Assert in Warzone: qtscript.cpp:214 (false), last script event: '<none>'
error   |12:59:14: [js_hackMarkTiles:4239] Label launchScavAttack not found
error   |12:59:14: [callFunction:211] 0 : __camUpdateMarkedTiles() at script/campaign/libcampaign.js:679
error   |12:59:14: [callFunction:211] 1 : cam_eventCheatMode(entered = true) at script/campaign/libcampaign.js:3764
error   |12:59:14: [callFunction:211] 2 : <global>() at -1
info    |12:59:14: [callFunction:214] Uncaught exception calling function "cam_eventCheatMode" at line 679: ReferenceError: labels.contains(label) failed in js_hackMarkTiles at line 4239
info    |12:59:14: [callFunction:214] Assert in Warzone: qtscript.cpp:214 (false), last script event: '<none>'

--- Starting log [C:\Documents and Settings\<User Name>\Desktop\Warzone 2100_portable-master\Warzone 2100 master\logs\WZlog-0512_010118.txt]---
error   |01:01:25: [js_enumArea:3774] Label scavBase2Cleanup not found
error   |01:01:25: [callFunction:211] 0 : camSetEnemyBases() at script/campaign/libcampaign.js:1025
error   |01:01:25: [callFunction:211] 1 : cam_eventGameLoaded() at script/campaign/libcampaign.js:4034
error   |01:01:25: [callFunction:211] 2 : <global>() at -1
info    |01:01:25: [callFunction:214] Uncaught exception calling function "cam_eventGameLoaded" at line 1025: ReferenceError: labels.contains(label) failed in js_enumArea at line 3774
info    |01:01:25: [callFunction:214] Assert in Warzone: qtscript.cpp:214 (false), last script event: '<none>'
error   |01:12:55: [js_enumArea:3774] Label scavBase2Cleanup not found
error   |01:12:55: [callFunction:211] 0 : eventStructureBuilt(structure = [object Object], droid = [object Object]) at script/campaign/cam1a.js:84
error   |01:12:55: [callFunction:211] 1 : <global>() at -1
info    |01:12:55: [callFunction:214] Uncaught exception calling function "eventStructureBuilt" at line 84: ReferenceError: labels.contains(label) failed in js_enumArea at line 3774
info    |01:12:55: [callFunction:214] Assert in Warzone: qtscript.cpp:214 (false), last script event: '<none>'
Also no artefacts are dropping, the timer doesn't start, the vid about the incoming scav's doesn't play either, so yeah the whole thing is totally broken.
User avatar
Berserk Cyborg
Code contributor
Code contributor
Posts: 938
Joined: 26 Sep 2016, 19:56

Re: Help needed testing 3.2.x Campaign games!

Post by Berserk Cyborg »

Bethrezen wrote:Ok so just gave that mod a go and it seems to be totally broken not sure if its just coz I'm running it on master warzone2100-master-20180204-051309-ab17b9b or what but I'm getting errors
Ah, forgot about that. I have to give you the new (relative to that master) labels which hopefully make it work.
updated-campaign.wz
Bethrezen
Regular
Regular
Posts: 661
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen »

Ah, forgot about that. I have to give you the new (relative to that master) labels which hopefully make it work.
Ok so that seemed to work properly this time.

So a couple of quick observations, first off I notice that the machine gun in its default state seems to of been nurfed,

In an unmoded game the default state of the machine gun is

Range - 6.0
Damage - 10
Rate of fire -120
Weight - 200

With the mod they have

Range - 5.9
Damage - 7
Rate of fire -150
Weight - 200

not really sure if it’s necessary to nurf machine guns base stats because to my mind the main issue with machine guns on alpha 1 is simply that they got 3 damage upgrades back to back which effectively doubled the machine guns damage which is probably a bit much and I don’t really think that they need that much of a damage increase at least not until you encounter the new paradigm which if I recall correctly was on alpha 05, up until then you are only fighting scav's and they aren't really that strong, so I would be tempted to leave machine guns base stats alone, and simply switch up the upgrades a bit so they don’t get so much of a damage boost.

Second I notice that you have to research the flamer turret before you can research the repair turret I'm not really sure that making the flame turret a pre requisite for the repair turret makes sense, so I would probably decouple them.

Flamers I think need to be a bit more effective against structures, they could also perhaps do with a bit more range, because the combination of the short range and relative ineffectiveness against structures resulted in them taking a lot of damage against enemy turrets probably more than they really should.

Also I can't be certain but I think they might have been taking friendly fire if that is the case then that needs to be turned off given that flamers are an area effect weapon.
There are two MG upgrades: first increases damage and range, and the other does the same (with lesser increases) and a bullet speed increase. More upgrades can be unlocked when we get the twin MG researched,
Ok so after collecting the MG upgrades the stats of the MG looked like this.

Range - 5.9
Damage - 11
Rate of fire -150
Weight - 200

So only the damage was increased so don’t think the upgrades are working right also the info in the intel screen for the MG upgrades only mention damage as well for both upgrades.

as far as MGs go on alpha 1 personally I think I'd go for

1 upgrade that increases the range from 6.0 to 7.5 matching the Heavy MG range
1 damage upgrade that takes the damage from 10 to 12 the (default damage upgrade is 3)
1 rate of fire upgrade that increases the rate of fire form 120 to 150

BTW what does bullet speed increase actually do since that info is not displayed in the intel screen or the design screen because I cant say I really noticed any difference.
Post Reply