Help needed testing 3.2.x Campaign games!

The projects speaking tube.
Add your two cents if you want to.

Re: Help needed testing 3.2.x Campaign games!

Postby Mauricio Yaman Yusuf » 24 Apr 2018, 15:50

where to report these wonderful documentation Rubbish?

2018-04-24_08-18-36.gif
User avatar
Mauricio Yaman Yusuf
Greenhorn
 
Posts: 9
Joined: 28 Jan 2015, 10:41
Location: México City

Re: Help needed testing 3.2.x Campaign games!

Postby Forgon » 24 Apr 2018, 16:02

Mauricio Yaman Yusuf wrote:where to report these wonderful documentation bullshit?

There is nothing to report.
Our official repository is found here, and differs from the fork you visited.
Documentation is also not final yet -- expect changes before the next release.

The PIE format is explained here.
For more detailed information study the source code.
Review my new green assignment crosshair cursor: ticket #4778
Forgon
Code contributor
Code contributor
 
Posts: 120
Joined: 07 Dec 2016, 22:23

Re: Help needed testing 3.2.x Campaign games!

Postby Mauricio Yaman Yusuf » 25 Apr 2018, 04:29

[edit]
Please don't be rude and please don't use profanity.

If you don't understand the repository then ask politely and I'm sure someone will be able to point you in the right direction.

If you want to know where to download the latest masters, then again all you need do is ask politely and I'm sure some one will be able to point you in the right direction.

Bethrezen
User avatar
Mauricio Yaman Yusuf
Greenhorn
 
Posts: 9
Joined: 28 Jan 2015, 10:41
Location: México City

Re: Help needed testing 3.2.x Campaign games!

Postby alfred007 » 29 Apr 2018, 11:46

At the moment that I write this post, I can't find commit 7487247 for windows at buildbot but only for Mac. The last commit for windows is commit 0182af5, but this is one commit behind master at github. Is there something not working or is that intended?
alfred007
Trained
Trained
 
Posts: 277
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Postby alfred007 » 29 Apr 2018, 12:16

Mauricio Yaman Yusuf wrote:
vexed wrote:For those of you playing the Campaign game using 3.2.x, you may have noticed that it is broken in a variety of places. :(

Well, we have a few contributors who have took it upon themselves to help convert the Campaign game.
You will find their work in this thread: viewtopic.php?f=35&t=11872

However, we need lots of testing, to make sure the Campaign is solid, as it was in the past, so, if you have time, then please help us test the Campaign.
Please read this post: viewtopic.php?f=35&t=11872&p=135811#p135790
and download the .wz file located there (download/file.php?id=16375), and then use that file as a mod, and start playing Campaign games like you would before.
We need to know:
A) If it is a direct copy of how the Campaign mission is supposed to work as it did in the past.
B) If you run into any bugs.
C) Test saving & loading on missions and away missions to make sure those work as well.
D) Try breaking things. :)
E) Performance issues.

Once these have been debugged, the plan is to out them into the repo in time for the next release.
Oh, and feel free to report bugs in that thread, or, you may report things in this thread.

Thanks for testing!


I do not understand your repository, it's full of sh*t, incredible that you can not even put a small text file that tells you part of the code you're reading, have a map or diagram of how it's built, I would honestly help you but I do not know if do it, in the past I observed personal quarrelling between several groups that wanted to do what they wanted, and I do not like to be arguing, although I have black character and mood, it is not a pleasure to be fighting for nonsense

On the other hand, even though it burns them to die to tell them their truths with franquesa, it is inevitable to lament that this beautiful game that gave me so many hours of happiness in my lonely childhood, has been in the hands of a handful that far from sharing knowledge for Grow and do this more sensibly, have one to be begging for a little attention and to explain to one where it starts and where it ends up.

I gave someone is interested in giving me a hand I'll thank you, otherwise I'll have to do what many are doing, look for a similar game where you can participate without so much envy, it is really unfortunate that this is losing followers and fans


There is no need for getting rude or respectless. All developer, code contributor and who else is also working on this project is doing this unpaid in his free time. To "understand" the repository you must have programming knowledge. If you don't have any programming knowledge it would be waste of time to explain what every part of the code means. And the fact that you don't understand the repository tells me (Tell me if I'm wrong) that you don't have much programming knowledge. Like me. During testing the game and with the help of Berserk Cyborg I learned a few things about the source and how the programm is built. Now I can define trigger areas and factory assemblies. And I understand a little bit of a few things how the code is working. But I would never ever be able to do what Berserk Cyborg or Forgon or the other code contributor are doing with the code. And even if they would describe me what they are doing I wouldn't understand it due to my not existing programming knowledge.
So what I do is testing the changes the programmers are doing, telling them if I find bugs or issues and giving suggestions what to change in the code. And if you want to help us to make this great game even better download the latest master from buildbot (http://buildbot.wz2100.net/files/) and play the game. And if you found any issues or bugs or if you have any ideas for changing post them here in this thread. And if you have a specific question and ask politely I'm sure you will get an answer. Talking about the repository is full of sh*t and writing in bold letters won't help.
alfred007
Trained
Trained
 
Posts: 277
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Postby NoQ » 29 Apr 2018, 17:40

I'll have to do what many are doing, look for a similar game

I'm really worried about that game :?
User avatar
NoQ
Special
Special
 
Posts: 6136
Joined: 24 Dec 2009, 11:35
Location: /var/zone

Re: Help needed testing 3.2.x Campaign games!

Postby Berserk Cyborg » 29 Apr 2018, 21:26

alfred007 wrote:At the moment that I write this post, I can't find commit 7487247 for windows at buildbot but only for Mac. The last commit for windows is commit 0182af5, but this is one commit behind master at github. Is there something not working or is that intended?
Buildbot is probably broken again.
User avatar
Berserk Cyborg
Code contributor
Code contributor
 
Posts: 527
Joined: 26 Sep 2016, 19:56

Re: Help needed testing 3.2.x Campaign games!

Postby alfred007 » 29 Apr 2018, 22:33

Berserk Cyborg wrote:
alfred007 wrote:At the moment that I write this post, I can't find commit 7487247 for windows at buildbot but only for Mac. The last commit for windows is commit 0182af5, but this is one commit behind master at github. Is there something not working or is that intended?
Buildbot is probably broken again.

Looks like buildbot is working again. I could download commit 7487247.
alfred007
Trained
Trained
 
Posts: 277
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Postby Bethrezen » 30 Apr 2018, 18:09

vexed wrote:
Bethrezen wrote:Unfortunately I can't test that because master wont load for me just updated to the latest version
warzone2100-master-20180211-143024-e02cc41.exe (11-Feb-2018 13:58) and when i try to start it I'm still getting the following error.

Image

The problem is that is a Vista DLL apparently and I'm running XP, so the question is why is master looking for a Vista dependency because it runs fine on XP or at least it did ?

If i had to hazard a guess it's probably changes made to get the resizing high dpi working in build bot so someone needs to check that and figure out what got messed up.


I looked over the code, and I see no reason at all why we had to move from a version of Qt that supported XP & Vista to what we are on now. We have added nothing new, that I can tell, that uses the newer API calls, so, I think we should roll-back Qt to a version that still works with XP / Vista, and not let our primary testers down.

Sorry Bethrezen. :(


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.


Don’t worry about it these things happen, it’s a bit of a pain sure but it's fine, I can just cool my jets for a while till this gets sorted out, talking of which any ETA on when that’s likely to be, or should I just keep doing what I have been doing and just keep checking back periodically ?
Bethrezen
Trained
Trained
 
Posts: 481
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Postby alfred007 » 01 May 2018, 00:29

I updated to master 7487247 and moved forward to beta 01. I found a weird bug. The CO transporter was invisible. As you can see at the two screenshots below one second is nothing to be seen of the CO transporter and one second later you see the unloaded units. When I tried to reload a saved game where the transporter is already announced the game crashed. When I reloaded a game where the transporter was not already announced no more transporters arrived.

wz2100-20180430_230414-CAM_2A.png

wz2100-20180430_230424-CAM_2A.png


The invisible transporters came in a game that I started from the end of alpha mission. Except for the invisible transporters, everything worked fine.
Attached are the logs for the game from the end of alpha mission including the crash report, the logs for a reloaded game where no more transporters arrived and two saved games.

I also reactivated an old idea of mine (viewtopic.php?f=1&t=12714&p=139550&hilit=all+rounder#p139550) for a first step for rebalancing some weapons. I took all cannons and Gauss guns to the new weaponEffect ALL ROUNDER in weapons.json. Also, I used the changes from Berserk Cyborg based on my idea of weaponmodifier.json and structuremodifier.json. Then I decreased the base damage of lancers to 120 and increased the base damage of the heavy cannon to 90. The reason is the overpowered lancer and the nearly useless heavy cannon during campaign. I used this changes since I started a new alpha campaign end of January. Lancers are still powerful but no longer overpowered. So I think we can reduce the base damage for lancers to 120. As you can see in the attached saved game I use Heavy Cannon Python Hover Tanks. In all my other tests I used Lancer Mantis Tracked Tanks. In this test for beta 01 the heavy cannons look like an useful alternative to lancers. I will move on with my the tests to see if they are now really no longer useless. The used mod is also attached.
Attachments
beta 01.zip
(153.61 KiB) Downloaded 10 times
alfred007
Trained
Trained
 
Posts: 277
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Postby Berserk Cyborg » 01 May 2018, 01:01

See http://forums.wz2100.net/viewtopic.php?f=1&t=12714&start=930#p140772. Until then, loading any savegame might cause all kinds of interesting side effects. Or crash as Beta 1 always does at the moment. You'll basically need to save/load exclusively on the pre-away missions, which seem to be unaffected.

I was thinking about how practical redoing the whole stats and research tree of the campaign would be (or at least doing an intensive review of everything and molding it with suggestions). Even bring in the stuff exclusively found in multiplayer. Its probably the one big flaw with the campaign right now.
User avatar
Berserk Cyborg
Code contributor
Code contributor
 
Posts: 527
Joined: 26 Sep 2016, 19:56

Re: Help needed testing 3.2.x Campaign games!

Postby alfred007 » 01 May 2018, 16:09

I reloaded a saved game from the end of alpha 12 and moved forward to beta 01. The transporter is still invisible.
alfred007
Trained
Trained
 
Posts: 277
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Postby Bethrezen » 01 May 2018, 23:57

Not sure if you would necessarily have to redo the whole tree but certainly there are a lot of weapons that need rebalancing, cannons being a prime example while heavy cannons have the same rate of fire as lancers they do only a fraction of the damage now part of that may be due to the fact that lancer damage is calculated incorrectly which results it actually doing double the damage that it says in the design screen which is incorrect behaviour so the damage calculation for that needs fixing, because it should not be counting both shots individually and then adding them up to get the total damage

the total damage done for both shots should be what it says in the design screen like it is for every other weapons, and if only 1 shot hits then target should only receive half the damage that it says in the design screen modified by damage reduction modifiers of course.

now in v1.10 you don't notice this bug but that's only because its masked by another bug that stops lancer research bonus being applied.

with regards to bringing in stuff from multiplayer there are definitely certain things from multiplayer that would be helpful but equally there are some things in the multiplayer tree that wouldn't be appropriate for campaign mode.

for example having light, medium and heavy cyborg body, would be helpful because right now cyborgs in campaign are useless they don’t have enough stopping power the rate of fire on there weapons is to slow and they are way way way to easy to kill for them to be any use at all they need at least double the armour and firepower so that they are comparable to tanks.

It would also be nice to be able to design cyborgs in the units design screen to so that they can be equipped with all the equipment that you have access to at any given point in the game because by the time you get to alpha 12 for example you have 13 weapons for your tanks yet only 4 cyborgs why I don’t know but really your cyborgs should be able to equip any of those 13 weapons and any of the none combat attachments as well so that you can create builder cyborgs, repair cyborgs, scanner cyborgs or cyborg commanders.

It would also be nice to get the wide spectrum scanner once you have all the other types of scanner tech or even better just get rid of all the other types of scanners and simply replace them with the wide spectrum scanner by default as that can do the job of any of the other scanner types anyway.

now in terms of things that wouldn't be appropriate to add to campaign mode the ultra heavy Wyvern and Dragon bodies come to mind.

some of the tech could also perhaps do with moving around as well the pepper pot comes to mind that weapon is automatically obsolete as soon as you get it on beta since you already have the howitzer and ripple rockets so what i would do is move the hell storm howitzer to where you get the pepper pot on beta and I'd move the popper pot to the level on alpha where you get the bombard alpha 8 i think it is if i recall correctly
Bethrezen
Trained
Trained
 
Posts: 481
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Postby alfred007 » 02 May 2018, 23:07

Bethrezen wrote: ...due to the fact that lancer damage is calculated incorrectly which results it actually doing double the damage that it says in the design screen which is incorrect behaviour so the damage calculation for that needs fixing, because it should not be counting both shots individually and then adding them up to get the total damage

the total damage done for both shots should be what it says in the design screen like it is for every other weapons, and if only 1 shot hits then target should only receive half the damage that it says in the design screen modified by damage reduction modifiers of course.


This belongs to the way you interpret the design screen. I made a fast test in the alpha/beta transition level. The design screen shows a ROF for lancers of 10. And lancers are shooting 5 times 2 rockets per minute. That are 10 projectiles. If you interpret the ROF as projectiles per minute and not as shots per minute the calculation is correct. For weapons with only one barrel projectiles per minute and shots per minute are the same. But there is a difference for weapons with two barrels like lancers, tank killers and scourge missiles. So the ROF multiplied with the damage gives the correct damage per minute for lancers too.
Only one thing confused me. The design screen showed a damage per projectile of 192. Two projectiles with a damage of 192 each multiplied with the modifier for tracked tanks of 125 results in a damage of 480 for every shot. But the tank got only a damage of 404. I shot at an own tracked python commander tank. Maybe this is the effect of the armor of the python body, but I don't know. Where I agree with you is the effect that the tank received the full damage no matter if both or only one rocket hit the target. This should be fixed but I don't know how much work this would be.

Edit: I looked again at the design screen and the python body have a kinetic armor of 38. If you subtract this armor from the damage for every rocket you get the difference of 76 between calculated damage of 480 and noticed damage of 404.
alfred007
Trained
Trained
 
Posts: 277
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Postby Bethrezen » 03 May 2018, 14:31

This belongs to the way you interpret the design screen. I made a fast test in the alpha/beta transition level. The design screen shows a ROF for lancers of 10. And lancers are shooting 5 times 2 rockets per minute. That are 10 projectiles. If you interpret the ROF as projectiles per minute and not as shots per minute the calculation is correct. For weapons with only one barrel projectiles per minute and shots per minute are the same. But there is a difference for weapons with two barrels like lancers, tank killers and scourge missiles. So the ROF multiplied with the damage gives the correct damage per minute for lancers too.


see and there in lies the flaw in the way lancer damage and the damage of similar weapons is calculated which is what results in lancer/tank killer etc vtols doing way way way to much damage and one shotting even your most heavily armoured units, because it counts both shots individually and then adds up the damage of both shots to get the total damage resulting in them doing double the damage they should.

This also has the effect of rendering prity much every other weapon pointless because no other weapon can even come close to matching the damage of lancers and similar this is why i reckon that the damage calculation for lancers and similar needs to be reworked. If the design screen says the damage is 192 before applying modifiers then that should be the total damage for both shots, doing it this way the damage output is brought back down to the expected level and when you look at the design screen it gives you the correct information which currently it doesn't.

so lets take some numbers from before you take off for alpha 12 for this test I'll use 2 python lancers on tracks and just to stop things getting complicated I'll use only rookie units since rank confers bonuses to damage reduction and fire power.

now in the design screen it says the damage value of these lancers is 256 and it says that the health is 1608.

when i force fire one unit against another the damage done is

1608 - 1032 = 576

now while 256 x 2 is 512 and not 576 this can be explained by the damage modifier because i think lancers get a bit more damage against tracked vehicles so we can ignore that small difference.

Anyway as we can see the game has added up the damage of both shots so instead of the total damage for both shots being 256 before modifiers it has in fact inflicted double the damage that it says in the design screen this is incorrect to further prove this we can compare lancers with its nearest equivalent the heavy cannon.

if we build a python tracked heavy cannon and fire it once at one of the lancers that single shot inflicts 134 damage now while it says in the design screen that the heavy cannon has a damage value of 133 and not 134 this is probably just a rounding error and can be ignored, more over this proves that the damage value that is displayed in the design screen is the damage that a single shot should do before any modifiers are applied, therefore as lancers and smiler fire 2 rockets then the value that is displayed in the design screen should be the damage done if both rockets hit the target.

if we extrapolate further we can work out roughly what the game is doing when it calculates lancer damage

256 per shot x 2 shots x 10 shots per minute = total damage per minute 5120

Obviously this calculation is what it would calculate before you start adding modifiers assuming 100% accuracy

now if the lancer damage calculation worked the same way it does for other weapons like the heavy cannon for example then the damage calculation would be

256 damage for both shots x 10 shots per minute = total damage per minute 2560

which is exactly half the damage and is in line with what the heavy cannon does

heavy cannon damage 133 x 12 shots per minute = total damage per minute 1596

Again this is what it would calculate before modifiers are applied assuming 100% accuracy

so from this i have just proven 4 things

1.) the damage calculation for lancers and other similar weapons is incorrect.

2.) the damage value for lancers and similar that is displayed in the design screen is also incorrect.

3.) Heavy cannons are underpowered and should probably be inflicting between 500 to 1000 more points of damage, since really the heavy cannon should be comparable in power to lancers.

now i realise that giving heavy cannons the same damage as lancers would make them overpowered since they have heaver armour so what i would do is take the damage that lancers do reduce it by 25% and then make that the new damage value for the heavy cannon why only 25% instead of the current 50% because cannons are also slower moving then lancers so by only having 75% of the damage of lancers and by being slower moving this balances out the extra armour that the heavy cannon provides.

4.) the damage value displayed in the design screen is the damage for a single shot before modifiers are applied and since lancers fire 2 shots then for lancers and similar the damage value displayed in the design screen should be for both shots, and if only 1 shot hits then the damage done should be half the value displayed in the design screen.

This gets even worse when you are talking about lancer vtols since they already do double the damage of ground based lancers because they only get 1 attack before they have to go and rearm and before you say it yes i know the damage of lancers vtols was reduced a bit to try and bring down there damage to a more reasonable level because they are overpowered as hell but irrespective of this the point still stands the damage calculation for lancers and similar weapons is bugged which is resulting in them doing double the expected damage the ironic thing about this is that this is a very old bug that has been around for years it's just never been fixed.

Now presumable the reason for this issue is that the game is using 1 damage calculation for all weapons when it should really be using 2, 1 for weapons like the cannon that fire only 1 shot and another for weapons like the lancer that fire 2 it's probably also the reason that the design screen shows the wrong damage value for lancers and similar because again its using the same block of code to display the damage value for all weapons when it should be using 2.

now in theory this should actually be a really easy problem to solve because all you need to do is to copy the damage calculation modify it to fix the issue and then insert both calculations in to an if else if statement and then just add a couple of conditions to make sure the correct calculation is used and you can do the same thing to fix the issue with the design screen so that it shows the correct damage value for lancers and similar, of course this is assuming I'm right about the cause of the issue that is.

In any even if any rebalancing work is to be done then the first step will be to fix the faulty damage calculation for lancers and similar so that the damage they are inflicting is within the same range as other weapons.

Now i realise that doing this will in all likelihood necessitate some rebalancing due to the fact that collective units would more or less be impossible to beat if lancers where only doing half the damage they do currently due to this bug and i know that because i have tried fighting them with tracked heavy cannons in the past and its a real slog because there units are so heavily armoured and you inflict so little damage, but the fact of the matter is the entire campaign needs rebalancing anyway because lancers are way overpowered due to this bug and most other weapons are way underpowered and are totally unless even against new paradigm units, and they are the weakest units in the game after scavs so if they are useless against the new paradigm then against the collective or nexus you may as well be shooting blanks.

Where I agree with you is the effect that the tank received the full damage no matter if both or only one rocket hit the target.


actually this is incorrect when a lancer fires if only 1 rocket hits then it only takes half the damage to put this into perspective i will again take some numbers from just before you take off for alpha 12 and again both units will be rookie tracked python lancers to stop things getting complicated again the damage value displayed in the design screen is 256 and the health value is 1608

when i force fire one lancer again another if only 1 rocket hits then the damage is

1608 - 1320 = 288

where as if both rockets hit then the damage is

1608 - 1032 = 576

now since the design screen displays a damage value of 256

if only 1 rocker hits then the damage done should be 144
if both rockets hit then the damage done should be 288

However we can see from the maths above the lancer is in fact inflicting double the damage that it should be, because the damage calculation is wrong.

I shot at an own tracked python commander tank. Maybe this is the effect of the armour of the python body, but I don't know


did the units you used in your test have any rank if so that could skew the results because rank confers various bonuses its why i used rookie units because removing variables makes the maths easier to do and easier to understand it also makes it far less likely that I'm going to make a mistake and get it wrong.

what i would say to you if you still think I'm wrong is to repeat my test with the same parameters

just before you take off for alpha 12 build 2 rookie python tracked lancers and then make note of there health

then have one lancer fire on another and make a note of the damage inflicted when only 1 rocket hits and when both rockets hit

next build a rookie python tracked heavy cannon make note of its damage then have it fire a single shot at one of your lancers and make note of the damage it inflicts

then take these damage numbers and multiply them by the rate of fire to get the damage per minute

finally compare the damage per minute of the lancer to the damage per minute of the heavy cannon when you do you will see what i mean when i say that the lancer is in fact inflicting double the damage that it should be because as i already noted above the damage per minute of the heavy cannon is 1596 where the damage per minute of the lancer is 5120

now if the damage calculation for the lancer was working correctly then the lancers damage per minute would only be 2560 not 5120.

as i already mentioned lancers are doing double the damage they should be and if you followed my methodology and successful replicated my results you'll see that I'm right you can't argue with the maths :)
Bethrezen
Trained
Trained
 
Posts: 481
Joined: 25 Sep 2009, 02:05

PreviousNext

Return to News and announcements