Welcome to the LnLP Forums and Resource Area

We have updated our forums to the latest version. If you had an account you should be able to log in and use it as before. If not please create an account and we look forward to having you as a member.

SITREP

Dave 'Arjuna' O'Connor

Panther Games Designer
Joined
Jul 31, 2014
Messages
3,415
Points
113
Location
Canberra, Australia
Website
www.panthergames.com
SITREP 1000 1 May 2015

New Build 5.0.11 underway at the moment. I have compiled a list of fixes under the Build 5.0.11 thread. Still outstanding are:
  1. HQ Holdout issue
    • I would like a save with the new 5.0.11 build that I can start from and see this issue.
  2. Out of Sync issue
    • I need to check to see if the changes I have made in 5.0.11 have addressed this.
  3. Estab Compile issue
    • Still to look at. Will do so next week.
  4. Keyboard shortcuts for dialogs
    • Currently awaiting advice from Paul
Have I missed anything?
 

Dave 'Arjuna' O'Connor

Panther Games Designer
Joined
Jul 31, 2014
Messages
3,415
Points
113
Location
Canberra, Australia
Website
www.panthergames.com
SITREP 1730 Fri 8 May 2015

It's been a tough week coding wise because of the problems with the Fire code. Miquel had done some nice work but also made some serious missteps. I thought I was finished with this pending testing, but when I started testing, the data logs (which Miquel has implemented and are much appreciated) these highlighted the need for some further mods. I've run out of time this week but will pursue this matter on Monday.
 

EdJaws

Member
Joined
Jan 11, 2015
Messages
76
Points
8
Age
72
Location
USA
I got a few years on you and I've slowed down quite a bit. I sometimes wonder why the hell you endeavoured to update CO with all the long, hard work it entails. A lessor man might quit but you obviously have passion for Command Ops.

Ed
 

Iconoclast

Member
Joined
Oct 26, 2014
Messages
130
Points
18
Location
germany
A lessor man might quit but you obviously have passion for Command Ops.

And this is why Theodor Fontane dedicated this poem to Dave:

Dave O'Connor!

"Who is Dave O'Connor?"

"Dave O'Connor was our helmsman true.
To solid land he carried us through.
He saved our lives, our noble king.
He died for us; his praise we sing.
Dave O'Connor."

From CO1 to to CO2
As mist sprays her bow like flakes of snow
Over Lake Erie the "LnL Publishing" takes flight
And every heart is joyful and light.
In the dusk, the passengers all
Can already make out the dim landfall,
And approaching Dave O'Connor, their hearts free of care,
They ask of their helmsman, "Are we almost there?"
He looks around and toward the shore:
"Still another bug to solve.... a half hour more."

All hearts are happy, all hearts are light --
Then out of the hold comes a cry of fright.
"more Bugs!" it is, that terrified shout.
From the cabin and hatch black smoke pours out.
Smoke, then fire and flames aglow,
And still 20 minutes to the next build.

And the passengers, in a colorful crowd
Stand pressed together on the bow.
Up on the bow there is still air and light
But the smoke at the helm forms a thick, dark night.
"Where are we? Where?" Peter must know,
And still 15 minutes to the next build. --

The wind grows strong but the smoke cloud stays.
To the helm Daz turns his gaze.
The helmsman is hidden by the raging fires
But through the bullhorn Daz enquires:
"Still there, Dave O' Connor?"
"Yes, sir. I am."
"Onto the beach! Into the surf!"
"Yes, sir. That's my plan."
And the people cry: "Hold on! Hallo!"
And still 10 minutes to the next build.--

"Still there, Dave O'Connor?" And the answer is clear,
Though with dying voice: "Yes, sir. I'm still here."
And in the surf, rocks, obstacles afloat,
Into their midst he plunges the boat.
To be saved, it's the only way to go.
Salvation: the installer of the next Build!

The fire is out. The ship's run aground.
All are saved. Only one can't be found.

The bells ring out, their notes all fly
From Forums and Magazines to heaven on high.
The city is still but for funeral bells.
For one service only the sad sound swells:
In the procession ten thousand go by,
Or maybe more -- and not one dry eye.

With layers of Changelogs the grave they soften.
Under more Changelogs they bury the coffin.
With golden script in marble stone
The city has its tribute shown:
"Here lies Dave O'Connor! In smoke and fire
He held fast to the wheel; he did not tire.
He saved our lives, our noble king.
He died for us; his praise we sing.
Dave O'Connor!"


At least that's what I remember how it goes.....or doesn't it?
 

Dave 'Arjuna' O'Connor

Panther Games Designer
Joined
Jul 31, 2014
Messages
3,415
Points
113
Location
Canberra, Australia
Website
www.panthergames.com
SITREP 1530 Thu 14 May 2015

It's been a gruelling week overhauling the APer fire code. It was left in a terrible state, which is why we had things like small 2 man units holding up entire regiments. I must apologise for the state it was in as I should have checked that earlier rather than just relying on Miquel. In the end I have done a complete root and branch overhaul. There is little left untouched. In the process I have done a lot of testing to ensure that the factors have the desired effect. These include weapon accuracy, enfilade fire, visibility, terrain, unit effectiveness, relative height, rout status, grenade bonus, target and firer deployment status, acquisition, target speed and rate of fire. Along the way I have also revised the retreat code, including the code that determines if it's too hot to stay put. To this I have added code that makes it more likely for a unit under bombardment to stay put if its dug in or better or if its in urban terrain.

I am just now running through some games and checking on the variable results using Miquel's very useful fire event data logs, which in the debug version of the game log every fire event. These have been invaluable in identifying the factors to tweak. One I have done that I need to make some mods to our autotest feature and then run some tests overnight. If these indicate acceptable casualty results then I will do a build tomorrow and upload it to LNL for release next week.
 

Kurt

Member
Joined
Jan 4, 2015
Messages
896
Points
28
Age
58
Location
England
Impressive work Dave , once the core game is sorted you should take a holiday mate.
 
Joined
Jan 23, 2015
Messages
38
Points
6
Age
70
Location
South West London
SITREP 1530 Thu 14 May 2015

It's been a gruelling week overhauling the APer fire code. It was left in a terrible state, which is why we had things like small 2 man units holding up entire regiments. I must apologise for the state it was in as I should have checked that earlier rather than just relying on Miquel. In the end I have done a complete root and branch overhaul. There is little left untouched. In the process I have done a lot of testing to ensure that the factors have the desired effect. These include weapon accuracy, enfilade fire, visibility, terrain, unit effectiveness, relative height, rout status, grenade bonus, target and firer deployment status, acquisition, target speed and rate of fire. Along the way I have also revised the retreat code, including the code that determines if it's too hot to stay put. To this I have added code that makes it more likely for a unit under bombardment to stay put if its dug in or better or if its in urban terrain.

I am just now running through some games and checking on the variable results using Miquel's very useful fire event data logs, which in the debug version of the game log every fire event. These have been invaluable in identifying the factors to tweak. One I have done that I need to make some mods to our autotest feature and then run some tests overnight. If these indicate acceptable casualty results then I will do a build tomorrow and upload it to LNL for release next week.

Nice work Dave.

If you want any specifics tested point me in the direction and I'll do my best to test it (break it ;))
 

Dave 'Arjuna' O'Connor

Panther Games Designer
Joined
Jul 31, 2014
Messages
3,415
Points
113
Location
Canberra, Australia
Website
www.panthergames.com
Yeh well my testing here is limited. I have endeavoured to get the APer casualties down to historical levels without neutering the effectiveness of the units. It's pretty close now using the tutorial. What I will need is a lot more testing and feedback from other scenarios. That is where you all come in.
 

EzraNehemiah

Member
Joined
Nov 14, 2014
Messages
31
Points
8
Location
British Columbia, Canada
Wow! The inner workings, revisions and development of this title is incredible. Something as a kid (Early Board Game Days) I only dreamed of.

Achieving absolution while pursuing perfection.

Bravo Dave!
 

john connor

Member
Joined
Oct 22, 2014
Messages
2,488
Points
63
Age
60
Location
Brussels
How interesting, Dave. I hardly think apologies are needed. My view is that CO1 - with all Miguel's stuff in - worked very well (it's odd that these issues are so prominent now when they weren't with CO1, but what do I know about coding?), and whilst, with CO1,you might now and then be bothered by a unit that held out too long etc, or other 'minor' things, the issues that really worried me then and now are those to do with how the AI handles attacks and defence, which I've already posted about. I can't remember whether you've actually actioned any of those types of issue yet? Or whether the 'gruelling' changes just made will have any impact on that?

The single thing that would put me off right now (if this were not a beta product) would be that there seem to be bugs which halt the enemy AI and prevent it from consistently putting in strong, co-ordinated attacks using the forces it has available, so that I frequently would get to the end of a scenario (especially a longer one) and find that I had either drawn or won (whatever, it doesn't matter really) but at end I discover that the AI had masses of forces available to attack in key places in a coordinated, concentrated fashion and just didn't do it, so it looks like the enemy AI has just been paralysed. Daz surmised that this might be a result of bugs which we see halting our own attacks and forces in various ways (there seem to be a number of ways in which this could happen, from 'halting, halting, halting' type situations, through to slipped attacks that are then just cancelled (and I guess, for the enemy AI not replaced), through to an over eagerness to stop on contact etc) and I just can't recall, sorry, whether you had got to these issues yet, or whether some of them were part of the changes to Miguel's code? I sent you saves of some of the situations which you said you would look at. Just need to know, if possible, so I know what to watch out for when 'testing' scenarios.

Peter
 

Daz

Member
Joined
Oct 22, 2014
Messages
861
Points
43
Location
England
One of the reasons Dave has had to revise the code was because of the effect that 2-3 man HQ could have on an attack.
These units could unrealistically hold up an attack for hours, and this is most definitely one of the reasons that the Enemy AI was having trouble completing and exploiting an attack, just as have we.

Whilst trying to take down the last few men in these problematic units, the others in the attack formation were getting continuously fatigued at an accelerated rate, then when the unit is eventually overcome the men involved are to exhausted to continue on, completely taking away the momentum needed for a successful breakthrough.

When he went deep into the code to find out what the cause was, he discovered that there were other problems with the defence code that was effecting not just the HQ's, but all the units.
Hopefully this will be a step in the right direction to address one of the issues with the Enemy AI attacks.

One of the things you have to remember though Pete is the Enemy AI may have lots of units left, but they are not necessarily strong enough to attack. Also a lot of the units left may not be line units.
If you have weakend the Enemy line units sufficiently, which can even be done in just one of their failed attacks, they may never have enough strength to mount another.
So you can effectively win the scenario in the first few hours, in some of the scenarios, even though they may go on for several more days.

What we don't want to see, is the AI continuously attacking, with insufficient strength, a well defended location and just wiping itself out :banghead: ( I think that is quite a good emoticon to sum up that line, in the absence of a dead horse to flog).

In some of the situations you mention above Pete, where at the end of the scenario the enemy has lots of troops left that are not attacking. Is it fair to assume that they may not have the strength in their Line Units to mount one successfully, so they have switched to the defensive to consolidate what they have left?
In this situation the initiative would be with you to switch to the offensive to try and wipe them out, would it not?
 
Last edited:

john connor

Member
Joined
Oct 22, 2014
Messages
2,488
Points
63
Age
60
Location
Brussels
It could be, Daz, that in some of the situations I have seen that that is a component of what is happening, yes. For instance, I'm prepared to believe that when I play Spearhead v Reich and by half way through the Axis AI is just completely incapable of using any of the units to mount concentrated attacks then - as they were all knackered to start with and have done something since then - it's possible they're recovering. Same might just be true when nothing happens in From the Meuse to the Rhine.

But what you suggest won't explain why nothing happens in test scenarios set up just to achieve this - and I posted the results of one such test scenario at the back end of my Bed Fomm AAR in the AAR section. That was an adapted Manhay set up with 100% rested, 100% elite units with just one objective to attack. In two run-throughs they DID put in an attack on one run, but on the second run they did NOTHING, absolutely nothing. They did not even move throughout the whole scenario. Now that was clearly a bug, and if it happens like that there then it can be happening all the time to cripple AI aggressiveness and effectiveness in any scenario at all. I already sent Dave a save of it.

Nor will your idea explain my experience trying to replicate the battle of Prokhorovka using CO1. In this battle, historically, the Soviets had massive forces at their disposal, a really massive superiority in men and materiel, including good armour, over the Axis forces. But their C&C and quality generally was poor. Consequently, 'crack' SS units were able to decimate their attacks and hold it to a draw, roughly. In some areas, the soviets lost 18 times more tanks than the Axis.

When I try to replicate this in CO1 (or CO2) it should be the case that I can set up (on the accurate map I drew) the exact situation I described above, with accurate historical objectives and the soviet AI will USE nearly all that armour (as they did historically) to mount large, multiple, concentrated attacks on various key objectives. But that was impossible to achieve. To get the AI to mount anything like a concentrated attack I had to set the soviet's (all of them) to 100% stats for units and leaders and 0% fatigue. Then I got something similar to the historical result in that there were around 3 apparently brigade level attacks put in. That still left around two thirds of the soviet forces unused in situations where I would give them no defend objectives (until around midday) so all the AI COULD do, if it was to do anything, was put in attacks - there was nothing for it to defend with those two-thirds left over. I don't mind that happening but it should be happening with the stats set to realistic levels, not maxed out. Set to realistic levels the AI became paralysed and attacked nothing. several small probes, no attacks. I spent two weeks working on this scenario, messing with every possible permutation. So I am quite sure that THAT is what is more likely happening (besides the bugs we've mentioned) when the AI does nothing. It's an issue with the way the stats work, I think. They should have a much bigger impact on getting the AI to behave in certain ways otherwise it's (for example) virtually impossible, I believe,to create scenarios were one set up works equally for both sides. I'm not talking about changing the AI to get it to flog itself, I'm talking about trying to get it to do something. In my little test scenario, mentioned above, I got the same result if I set the Axis AI forces to realistic stats - paralysis.

As with so much to do with all this 'unofficial testing' that we're all doing, you have one arm tied behind your back the whole time because only Dave has the ability to see properly what the enemy AI is doing, so, naturally, I could be wrong.

But try playing through Manhay as Allies. That is meant to be a historical scenario, with realistic stats. You won't get anywhere near a historical play-through, I think. The Axis AI will be very ineffective against you (even if you don't try very hard). You'll get closer if you play as Axis. It shouldn't be like that. At the beginning of that scenario the AI as Axis should be able to mount the same 2 or 3 concentrated attacks on key objectives that you will as a human choose.
 
Top