Odd building entry bug

[Lt Bull]

Hello,

 

I have some memory of perhaps some similar issue like this relocating to building entry by infantry being talked about in the past.

 

I discovered this some time ago while playing a PBEM of the scenario MG Lonsdale's Block.

 

I have tried to replicate the problem on a "blank" new scenario file but the problem doesn't present itself.

 

At this stage the problem just seems to be specific the actual MG Lonsdale's Block.btt file for some reason.

 

There are two four-story buildings types that happen to both face west.  

 

In the editor they are termed Independent Buildings>Other>number 2.

 

 

This building has doors on opposite sides of the building (front and back) that normally should allow infantry to enter via either door.

 

For some odd reason in this scenario, that is not the case.  The building can only be entered via the rear/eastern facing door.  

 

 

During the PBEM this actually was very costly and surprising for the Brit player as the westerly located Brit forces that advanced towards the building to occupy it suddenly were faced with watching in horror as their infantry instead ignored the covered western entrance and ran around the building to attempt to enter the building from the exposed eastern side, only to get cut down of course by the enemy in the east.  The first time it happened I blamed myself for perhaps giving the wrong order or something.  My repeated attempt was met with the same disastrous result.  In my PBEM those buildings were otherwise quite key defensive positions to anchor the northern flank.

 

I have used the scenario editor to both delete and replace the same buildings and place others elsewhere with the same facing.  In all cases I could not get infantry to enter the building via the front/western facing doorway.

 

The other odd thing however is that if the building is orientated facing any other direction, infantry are able to then normally enter from either doorway.

 

I have attached the original scenario file in which I have added the same building on some open ground in different facing orientation for the purposes of testing with infantry conveniently positioned nearby.  Just need to select Brits, give orders and run for one turn.

 

 

MG building bug.btt

 

Here is a scenario file set up from scratch with the same building facing in all four possible directions that does not experience the same problem.

Building bug test.btt

 

[Paulus]

Hi Lt Bull,

 

If i remember right, it was an elevation (of the ground) issue. 

I.e. one of doors outside the min-max range, in relation to the land.

Good luck!

 

Regards P

[Lt Bull]

Hi Paulus, thanks for that.

 

Inspection of the elevations around the building tile in the example I provide reveals the elevations to be uniform so that issue doesn't seem to be the one causing the problem.

 

 

Not sure how popular this scenario is but it apparently is part of the CMMG module.  I can not say if this problem existed at the time of release at whatever build version that was then but it seems a bit odd that no one else would have reported the problem long ago.  Has this problem only appeared along with the latest patches/upgrades?

[RockinHarry]

I tested the file and can confirm the problem. Looks like the westerly front wall is swapped with the right hand side wall. The ptroopers actually do not enter from the rear side, but rather through an invisible door at the mentioned side wall.

[Vanir Ausf B]

Good catch. Reported.

[Lt Bull]

The ptroopers actually do not enter from the rear side, but rather through an invisible door at the mentioned side wall.

 

I see now, yes that is actually where they are entering and not from the eastern side as I thought.  This just makes it even more random/bizarre/unexpected.

 

 

Good catch. Reported.

 

Have you any idea why something as apparently random and uber specific like this could possibly be happening?  How can this problem apparently be affecting a particular scenario?  Keep in mind that it could be occurring in other sceanrios yet no one has yet picked up on it.  I might just do a random check amongst other MG scenarios to see if the problem with that building occurs in those sceanrios.

 

Surely if it has always been like this from the start someone else playing MG Lonsdale's Block would have picked up on it and reported it.  Those buildings are front-line buildings and probably commonly occupied by either force at some stage of the scenario.

Edited September 2, 2015 by Lt Bull

[womble]

IIRC, this sort of specific map-based mispositioning of entry points and/or windows has arisen from maps being created in earlier versions of the engine than the final release. I have no good explanation of why this sort of thing is affected by different engine builds, but I'm sure that was the explanation given for some of this kind of issue.

[Lt Bull]

I just checked by opening up another tock MG scenario (Boy's Against Men) and the same thing happens.

I placed four of the same building on the map each facing one of the four possible directions.

 

However I do need to clarify/expand on what RockinHarry correctly pointed out was actually happening.

 

On both maps, if that particular building is facing either WEST or EAST (not just west as I originally thought was the case), then the building entry point seems to appear in the the following location:

 

If the building is facing west, the door on the western facing side (the front) does not function as an entry point. A virtual entry point appears to now virtually exist on the southern face of the building, close to the eastern corner.

If the building is facing east, the door on the eastern facing side (the front) does not function as an entry point.  A virtual entry point appears to now virtually exist on the northern face of the building, close to the eastern corner. 

 

When the building is facing east, it is harder to notice that the troops are not actually using the door on the eastern face.  The door they actually use is literally just a short trip around the corner on the northern side of the building.  Whereas when the building is facing west, the troops have a longer march to get to the virtual door at the eastern end of the south facing wall.

 

IIRC, this sort of specific map-based mispositioning of entry points and/or windows has arisen from maps being created in earlier versions of the engine than the final release. I have no good explanation of why this sort of thing is affected by different engine builds, but I'm sure that was the explanation given for some of this kind of issue.

 

Just my speculation, but it just seems like its code specific to that building in conflict with the game updates.  Probably something very minor yet  different to how the other buildings were coded.

[Vanir Ausf B]

How can this problem apparently be affecting a particular scenario?

 

It isn't. I reproduced it in a test scenario I made myself.

[Lt Bull]

It isn't. I reproduced it in a test scenario I made myself.

 

Hi Vanir, you are correct but I did acknowledge this in my last post, when  I found editing another MG scenario, MG Boy's Against Men, and found the same issue could be reproduced by adding that same building type (note: the scenario never originally featured that building).

 

Here is that file:

MG Building Check2.btt

 

Yes the problem is not just affecting A particular scenario, but it is affecting SEVERAL particular scenarios in which the same building is featured on the map facing east or west or could be featured on the map facing east or west.

 

The problem seems to be restricted to SOME scenarios, but not ALL scenarios.

 

You say you reproduced it in a test scenario you made yourself.  Did you make it using a fresh new default blank template from the editor?

 

It may not have been clear but this was one of the first things I did check originally and found the problem DID NOT present itself.  Here is the actual test scenario I made from scratch using the blank default template.

 

Building entry test.btt

Edited September 2, 2015 by Lt Bull

[Vanir Ausf B]

Did you make it using a fresh new default blank template from the editor?

 

Yes.

https://www.dropbox.com/s/b6b9psp776wakup/building%20test%20two.btt?dl=0

 

I did try your test and indeed the problem isn't manifest there for some reason.

[Lt Bull]

Wow, yeah, I saw the problem using your file.  What is going on?

 

After running your test, I realised that my definition/description of the problem could do with a refinement.  In your tests you actually positioned the infantry so that they were placed outside the rear of the building.  All my tests were focussing on infantry placed in front of the building however.  Apparently this is totally inconsequential to the problem.

 

In some scenario files for reasons yet to be identified, if the building is placed facing either WEST or EAST, then both the front AND rear doorway entry/exit points do not function at all.  Instead, a single "virtual" doorway appears to function as an entry/exit point to the building on one of the visually fully bricked up walls of the building.

 

If the building is facing WEST, this virtual entry point exists on the southern face of the building towards the eastern edge (as if the front face of the building was actually facing SOUTH)

If the building is facing EAST, this virtual entry point exists on the northern face of the building towards the eastern edge (as if the rear face of the building was facing NORTH)

 

note:  when looking at the front face of the building, the graphic and normal location of the functioning door is offset to the right; and when looking at the rear building face the door is offset to the left.

 

It does seem that the code that controls the graphics/3D model of the building is separate to the code that actually tells the game where the actual entry points (doors) are to enter/leave the building. ie. the "what you see" aspect of CM code as opposed to the "what you get" aspect of CM code are not naturally linked or related.  The two must be manually forced to coincide.  I don't think this is anything new to what we already know about CM.  It does however explain why things like this can exist.

 

The fact that Vanir and myself apparently both created what appears to be the same tests using a fresh blank scenario template using the same buildings only to find Vanir's tests showing the problem and mine not showing the problem however is the most puzzling.

[Warts 'n' all]

I ran my own test using the Germans rather than the British, just to see what would happen. It seemed to me that the position of the door was vital. If the door is on one end of the facing wall then my troops went around to the "non-existent"  "side" door. Whereas, if the door is either slap in the centre, or where the 2nd or 3rd window would be, i.e. slightly off centre, then my troops went in via it. I took my test as far as having troops run into a building, run out the other side and then back in again. and repeated it over three or four different buildings of the same type on another part of the map, always with the same results.

 

Just for the record, I'm still running 2.12. I had no problems like this during the "Road to Nijmegen" campaign. Nor in any of the single battles that are set before "Lonsdale's Block".

[Lt Bull]

Hi Warts,

 

I can't seem to understand if your are trying to report some new observation that is different to what has already been observed or if you are reporting the same behaviour using a different way to describe it.  To keep it easy, perhaps try to reference the facing of the building (determined by observing the direction indicated within the scenario editor, which happens to be in this case the white face of this particular building that has a door offset to the right).  It can either be facing one of four directions (N,S, E, W) and the problem has been found to occur only when they face east or west on some maps/in some scenarios.

[c3k]

Lt Bull,

 

Excellent (and useful) write up! This is how bugs get squashed.

 

Thanks for your detailed work.

 

Ken

[Warts 'n' all]

Hi Warts,

 

I can't seem to understand if your are trying to report some new observation that is different to what has already been observed or if you are reporting the same behaviour using a different way to describe it.  To keep it easy, perhaps try to reference the facing of the building (determined by observing the direction indicated within the scenario editor, which happens to be in this case the white face of this particular building that has a door offset to the right).  It can either be facing one of four directions (N,S, E, W) and the problem has been found to occur only when they face east or west on some maps/in some scenarios.

Sorry for my lack of clarity. It's my own fault for carrying out my little test just before heading off for a long weekend. Primarily I wanted to test whether the problem existed before the v3 upgrade. I can confirm that it did. Also, I only found it happening if the building was on an East-West axis.

 

Out of curiosity I looked at buildings 1, 3, 4 and 5 as well and had no problem with any of those.

[Lt Bull]

On 04/09/2015 at 2:54 PM, c3k said:

Excellent (and useful) write up! This is how bugs get squashed.

Apparently not.  We are at v3.11 and I have just checked that this bug still exists for this building. :/

[Bulletpoint]

-

Edited June 18, 2016 by Bulletpoint

[Lt Bull]

On 02/09/2015 at 8:38 PM, Vanir Ausf B said:

Good catch. Reported.

Hi Vanir Aust B,

Did you actually report this?  Maybe you know why this bug is still with us.

Surely it would take no more than an hour or so (if that) of coding to fix this.

Edited June 19, 2016 by Lt Bull

[Vanir Ausf B]

Yes, I did actually report it. No, I do not know why it was not fixed nor if there is any intention of fixing it in a future patch. Only a uniformed BFC employee could answer those questions.

[Lt Bull]

Its been almost five years since I started this thread and would have thought that the issue(s) originally highlighted would have been fixed long ago via a patch/update etc.  This sadly is not the case.  The issue described is not a "cosmetic" issue with no effect on gameplay. Contrary, it has the potential to turn what players would think is a relatively safe move order for infantry in to a order that may result in the entire enemy unit being decimated, as I had experienced when I first noticed the issue all those years ago.

I am revisiting this thread and the issue that was discussed because I was just curious to see what, if anything, was achieved in first highlighting the issue almost five years ago. I also like to think even trying to address such issues on these forums is not just a complete waste of time and effort. If anyone can point to a thread where Battlefront had at least previously acknowledged this issue, that will be good.

 (I should also add that I did actually stop playing CM around that time out of a frustration that gameplay issues like this weren't being addressed, let alone acknowldged by those in a position to do something about it. After coming back to CM after a many year hiatus, I really was surprised that this issue was never fixed. I have stopped playing CMBN because of a new odd suicidal TacAI behaviour issue (apparently introduced after a recent update/patch) that can result in infantry defending and under fire behind a line of hedgerows deciding to break cover and run laterally along the hedgerow until they reach the infantry-sized gap in the hedgerow and start running through the gap towards the enemy/incoming fire, invariably to their death (read all about it and see for yourself here))

Perhaps this post (unlike Battlefront), will warn both players and probably more importantly/practically, scenario/map designers, that certain buildings from the Scenario Editor if used in scenarios in certain orientations will definitely result in the kind of unexpected infantry building entry/exit behaviour discussed above (and more comprehensively below for your convenience) that really can turn players off.

I cannot confirm (have mnot searched) if there are equivalent issues with certain other buildings or in  other CM titles, but I will qualify that they definitely do exist in the building types I discuss below in CMBN.

I have just reviewed and tested ALL seven types of "Independent>Other" buildings available in the CMBN Scenario Editor.  I have created a scenario file and two game save files to download to see for yourself that features all seven buildings in all four possible facing orientations (north, south, east, west) with all the infantry already setup and given move orders to enter the building from one end and exit it on the opposite side.  The buildings are laid out as follows in the scenario/save file(s):

Typically all of these buildings are assigned a "direction" by the Scenario Editor, and visually/cosmetically all appear to have two doors: one on the "front" and one on the "rear" assigned faces of the building (through which it is expected infantry can/should and be only able to enter/exit from).  The side walls of all seven buildings clearly have fully bricked side walls devoid of visible.

eg. rear view of Independedent>Other building "C".  Note location of door on right of rear face, alongside the left edge of the building.

It is expected that if an infantry unit is located just outside the front or the rear face of the building and given a move order waypoint located inside the building, then the infantry unit will take the shortest route to the waypoint and move towards and through what is nearest respective door, located on that front or rear building face. Similarly, if a unit is already within the building and given a waypoint directly out the front or rear of the building, the infantry unit will exit the building using the respective doors in that direction.

If you run the save game files provided, they are already setup with movement waypoint orders assigned to infantry squads located at the front and rear of the buildings: the first waypoint is in the building, the second is on the opposite side of the building. File 001 has units positioned north and west of the buildings.  File 002 has units positioned south and east of the buildings. Links to test files and the scenario test file itself:

Scenario file: CMBN Indpt Bldg Inf Enter-Exit test.btt

Save file: Indpt Bldg Inf Enter-Exit test 001.bts

Save file: Indpt Bldg Inf Enter-Exit test 002.bts

A few key points:

• All the buildings tested definitely have one or two entry/exit points, though not necessarily where they are otherwise graphically indicated on the front and rear of the building.  It depends on the building and it's facing.
• When some buildings are placed on the map in certain orinetations/facings (at the map designing phase stage via the Scenario Editor) it will determine if one or both of the graphically represented doors on the front and/or rear of the building will cease to operate as entry/exit points during the game.  In these instances, an apparent "invisible" side door (or entry/exit point) instead will apparently appear to function along one of the non-front/rear faces of the building, located close to one edge/corner of the building face.  The location and existence of these "invisible" side doors is predetermined by the building orientation/facing.
• The test files feature 10 man squads.  Using smaller squads may show more consistency in whether ALL pixeltruppen enter/exit a building via one entry/exit point, or whether the pixeltruppen will enter/exit the building using both entry/exit points during the same move order.
• Random localised positioning of each pixeltruppen seems to be a factor in some cases determining whether all, most or some of the pixeltruppen belonging to a squad entering/exiting a building during a move will use one or two of the existing building entry/exit points.
• Unless a player uses the Scenario Editor (or the save files provided in this thread) to learn to recognise/identify the 7 types of Independent>Other discussed in this post, they will invariably be unable to recognise them in any CMBN scenario they choose to play that features them.
• The comprehensive table of results of testing is available as a PDF and Excel file at links below:

Excel: Bull's CMBN Independent Houses.xlsx

PDF: Bull's CMBN Independent Houses.pdf

Preview of table:

The table text and cells are colour coded for each situation to aid in interpretation as follows:

RED text indicates (and warns players) that it has been demonstrated that it is possible (though not  guaranteed) that at least some pixeltruppen MAY avoid the nearest door and instead, if entering a buildings, route around the sides of the building to instead enter the building via the indicated door on the OPPOSITE side of the building from where the unit started it's movement from, or i exiting a building, use the door on the opposite side of the building to where the waypoint was placed.  Planning/expecting to enter/exit a building via a door on the near side but finding pixeltruppen entering via a door on the direct opposite side of the building is probably more likely to be of a tactical concern/disaster than say if it entered/exited the building via one of the "invisible" side doors, that's why I have highlighted the text in red alerting players to that possibility for that situation.

Backgrounds of shades of GREEN indicate that all the doors indicated on the front and rear of the building do actually work as advertised and no "invisible" doors exist.  It is a darker GREEN if in the limited trials conducted, no instances of the "wrong" door being used by any pixeltruppen in that situation was observed.  This would be updated  if more testing at least reveals one case of a "wrong" door being used. Note that for larger sized infantry squads, it is no guarantee that all pixeltruppen will use the right (nearest) door for each situation listed (see notes on RED text).  Note that this possibility is probably reduced (possibly to zero), the less pixeltruppen in the infantry team. My guess is when six or less pixeltruppen exist in a team.  Further testing can confirm..

Although some cases of entering/exiting the buildings are listed with darker GREEN backgrounds and as "All enter OK" or "All exit OK" (meaning it was observed in the limited trials that all pixeltruppen enter or exit through the nearest graphically represented door as expected, the ideal case), as alluded to above, it has been noted that repeated testing can turn up cases where at least some of the pixeltruppen involved in the move order will use the second entry/exit point of the building, be it the one on the opposite side of the building, or one of the apparently "invisible" side doors that apparently exist for some buildings when facing a certain way.

Backgrounds of shades of ORANGE indicate the existence of at least one "invisible" side door through which infantry can/and will apparently use to enter the building depending on the circumstances.  The darker ORANGE background indicates that either NONE of the graphically indicated doors on the building are functional in that situation, and instead the building features one or two "invisible" side doors, one on each flank (or side) of the building, or only one "invisible" serves as the only entry/exit point to the building.

In summary, a review of test results:

Regardless of which of the seven types of Independent>Other building feature in a sceanrio (regardless of their orientation), players can expect to be "surprised" by the path and subsequent entry/exit point chosen by each pixeltruppen to enter/exit the building during a single move order if the infantry team has greater than typically six pixeltruppen, if they expect a) infantry to ALWAYS use the nearest entry/exit point and b) expect the only functional and possible entry/exit points of buildings to be where they are graphically indicated.

The only  Independent>Other building that feature front/rear doors/entry/exit points functioning as advertised regardless of orientation is building "G". Of the remaining six buildings, all will feature front/rear doors/entry/exit points functioning as advertised if in the following orientations/facings:

 

I can only suggest regular players of CMBN scenarios to be at least aware of these buggy Independent>Other buildings, especially on maps that are likely to involved and rely on very precise "house-to house" fighting and manoeuvring.  They can really unexpectedly wreck you plans especially if they are embedded somewhere near critical terrain/victory locations.

To the scenario designers/map makers, I would hope they see the sense in completely avoiding the use of all building type/orientation combos that are not listed in the green column of the table above. Note that even using those buildings/orientations featured in the green column, it has been shown that at least some pixeltruppen will nevertheless choose to enter/exit from the opposite side door from which one might otherwise expect them to use when the unit size is greater than 6.  Perhaps some scenario/map designers might even feel inspired to revisit previously released sceanrios/maps that feature the problem buildings and modify the maps accordingly.  Of course vigilant capable players could do this themsleves.  A scenario comes to mind already...the one that I was playing when I first encountered this issue almost 5 years ago...Lonsdales Block. I clearly remember where that damn building was that led to the decimation of a complete para squad trying to enter it.

 

[pkanarki]

But why don't you find the joy in that offcourse BF has fixed the bar rifle graphic where the magazine was slightly positioned maybe 0.002 pixel too far to the left, which almost ruined the whole gaming experience. 

 

To be honest. CMBN  is a serious joke and wreck by now.

[Warts 'n' all]

Sadly, the problem with the building type in the picture was never fixed. And they even crop up in Final Blitzkrieg.

[c3k]

@Lt Bull, great post.

It will be used, internally.

[John1966]

Due to my frustration with doors I searched "doors" and came up with this.

My pixeltruppen ignore perfectly good doors in front of them all the time only to run round the building into a hail of fire. Particularly frustrating moment last night when I patiently waited for the engineers to arrive to blow a hedge so my guys could get into a building through the "safe" side. Guess what they did?

But reading some of this thread, is this largely a problem in CMBN rather than the later releases? I note someone saying it happened occasionally in CMFB but "occasionally" is preferable to the regular frustration I currently seem to be experiencing.