Quartermaster

Can you tell me what game uses this mechanic? If I have the time this weekend I will try to play it so that I can get a feel for what it would be like to play with this mechanic. I'm not going to debate with you anymore until I have an opportunity to try it out as that would remedy my suspicions if it works well.

To make this clear let's pretend that we are cooking and need to measure. Which would you prefer to use if you have to cook quickly, two measuring cups that measure 1/2 cup and a full cup (adding more than 1 burst option) OR having 1 cup and 3 smaller cups, but the smaller cups can only be used to remove things from the full cup (your proposed method). The second option is innately more complex and difficult to use. Yes there are more potential combinations that can be used to measure the perfect amount, but how how often do you need the 'perfect amount' and is it worth the amount of complication? I lean towards 'No' which is why I am arguing against it. I can understand that some people might like it the way you are proposing, but I don't. If you can describe a way to do full auto using your method that has controls that are as intuitive as the current burst fire method or one that would also allow multiple burst modes, than I will be all for it.

Yes but it seems like this will be difficult for new players, more prone to error than the current method, and less controllable than simply adding a second burst option for some weapons (i.e., the machine gun, carbines). Fiddling around with the reserve AP slider (I am assuming you are talking about the current version) to simply set aside rounds will be prone to error. It will be easy to forget to set it and secondly there are limited options of which stop points you have for reserve AP and they are set by weapon. With the limited options it will be a more complicated system that is not as easily controllable as simply adding multiple burst aiming for certain weapons. Thus I do not se the advantage of this unless the AP reserve slider is also redone.

As the title states, I threw a flashbang near a Caesan next to a Gas Tank and it went up in flames. Not sure if it is intentional or not so I am reporting it.

My problem with the method you are discussing, Gauddlike, is that it either lacks as much control as burst fire currently does or it becomes really difficult to control depending on what route is taken. I really like the idea of how you are calculating the AP for that many rounds fired, but the loss of mobility after firing worries me. For example if you don't move to start with and fire, you have lost the entire round of combat and cannot move. Personally, I would prefer the ability to fire a burst and then duck behind cover. On the other hand if you do add a method that allows one to control how many shots you fire it could get complicated depending on the interface. To simplify that, it just seems easier to have two burst fire modes for some weapons and we end up back with the current method.

I have been looking at the files that one can extract and thus presumably mod via the 'Game Mods' method which is currently implemented. Is there a way to add actual weapons with spectres to the game through this as it stands? Please let me know what your experiences with this method of modification have been so far. For those that need background information here is the wiki post describing how to use the Game Mods method: Method

The code is now updated to V14. Let me know if you have any troubles. I still have not gotten around to adding the function which will allow you to select the directory. I am working on my platform paper to get fully accepted into the doctorate program at my school so I have not been putting very much time into this. When I get the opportunity or take a break from real work I will try to add it in. As a note I added some Try... Catch statements so you might get some Message Boxes which will pop up and let you know where the error has occurred.

Depends on when the upload is submitted. Desura's HQ is in Australia so it is around 6 AM right now, so probably in a few hours.

In reply to Gazz: Yes I am not modelling "walking back" the rounds after it starts jerking, because I am not trying to model reality exactly. Even if you want to make the reality argument, then the number of rounds that a Xenonaut could fire in a burst (even with the new waste your entire turn spraying bullets method) would not give you enough time to walk it back. With 60 TU, one could at most fire about 10 rounds in a turn (10 TU for the first + 5 TU/shot). In real life you might at the sixth round have an opportunity to start walking it back, but I didn't think that the additional effort of adding an additional parameter to each weapon (to make a third order polynomial which your squiggley represents) would be worth the time and effort. There is probably a way to add a constant that would be consistent across all weapons, but I would have to think about it a bit. In reply to Agent Dark: I agree in that I really do not want full auto for weapons other than the machine gun and maybe the carbine. In which case the addition of a mode of fire or two for burst is not unreasonable, because you don't need any of the single fire options for the machine gun and at most one or two for the carbines. So you end up with the same number of fire options as there are for any other weapon.

Burst fire would be more controllable if you simply set it up so that they are different burst modes that just fire different numbers of rounds. Simply if you have multiple modes you can set it up so that the dynamics are the same (first shot = single round + 5 for each additional). If you allow two forms of burst (short and long) it should work about the same as letting them fire a specific number of rounds and be easy to control.

My apologies that this will be a bit technical, but I wanted to simply make a recommendation regarding how to do this. Rather than having a fixed decrease in the accuracy with each shot I would recommend using a multiplier that uses values less than 1. This would require at least 1 additional parameter for each weapon (I think the existing parameter of recoil would be fine). Essentially the first shot in a burst would have the same value as whatever the normal accuracy would be for a single shot, the second shot's accuracy would be the first shot's accuracy times the multiplier, the third shot's accuracy would be the second shot's accuracy times the multiplier, etc. Algebraically this could be set up so that each shot will have the following accuracy: Base Accuracy * Recoil ^ (Shot# - 1), where Base Accuracy is the set accuracy of the burst, Recoil is the value of the recoil parameter, and Shot# is which shot in the burst it is. There are two good reasons for using this method. The first is that the drops in accuracy are not linear. What you will see if you plot the accuracy points, is that the first to second shot creates the largest drop in accuracy (which is true with burst fire). The rate in the decrease in accuracy becomes slower over time which could be thought of as the soldier compensating somewhat for the burst. Secondly, this method asymptotes at 0 which means that it will never become 0 no matter how many shots are fired which prevents a number of errors from occurring. There are 2 things I would think about adding to this simple method of burst fire accuracy. The first would be a additive modifier to the recoil parameter (that can max at 1) when the firing unit has a piece of cover that they could prop the weapon on when firing at the target. This could be a uniform addition across weapons or you could set it up as a parameter for individual weapons. You may want to do the latter so that propping up carbines is not as beneficial as it is for a rifle. The second thing I would think about adding would be a minimum burst accuracy value for each weapon so that the asymptote is higher. The easiest way of implementing this would be to have it check the accuracy of each shot and if it is below the minimum value, it is set equal to it. An alternative would be to change the earlier equation to: ((Base Accuracy - Min. Accuracy) * Recoil ^ (Shot# - 1)) + Min. Accuracy The downside to this method is that Min. Accuracy and Recoil do not function independently so the rate of accuracy decreasing per shot would be a little more difficult to look at.

Thank you for doing the write ups on how the AI is going to work! It looks really good and I like how you are explaining.

The reason for not just assaulting the alien craft and ignoring everything else is multifold. First you will lose the valuable resources that the aliens are carrying (about 30k in weapons & alenium in the form of clips). Secondly, they will spend the time that you are capturing the ship cutting down the civilians weakening your score with that country. Additionally, depending on how good the AI ends up being, you may have to worry about a coordinated counter-attack, delaying tactics by the forces defending the UFO, or ambushes. Just because you are bee lining to the ship does not suddenly make scout armor useless. The main benefit of scout armor in an assault is that it lets you see and fire at aliens before they can do the same to you. Thus the visibility of scout armor is in a sense a different kind of defense (in that you avoid fire). The reason to display it on the overhead but not remove fog of war in the regular display is that the way the sub maps functions is such that you design multiple buildings with the same exterior, but different interiors. Since one is randomly chosen from the list there will be times when you are able to cut through a certain building for cover and others where it just prohibits your movement. Having it only on the overhead map maintains a lot of the aspects of not knowing what is there while simultaneously giving you a rough enough sketch to formulate a battle plan with.

I do not agree with the notion that knowing where the alien craft is would significantly reduce the suspense. What builds up suspense in my opinion is not knowing where the aliens are and being unable to anticipate their location. To a certain extent knowing where the craft is does reduce the number of places that a certain proportion of the aliens might be, but not to a great enough extent to eliminate suspense. To clarify, I am only talking about the large overhead map (which is not developed yet) rather than the current view which should remain as is. Regardless of the reality of whether or not one "saw the map on the way in," I feel that knowing the general location of buildings on the map would add to the strategic nature of the game. It would add to the game because it encourages squad movement rather than the sweep method of looking at the whole map. If the sub-objectives are added to the game, like protecting the civilians in the hospital from the alien invaders, the commander would have the option of splitting his forces to deal with each threat appropriately (the hospital assault and the alien craft). Not knowing where things are only makes one waste time examining the whole map. Personally I feel like there should be a pressing reason to assault the alien craft (something like the Iceland Incident) to create more suspense.

I can't seem to recall if an area map will be available for the game (I'm meaning the top down view of the area), but if so I think it would be appropriate to have that show where all of the buildings and the UFO are located regardless of whether or not your soldiers have been there yet (albeit lacking a bit of detail). Seeing buildings on the area map will also be useful when we have secondary goals, like protecting civilians in a hospital. Since there is the dynamic where we can capture UFO by occupying it, it seems appropriate that we should be able to select tactics that suit our needs rather than being forced to fan out and sweep the map every time. For example if we are facing a more advanced enemy than the soldier's tech level could handle, it would be more appropriate to bee line for the UFO rather than risk being spread out and losing soldiers due to them being isolated and lacking power individually.

This is almost certainly the error. I'm at work for the next 6 hours, but when I get home I still have the files from before the update on my other computer. I will try to copy them to the latest version to see if the program will work once the maps are present.

I was originally going to do that but the program uses the root directory to determine where all of the files are that it uses. This was simply a shortcut method that I implemented to get it working for some folks so I minimized the amount of work when I put it in. I would have had to do more work at the time to have it cut down to the correct file. When I add in the interface for looking for and selecting a folder it will be much more clear cut and vb.net's default interface is setup to ensure that the folder already exists.

Can you try it again, but add a '\' to the end of the destination? The way the code functions is that it looks for the weapons_gc.xml file based on the location typed in. So when you type in what you have it is looking at the following location "G:\Games\Xenonautsassets\weapons_gc.xml". As you can see the lack of a '\' is problematic. For the next version I will try to change the setup so that it uses a folder browser rather than relying on the user to type in the address.

I just began hosting the XenoArmory files on Xenolith (we really like Xeno) as it will give me some advantages when it comes to updates that I make. I am still going to be hosting the files on here through my Box.net account, but Xenolith will give users a heads up when I make updates and improvements to the program. With that said I have added some try...catch statements which will now provide information to the user when the program is having trouble locating files or performing various functions. As these errors pop up, please let me know so that I can try to figure out how to change the program so that the errors do not mess up the program for you. Please see the first post for the updated files.

I didn't get the chance to finalize the changes last night, but I will try to get them posted later today.

As a heads up, it will not be until the weekend until I can look at this stuff. This weekend I will try to build in a number of try catch statements so that I can start catching the errors and collecting data to find out what isn't working right for you and others. Thank you for letting me know that you are having a problem and I will do my best to correct it.

Looking forward to seeing the game pass these marks!

When at the Goldhawk Interactive webpage if you click on the forum button it goes to http://www.goldhawkinteractive.com/forum which gave me a 404 Not Found message.

Looks good so far. The only suggestion I would make is that it is better to edit the strings.xml, weapons.xml, and items.xml through Excel. As Excel files they have some tag that keeps track of how many items are in it so what happens if you add items to the list is that sometimes they get ignored by Xenonauts even though they look like they are there. My guess is that it the program looks at how many items it says are in the list and then only goes through that many.

I think we are on the same page based on your description. I am using the terms that the submap editor uses for the orientation of the images on the screen and. As to what you are referring to in regards to what images overlap other images on the screen that is a function of the sort rank (value maybe?) parameter for a spectre. Soldier images are 0.5 as are most tiles. Lower values indicate that something should go in front of another object. When the value is identical for two spectres on the same tile it has a tendency for one of them to get cut off which is what results in the odd looking walls and stairwells. There is currently an issue (in my opinion) of how the sort values are set up for walls that needs resolved yet. Essentially someone needs to sit down and figure out what the proper order is for spectres to be put on the screen. For example all walls that are on the sides of tiles that are toward the top of the screen need their sort order set to something higher. There is no need for "two" layers for a single tile, because that issue just needs sorted out with the sort rank parameter of the image spectres in order for the images to function correctly.