I'd be interested in seeing how you manage to get this gameplay working in a squad based fashion. One suggestion for squads is to have a small number of units considered one unit. Or you could have each unit move and act independently but give players the choice to have your other squad members controlled by AI. This allows you to maintain micromanagement of 1 unit, in case that simplifies gameplay for the players.
Not sure how I will experiment with it exactly. Only thought about AI only Squads so far, but actually controlling a whole squad is a pretty cool idea. I'll see if I have enough time to work in that direction, haven't thought about it yet :) thanks.
You can play single-player maps on b.net without lag, just download it through create game (or join game) and just go to Single Player and then Play vs. AI (something like that, forgot the name), and it will appear in the list.
And the vs. AI maps are completely lagfree? That would be very good and very important. :)
Just finished another ability concept, basically only the target lock thing is finished, the whole visuals and time it takes to lock on the target and audio etc. is still WIP (as for most of my stuff). Also it still has a bug which I didn't show in the video, but I'm pretty sure I'll be able to solve it relatively easy. (edit: actually I'm not sure if it is even a bug or if I want the ability to behave like it does now. I'll probably just discuss it with my personal tester buddies ;) )
edit:
In contrast to the video I also already changed it in a way, that the ability will lock faster/slower depending on how well you aim.
This is my current project and first map for Starcraft 2. I started the map just to get a good grip of the editor and to learn the basics, but during that process I continually expanded it and today I think that I can actually make a viable and good map out of the concept if I work hard to make it better. So I will regularly post the state of the map here, to hopefully get a little bit of feedback (I know its hard!) and to keep my motivation up. Also to clear my mind a bit because this thread is my personal reference for a concept. ;) Haven't really written anything down somewhere else.
First and most important before you watch the video: This will be a SINGLEPLAYER map because of lag issues and my interest in mission design, so every enemy is computer driven. The current map is just a testing terrain that I will use until I start with actual mission design.
Now lets talk a bit about the concept. Nothing in it is set in stone, so if you have cool feedback (or maybe just want to desperately see something implemented as a player because you are a big fan of unit X or concept Y) just say it, there is a good chance I will do it in some way or the other.
Concept
Top View Shooter, WASD + Mousecontrol
Singleplayer map, if there is no way to publish it without lag issues (even as a Singleplayer map, I don't know about the limitations of Bnet exactly) in Bnet, I will just upload it as a downloadable map to use for the Editor.
Every unit type is designed with purely data and a trigger AI that will be applied once the unit gets placed on the map. This way I can hopefully split up my work as much as possible, so I don't have to think about specific data and unit problems as much once I am in the mission design process (if I will ever get there ;) ). The AI really needs less computing time than I expected, currently I can easily place 30-40 units on the screen without lag.
The AI can be enabled and disabled if you want the units to do specific actions at a certain time.
I also want to make a lot of use of friendly AI units. Recently I thought a lot about how and if I could make it work to command unit squads on the screen, maybe even balance it in a way that I could divide gameplay a bit - More a Commander type of feel for people who like it, and more an Action/Shooter type of feel for other people. I'm not sure if I can make it work, and if it feels shit I won't do it, but I will definitely experiment with concepts like this.
I will probably work on data and AI stuff until HotS comes out. I have a lot of ideas for different unit types and abilities, so I even doubt it will be enough time. After HotS I will start to work on a UI and terraining.
Every unit should feel more awesome than it does in the basic Starcraft 2. I want to give especially the bigger units a really broad variety of cool abilities that really make them shine. Small units should feel expendable, but still have some little unique stuff (like the zergling jump). The general strength of units should be comparable, so it should be hard for a goliath to kill an Ultralisk etc.
Current State
The video shows a really early version of the game, I'm continually changing stuff for pretty much everything.
The map currently includes:
Working WASD system, goliath turret moves in direction of the mouse.
Camera movement in direction of the mouse to cover more area and to get a more interactive feel. Angle of the camera is changable, not sure about the best angle yet.
Cover system to protect against missiles and to cover behind for both AI and player.
Goliath: Various offensive abilities, one single pitiful defensive ability. Most of the offensive abilities will get major changes, currently I'm pretty much only satisfied by the basic machine gun thing and the basic rockets with the slow effect. There are a lot of abilities to come. Which of them make it into the final game I can not say at all atm. There will probably be a maximum of 6-8 abilities that I will use ingame at the same time. If I end up with too many abilities that all are kind of cool I will make them an option at the start of the map in exchange for other abilities (maybe as a banking system to unlock stuff).
Hydralisk: Finished (but still has to be polished) AI, some abilities still missing. I agree that the AI isn't perfect, but I want to finish my map at some point and I can't make perfect AIs that respond reasonable at every point in time (I'm just not good enough). I think it is decent, leave some comments about it from what you see. The Hydralisk is designed as a mainly space controlling ranged unit that just spits out the attacks really wide and fast. I want to give them another offensive ability which spits a pool of slime to a location that hinders movement and does damage for a period of time, but it isn't finished yet. Also every zerg unit obviously has to get a burrow ability for sneaky ambushes and defensive measures, but I also haven't done that yet.
Some minor stuff, I won't go into much detail there. The Zergling AI (they are really dumb zerglings), the marine weapons etc.
Also I already changed some stuff from what you see in the video. The Goliath is much slower now and I removed that annoying Movement sound.
If you have good ideas or an opinion about my map, think it is a bad idea or any other form of feedback just post it here, I will definitely read and think about it. If you don't have specific opinions, share your opinion about these questions:
Do you like the classic Terran vs. Zerg theme or do you think I should finish Protoss or Terran enemy units first?
Would you play a Singleplayer Top-View-Shooter for the SC2 engine or do you think the concept is bad (and why)?
Do you think the damage indicator for enemy units (they turn white/bright for a brief moment) is a good design idea or should it be removed?
Did you realize how I was too lazy to add water to the whole map? ;)
Also a general thanks to all the good tutorials and the overall helpfulness on this site. Especially to Kueken & DrSuperEvil who are superheroes (no really, they are!), share their knowledge and always help if possible. Also of course people like BorgDragon and countless others!
Why does the current form remind me of a certain awesome superunit from Supreme Commander?
No idea, never played it :D Also I didn't have any concept at all, just turned out like that. I guess its just a very cliche robot ;) I didn't plan to show a video or anything of it in the first place, I just experimented, but I was very satisfied how decent it looked with quite little effort put in so I uploaded the video.
Basicly the long term aim of the project is to make a skeleton with good and versatile animations you can just hang stuff off to make any impressive composite unit. Has alot of potential for RPG bosses.
Yeah I agree, and thats why it would be so cool if someone is still working on it. There is really a lot of potential once a basic set of animations is done (it doesn't even have to be a huge amount of animations with 2 billion joints), because applying the visual appearance isn't really hard once you understand how the moving parts are done and which joints move what parts of the unit.
It took me like 5 or 6 hours to get this kind of model (including changing the animations a bit), and it was the first time I did something with Forward Vector SiteOps (and I had to get used to the Rotation ones too) at all, so if I had to do this thing again it would probably take me 2-3 hours max. Sure it still looks pretty bad (especially the back and the animations) and I probably wouldn't use it in a map, but in my opinion you can very easily improve it and if you actually want to do some kind of huge mech robot maps it could easily be done this way in a reasonable amount of time :)
Don't let this project die! I hope somebody is still working on this, it has SO MUCH potential!
For motivational purposes I barely slept last night to make my personal Toss-A-Tron. Beware!
Also thx to DrSuperEvil for pointing me towards the awesome data threads :)
edit:
not sure if I will continue working on this, because it really has no purpose for me and I have another project that I want to focus on :) But next thing should be to implement the Turn Sensor I guess (besides making more animations and sticks in general, I just added a single one for each arm, thats really not enough for good animations)
Imo this topic is pretty much cycling around itself right now, but I will still post my opinion here.
I am kind of in between the worlds at the moment. I have started with the Galaxy Editor some months ago and I still feel like a noob, but I think I can wrap my head around most of the basic data stuff and if you tell me to do a unit in the data editor with the custom abilities X, Y and Z I could find a way to do it for many/most ability concepts and visuals (if you would take for example abilities like we know them from various MOBA games).
On other parts of the editor (mainly UI, Cinematics) or simply knowing what types of Events/Actions exist in the trigger editor, or knowing what types of doodads exist and how they look, I'm simply bad. So I will ask basic questions here (for example the difference between Functions and Action Definitions, I didn't know that ADs can't return a value. It is completely logical, but everyone knows these kinds of mental blackouts) that might seem stupid, just because it will save me a ton of time. The general purpose of questions is to get information. Whether someone else is willing to spend his time and give me the information or not is a completely different thing and therefore no one should blame people just because they don't want to share information. Sure it would be nice if everyone did, but especially in the 21. century things like copyright and the material value of ideas get more important every day. So there are multiple points of view on this topic and I don't want to say my opinion is the only right one. If you scroll through the battle.net maps you will get a good idea of how many people lock their maps and don't want to share anything. I haven't tried the top 20 most played maps, but I bet most of them will be locked (probably all of them except the Blizzard ones). Do I like it? Probably more than not sharing information, because it is actual work that was done and not just knowledge. But I still think every map that has not like a bazillion custom models and 10000 hours of work in it would be reasonable to share. In the end everyone who uses the editor is happy if there are more mappers and artists who get inspired by seeing and being able to analyze other peoples work.
On the other hand there will always be cheaters and douchebags who exploit it on all levels, last year the Minister of Defense of my country (Germany) had to resign because he copied most of his dissertation. So I don't want to pretend everyone does what is best for the community or is really willing to learn. But I personally prefer to think like that about other people, because it just makes me a much more relaxed human being. :)
To get back to the topic, of course all of that is no reason to make the one that asked the question angry, if someone is like "go figure it out yourself" or "that's easy haha nap" moderators should just intervene and delete it, because it doesn't help anyone. If I go to a math forum to get help it would be horrible if people react like this. At the end of the day we all just want to get better at the editor or get things done there, so either help (in whatever rude way you like, as long as your help contains information. But I wonder how you can provide help in a rude way? Can't really imagine that.) or don't help and keep your mouth shut, which is a perfectly reasonable and understandable thing to do. :)
There are no stupid questions, there are only stupid answers. Especially concerning something as intransparent as the Galaxy Editor.
I just wanted to thank you for making the Terraining Thursday episodes. I think they are really good, and this one in particular was a nice topic. Its a pity there aren't more viewers, but don't give it up because of that :)
Though I agree with 95% that is said, I don't think that a more complex editor that can do a lot of stuff means that every map has to be ultra complex and use the possibilities. That you can DO a lot of stuff doesn't mean that you HAVE to do it. The easy to understand and simple gameplay ideas are often the best, especially for custom maps.
The whole Nexus Wars genre is really boring and I'd personally love to see it vanish out of Bnet though :) But tastes are different.
Aw, you were faster :)
You don't need the Apply Behavior effect though, you can create an ability that directly applies a behavior (of type "Behavior").
Hey this is a good point about the search bar on sc2mapster. It is pretty bad sometimes at finding results. For some reason almost all of the good results get scattered amongst 5-6 pages, with completely irrelevant results making up most of the list.
Anyone else found that? It's not a big deal but it seems inaccurate.
Can't emphasize this enough.
Sometimes I know exactly what thread I am searching for, I write 90% of the EXACT thread title into the search bar, only to find the thread I'm searching for on page 2 or 3, topped by junk threads which contain no information at all. Happens rarely, usually if you know what you are searching for you will find it pretty fast, but I swear it happened to me like this.
And also there are just a lot of threads which contain a question but never got solved. This is especially true for older threads from may 2010 until like october or november 2010.
damnit, its a shame that it isnt possible. when do the lags start on battle.net. with 1 player? with 2 or 3 ? or with more than 3?
so... lets wait for HOTS :)
As soon as you have to communicate with the server it will lag.
The number of players doesn't play a significant role I guess (but it might be the case that the server responds to the player with the lowest ping, so if someone has a shit ping it will be shit for everyone in the game).
edit:
And btw I don't think they will change their network code for HotS. But we can hope I guess :) It would give so many possibilities of new maps that would instantly be a lot more fun. Like car racing, direct input RPGs, Shooters etc.
But I think it would be very complicated to let a player choose how the server communicates with the map, because they won't change how it effects the standard RTS type of SC2. I doubt they can/will change anything with HotS.
Won't be possible anywhere in the near future to get a lagless direct input game via battle.net. It is the network code that makes this problem, not something we can change in the editor.
I personally am working on a singleplayer (so no multiplayer lag) top-view-shooter though, don't know if it will ever get finished, but I'm very motivated and I think there's a good chance that it will turn out to be a good kit to create some intense and challenging singleplayer maps. Not anywhere in the near future though, I'm not in a rush and I want to provide quality work that other ppl can use if they want to, without shivering in pain because of my shit work ;) I'm aiming for getting the data work and AI stuff done that it feels like a good map until HotS, and look into the level design and UI after the addon.
As soon as I get my new PC somwhere in the next month I'll start a thread here and provide some more information what I am aiming for etc. with videos.
Also there is an multiplayer FPS game somewhere on this site which looked very ambitious and started recently. I doubt they overcame the lag problem, but the project itself looked quite good. I forgot the name sadly.
If you want a random offset for the same effect (doesnt matter which effect) I would always use a "Create Persistent" effect which runs the desired effect as a periodic effect with period count 1, random offset checked and the random offsets added into the "Random Offset" value list in the Persistent effect :)
You could also create a ton of effects and use the "Set" effect as above I suppose, but that gets really messy once you need more than 2 or 3 Offsets, doing it with a Persistent on the other hand takes like a minute to set up and only needs 1 additional effect.
Bommes - your version will either
1 - Create 1 of each unit.
or
0 - Create no units at all.
Many ways to skin a cat :p
No, it won't, because the min and max count is set to 1, which means only 1 effect of the list will be run :)
Because Random is checked it will take a random one, if random is unchecked it starts from the top of the list and goes downwards.
I've never created any units with data so far, but there should be different ways to do it. I'd try:
First of all, make 2 unique effects of type "Create Unit" which create the 2 units you need (add the unit to "Unit: Spawn Unit" in the effect).
After that, create an effect of type "Set", add the both effects to it (in the "Effect: Effects" field), check the "Effect - Random" value and set min and max count to 1.
That's it. I haven't tried it, but this should work. (to spawn the unit just run the "Set" effect somewhere, from a behavior, an ability or w/e you prefer)
0
Thanks for all the nice comments :) Appreciating it!
Not sure how I will experiment with it exactly. Only thought about AI only Squads so far, but actually controlling a whole squad is a pretty cool idea. I'll see if I have enough time to work in that direction, haven't thought about it yet :) thanks.
Haha, didn't know something like this exists. Pretty cool!
And the vs. AI maps are completely lagfree? That would be very good and very important. :)
Just finished another ability concept, basically only the target lock thing is finished, the whole visuals and time it takes to lock on the target and audio etc. is still WIP (as for most of my stuff). Also it still has a bug which I didn't show in the video, but I'm pretty sure I'll be able to solve it relatively easy. (edit: actually I'm not sure if it is even a bug or if I want the ability to behave like it does now. I'll probably just discuss it with my personal tester buddies ;) )
edit:
In contrast to the video I also already changed it in a way, that the ability will lock faster/slower depending on how well you aim.
0
MechCraft
Latest Update:
This is my current project and first map for Starcraft 2. I started the map just to get a good grip of the editor and to learn the basics, but during that process I continually expanded it and today I think that I can actually make a viable and good map out of the concept if I work hard to make it better. So I will regularly post the state of the map here, to hopefully get a little bit of feedback (I know its hard!) and to keep my motivation up. Also to clear my mind a bit because this thread is my personal reference for a concept. ;) Haven't really written anything down somewhere else.
First and most important before you watch the video: This will be a SINGLEPLAYER map because of lag issues and my interest in mission design, so every enemy is computer driven. The current map is just a testing terrain that I will use until I start with actual mission design.
Now lets talk a bit about the concept. Nothing in it is set in stone, so if you have cool feedback (or maybe just want to desperately see something implemented as a player because you are a big fan of unit X or concept Y) just say it, there is a good chance I will do it in some way or the other.
Concept
Current State
The video shows a really early version of the game, I'm continually changing stuff for pretty much everything.
The map currently includes:
Also I already changed some stuff from what you see in the video. The Goliath is much slower now and I removed that annoying Movement sound.
If you have good ideas or an opinion about my map, think it is a bad idea or any other form of feedback just post it here, I will definitely read and think about it. If you don't have specific opinions, share your opinion about these questions:
Do you like the classic Terran vs. Zerg theme or do you think I should finish Protoss or Terran enemy units first?
Would you play a Singleplayer Top-View-Shooter for the SC2 engine or do you think the concept is bad (and why)?
Do you think the damage indicator for enemy units (they turn white/bright for a brief moment) is a good design idea or should it be removed?
Did you realize how I was too lazy to add water to the whole map? ;)
Also a general thanks to all the good tutorials and the overall helpfulness on this site. Especially to Kueken & DrSuperEvil who are superheroes (no really, they are!), share their knowledge and always help if possible. Also of course people like BorgDragon and countless others!
17.06.2012
0
No idea, never played it :D Also I didn't have any concept at all, just turned out like that. I guess its just a very cliche robot ;) I didn't plan to show a video or anything of it in the first place, I just experimented, but I was very satisfied how decent it looked with quite little effort put in so I uploaded the video.
Yeah I agree, and thats why it would be so cool if someone is still working on it. There is really a lot of potential once a basic set of animations is done (it doesn't even have to be a huge amount of animations with 2 billion joints), because applying the visual appearance isn't really hard once you understand how the moving parts are done and which joints move what parts of the unit.
It took me like 5 or 6 hours to get this kind of model (including changing the animations a bit), and it was the first time I did something with Forward Vector SiteOps (and I had to get used to the Rotation ones too) at all, so if I had to do this thing again it would probably take me 2-3 hours max. Sure it still looks pretty bad (especially the back and the animations) and I probably wouldn't use it in a map, but in my opinion you can very easily improve it and if you actually want to do some kind of huge mech robot maps it could easily be done this way in a reasonable amount of time :)
0
Don't let this project die! I hope somebody is still working on this, it has SO MUCH potential!
For motivational purposes I barely slept last night to make my personal Toss-A-Tron. Beware!
Also thx to DrSuperEvil for pointing me towards the awesome data threads :)
Download Map
edit:
not sure if I will continue working on this, because it really has no purpose for me and I have another project that I want to focus on :) But next thing should be to implement the Turn Sensor I guess (besides making more animations and sticks in general, I just added a single one for each arm, thats really not enough for good animations)
0
Imo this topic is pretty much cycling around itself right now, but I will still post my opinion here.
I am kind of in between the worlds at the moment. I have started with the Galaxy Editor some months ago and I still feel like a noob, but I think I can wrap my head around most of the basic data stuff and if you tell me to do a unit in the data editor with the custom abilities X, Y and Z I could find a way to do it for many/most ability concepts and visuals (if you would take for example abilities like we know them from various MOBA games).
On other parts of the editor (mainly UI, Cinematics) or simply knowing what types of Events/Actions exist in the trigger editor, or knowing what types of doodads exist and how they look, I'm simply bad. So I will ask basic questions here (for example the difference between Functions and Action Definitions, I didn't know that ADs can't return a value. It is completely logical, but everyone knows these kinds of mental blackouts) that might seem stupid, just because it will save me a ton of time. The general purpose of questions is to get information. Whether someone else is willing to spend his time and give me the information or not is a completely different thing and therefore no one should blame people just because they don't want to share information. Sure it would be nice if everyone did, but especially in the 21. century things like copyright and the material value of ideas get more important every day. So there are multiple points of view on this topic and I don't want to say my opinion is the only right one. If you scroll through the battle.net maps you will get a good idea of how many people lock their maps and don't want to share anything. I haven't tried the top 20 most played maps, but I bet most of them will be locked (probably all of them except the Blizzard ones). Do I like it? Probably more than not sharing information, because it is actual work that was done and not just knowledge. But I still think every map that has not like a bazillion custom models and 10000 hours of work in it would be reasonable to share. In the end everyone who uses the editor is happy if there are more mappers and artists who get inspired by seeing and being able to analyze other peoples work.
On the other hand there will always be cheaters and douchebags who exploit it on all levels, last year the Minister of Defense of my country (Germany) had to resign because he copied most of his dissertation. So I don't want to pretend everyone does what is best for the community or is really willing to learn. But I personally prefer to think like that about other people, because it just makes me a much more relaxed human being. :)
To get back to the topic, of course all of that is no reason to make the one that asked the question angry, if someone is like "go figure it out yourself" or "that's easy haha nap" moderators should just intervene and delete it, because it doesn't help anyone. If I go to a math forum to get help it would be horrible if people react like this. At the end of the day we all just want to get better at the editor or get things done there, so either help (in whatever rude way you like, as long as your help contains information. But I wonder how you can provide help in a rude way? Can't really imagine that.) or don't help and keep your mouth shut, which is a perfectly reasonable and understandable thing to do. :)
There are no stupid questions, there are only stupid answers. Especially concerning something as intransparent as the Galaxy Editor.
0
@Mozared: Go
I just wanted to thank you for making the Terraining Thursday episodes. I think they are really good, and this one in particular was a nice topic. Its a pity there aren't more viewers, but don't give it up because of that :)
0
As a workaround you could use multiple behaviors that trigger each other after they end.
Or you could handle this internally in the actor events and work with timers that start when the behavior starts etc.
0
Though I agree with 95% that is said, I don't think that a more complex editor that can do a lot of stuff means that every map has to be ultra complex and use the possibilities. That you can DO a lot of stuff doesn't mean that you HAVE to do it. The easy to understand and simple gameplay ideas are often the best, especially for custom maps.
The whole Nexus Wars genre is really boring and I'd personally love to see it vanish out of Bnet though :) But tastes are different.
0
@TrenchaunT: Go
Aw, you were faster :)
You don't need the Apply Behavior effect though, you can create an ability that directly applies a behavior (of type "Behavior").
But doesn't make a huge difference I guess.
0
Can't emphasize this enough.
Sometimes I know exactly what thread I am searching for, I write 90% of the EXACT thread title into the search bar, only to find the thread I'm searching for on page 2 or 3, topped by junk threads which contain no information at all. Happens rarely, usually if you know what you are searching for you will find it pretty fast, but I swear it happened to me like this.
And also there are just a lot of threads which contain a question but never got solved. This is especially true for older threads from may 2010 until like october or november 2010.
0
As soon as you have to communicate with the server it will lag.
The number of players doesn't play a significant role I guess (but it might be the case that the server responds to the player with the lowest ping, so if someone has a shit ping it will be shit for everyone in the game).
edit:
And btw I don't think they will change their network code for HotS. But we can hope I guess :) It would give so many possibilities of new maps that would instantly be a lot more fun. Like car racing, direct input RPGs, Shooters etc.
But I think it would be very complicated to let a player choose how the server communicates with the map, because they won't change how it effects the standard RTS type of SC2. I doubt they can/will change anything with HotS.
0
Won't be possible anywhere in the near future to get a lagless direct input game via battle.net. It is the network code that makes this problem, not something we can change in the editor.
I personally am working on a singleplayer (so no multiplayer lag) top-view-shooter though, don't know if it will ever get finished, but I'm very motivated and I think there's a good chance that it will turn out to be a good kit to create some intense and challenging singleplayer maps. Not anywhere in the near future though, I'm not in a rush and I want to provide quality work that other ppl can use if they want to, without shivering in pain because of my shit work ;) I'm aiming for getting the data work and AI stuff done that it feels like a good map until HotS, and look into the level design and UI after the addon.
As soon as I get my new PC somwhere in the next month I'll start a thread here and provide some more information what I am aiming for etc. with videos.
Also there is an multiplayer FPS game somewhere on this site which looked very ambitious and started recently. I doubt they overcame the lag problem, but the project itself looked quite good. I forgot the name sadly.
0
@playpong: Go
If you want a random offset for the same effect (doesnt matter which effect) I would always use a "Create Persistent" effect which runs the desired effect as a periodic effect with period count 1, random offset checked and the random offsets added into the "Random Offset" value list in the Persistent effect :)
You could also create a ton of effects and use the "Set" effect as above I suppose, but that gets really messy once you need more than 2 or 3 Offsets, doing it with a Persistent on the other hand takes like a minute to set up and only needs 1 additional effect.
0
No, it won't, because the min and max count is set to 1, which means only 1 effect of the list will be run :)
Because Random is checked it will take a random one, if random is unchecked it starts from the top of the list and goes downwards.
0
I've never created any units with data so far, but there should be different ways to do it. I'd try:
First of all, make 2 unique effects of type "Create Unit" which create the 2 units you need (add the unit to "Unit: Spawn Unit" in the effect).
After that, create an effect of type "Set", add the both effects to it (in the "Effect: Effects" field), check the "Effect - Random" value and set min and max count to 1.
That's it. I haven't tried it, but this should work. (to spawn the unit just run the "Set" effect somewhere, from a behavior, an ability or w/e you prefer)