TF2M HUGE Collaboration project?

Asd417 said:

I am currently working with Googujajoob to make an arena map and while I was working with him, I learnt a lot.

Mapping which always felt daunting to finish, suddenly seemed less overwhelming.
I was encouraged as a newbie mapper to start and finish a map.
So I started cp_sittingrock(wip name) and so far it is turning out pretty good.

So the Collab. Why so suddenly?
Well, I think collaboration can replicate my experience of being encouraged for new members of this community!

What I am thinking of is:
-10-15 people collaboration of big map (multistage cp or payload. or even TC? idk)
-Aimed to make a FULL map which goes through the whole process of map development.
-dedicated steam chat for people in the collab to easily discuss the map.
-Possibly a map with a whole new theme. (Or a custom theme with a few custom props)
-A collab where people with different skillsets join together and possibly share/teach their skill
-A collab where named professionals give real-time tips about a very specific question (achieved by steam chat)

What do you guys think? I think it would be a great chance for the new members to learn and a good fun for the pros.

Click to expand...

While I think your idea sounds like a great project to do and learn from, I think there would be a big problem in communication. A collab between 10-15 would be hard to organize and there will be probably problems with deadlines, because somebody always will have problems finishing his part on time, wether he just forgot about it or had other problems. However I think, that scaling the size of the groups a little bit down(3-4) would make it way easier to finish a project together and would lead to much better results, because it would be much easier putting your work together and making it consistent in balancing and detailing.

Okay, gather 'round, kids. I got a story for you all.

Once upon a time, a group of 10 or so, myself included, tried something like this. We kept the communication in a private conversation so there probably aren't many others who know of it. But there was a big lesson to be learned, and it turns out (unsurprisingly) that this kind of mass collaboration isn't easy in the slightest. I'll share with you the main reasons the project never saw the light of day so that you might be saved from the same headaches, should you choose to continue:

• The distribution of skill between all the participants was too wide. In theory, a project like this sounds like a fun environment for new people to work with "the pros" and learn how they do things, but you have to consider that the pros will also have to deal with the inexperience of the newbies. There's a reason experienced mappers are picky with single collab partners - it's not going to be any different here. Everyone is still expected to pull their own weight. Which brings me to the next point -

• Some of the participants were still making newbie mistakes. This was probably the worst offender. We went in hoping it would be a learning experience type of project, but it turns out that most people just don't have the patience to teach when the skill gap is too great. By the time the first "pro" got their hands on the map, it was full of off-grid and sloppy brushwork - as a result, several people were spending their allotted time fixing and working around the mistakes of the previous contributors instead of implementing their own changes. Three people left the project as soon as they learned about this.

• We didn't have a log of who made what. This made it rather difficult to track down the person who made a certain mistake so we could tell them not to do it again, or talk about someone else's design choices. On a larger scale, this meant that no one was going to learn anything because you couldn't tell what was "newbie" work and what was "pro" (aside from the obvious, I mean). In hindsight it was suggested that everyone should have placed their own work in a separate visgroup with their name.

If you really want to go ahead with this project, I strongly urge you to be picky about who you let on board to avoid the same frustration we had. Every participant should know basic concepts like clean brushwork and scaling. Be extremely meticulous when logging your changes so that you can properly discuss everyone's contributions. And most of all: don't pass the damn map in a circle like we did. Bounce it between everyone and don't keep secrets about what you're doing. This isn't like the 3 Word Story where you can just tack something on the end without looking at the rest of it.

Good luck.

Adding on, i believe there to be at least 2 projects like this before, neither of which reached an a1 even. It will not be like 1/15th of the work of making a map alone, it may well be more than 1/1 of the work with all the organisation and communication.

There was a huge segmented egypt themed arena map plan. I don't even know how long time ago it was but as far as I know it went nowhere.

I really liked to be part of a bigger project but for the reasons @aui listed, it will be very hard to actually make happen. About skill levels, I think there is a relatively low level when mapper understands basics of how to place brushes efficiently without making huge mess and going offgrid. But there is still newer mappers that do that. Rejecting someone who feels like he has the skill will feel bit bad.

Also 10-15 mappers is a lot, even for tc map because the most important part is the actual layout, not really how you then make it. Trying to make layout with 10-15 people can easily end up being mess because no one can really take the leader role and say how things go. Luckily, we already have solved that problem; single mapper and feedback. Let's say that isn't a problem and the group gets good layout. Who will start it? Who will continue it? How files will be transferred? How to be sure everyone has the correct version and that their ideas will be heard? Organizing won't be easy by any mean

pd_snowville was a 4-ish person collab and it ended up working out okay ish.

if you are collaborating, you really want to be using manifests or instances to break down your vmf into chunks, this facilitates splitting the workload. You also need to be using version control software to host your project and track every changes made. version control also allows you to rollback to earlier versions when needed. it's also a self documenting changelog.

In snowville's case, unfortunately we were working under a tight time constrain and didn't have the time to teach much of anything to newcomers; I feel like they got kinda left behind near the end.

So, to be inclusive of newbies, I feel like everyone who's in the upper 50% of the skill pool needs to learn to delegate better. what I'm finding out is I kinda enjoy the gruntwork associated with source because I can just slip into a meditative state and get things done but really I should have been passing down those tasks and reviewing them.

Main takeaway is collaborating is haaaard for everyone. sadly the tf2m community has next to no collaboration knowledge. I'd say it is the one thing that is missing in terms of allowing people here to become fully fledged level designers. It's definitely something we should pursue but basically everyone here, pro or noob, is starting from zero. That's your main hurdle.

ps: I was considering doing a big collab for a multistage koth and the idea is still at the back of my mind but I can't quite commit to a project like that atm.

the tl;dr of all the long posts.

• Can TF2M do a big collab project? No, not unless there is a "pro" who can lead it and people who are 100% behind them.
• Skill gap is wide, and most people now aren't even at the base-level of quality you'd want.
• Think of how long you think it'll take to do this, or how easy it might be, and x100. Then you're close to the min amount of experience, time and resources required for this.

fubarFX said:

sadly the tf2m community has next to no collaboration knowledge. I'd say it is the one thing that is missing in terms of allowing people here to become fully fledged level designers

Click to expand...

Considering Hammer Editor's capability and support for collaboration that isn't big surprise

It's interesting how quickly new people come up with stuff as ambitious lately.

Hyperion said:

Considering Hammer Editor's capability and support for collaboration that isn't big surprise

Click to expand...

no you're just bad, hammer actually shits on many other editors in terms of support for collaboration. people just don't know how to make it work and that attitude isn't helping

fubarFX said:

no you're just bad, hammer actually shits on many other editors in terms of support for collaboration. people just don't know how to make it work and that attitude isn't helping

Click to expand...

I don't really know what attitude you mean. There is editors that handle collaborations real time, not saying Hammer should be one tho

Just the attitude toward hammer being bad. it's bad, for many reasons, but support for collaboration is not one of them.

Adding point :-
A collab in a way where one person makes the map, someone models and someone does particles. For instance that would work but having several people working on a single map is bad with Hammer

Recently Yrr and I made a map (cp_extol) by passing a .vmf back and forth when we got stuck on something. We didn't have a lot of time to test the system, but it worked pretty well and didn't require any fancy techniques - we just shared a Dropbox folder and saved as a new file every time the map traded hands.

In our particular case, Yrr made B and got stuck with making A, I made a connector and part of A, and she finished it off. I can't speak for Yrr, but I felt like being able to bounce off of someone else's brushwork made developing the map easier than working alone (or taking feedback from words/screenshots/paintovers).

It's not a super focused method of working, but it doesn't take a ton of commitment to make work. It probably works best if one person is in charge so they can just solicit ideas and take the project wherever they want without waiting on slowpokes.

Idolon said:

Recently Yrr and I made a map (cp_extol) by passing a .vmf back and forth when we got stuck on something. We didn't have a lot of time to test the system, but it worked pretty well and didn't require any fancy techniques - we just shared a Dropbox folder and saved as a new file every time the map traded hands.

Click to expand...

That is the same method that me and hyperion used when making arena_ascent. It worked out pretty nice, the map would never had have been an a1 even if i couldn't have bounced it off hyperion for him to take a stab at it. The map never went anywhere but the method is totally a good thing.

However, im not sure how well this method would work with more than 2 people. It means more people to bounce ideas off of so you wouldnt get stuck for as long / as often. But it also could be bad in ways such as people not understanding the geometry or what it is trying to achieve as much if its just being bounced around whenever someone gets stuck. It could be explained to them but they may still not understand how it works, as geometry can be pretty complicated. Then, as they dont understand (or dont understand quite well enough) they may just make whatever in the area the other people were stuck on, which could make the map unfocused and not work together well.

What im trying to say is that maybe this method with a large amount of people could get hectic and result in a map thats sort of just stuck together like Frankensteins monster.

Asd417 said:

It seems everyone is saying that in order for this to work. we need a very clear direction participants should at least be all intermediate mapper.

Click to expand...

Everyone is saying you need far more than just a clear direction and intermediate mappers. Your definition of intermediate mapper is probably still below the level that people are saying you need minimum.

The main thing is that people don't want to fix other people's work. I was part of a big collab payload map a while back and after a few people worked on the map, everyone who saw it started quitting the project cause it wasn't good enough and they didn't feel like fixing problems other mappers made. It was an annoying experience since I was just glad to be working with some big names from TF2maps but that's the way things go.

More mappers = more problems.

I've collaborated/tried to collaborate about 5 times over the years. Biggest problem for 1 of those cases was that the other part was too unexperienced with making maps. He liked to use big grid size, made a lot of very thick brushes and the layout was more like straight forward than something that i liked to do.

My first ever was trying to make a team of people who would give feedback and make suggestions, which i and someone else would then implement to the map. With this, i tried to avoid "beginners messing up" as they make mistakes that are needed to fix. Turns out that communication was the problem and myself too, couldn't just stand the pressure of making all those people wait while i make something up and there wasn't many ideas to work from either.

In another case, i wanted to encourage this person to finish his map, as it looked pretty much ok for me to continue work on. But i ended up fixing it a lot which led to frustration + the guy wasn't working with the map at the time since he was moving and was going to be out of the net for over a month. So that didn't work at the end either.

One case i walso wanted to help someone to make out a working map as he was dead ending with his. Again I ended up fixing stuff, but there wasn't too much to fix for. It was more like optimization gig and some feature implementation. But turns out the other person already had lost interest in it and it just didn't work out. Maybe someday he or i will return to it and make it working map. Maybe not.

I have one example of people who are very advanced in making maps, models and such. The goal was to finish a project with a story. First the engine pick was an issue. Some wanted Source, the others Unreal so team got split into two. But it simply ended, for several reasons because the goal was too big and there wasn't really much agreeing and communication between people.

I did do a successfull collaboration with a real life friend of mine, who was making CS Source maps at the time that i was years ago+ tf2 maps too. I kind of showed him a map that i wanted to make, but i ended up dead end with it. So i gave him the vmf, and after few days he gave it back. I found new ideas to work on it, we talked about it and kept switching the file forth and back, basically 50% made of it by each of us. At the end, the project was fruitfull and it turned out to be my best collaborative experience.

As a conclusion, tl;dr: as others said, having a large team of people just complicates things and makes it hard to finish up. Someone tends to take a lead and needs to take it too in order to get things and team working together. But if that lead turns out to be someone that turns out to be hard to work with, whole thing comes crashing down or if the team has skills enough but can't agree on anything, change what others think is good and turn it to worse, it just leads into problems.

Sometimes it's better to work alone to get out something you like, instead of something that everyone would be happy. If anyone is even considering collaboration, find someone who is near your skill level and start working on that base alone. It's not going to be easy, unless you work with someone who for example you know from real life or/and has the skills near the same level. Find someone you can work with and set up goals to reach. Plan ahead, leading team is not easy.