Modders, How do you deal with Feature creep?

[Tentacus]

I've been working on a large AAF mod for the better part of a year now. (though I've released several smaller mods in that time). I've been in what I thought was final testing several times, but always seem to find a new feature I want to add during playtesting. Anybody else have this problem, and if so how do you deal with it?

[Invictusblade]

To a lesser extent, Family Planning Enhanced More Creatures is a less advanced version of that.

 

I was fairly happy with it, but I wanted to add some form of follower system

which turns out to a lot more easier than I thought until I noticed the bugs so I backtracked the scope.

 

One lesson I learnt in game design at University is

"You have to hate what you are working on"

so you won't want to add features instead you have to try to finish the bastard. (then maybe added one or two extra's)

 

[Reginald_001]

Make a release schedule, set a date and stick to it. Communicate the date to the public, so you'll have created an expectation.

Then live up to that expectation. Push new ideas to a next release.

 

It's the only way.

[Guest]

RSE man. Just look at original RSE. Feature creep to the max. Eventually the features will clash. That's your cue to stop. Which I obviously did not do back then...

[Tentacus]

12 hours ago, Reginald_001 said:

Make a release schedule, set a date and stick to it. Communicate the date to the public, so you'll have created an expectation.

Then live up to that expectation. Push new ideas to a next release.

 

It's the only way.

That sounds like a nightmare... but you're probably right.

6 hours ago, Flashy (JoeR) said:

RSE man. Just look at original RSE. Feature creep to the max. Eventually the features will clash. That's your cue to stop. Which I obviously did not do back then...

Heh heh, Yeah I'm about at that point now... In fact this prostitution mod even overlaps with my Mercenaries mod as you can safely travel in Gunners territory and hire Gunners as mercs.

[Carabosse]

On 5/30/2019 at 1:49 AM, Tentacus said:

I've been working on a large AAF mod for the better part of a year now. (though I've released several smaller mods in that time). I've been in what I thought was final testing several times, but always seem to find a new feature I want to add during playtesting. Anybody else have this problem, and if so how do you deal with it?

 

The last update to my Fallout mod was last summer. The next update was meant to be released at the end of October, then November, then I wanted to drop it just before Christmas. You get the picture.

 

It became clear early this year that as the next update is such a fundamental change and brings with it so many new tools for me to play with, I was better off doubling down on making it as good and complete as possible rather than rushing it or releasing something I wasn't happy with. My aim when the next version is released is to almost exclusively focus on story from that point and stick to a roughly monthly update cycle (which would be much easier at that stage). We'll see how that goes. 

 

Also I have to remind myself now and then why I bother making mods at all. It's because I enjoy the process, that's where all the fun is for me. I actually hate releasing mods. That sinking feeling you get just after uploading a file, wondering what you've forgotten or how badly your mod is going to break someones game. I feel a bit sick just thinking about it. 

[WandererZero]

On 5/29/2019 at 7:49 PM, Tentacus said:

how do you deal with it?

Write a spec. When I say write a spec, at least within this context...I mean, outline what you want the mod to do, it's features at the outset, and the purpose. It doesn't have to be a biblical tome. One page is typically enough. 

 

At the bottom of the spec, put the following down: 

 

1. 

2. 

3. 

 

Leave them blank. Fill them in later. But, once you do so...the spec is complete, and you can't go beyond it for the first release. You don't have to have 3. When I supervised developers, depending on the team & task...and how they worked, I sometimes only gave them 1 or 2. Write the spec, if you find an obvious thing that should be added...fill in one of the blanks...and as I said: once they're filled the initial spec is complete. You don't go past it. 

 

You'd be surprised how far a small amount of organization will take you with things like this. 

[Tentacus]

Thanks for the feedback all. I just decided to nuke 2 days of work because I was getting way into the weeds. I'll call it feature complete and get back to just testing.

[Nebuchadnezzer2]

On 6/1/2019 at 11:18 AM, WandererZero said:

Write a spec. When I say write a spec, at least within this context...I mean, outline what you want the mod to do, it's features at the outset, and the purpose. It doesn't have to be a biblical tome. One page is typically enough. 

 

At the bottom of the spec, put the following down: 

 

1. 

2. 

3. 

 

Leave them blank. Fill them in later. But, once you do so...the spec is complete, and you can't go beyond it for the first release. You don't have to have 3. When I supervised developers, depending on the team & task...and how they worked, I sometimes only gave them 1 or 2. Write the spec, if you find an obvious thing that should be added...fill in one of the blanks...and as I said: once they're filled the initial spec is complete. You don't go past it. 

 

You'd be surprised how far a small amount of organization will take you with things like this. 

Think I need to do this sort of thing for 'everyday' stuff.
Tangents on tangents and oh look, 6 hours later, I did half of what I wanted...

[Carreau]

I keep a list of features I’d like to add. If I come up with a new one, I put it on the list. When I’m making a release push, I force myself to stay focused knowing I’ll add new features over time. 

 

The downside to this is I never play the damn game. I’m always churning out updated content.