Testing mods before upload/ Best Practices

[ryanfss]

Just looking for tips or maybe even land mines to avoid when uploading a new mod. Also whats the best process for testing? Reason for asking is what I thought should be a simple process for a seemingly working mod. When i say tested I archived the files, removed them from data folders after compressing them, open up and create new game, save, quit. Then I made sure the scripts were removed for the scripts folder (user) Then I extracted the files back into my game and installed then uninstalled successfully. The files I uploaded were the same ones I extracted back into my game so they were untouched.

 

I've uploaded 4 times now and it never works. I've rewritten the scripts, etc. but to no avail. I've scrapped it altoghether and am starting over with a fresh install of the CK. Kinda discouraging really to put that much effort into something and not be able to get it on the Nexus.

 

Should I be testing on my machine? Virtual server? Sending to someone to test first? Any and all tips/pointers would be great. I'm afraid if I go through the process again with the same results, I might end up in Maryland looking to punch the first person I see in a lab coat and Bethesda Nametag.

[caleb68]

compiled user scripts get moved into the Scripts/ folder as pex files, the source/user folder is just that, the source files, you wouldn't need to remove those in order to test the mod.

 

When building your Archives, the main archive must be yourmodname - Main.ba2 and the textures archive must be yourmodname - Textures.ba2, Make sure to match CaSe with the mod for the ba2 names.

 

if your trying to test your archives, and have no other mods installed:

 

remove the Materials, Meshes, Textures folders from the Data folder.

Move any scripts you created from the Scripts folder, but you can leave the 'source' folder.

Move any sounds you created from the Sound folder

Move any Videos you created from the Video folder

 

load up fallout, make sure your mod is checked in the mod list, and try it out. If you've done everything right, it should work just fine for you.

 

I make a folder called DataTemp in my data folder, and move everything into that folder to test the archives.

 

If your mod includes NPC's and they are just using default resources for their looks / hair bits, you don't need to include the character head & texture exports in your archive, this will help cut down on your mod size a little as well, you only need to include those if they are using non-default resources (i.e. user created hair/makeup/morphs)

 

Alternatively, if you have a second machine, toss a copy of fallout on it, and use it for testing :) ( I would but im being lazy lol ) that would make sure your only testing the mod and its archive files.

 

After I know that works fine, I send the file off to the tester(s) for the mod and let them hammer away at it and see if they can find anything broken.

[ryanfss]

Good Stuff, thanks. I assume these testers are people you know or recruited to help? That would have been the thing to do. Good info, thanks again! Also, I did state that wrong - the pex files were the one I removed

[montky]

that is an excellent question, and I am interested to hear more

from other modders and folks about how they test their mods and all too.

I have always wondered what goes into the whole process of making a mod, from start to finish,

and this seems like an ongoing part of the cycle.

 

 

 

without seeing the issue, I'm going to guess that it is some 'pathing error' -

you're making the mod dependent on a pathway that isn't present in other machines etc.

it's usually something like that, it can happen to folks at random.

if folks have different install pathways, then, your pathway might not match to standard convention or to their particular install.

 

It's a good idea, if you have access to a Local LAN party club or such,

to take along your mod to the local LAN party, and try them out there.

this is great for a few reasons;

i) you get to meet folks and how they install or run things

ii) you get to test a random sample of installs and configs all at once. you get access to different installs and configs, high end, medium, low end.

iii) if issues happen, you get to see them resolve in real-time.

iv) you have a lot more control over this kind of 'limited manual release", especially if you're working with sensitive stuff you don't want out there.

 

so, you're testing via downloading or introducing the file, and installing etc.

its a great idea to test on a machine that isn't the one you made the mod on.

 

the next option is, putting out a call for beta-testers and the such like;

folks you trust on web domains such as NMM, Moddb, ThisWeekInTech, CodeStackExchange, NewGrounds/Miniclip etc.

I say, trusted, as some folks will try and use that for vulnerabilities,

or claim your mod 'bricked their machine' or gave them hyperviruses hehe.

 

in terms of rapid testing, and bug-testing,

I like to test across the full gamit of installs and configs,

and try and anticipate likely sources of errors or mod conflict ahead of time.

I like testing for low end, medium end, and high-end.

as, they have very different factors as to why certain errors or conflicts occur etc.

 

you've gotta control for OS - which version of what are they running?

what is their hardware like?

what is the mod order - do they have dependencies in the right order? are there any missing components? do multiple mods change the same thing?

what are the symptoms of the conflict etc

random CTDs are the worst to try and troubleshoot hehe.

 

 

 

I hope that gives some options and insight;

I'm hopeful it's just a pathway error, or a version mismatch error,

which is preventing it from working across other installs.

[ZerasCETA]

Sounds like you've tested the file structure extensively, so I doubt that will be the bug, if any. In my experience, just release it as version "1.0a" and explain that its currently in the alpha testing stage and that you're looking for feedback on potential conflicts and issues and hunt down the reports you get. Most will be user error in one way or another, as all modder's see, but some will enlighten you to real issues you need to fix. Once you feel confident it's ready, move it to 1.1r and beyond.

[Greslin]

I'm a one-man band who writes small niche mods. So I have an alternate take on this whole testing business.

 

I've been coding since the early 80's (yes, I'm that old :)), so some old habits got locked down before I even hit puberty. One of the main things I've learned over the years is not to code larger than I can effectively test. In FO4 modding terms, that means to be wary of overhauls.

 

When I have a good idea for a mod, I rush to implement a limited subset of that mod and release it to the world. Get something working and start generating user feedback. Then make it better, lock down features, fix bugs, keep releasing. Expand the mod's scope only as quickly and as far as you can effectively manage bugs and feature creep. But always, always, always start small.

 

Beyond that, my main concern on any release is not screwing up my users' game saves. As long as game saves won't get wrecked and the mod can be easily removed, I'm content to ship it as alpha or beta and let the bolts come loose in public.

[caleb68]

Good Stuff, thanks. I assume these testers are people you know or recruited to help? That would have been the thing to do. Good info, thanks again! Also, I did state that wrong - the pex files were the one I removed

The testers are usually people who asked me to make something specific for them, But I do have a couple people who like testing different stuff I'm working on, show what they had problems with, toss in suggestions to improve them, etc. which is nice, Its more like they recruited me then me recruiting them heh. But having a couple people who are willing to help you test stuff and not leak it to the world before your anywhere near done is nice to have :) i.e. like this one I'm working on, 3 months of work, and its still no where near done, its still in alpha state and needs alot of construct/scrap recipes made, as well as more testing on functionality.

 

Sometimes finding a bug can be difficult, and asking for help is about all you can do because you overlook / can't see your mistake (Just made a post before this asking for help with a terminal issue).

[ryanfss]

 

Good Stuff, thanks. I assume these testers are people you know or recruited to help? That would have been the thing to do. Good info, thanks again! Also, I did state that wrong - the pex files were the one I removed

The testers are usually people who asked me to make something specific for them, But I do have a couple people who like testing different stuff I'm working on, show what they had problems with, toss in suggestions to improve them, etc. which is nice, Its more like they recruited me then me recruiting them heh. But having a couple people who are willing to help you test stuff and not leak it to the world before your anywhere near done is nice to have :smile: i.e. like this one I'm working on, 3 months of work, and its still no where near done, its still in alpha state and needs a lot of construct/scrap recipes made, as well as more testing on functionality.

 

Sometimes finding a bug can be difficult, and asking for help is about all you can do because you overlook / can't see your mistake (Just made a post before this asking for help with a terminal issue).

 

Well, you want to give this one for a spin? It's really a very small mod compared to most - single build menu with some sub menus. Really not much to test aside from the Navmesh with settlers and such. But I also thought I had it nailed down the first time and it didn't go so well. This is my first mod so it spooked me a little. If not interested, np. But thanks again for taking the time to respond.

[ryanfss]

I'm a one-man band who writes small niche mods. So I have an alternate take on this whole testing business.

 

I've been coding since the early 80's (yes, I'm that old :smile:), so some old habits got locked down before I even hit puberty. One of the main things I've learned over the years is not to code larger than I can effectively test. In FO4 modding terms, that means to be wary of overhauls.

 

When I have a good idea for a mod, I rush to implement a limited subset of that mod and release it to the world. Get something working and start generating user feedback. Then make it better, lock down features, fix bugs, keep releasing. Expand the mod's scope only as quickly and as far as you can effectively manage bugs and feature creep. But always, always, always start small.

 

Beyond that, my main concern on any release is not screwing up my users' game saves. As long as game saves won't get wrecked and the mod can be easily removed, I'm content to ship it as alpha or beta and let the bolts come loose in public.

Good advice, thanks!