Improbable weapons mod

[mugling]

Has been floated as an idea before. This PR moves some of the more improbable weapons to a mod. Also provides a suitable outlet for contributions that are not a good fit for core. Save game compatibility is preserved via the new legacy function and the items themselves are retained verbatim as mod content.

[Xnfinity]

What point is it to move the lesser used items out of the game in to a mod?

All it would do is add a another value to most peoples automatic added list and just be left out and forgotten by everyone else, seems like a waste

(ps you forgot the fire weapon louisville slaughterer)

[Chezzo]

Thumbs up for the comment, thumbs down for mentioning my beloved sacred Louisville Slaughterers.

[roushguy]

Part of the fun of Cata is using those improbable weapons. I mean, for Pete's sake, you can fight Shia LeBeouf. There are all sorts of jokes and such in the game and it's all wrapped up in the game's premise.

[Treah]

I would not say that the Louisville Slaughterer is improbable... I don't think that the flaming machete is improbable either.

[Chezzo]

I think all of these are probable, (except the chainsaw lajatang, I will concede.) Here's the shishkebob: https://www.youtube.com/watch?v=Zithf1B1Cuo

And I think they are all really really fun! What is the point of simulating the zombie apocalypse if we cannot strap fuel tanks to our swords?

[SeniorScore]

#FlamingThingsMatter

[Coolthulhu]

What is the point of simulating the zombie apocalypse if we cannot strap fuel tanks to our swords?

But you can, with a mod. Mods are good for this kind of shit because then they can become crazy and unbalanced, while core content has to have standards.

[Chezzo]

Using a mod that only adds fun things and adds no difficulty feels like cheating to me.

[mugling]

I would not say that the Louisville Slaughterer is improbable

It is not included in this mod

Mods are good for this kind of shit because then they can become crazy and unbalanced

We have a lot of requests to merge improbable weapons. Contributors tend to get upset when they are rejected - this provides a solution. Hopefully we can also attract a volunteer maintainer for the mod.

[kilozombie]

Why do we have to move every damn thing in the game to a mod?

Each time I get a new version and make a new world I'm forced to figure out what the hell all these miscellaneous random vague mods do. I just wanna play the game in a complete package. C'mon.

[mugling]

Why do we have to move every damn thing in the game to a mod?

Hyperbole. That isn't happening at all. A mod was created for new makeshift items and this PR moves a handful of items to a second. Both are intended to aid maintenance and encourage new contributions.

Each time I get a new version and make a new world I'm forced to figure out what the hell all these miscellaneous random vague mods do.

I see no problems with the descriptions for either.

I just wanna play the game in a complete package. C'mon.

Experimental branch often has many significant changes (far more than this PR) so if this change bothers you it might be best if you gave it a miss.

[kilozombie]

I play experimental. I have since the game was developed by more than one person. I don't enjoy how many things are in these tiny menial mods and I never have-- they barely contain any items or monsters.

Friend of mine thought the 'new makeshift items' and I had to clarify that all it added was one bayonet. The description of Crazy Cataclysm doesn't really explain the sort of monsters that are in it-- is it impossible for us to decide on the tone of our game, that we're OK with zombie dancers or Cannibal Shia Labeouf...? It's the little things that just add tedium to going into the game and playing and getting into the game as a new player. It feels like soon enough half the features in the game will be added to mods and blacklists. Want no vehicles? Want to remove welding rigs because they're imbalanced? Want to make trees 20% stronger but add zombear honey in them??? Here's all these choices, new player!

You say "if this change bothers you, don't play the experimental" like every PR you put on here is guaranteed to pass. This one hasn't yet.

[Zireael07]

Which items are moved apart from fire weapons?

[mugling]

@kilozombie this is a technical forum. Please stay on topic, stick to the facts and don't repeat your original argument at length. There are precisely two new content mods both of which are recent additions and have already begun to include new content.

You say "if this change bothers you, don't play the experimental" like every PR you put on here is guaranteed to pass. This one hasn't yet.

This is a trivial change and does not require the above vitriol. You have lost precisely nothing. It has been previously discussed, has the support of another developer and is backed by an implementation. Easier maintenance and attracting new contributors are significant advantages.

[mugling]

Which items are moved apart from fire weapons?

• firemachete_off
• firekatana_off
• zweifire_off
• broadfire_off

Note cs_lajatang_off is also moved whereas battletorch remains in core

[Zireael07]

I'm ok with all the moves except for chainsaw lajatang which does not seem that improbable.

Looks like it's all fire weapons, so I suggest putting the lajatang back and renaming the mod to 'fire weapons'. Less problems that way - both for maintainers and players.

[mugling]

Looks like it's all fire weapons, so I suggest putting the lajatang back and renaming the mod to 'fire weapons'. Less problems that way - both for maintainers and players.

Subverts the goal of adding new improbable weapons (for example this mod could have resolved #15329)

[Zireael07]

In that case add some more 'improbable' stuff (the linked thing is a good case) and keep the existing name.

If it's just fire weapons + 1 other, people and maintainers both will wonder 'what's the case' in a couple of months.

We already have some mechanics/design decisions very few players/maintainers understand because the project is fairly old and decisions were made years ago, some even by Whales.

[mugling]

In that case add some more 'improbable' stuff (the linked thing is a good case) and keep the existing name.

This can come in time - I'm reluctant to do a necro on that old issue. Makeshift items mod tarted with one item and now has 5 within a week. Possibly some of the items from #18187 could go in?

We already have some mechanics/design decisions very few players/maintainers understand because the project is fairly old and decisions were made years ago, some even by Whales.

Exactly. I'd prefer to add unbalanced content to mods not core as this prevents proliferation of this problem. Otherwise the same pattern repeats itself, contributor adds content with marginal appeal, contributor leaves, content bit rots.

[Mecares]

Possibly some of the items from #18187 could go in?

Sounds good to add them to the mod, if they fit there better than into the core game, i would be glad if they make their way into the game in some way.

[Snaaty]

Good idea. I do feel that several of the mods could benefit from a more lengthy description that gives more detail regarding what the mod includes.

[ghost]

I'm against this from a practical standpoint: Making it into a mod would be ideal, but the state of mods right now isn't good. A lot of different mods I've tried have had problems, thrown debug errors, and have had unforeseen consequences and interactions.

There's no way to distinguish between the buggy mods from the rest besides trial & error or word-of-mouth. Making more things in the game into mods is something I'm against until the built-in mods are held to a higher standard.

[mugling]

I'm against this from a practical standpoint: Making it into a mod would be ideal, but the state of mods right now isn't good.

It's the same JSON whether its loaded from data/json or data/mods

A lot of different mods I've tried have had problems, thrown debug errors

Not in core they don't - I regularly check this via --check-mods and we have exactly one long-term issue with Generic Guns (#18326). Arbitrary code you fetch from other sources doesn't count.

Making it into a mod would be ideal

The alternative is removing the content given that it's of marginal appeal, horribly balanced and the C++ implementation is fairly ugly. Moving it to a mod gives it a chance to be redeemed and further maintained.

[ghost]

You're kinda proving my point about the low standards mods seem to have. If this stuff isn't up to coding standard, then it should either be revised or removed, not made into a mod.

And for the record, I've been having more problems with other mods, Generic Guns is just one I really want to see get fixed. Vehicle Additions loves throwing turret errors at me. PKs Reimagining recently had an untested merge go through that 100% broke the mod for a bit. And don't even get me started on the magic mods... I gave up on those completely.

[ghost]

Also, there's this issue with Generic Guns as well #18331 Ran into it pretty quickly.

[mugling]

Also, there's this issue with Generic Guns as well #18331 Ran into it pretty quickly.

There are also 1200+ issues open against core so I'm not quite sure what your point is?

Quoting @Coolthulhu:

But you can, with a mod. Mods are good for this kind of shit because then they can become crazy and unbalanced, while core content has to have standards.

We have to maintain core. This makes the job easier whilst depriving players of absolutely nothing.

[mugling]

Those aren't improbable weapons though

[ghost]

They're improbable in the sense you probably wouldn't run into them where the game is set.

[Rivet-the-Zombie]

TEC-9s aren't supposed to exist outside prop departments

It's purely anecdotal, but I've fired an unmodified (semi-auto only) TEC-9 before.

[mugling]

This PR simply makes 5 infrequently used items opt-in as opposed to opt-out.

We now have 16 participants...

[ghost]

I think it's more about the philosophy of the decision than the items themselves

[mugling]

I think it's more about the philosophy of the decision than the items themselves

Exactly, pure bikeshedding

1. Futile investment of time and energy in discussion of marginal technical issues.
2. Procrastination.

We have far more serious technical issues to resolve but they don't receive even a fraction of this degree of attention. Instead we consume enormous amounts of time dealing with trivial issues such as this because everyone feels qualified to pass their opinion, in some cases ad infinitum.

The concept of a mod for such content was settled in #15329 and this PR provides an implementation. Development time is finite and it's reasonable to expect that once a consensus is reached amongst developers that an implementation will follow. How else are we expected to make any progress?

Waiting for developer review/merge

[Coolthulhu]

I'll keep it hanging for a while, we'll negotiate it on forums.
While I'm for moving to mod, it's pretty much 50%/50% in playerbase.
It's bikeshedding and voting mostly, but sometimes the Wrong Thing is the way to go.

If we want an excuse to keep this in mainline, it would be that it has an iuse_actor, which would otherwise bitrot.

[mugling]

The concern I have is that this issue was settled in #15329 so it's not unreasonable that an implementation should follow. The fine details need working out but almost none of this discussion has been useful.

Many of the accounts both here and on the forum are only very recently created, in some cases in the last few days. The idea that at any time someone can register, ignore the previous consensus, and then proceed to derail a PR is problematic.

Online votes count for little as it is difficult to prevent sock puppetry. They also fuel bikeshedding. The two classical remedies are moderation and forking. Both approaches work because those that are bikeshedding actually have nothing concrete to contribute. The recent sharp downturn in productivity is a worry. If the situation persists it may be worth exploring the possibility of a super-experimental branch with periodic backports to master.

EDIT: Unlock subject to further posts being concise, on-topic and not previously repeated

[Coolthulhu]

Voting thread:
http://smf.cataclysmdda.com/index.php?topic=13275.0

Online votes count for little as it is difficult to prevent sock puppetry.

Still, voting allows those who don't have arguments to display opinion, without having to wrap them in a semblance of argument.
Also, it's less about voting on an issue, since development isn't a democracy, and more about giving the losing side a chance. If someone invents some compromise (unlikely, but still) or the fanbase really wants to keep the flaming weapons mainlined, we can consider some other option.
No need to force one correct option, even if it is objectively correct, for as long as it is only a minor thing. If flaming weapons didn't have an iuse_actor (and was done using the old horrible iuse functions, 2 per flaming weapon), I'd force this through and wouldn't look back, but nowadays it's just the issue of where the jsons are located.

As for sockpuppets - we could ask a forum moderator to list users. We could then check those for old timers and new accounts made after the poll.

[mugling]

If someone invents some compromise (unlikely, but still)

A better argument should always win so as above I reopened comments for that reason. Hopefully anything insightful isn't immediately crowded out.

As for sockpuppets - we could ask a forum moderator to list users

I'd argue only those with 100+ posts are eligible. Interestingly I wouldn't be eligible to vote according to that rule. It's hard to defend the idea that anyone can at any time register an account and suspend development to force decision by committee.

I think it may be time for a new release model. It's not by design but we are now effectively doing continuous integration of a rolling release. This actively encourages some of the recent problems (this and others). It may be more appropriate to consider a curated weekly build? For example we could have avoided the recent vitriol about the volume unit changes in that manner.

[Coolthulhu]

It's hard to defend the idea that anyone can at any time register an account and suspend development to force decision by committee.

It's not about anyone, but about a group of people.
I don't think anyone cares about flaming weapons enough to make 30 accounts just to shit up voting.

[mugling]

I don't think anyone cares about flaming weapons enough to make 30 accounts just to shit up voting.

No, I don't think that's likely, and that's not what I meant by:

It's hard to defend the idea that anyone can at any time register an account and suspend development to force decision by committee.

My worry is that the issue was discussed by the developers some time ago. Many of these accounts are from brand new users, in some cases new to both GH and the forum. We can't reopen discussion every time someone new joins.

So anyway, weekly builds?

[kilozombie]

Weekly builds from jenkins would probably ease a lot of current uproar on these things, if only for people that don't compile themselves.

It's not like the majority of people update any more than ~1 time a week, if that. Personally I only update once every few weeks or if something very interesting gets added.

[Chezzo]

I don't see the discussion in the issue #15329 you posted, @mugling.

[mugling]

Don't we have a mod for weapons like this that don't work in the real world but sound cool?

Could be resubmitted as part of a larger mod if there was sufficient interest?

There's a number of developers in that thread. The idea for a mod comes from the lead developer.

[BorkBorkGoesTheCode]

The only weapons I can see are the acid and scrambler grenades. If the acid grenade gets put in the mod, the lab challenge will need to be adjusted to be winnable.

And by the way, the TEC-9s in game aren't semi-automatic.

[mugling]

Series of unproductive posts moderated. If anyone has anything new to say about what should or shouldn't be in a mod then please do so at the forums and we can collect opinions in summary.

[illi-kun]

Re: weekly releases
I'm strongly against it because our current release model is reasonable: you have a choice between the fresh version with the latest features (but also with bugs) vs. stable bugs-free version. Yes, we didn't release the stable version for years but it's an issue which isn't solved by switch to a new release model.
BTW, could you explain the significant diffetence between experimental, super-experimental, and weekly builds? All of them are unstable, and we can't be sure we able to implement significant features and fix the related bugs in a week cycle.

BTW, afaik, @kevingranade doesn't like such discussions because, well, he is a manager and that's what managers decide.

[mugling]

Almost everyone is tracking experimental and each PR usually results in a separate build. This means that the userbase are all have slightly differing versions. This makes bugfixes very difficult as we first have to clarify which build is in use.

The null weather bug is a good example, it affected only a very few specific builds and was fixed quickly, but we still continue to get bug reports for it now. Often the first reply to an issue "please state your version as this may have already been fixed". You cannot blame the userbase for this - we asked for bug reports. Having less builds in the wild would help though, especially so if they were released on a predicatable cycle and we can assume everyone updates at roughly the same time.

The volume units are another good example. We've been patching at that ad-hoc since it was released. If we had a weekly release goal I suspect the approach would have been more structured.

Finally we pushed bad JSON yesterday that broke the build. The current model doesn't encourage thorough testing and as above each PR result in a separate build. A weekly build could more easily exploit the unit tests, --check-mods and other automated tooling.

I don't think productive discussions as to differing options are harmful in any sense. It's a fairly significant change and in any case there might be flaws in such an approach.

[mugling]

Needs rebase