Version 1.0 is content-complete

[blerkz]

As Spock said, the needs of the many outweigh the needs of the few. Some numbers to put things in perspective.

About 145,000 people are subscribed on Workshop to my most popular mod. I have about 4,000 unique downloads for the A18 version on Nexusmods, and that includes people that downloaded it before A19 came out. So if I don't update on Workshop, 97% of people need to manually dig through Steam's (terrible) workshop interface to find a new version of their favorite mods, and they have to remember what each mod was called to do so.
This is not a reasonable burden to put on the vast majority of players, just so the 3% that stick to an old version of Rimworld are not inconvenienced. That balance seems way off to me.

I'm not unsympathetic to people having their savegames broken (although my mods can be removed without breaking things) but it's far more reasonable for that 3% to spend the two minutes installing the old version from Nexusmods (available simply by clicking on a link on the mod page in-game) and the 97% can just play the enjoy the latest version of Rimworld with all of favorite mods conveniently updated.

We all wish Steam had proper version support but it doesn't, so we have to inconvenience some of the players one way or the other. Better it be a small percentage.

[Tynan]

..

The goal of doing non-destructive updates is to not break savegames. So yes, it's true that if we ignore that concern entirely, the time savings of reactivating mods versus re-subbing to mods wins out. Not sure why we'd just ignore that concern entirely, though. The many people who came to me furious about last time certainly didn't have the option of ignoring it.

Re-subbing to mods immediately after 1.0 release won't take "hours and hours". I individually checked all 100 mods on Workshop last night; it took 10 minutes. There just won't be that many mods in the early days.

Loss aversion: Non-destructive updates mean we never take away anyone's mods. Nothing is "taken away from you". People can keep playing with their old mods as long as they want on Beta 19. Not taking anything away is the whole point.

I'm not sure why you put "'lazy' modders" in quotes, Fluffy. I've never said such a thing and I never would.

Making local copies is an alternative that works for individuals, but there is no way to make 95% of players do that because there's no way to reach (e.g. spam) them. It's not an alternative, it's a marginal mitigation strategy. I'm taking about normal players, not the 5% who interact on forums.

I can't casually push a patch that pushes messages on people because it'll likely break a lot of mods immediately since the code doesn't compile in a very stable way. Too dangerous.

I'm not assuming I can reach every single modder. But I can reach the ones doing big complex mods that touch code in complex ways. Simple code mods and XML mods will probably all be cross-compatible without changes, so they don't need to be included. Even if it's not every modder, more is better than less.

Terms like "bullshit" are unnecessary, please be civil.

I'd love if there was a straightforward technical solution that would just solve anything, but it's not there so we have to work around imperfect technology. If it becomes necessary in future I hope to look at possibly implementing something better (or cajoling Valve into getting their shit together on the Workshop).

In any case, you're allowed to do what you want. As I said, I'm just putting the request out there; I'd be remiss if it didn't. Especially after last time, when players made it very clear that they were extremely unhappy with what happened. My impression is that a lot of modders weren't really aware the amount of rage that they caused with the in-place updates. You deserve to know, at least.

I am going to point people to NexusMods and the forums to recover old versions of mods that get taken off the Workshop. This will be a weak mitigation.

EDIT: As for mod lists, I'd expect them to be more useful with non-destructive updates. "Best B19 mods" list will thus work forever since the mods don't change. Currently what happens is the mod list ends up full of mixed-version mods (e.g. it's broken) until the author can update it. And if the author is updating it by checking each mod for its update state, the author can equally just make a "Best 1.0 mods" list alongside. That said, I haven't looked deeply into modlists so there may be a detail I'm missing here.

Numbers
There's been some talk of numbers, and the numbers are important so let's ground them in the data we have.

RimWorld peaks at about 10,000 players each day so I'd guess there's maybe 50,000-100,000 games being played right now. Some significant percentage are modded and that's how many will get permanently trashed by destructive updates. (A few will be able to recover from NexusMods/forums, etc, but there will be tens of thousands that are simply screwed.)

I guess the question is, how many tens of thousands need to be screwed over before this concern wins out? I can see several possible answers, but considering it in the whole I think non-destructive updates are the best recommendation.

Despite the slight inconvenience of re-subbing vs re-activating, it's a lot better to just not take away people's game or their mods. Put simply, totally screwing over a significant group of people to offer a minor convenience to a larger group is the wrong tradeoff IMO.

But, as I said, it's just a recommendation and an attempt to transmit feedback from players. Do with it as you will.

[YokoZar]

I'd love if there was a straightforward technical solution that would just solve anything, but it's not there so we have to work around imperfect technology. If it becomes necessary in future I hope to look at possibly implementing something better (or cajoling Valve into getting their shit together on the Workshop).

What's the particular reason why we can't have mods support more than one version simultaneously?  If the 0.19 and 1.0 versions were bundled into the same mod and a tiny amount of metadata told the game which one to load, wouldn't that make everyone happy?

[Tynan]

I'd love if there was a straightforward technical solution that would just solve anything, but it's not there so we have to work around imperfect technology. If it becomes necessary in future I hope to look at possibly implementing something better (or cajoling Valve into getting their shit together on the Workshop).

What's the particular reason why we can't have mods support more than one version simultaneously?  If the 0.19 and 1.0 versions were bundled into the same mod and a tiny amount of metadata told the game which one to load, wouldn't that make everyone happy?

There's no reason we can't, it just never came up as a feature request before today. It's in the list But a bit too late to solve this local issue.

[Fluffy (l2032)]

@All; I've made my thoughts known. I'm going to stop posting here as I'm getting overly frustrated, and it's not helping the discussion. I'll contribute directly in other - more productive - channels.

Tynan, sorry for the 'bullshit' remark; you know where it came from .

[Bobisme]

My 1 cent....

At present all mods are at B19? 
all mods are looking to update
you, right now have a copy of B19, why not keep that copy, uplaod it to workshop and then update your mod with the link to B19 in the 1.0 description. (or github* )
This may have been posted already, i don't see the issue
I mean you have B19 ALREADY and you ARE about to update to 1.0
Why not just link B19?
If someone wants to continue to play in say, B19 and they see 'mod not working', they go to your steam work shop mod, see the (go here for B19) and they're happy.
and the others that go with the flow get there uninterrupted latest modded game?

the game updates, they should expect this to happen when using mods.
ppl that use mods have enough sense to know that the mod works for a version and a new one is coming.
....Though, i bet a lot also pray a modder doesn't just hit *delete version*


Love the mods, Love the game

[Fluffy (l2032)]

sorry, have to comment on this, but that's kindof the worst of both worlds Bobisme; users still have to look for different versions, and modders still have to maintain multiple steam releases.

[Dingo]

I can't casually push a patch that pushes messages on people because it'll likely break a lot of mods immediately since the code doesn't compile in a very stable way. Too dangerous.

Make a mod that will get added to the Mods folder in a small update. Have code in it that presents a pop-up message in the main menu and lets players click a button to copy all mods. Zero assembly changes and you can remove that mod for 1.0 when you release.

If I came up with this solution as I was reading this line I have to assume you've not spent time considering alternatives.

[yuri12344]

~Not a modder (at least not on Rimworld), just a user of tons of mods.

I like the update in place argument, but it only breaks the scale for me by a tiny margin. The way I went from B17 > B19 was looking at my subscribed items on steam for Rimworld and then systematically going down the list, opening each mod page. This wasn't necessary for B19 labeled mods that were updated in place, but I did it anyway, just to read the latest comments and see if there were any game breaking issues with the update. This honestly needs to be done by anyone who is modding their game, so I don't see it as a time waste or an inconvenience. In that way, the only difference between in place and uploading to a new page is that I would have to open a new window, unsubscribe from the first and subscribe to the next (an extra 5-10 seconds). Some outdated mods that hadn't updated in place (or updated at all), I simply left myself subscribed to so I would have them on my in game modlist should they be updated in place, or so I could look back at them later and see if the author linked to a new version.

Now, as far as people still playing the old version... I play quite a lot on each of my playthroughs, but coming as someone who has played a ton of hours of modded games (Bethesda suite, etc) and being used to losing quite a lot of progress to silly game corruptions and the like -- it honestly doesn't affect me losing my time by my save being broke. I'll end up not playing Rimworld for a while anyway after the new update (hence my transition from B17 and not playing during B18 because I didn't want to take on the slog of finding updated mods). I considered coming back to the game several times, but being midway through B18's life cycle, I figured there would be a patch coming along any day to ruin the work spent putting everything together -- so I just waited for the news of B19's release to begin again. I see 1.0 being very similar for me.

I suppose this is where the counter argument to updating in place comes into play, since in order to play the game the way I want to (with certain mods) and not having to wait several weeks to a month first for updates, I would need to continue to play on B19, whilst making sure that update in place authors mods were accounted for each time I load the game. But honestly, it's not that big of a deal. When you load a modded game, it tells you if you don't have the mods required so:

1. I find out what is missing.
2. I go to the workshop page.
3. If the author has a link to the older version, I go there, manually download (or subscribe, depending on the author)
4. Keep playing.

If they don't keep the old version somewhere -- well, that sucks, but that's not what the mod authors in this thread are advocating for. They actively keep github links and old releases in their threads on this forum and it isn't hard to find these things with simple google searches.

(This is all assuming I don't back up all of my mods manually... which I've already done.)

Quite honestly, I feel that authors have put in quite enough of their time setting their mod up the first time. As a measure of the gratitude I feel toward them for their awesome content, I see no issue at all with having to do the minuscule amount of work to keep my own mod list tidy. (and my list is over 100 mods)

[erdrik]

...
Make a mod that will get added to the Mods folder in a small update. Have code in it that presents a pop-up message in the main menu and lets players click a button to copy all mods. Zero assembly changes and you can remove that mod for 1.0 when you release.

If I came up with this solution as I was reading this line I have to assume you've not spent time considering alternatives.

Or your solution won't work as well as you think it will. Not saying that is actually the case, but it is a much better assumption than insulting the Dev.

[RemingtonRyder]

Setting up a new mod on Workshop isn't really a chore. I don't know why people make it seem like it is. That being said, I'm used to it, so I can see why it would seem daunting to someone who is used to just updating in place.

There's a mod called PublisherPlus which improves the upload process, though I dunno if it works with 1.0 yet.

[bigheadzach]

Setting up a new mod on Workshop isn't really a chore. I don't know why people make it seem like it is. That being said, I'm used to it, so I can see why it would seem daunting to someone who is used to just updating in place.

In the past I can see how it would be, given that there are 19 different versions of the game all with their own idiosyncrasies and workarounds that mods must implement (and some mods made irrelevant by the updates, even).

At some point we need to draw the line and end support for Betas less than, say, 18, and tell people to just play the new game. I can't believe that enough people still play builds more than 2 versions ago, what with all the things that got brought into the game afterwards.

Reducing the chore of having to manage 10 versions to more like 2-3 seems like it's worth the hassle of having distinct Workshop items for each.

I mean, Microsoft had an entire media campaign that literally said "STOP USING IE 6" because it was wasting their support budget trying to deal with those folks.

Unless the claim is that nothing significant changes from B19 to 1.0 that a B19 mod would poop the sheets over. I'm sure plenty of modders are already running their B19 packs on the 1.0 build just to validate/refute that claim.

[RemingtonRyder]

Yeah. My approach to multiple versions on Workshop has been to hide some of the really old versions, which means that they aren't searchable, but they aren't immediately yanked from users' mod folders. Then after a while I can delete those old versions. If someone is still playing Alpha 15 or something, they can pick up a direct download of one of my mods on my site (which actually goes all the way back beyond the Steam release).

I also turn off notifications for old versions, because I need to focus on issues with new versions more the old.

This approach means that I still only have 48 items on Workshop (some of those are hidden). Of those I'm actively supporting maybe a handful, with an occasional glance at the older versions in case someone lands on the wrong version.

[Strawcave]

Just a player who often plays with mods here, but I can relate to what Tynan is saying - I have had saves break between versions because I forgot to backup and the prospect of hunting down all the right old versions through GitHub and forum posts didn't (for me) seem worth it since Randy might just kill me within 5 minutes of getting that save fixed anyway, and I actually enjoy starting over. That said, I think there are pros and cons to both of these arguments, but probably a bit of bias for most of the people here on the forum. Most of the people here probably understand modding enough to know to take certain steps to avoid this. But, for better or worse, steam workshop makes modding so 'easy' that a huge amount of players who otherwise probably wouldn't mod their games at all can just shop down a list of cool things and click the ones they want. I do myself have friends like this. And when the version changes and their mods get all effed up, and their understanding of mods doesn't go much farther than 'scroll, subscribe, and play' they don't understand what is going on and probably get really frustrated.

Whereas before in the old days of modding you had to have a bit of an understanding of what you were actually doing, steam has made it essentially 'mods for dummies', and honestly I can feel for both people here. On the one side modders are doing this unpaid and making some amazing stuff that makes everyone's game better, but the integration in steam workshop is seamless (but certainly not perfect) enough that mods can start to come across as almost part of the game, or 'toggleable expansions' in a way that when they break, someone who doesn't know better might dump their losses on Tynan. And we probably won't hear from them here, so I do feel like Tynan is doing his due diligence to give voice to a group that exists but isn't as participatory in the forum, the githubs etc.

Just my two cents!

[AileTheAlien]

Whereas before in the old days of modding you had to have a bit of an understanding
[...] 
steam has made it essentially 'mods for dummies'
[...]
mods can start to come across as almost part of the game

I've been wanting to share my thoughts, and this comment is about as good a time as any.

I don't feel like it's unreasonable, to expect people to understand that a mod is made by someone (or a group) who is not the official game developers. Steam shows that, for example, Expanded Prosthetics and Organ Engineering is created by somebody named Ykara. This is clearly different from Ludeon Studios, first just by the name being different, and secondly because when you click on a game developer in steam, it lists all the games they've made. Clicking on a Steam user brings you to a page which looks very different - their Steam user-profile. I don't expect people downloading mods via Steams easy integration to have technical know-how of the modding process or ecosystem, but I do expect them to understand that the modder is a person separate from the game developer / company. If a mod auto-updates and breaks a person's save-game, then the problem is between the modder and that person.

Paraphrasing someone earlier in this thread, nobody would expect the Ford motor company to fix a problem on a truck, caused by an after-market product like a winch!