Clarification on code flagged for safety review

If it’s my code, in my private game, I should be able to write whatever I want in my code.

When it comes to places that are accessible by others, there are still issues as said before.

  • I shouldn’t have to be scared to define variables under names that may be flagged even though they have everything to do with my game and nothing to do with someone’s personal information.
  • What about the privacy of things like API keys?
  • How can developers be expected to just trust this “specially-trained team” when we know nothing about them, and have consistently seen issues with moderation on other content in the past?
11 Likes

Based on Section 13 of the Roblox Community Rules, “Offsite website links, services, and additional 3rd party content are not permitted on Roblox with the following exceptions: YouTube, Twitter, Twitch.”

If developer code is subject to these restrictions, then any game using an external analytics service or web server is technically in violation of the Community Rules. In other words, many of the top games on the platform can immediately have their proprietary code sent for review and have potentially sensitive API keys and other data leaked. Will this happen? Probably not. Nevertheless, the possibility of moderation abuse is very high and this policy, in general, is rather alarming.

16 Likes

I’m sorry, but no. I think this is probably about as invasive as you can get. If I’m writing server-sided code in a non-team-create game then I fully expect to be the only one reading said code.

The increasingly ridiculous moderation on this platform is giving me an intense desire to move on to other things. Maybe I’ll try my hand at Unity or the Unreal Engine.

38 Likes

Looks like I’ve gotta censor myself in my own closed-source code. :woozy_face:

12 Likes

I 100% agree with @Maximum_ADHD and @Crazyman32 both. and a bunch of others.
But when it comes to my code, I don’t want to have to worry about moderation while writing. I already have enough to worry about when it comes to making sure my code works with my other modules and whatnot. – not that big of a deal, but would be quite annoying.

My biggest concern (which from what I’ve read is a lot of others concerns too) is security; Clone mentioned Roblox has access to all of our data anyways, I imagine that is slimmed down to those with high power within Roblox Cooperation(At least, I’d like to think so).

But when @Roblox trusts their developers so little that they have to add something like an automatic code checking system that could lead to flagging and causing a team within Roblox to check out our entire code system. To make sure we aren’t using discriminatory language or personally-identifiable info? Seriously? It’s not like I’m going to post my code on youtube and be like “Look what Roblox allows me to type in their studio!”

Hell, this may not even have to do with trust; Don’t get me wrong. Take safety precautions, and especially with all of the super shady developers which Roblox does have.

Possibly, if it’s really that much of a concern, may I recommend emailing the said developer and requesting access to their code or something of that sort. I don’t feel it’s right for something like this to be a thing honestly. I am the only one that ever sees my code (Outside of maybe teamcreate). In my opinion this should not be an issue. But obviously it is.

Don’t get me wrong, I don’t write anything discriminatory towards anyone or anything; but still, the Roblox chat filters things such as Butt… so, If I write the word Butt in my code will it be flagged?
but I have big plans for games here on Roblox, and If I get a banned or even a warning because I use something like

local Birthday = {Day = “5”, Month = “September”, Year = “1998”}

I honestly may take my programming elsewhere. I love Roblox, I have since I was 10 years old. Been here 11 years now. Roblox is the biggest impact I have had on my life and I’d hate to leave over something like this. It would be on me, right; But I don’t want to feel worried that I may get banned about something as little as things (again) within my code that only I can see.

I’ve said what I said. Agree, disagree; Whatever. I feel this is wrong.

Edit:
Just think about it; What if someone within the “specially-trained team” decides to take advantage of this situation, make a fake account, take said code and upload it to a game to profit off of. This is easily a possibility. Plus not only that, Our code could easily be sold or posted which again is a HUGE security risk. What happens if our code is out? It would be MUCH easier for those who write exploits to actually make exploits for said game. As they stated…

What if one of their mistakes were hiring a rat?

20 Likes

Is this real? Who at Roblox thought it was a good idea to moderate code that only the developers themselves should be reading, and potentially exploiters. Aside from the obvious security threat, this is a privacy threat. I do not want someone who is likely just an off-site moderator being paid close to minimum wage being able to read my code, and see any private API keys, etc.

If you want to moderate a game, how about you try joining the game to see if there is explicit content.

Roblox should not be playing the role of the thought police.

Edit:
Automated flagging? Are you kidding? So if I curse in my private code in one of the comments, that gives some random moderator permission to come read everything. This is a joke.

20 Likes

You should not be making your own filter in the first place. Roblox has API’s for this so you don’t need to.

3 Likes

I disagree. The roblox filter is known for being bypassed constantly by the most stupid word combinations. Remember a few months ago when you could say the n-word within the chat by saying “jew” before it?

4 Likes

inb4 this is going to be like youtube’s moderation system where you can upload gore but you can’t swear

Considering that the current moderation system can’t catch Synapse Xen obfuscation for months at a time, that’s probably what’s going to happen. We can debate privacy all we want, but it’s clear that this moderation system won’t work at all.

Will I be flagged for writing something like this?

local t = {}
t.__index = t
t.__metatable = "sal didn't use getrawmetatable making him tonight's biggest loser"

Probably. Will I be flagged for Synapse Xen obfuscation, though? Not until someone reports it, as clearly shown by the hundreds of backdoors currently in the Roblox library. Even after hundreds of revisions, you can’t make a perfect system, and “quickly” is never quick enough when your game is well-known.

I would consider moving to Unity or Unreal if it weren’t for the fact that Roblox is much easier to advertise on.

11 Likes

I intensely disagree with this code-review system, but the original post did say this:

5 Likes

There is a problem though, it is now 100% required to use the filtering api on all types of communication and player-generated strings. Would moderators take action against the creator for code that, at the time of it’s creation, followed all rules? (Which is something that already happens with images and sounds.)

3 Likes

This was always a requirement far before this system was implemented and I can find and send you posts proving this if you don’t believe me.

Yup, you get banned for things that were previously allowed. I have a massive post ranting about this.

2 Likes

You mean having some kind of filtering system, not using roblox’s api, right? Since (at least as far as I’m aware) it wasn’t implemented until sometime in 2016, correct me if I’m wrong.

2016 is far before this system was implemented. This code-review has only existed for a few months. The word “always” in my statement may be the confusing part, but my focus was more on the “before this system was implemented.”

Edit: I mistakenly read your “it is now 100% required” as referring to this review forcing you to do this.

1 Like

Not the “safety review” for scripts, the filtering API. Sorry if I was unclear.

2 Likes

I’m really unhappy that this even exists to begin with. My code is my code, whether or not I include anything technically against TOS. Just because my code might be considered “against TOS” does not mean anything to a player of my games. I am the developer of my game, whether or not my code contains innapropriate language, “personally” identifiable information, or anything else, nobody is supposed to see my code. This code is only for me to see.

This is equivalent to saying players who type inappropriate content into text boxes are apt to be banned. Nobody is seeing what they type, that’s personal to them in the same way nobody is seeing my code. Are they intentionally doing it? Probably. Does that make it unsafe for them to type inappropriate content in a text box? Well depending on context this could mean literally anything. With something both as abstract and literal as code is, it’s utterly impossible to judge a developer’s intentions on their code. Having moderators review something they may have no knowledge of is completely ridiculous.

Finally, Roblox cannot track free model usage effectively. Users will start targeting games by including potentially inappropriate content in their models. These models can be reviewed, but the entire basis of this form of botting is that moderators can never instantly review anything. Developers could make slight adjustments to code, and there’d be no way to know if they see something innapropriate.

This is not okay in my book. There are too many issues, and it’s laughably unclear what any of this could mean. How can anyone rely on Roblox as a platform with this type of potentially nonsensical moderation? How is Roblox even going to clarify the amount of issues here without locking their moderation in place? Surely Roblox is aware of the issues with this to begin with, but how Roblox plans to make this form of moderation reliable for anyone is completely unclear and I feel that this just cannot be clarified.

35 Likes

Hey guys, on the bright side, now you have a way to get Roblox’s attention when they’re ignoring our feedback.

Write a violent manifesto and sprinkle in curses so it gets flagged, and have a message at the bottom. They’ll finally read it then!

/s

42 Likes

Who thought this was a good idea? Roblox’s automated moderation system is bad enough as it is, why make scripts another potential cause for a random ban? More focus should be put towards improving the moderation system itself rather than adding more onto the list of horrendous flaws it has. It’s already nearly impossible to have a conversation with someone else on Roblox without any of my messages getting filtered when none of them contained anything explicit to begin with. If this “specially-trained team” is the same group of people – if they even are people – responsible for the multiple unjustified bans/warnings that my friends have received (one of them got their account locked for 3 days for putting a single swear word in comment in a game with tens of thousands of lines of code!) then this system is an absolutely horrible idea. As you can see from nearly 200 of the previous posts, it’s already received a lot of community backfire. Now what is the community response going to be like when this system gets fully implemented?

Oh, and this is also not to mention that this system feels like a gigantic breach of privacy. My code is for my own eyes only, and I would rather not have some intern at Roblox combing it for bad words.

11 Likes

I didn’t even think about this. This is a gloriously ridiculous way to moderate games I feel. You’ll either end up with inconsistent, unclear moderation that effects everyone on a mass scale thus flooding moderation, or you get ridiculously slow moderation that isn’t effective and is easily abusable by bots. This doesn’t just pose huge problems in terms of clarity, it creates an incentive to include bad content for the people who even have bad intentions to begin with. Sure it makes sense on a basic level, but this is just such an unpredictable way to moderate for developers, or Roblox themselves that it’s wasting resources and it’s creating a gateway for even worse behaviour and impact on the platform.

12 Likes

So if it passes moderation, Would the system just keep flagging it, or Flag it every time a edit is made?