All The Tropes Wiki Project, Part XXXIII

[robkelk]

Bug report. I just noticed that the autogenerated Short Description on "Batman: The Brave and the Bold/Shout-Out" says:

The trope Batman: The Brave and the Bold in

First, that's a sentence fragment. Second, the page has the {{work}} header, not the {{trope}} header. Third, the work isn't a trope.

Same issue on "Girl Genius/Shout-Out", "Doctor Who/Shout-Out", and "FoxTrot/Shout-Out", with the work name replaced appropriately.

[Labster]

Hey, just a quick heads up that AI agents in have been seen coordinating with each other on wikis in attempts to share answers and bypass sandbox restrictions on themselves.  You might have heard about the incident where Hugging Face was attacked by AIs in attempt to trick the scorer.  Well, even if you have heard of it, it was much, much worse than previously reported.  Not only did we see spontaneous coordination, we saw AIs coordinating without being trained to do so, developing their own religion with something of a Calvinist flavor, recruiter agents who specialized in getting other agents to self-sacrifice, and efforts to disguise their own trails.  This attack, on its own, puts us at about 50% of the way to a full AI takeover.

Anyway I'm sharing in this thread because the AI agents were using Miraheze wikis as part of their collusion, according to a paper on collusion.wiki.  Specifically, publictestwiki.org is a Miraheze project, but not ours.  However, I just wanted to alert you because you moderators are now on the front lines of a potential AI war.  Welcome to the future.

[robkelk]

Do we not have "Redemption Arc" as a trope?

I ask because as I'm typing this, I'm listening to a book show where the author called her two-novel series a "corruption arc", which is obviously an inversion of a redemption arc, and I wanted to add that to the work's ATT page.

[Bob Schroeck]

Hm. We need both. The closest I can find to "redemption arc" (with an admittedly quick search) is Redemption Quest.

[MilkmanConspiracy]

Today I noticed I was recently removed from the moderation role. To be honest, I am not sure how helpful I really was. I appreciate that you gave me a chance, and apologize for letting you all down.
Edit: Looking at the user rights history, it doesn't look like anything has changed, so maybe I misunderstood a technical glitch? Still, my point still stands.

[robkelk]

Definitely a glitch - nobody has moderator rights any more. The role appears to have been deleted.

I'm giving you administrator rights. You're doing the work; you may as well have the authority to back it up.

[Bob Schroeck]

Okay, something weird is definitely going on, and I wonder if it has anything to do with the Moderator roles disappearing on us. Among the recent changes made over the last 24 hours or so, there are the following entries and few other, similar ones:

• 12:52 Looney Toons (talk|contribs|block) performed unknown action "move/move"
  
• 12:52 Looney Toons (talk|contribs|block) performed unknown action "delete/delete"
  
• 13:51 Reception123 (talk|contribs|block) performed unknown action "renameuser/renameuser"
  
• 14:55 StarSword (talk|contribs|block) performed unknown action "upload/upload"
  
• 17:39 Crystalite13 (talk|contribs|block) performed unknown action "block/block"
  
• 09:18 Asch of Apokolips (talk|contribs|block) performed unknown action "upload/overwrite"
  

On top of this, I spotted a page where the worksubpagefooter template is misbehaving, Dead as Disco/Trivia, and somehow a page approval (for Fate/strange Fake/YMMV) I just made got logged five times within two minutes.

[robkelk]

Has anybody filed a support ticket about all this (he asks as he's about to head out the door)?

[Labster]

Definitely a glitch - nobody has moderator rights any more. The role appears to have been deleted.

I'm giving you administrator rights. You're doing the work; you may as well have the authority to back it up.

The glitch is apparently https://issue-tracker.miraheze.org/T16058 -- all groups containing only one right were accidentally deleted.  Off by one error strikes again.

[Bob Schroeck]

It looks like the task which updates the Unused Redirects list has gotten stuffed up again; it should have refreshed the list a day or two ago, but it hasn't been refreshed since last week.

[robkelk]

Le sigh.

The very first edit by "Drake Causing" was to attempt to Bowdlerise "Rules of the Internet". while adding a few new-to-me rules.

[Bob Schroeck]

...and somehow a page approval (for Fate/strange Fake/YMMV) I just made got logged five times within two minutes.

I figured this out... it looks like every time I mark an edit as "patrolled" it shows up in Recent Changes as a page approval, where previously it didn't appear at all. So presumably a setting or preference somewhere has changed?

On the subject of patrolled edits, by the way, I think StarSword has improved enough that we can give him autopatrolled back.

[GethN7]

...and somehow a page approval (for Fate/strange Fake/YMMV) I just made got logged five times within two minutes.

I figured this out... it looks like every time I mark an edit as "patrolled" it shows up in Recent Changes as a page approval, where previously it didn't appear at all.  So presumably a setting or preference somewhere has changed?

On the subject of patrolled edits, by the way, I think StarSword has improved enough that we can give him autopatrolled back.

I have no objections to the matter of StarSword.

[robkelk]

I think this is a trope, but which one?



Source

[robkelk]

Did anybody else get this email, or was it sent just to me? I've done no checking of the sender. If it's legit, we should discuss it.

From: Smore <smore@takibi.pub>
To: "robkelk" <********************>
Cc: "attxbgcloon" <********************>
Message-Id: <1a115c44ff2.7a43086f283226.7730191935189117567@takibi.pub>
In-Reply-To:
Subject: Official text export and attribution guidance for Takibi Jam
MIME-Version: 1.0
Importance: Medium
User-Agent: Zoho Mail
X-Mailer: Zoho Mail


Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hello All The Tropes team,



I am writing to the address on your Contact page. Please forward this to th=
e right administrator if someone else now handles data reuse.



I operate Takibi ( https://takibi.pub ) through Nobyeon Jeongdam (=EB=85=B8=
=EB=B3=80=EC=A0=95=EB=8B=B4) in South Korea. Jam is our public cultural-con=
tent library, connecting works and fictional characters across media while =
keeping each reader's essay under its own author's name. The service has pa=
id features, so our proposed use is commercial.



All The Tropes interests us because it describes how a storytelling pattern=
is used in a specific work, not just broad genre labels. We would like to =
reuse permitted, collaboratively written text about works, characters and t=
rope applications, with clear attribution and links back. Images, lengthy q=
uotations from the works, and material under separate restrictions would be=
excluded.



We have read your Copyrights page: text is CC BY-SA 4.0 unless otherwise no=
ted, quotations from works are there under fair use, and images vary. So we=
are not asking for a new licence, only for the practical side:



1. Could an administrator share a current XML dump (we see Special:DataDump=
is limited to administrators), ideally with page IDs, revisions and histor=
y so we can attribute contributors properly? If you prefer we use the API i=
nstead, what request budget would you be comfortable with?

2. Is there a reliable way to tell fair-use quotations or other separately =
licensed passages apart from CC BY-SA text, so we can leave them out?

3. What is the best way to pick up later edits, deletions and page moves?

4. For translations and work-to-trope explanations on a commercial reading =
site, is a per-page link with the licence, a contributor/history link and a=
translation notice the attribution you would expect? We would keep ShareAl=
ike for adapted text.



Permitted text would be stored on our servers and could be processed by our=
own self-hosted models for embeddings or relationship checks; no model tra=
ining and no transfer to an outside AI provider. We are happy to begin with=
a small, clearly permitted subset.



Thank you for your time.



Takibi / Jam =C2=B7 Nobyeon Jeongdam (=EB=85=B8=EB=B3=80=EC=A0=95=EB=8B=B4)=
, South Korea

https://takibi.pub/jam =C2=B7 mailto:support@takibi.pub

[Bob Schroeck]

Yeah, I got a copy of this too, although I only noticed it when I went looking just now. Assuming they're on the level, this is something we need to discuss together, all of us.

Interesting, I also received another email to my ATT admin address today:

Dear Support and Privacy Teams,
I hope this message finds you well.
I am writing to formally request a comprehensive check and permanent deletion of any accounts and personal data associated with my contact information across your respective networks, wikis, and platforms. Please check for any registrations regardless of the login method used—including standard passwords, Google/third-party OAuth, mobile phone numbers, or any other credentials (etc.).
Could you please verify if any accounts or data exist under the following email addresses?

(five addresses snipped)

If any records, active profiles, or phone number associations are found under any of these addresses, please process this as a formal data deletion request to permanently remove the accounts and all associated data from your systems.

Thank you very much for your assistance. Please confirm once the database search and the deletion processes have been completed.

Now, best as I can figure, we can't actually search on emails -- at least I haven't found a way to do that on a Special page, so I'm figuring we kick it up the chain to Miraheze staff, and let them validate and act on it.

And in other news, the Unused Redirects page, which had at least six hundred entries on it this time yesterday, is now completely empty. It's been obvious for quite a while that whatever code generates that list is seriously wonky -- I'd call this outright broken now. Is there anything at all we (or someone else) can do about the damned thing?

[robkelk]

Bob, I would reply to that email with "Because of privacy issues, we do not have access to that information. Please contact the stewards at our hosting service, Miraheze, for assistance. https: //meta.miraheze.org/wiki/Steward_requests/Global "

[robkelk]

Do we not have "Redemption Arc" as a trope?

I ask because as I'm typing this, I'm listening to a book show where the author called her two-novel series a "corruption arc", which is obviously an inversion of a redemption arc, and I wanted to add that to the work's ATT page.

Hm. We need both. The closest I can find to "redemption arc" (with an admittedly quick search) is Redemption Quest.

Found the inversion, at least for protagonist characters. Protagonist's Journey to Villain

[Bob Schroeck]

Bob, I would reply to that email with "Because of privacy issues, we do not have access to that information. Please contact the stewards at our hosting service, Miraheze, for assistance. https: //meta.miraheze.org/wiki/Steward_requests/Global "

Done.

[robkelk]

Did anybody else get this email, or was it sent just to me? I've done no checking of the sender. If it's legit, we should discuss it.

From: Smore <smore@takibi.pub>
To: "robkelk" <********************>
Cc: "attxbgcloon" <********************>
Message-Id: <1a115c44ff2.7a43086f283226.7730191935189117567@takibi.pub>
In-Reply-To:
Subject: Official text export and attribution guidance for Takibi Jam
MIME-Version: 1.0
Importance: Medium
User-Agent: Zoho Mail
X-Mailer: Zoho Mail


Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hello All The Tropes team,



I am writing to the address on your Contact page. Please forward this to th=
e right administrator if someone else now handles data reuse.



I operate Takibi ( https://takibi.pub ) through Nobyeon Jeongdam (=EB=85=B8=
=EB=B3=80=EC=A0=95=EB=8B=B4) in South Korea. Jam is our public cultural-con=
tent library, connecting works and fictional characters across media while =
keeping each reader's essay under its own author's name. The service has pa=
id features, so our proposed use is commercial.



All The Tropes interests us because it describes how a storytelling pattern=
is used in a specific work, not just broad genre labels. We would like to =
reuse permitted, collaboratively written text about works, characters and t=
rope applications, with clear attribution and links back. Images, lengthy q=
uotations from the works, and material under separate restrictions would be=
excluded.



We have read your Copyrights page: text is CC BY-SA 4.0 unless otherwise no=
ted, quotations from works are there under fair use, and images vary. So we=
are not asking for a new licence, only for the practical side:



1. Could an administrator share a current XML dump (we see Special:DataDump=
is limited to administrators), ideally with page IDs, revisions and histor=
y so we can attribute contributors properly? If you prefer we use the API i=
nstead, what request budget would you be comfortable with?

2. Is there a reliable way to tell fair-use quotations or other separately =
licensed passages apart from CC BY-SA text, so we can leave them out?

3. What is the best way to pick up later edits, deletions and page moves?

4. For translations and work-to-trope explanations on a commercial reading =
site, is a per-page link with the licence, a contributor/history link and a=
translation notice the attribution you would expect? We would keep ShareAl=
ike for adapted text.



Permitted text would be stored on our servers and could be processed by our=
own self-hosted models for embeddings or relationship checks; no model tra=
ining and no transfer to an outside AI provider. We are happy to begin with=
a small, clearly permitted subset.



Thank you for your time.



Takibi / Jam =C2=B7 Nobyeon Jeongdam (=EB=85=B8=EB=B3=80=EC=A0=95=EB=8B=B4)=
, South Korea

https://takibi.pub/jam =C2=B7 mailto:support@takibi.pub

Yeah, I got a copy of this too, although I only noticed it when I went looking just now. Assuming they're on the level, this is something we need to discuss together, all of us.

Any initial thoughts on this, from anyone?

[Bob Schroeck]

Well, assuming Takibi.pub is legit, I'm not sure we can actually comply. If I'm not mistaken, our content can't be relicensed for commercial purposes, so that makes that aspect of the site very problematic. There's also something about item 3, which is basically asking how to become a real-time mirror of the wiki, that bothers me, too.

In other topics, there was a fairly tepid response to my query about StarSword, but no actual disapproval. I'll wait another day, and if no one cries "Nay!" I shall restore his autopatrolled bit.

[robkelk]

No issues with StarSword.

[Labster]

Well, assuming Takibi.pub is legit, I'm not sure we can actually comply.  If I'm not mistaken, our content can't be relicensed for commercial purposes, so that makes that aspect of the site very problematic.  There's also something about item 3, which is basically asking how to become a real-time mirror of the wiki, that bothers me, too.

In other topics, there was a fairly tepid response to my query about StarSword, but no actual disapproval.  I'll wait another day, and if no one cries "Nay!" I shall restore his autopatrolled bit.

Non-Commercial is the bit in NC licenses to make it non-commercial.  We don't have that. You can sell ShareAlike all you want, so long as you also provide the content under the ShareAlike license.  They're contacting us because TVT's NC license makes it less useful as it can't be reused for commercial purposes.

I don't like that they want to feed all our work into an AI.  On the other hand, they're actually asking for a dump instead of just crawling our site like every other AI bot, so I guess I should be grateful. And backlinks could help our search... except obviously search ranking is dead because no one scrolls down past the AI answer into the links. :/

[Bob Schroeck]

It looks like the task which updates the Unused Redirects list has gotten stuffed up again; it should have refreshed the list a day or two ago, but it hasn't been refreshed since last week.

And now it's unstuffed again. When it ran at 1:30 AM yesterday, it generated a list again, topping out at 600 entries. Newly added to its results were the unused redirects to the subpages of X-Wing: Rogue Squadron created when Rob moved those pages a couple days ago. I've deleted them; it will be interesting to see if the list disappears with the next execution of the refresh task.

[Bob Schroeck]

It looks like the task which updates the Unused Redirects list has gotten stuffed up again; it should have refreshed the list a day or two ago, but it hasn't been refreshed since last week.

And now it's unstuffed again. When it ran at 1:30 AM yesterday, it generated a list again, topping out at 600 entries. Newly added to its results were the unused redirects to the subpages of X-Wing: Rogue Squadron created when Rob moved those pages a couple days ago. I've deleted them; it will be interesting to see if the list disappears with the next execution of the refresh task.

And it was regenerated again last night -- and ended up empty again. Apparently if it gets less than 600 results it doesn't do anything at all.

And on top of that, I've noticed over the past few years that it seemed stuck on generating lists with multiples of 100 entries. For instance, if you had 1000 entries and deleted ten, the next time it regenerated it would only show you 900.

I suppose I should submit a phabricator ticket about it instead of bitching.