All The Tropes Wiki Project

[OpMegs]

Spud and I've discussed this particular trope quite a bit over time, mostly because, as he's explained above, it's not exactly clear. Frankly, despite having watched Big O before getting into TV Tropes, I wouldn't have gotten the reference. Certainly, the number of people who would know immediately what was being referenced is a lot smaller than the initial trope namers would think.

There's a few cases where trope names were changed just because unless you were into a given fandom, the name made no sense. Takahashi Couple becoming Belligerent Sexual Tension is a big example, and the various -dere tropes were quite literally in another language unless you were an anime fan. TVTropes taught me the terms, but that doesn't mean it was any less opaque.

I suppose it really depends on what kind of crowd you're going for. If this is just going to be a Tropeflower for itinerant Pilgrims fleeing King Edward's tyrannical trope-censorship in the New World of Tropes, then keeping the old, more idiosyncratic names is fine. But if you want to draw in newbies and folks that haven't done this before, more inclusive methods will definitely help. This doesn't mean killing the old trope titles. One of the things I really liked about TV Tropes' system was the fact you could have multiple URL titles go to the same page. A lot of variations on one Trope can link to the main page, and a simple "This trope is also occasionally referred to as a 'Takahashi Couple', due to said prolific manga writer's tendency to make nearly all romantic couples in her works according to this archetype" both acknowledges the original (and explains it for anyone who runs into an older edit that has the original term) while still not making the base article particularly obfuscating to non-longstanding tropers.

Which is the superior option really depends on the wiki's basic policy and aims, however.
---
"Oh, silver blade, forged in the depths of the beyond. Heed my summons and purge those who stand in my way. Lay
waste."

[Labster]

The wiki's basic policy and aims will be to provide a place for people to identify patterns in all forms of media, without commercial considerations, and to share this knowledge with others.  As part of its educational goals, it should document writing tools as seen by both writers and fandom, and seek to explain them to the general public with clarity and humor.  The wiki will be free and open, with its content decided by the community.
-- ∇×V

[Morganite]

I just don't feel like redirects are that great a solution. You're still going to have every page showing the trope as one or the other, and you'll still get confusion in the end.

There is something that occurred to me that, if technically feasible, that might help. There's various ways to have some text appear when hovering over a link. How about using that to have a short-short description of a trope appear when hovering over one? (One would, of course, want the server to not be making a database pull for every link in a page, but I'm sure there's some way to handle that gracefully. And it's also probably a good idea to have this editable from the same place as editing a page.)

And for index pages, assuming you're going to have something that automatically generates them (which I think is a good idea), have them inserted next to the trope names. Which would make life easier for people in situations like "I know it's a magic trope... *looks at list* ... This could take a while."

Possibly also some other things this could come in handy for, but it depends on some other things.

And there is a question I've got - should tropes stay in the top level (or something that acts like it), or should they get a "Tropes" namespace?

-Morgan.

[Labster]

In terms of the hover/alt text for quick descriptions of tropes -- I had already been thinking about that idea.  Or trying to come up with some other mechanism to display the laconic of the trope, like a right-click, or something.  I will note that you have to do a db request for every single link anyway, to determine whether it's a red link or blue link.  It might be as simple as requesting another field when the parser runs -- but I'm not the one writing the software, so we might need to wait on this idea.
As for the index pages idea, that's already done.  We're thinking on the same wavelength here.  Each page will have a summary field, which is editable at the same time as the page, along with a tags field.  And then we have a macro to generate a list from the tags, and another that generates a tags list with the summary included.  Combine this system, and you have instant labeled indexing from a single edit.  (Assuming the index page already exists, of course.)  This does reverse the indexing system to be done at the page level again, but it's stored in the metadata rather than the page.
It also means that the Laconic descriptions can be moved into the page metadata as summaries as well.  I'll have to think about whether that can be done automatically.   Which, hidden as Laconic pages are currently, will likely be a massive improvement.
Unless someone has a concern about this sort of structure.
So far as the Tropes namespace -- I've thought about that one too, but I'm not sure what the advantage of it is.  Leave the root namespace empty except for redirects? It certainly can't be done automatically, given the massive amount of junk in Main/ currently, so a decision on this can wait.  What we can do is tag trope pages with "trope".  I'm not against it, but right now, I can't think of any arguments in favor of it.
-- ∇×V

[nylor]

Hmm. Would it be technically feasible to have tropes' names equipped with some kind of vote mechanism? Then the most popular name for the trope could reign. It could cause some squabbling, I suppose.
In technical terms, the 'Tomato in the Mirror' trope would have a unique database ID number - and that number could have several associated 'names'. And the call to build the page would look up the ID number, and select the highest voted name for that ID. The trouble would be in editing the pages - to avoid the editor having to know the trope's ID number, the editor would have to provide some way to select from existing tropes or possess the smarts to say 'Ah, he added the 'Wibbly Wobbly' Trope - that's an alternate name for trope 29, 'Tomato in the Mirror'.'
That's just brainstorming and may be far more headache than it's worth, though. For the initial build, the existing trope names should be fine.

[Morganite]

I will note that you have to do a db request for every single link anyway, to determine whether it's a red link or blue link.

What.

That strikes me as the sort of function that begs for some sort of consolidated lookup table. (As does the addition being suggested here, in my mind.)

Although I'm a bit wary of just using the Laconic entry and calling it good. A lot of the ones I see strike me as going for comedy through abbreviating things as much as possible. (And ones for series are usually even worse.) Not really ideal for the purpose at hand.

(snip stuff)

Unless someone has a concern about this sort of structure.

Sounds pretty good, although I've got my doubts about using Laconic entries as summaries. Enough doubts that the level of redundancy I'm engaging in at the moment feels sadly inadequate.

Well, there are a couple other things. As an example, take Love Tropes, because it has examples of both the things I'm thinking of, and not just because I'm a Love Freak. For the most part, it's a list. But there's also a bit of introductory text, and links at the side to other related indices. I think both of these are good things, and think it'd be a good idea to have a way to do those and have the auto-generated list get stuck in the appropriate place.

On the other side, there's things like the line for "Love Triangle" having the subtropes for specific types of love triangles listed under it. This also strikes me as good. I'm thinking this could be done by an expansion of the tag syntax - have a format for "In this list, as a child to this other entry".

So far as the Tropes namespace -- I've thought about that one too, but I'm not sure what the advantage of it is.  Leave the root namespace empty except for redirects? It certainly can't be done automatically, given the massive amount of junk in Main/ currently, so a decision on this can wait.  What we can do is tag trope pages with "trope".  I'm not against it, but right now, I can't think of any arguments in favor of it.

Well, there's something that occurred to me, but that I'm not sure is really a great idea or not. (Or even technologically feasible.) But I kind of don't like the way things seem to end up working with redirects and stuff right now. And then I thought, if there wasn't anything that lived in the root namespace, combined with summaries that can be pulled at need... you could automate the whole thing. If someone requests "/FortAwesome", and there's only one page named that, then it just redirects there, but if there's more than one, it automatically generates a disambiguation page asking if you want the anime Fort Awesome, the unrelated novel Fort Awesome, or the trope about suddenly being put in charge of an awesome fort.

nylor: I don't really like any arrangement for changing trope names that doesn't include a discussion process. For one thing, there's some reasons for keeping certain names that may not correlate with popularity among the people likely to be doing the voting.

-Morgan. Did I mention that a lot of Laconic entries don't make good summaries?

[Labster]

Okay, fine, a hash table lookup, not a db lookup.  Or perhaps a cache lookup.  But the software needs to look for each link anyway, so checking another slightly larger hash table is not going to add any significant complexity to the renderer.
Okay, Laconics don't make good summaries.  But are they, en masse, better or worse than an auto-generated summary from the first two lines of the article?  This is a question of degrees here -- we can always rewrite fullwise the ungoodthink later.  Er, I mean, revise the poorer entries.  Looking at about 100 entries at random, I only saw three that I wanted to exclude outright, and another couple that needed revised.  5% isn't too bad.  But this is definitely a YMMV sort of thing, because not many of the laconics are truly good either.  The question is -- is a blank box a better starting point for writing a summary than the Laconic pages?
Note: The autogenerated summary will be a blank box in the editor until someone writes a real summary.  And the laconic importer wouldn't take the "click here" lines, which cut down on the silliness substantially.
A index page will look something like this:
-->A quote, natch
Some description of the index
//// relatedpages
* ["An index"]
* ["Another Index"]
\\
[[labeledlist:love_tropes]
%% which expands to a bunch of index items
----
[- stadler and waldorf, etc.-]
So the only thing you lose is the ability to nest.  And ordering.  Maybe I'll think about a way to preserve that.
Automatic redirects?  Huh.  Maybe.
-- ∇×V

[nylor]

And I just want to add, for the record, that I don't mind oblique names at all. Like I said before, if I see an odd one (like "Tomato in the Mirror") - I'm more likely than not going to click the link to go to that trope's page, and thus, be enlightened. I would hope most readers would do the same, as they were likely driven to the site by feelings of curiosity in the first place.

[robkelk]

As long as I can find the tropes by their generally-accepted academic names (if such names exist) as well, the "oblique" names are okay...
--
Rob Kelk
"Governments have no right to question the loyalty of those who oppose
them. Adversaries remain citizens of the same state, common subjects of
the same sovereign, servants of the same law."

- Michael Ignatieff, addressing Stanford University in 2012

[Morganite]

Okay, fine, a hash table lookup, not a db lookup.  Or perhaps a cache lookup.  But the software needs to look for each link anyway, so checking another slightly larger hash table is not going to add any significant complexity to the renderer.

Oh, okay then. It's just that, having been lead to believe that db pulls are one of the most 'expensive' operations, I was pretty startled.

Okay, Laconics don't make good summaries.  But are they, en masse, better or worse than an auto-generated summary from the first two lines of the article?

Eh, kinda six of one, half a dozen of the other. Laconic'll give you injokes that only make sense if you've read the main article, grabbing the first two lines will sometimes give you nothing in particular because the begininning is just leadin to the real description. Though I feel like there's more of the former than the latter. I'd kind of hope whatever gets used, there'll be a tag to make it easier to go through and check ones with auto-generated summaries, because I'm thinking a lot will need it.

Also, there seem to be a good number of pages with no Laconic entry anyway.

If you can get something to keep nesting, I think that's a good idea.

-Morgan.

[Labster]

Database pulls aren't so expensive when you're doing them properly, and you have indexes on the fields in question. Indexes get lookups down to O(1+k/n), which is much faster than linear.. DB writes are the expensive operation (and even more so if the tables are indexed). But there are other tricks to speeding up data retrieval, like caching in memory as well.

Here's the thing. If Laconic is just for joking, I'm not sure if importing the pages is even worth it. This might be one of those "and nothing of value was lost" choices. Some of them are good summaries, most of them are passable summaries, and some of them are jokes, and a few are even good jokes. So what is it supposed to be?

VOTE

Given that the software will have a summary field, Laconic pages won't be needed for descriptive power. I see a few options here:

A1. Import the Laconic namespace as subpages, and use it for one-liner jokes. All summaries will start as autogenerated at first, and all pages will be tagged with 'needs_summary'.

A2. Use the existing Laconic entries as summaries where available, and tag all other pages with 'needs_summary'. Don't import Laconic as pages.

A3. As A2, but tag every page with 'needs_summary', so the laconic ones will be rewritten too.

A4. As A3, but go ahead and import the Laconic pages to subpages, too (so they'll be on both Page/Laconic and Page's summary).

A5. Don't import Laconic, autogenerate all page summaries, tag all pages 'needs_summary' (option Nuke Laconic from Orbit)

Again, in this context, "autogenerate" means "let the software take the first two lines of the article as a stand-in summary, until someone fills out the blank summary field". Also note that there are 11,302 laconic pages, which means ~1/15 of entries have a laconic -- the majority will be autogenerated in any case. And each option takes roughly the same amount of time to implement in the importer, so that's not an issue.

As for me, I'm leaning A3, and somewhat against A1 for now. But I'd like everyone else to vote or offer an opinion.
-- ∇×V

[Bob Schroeck]

A3 seems like the best course to me.
-- Bob
---------
Then the horns kicked in...
...and my shoes began to squeak.

[Morganite]

Given the range of stuff that it's used for, I'm not sure anyone ever had a solid idea of what Laconic was supposed to be.

Anyway, here's what I think, split out more...

-Import Laconic. Because I'm always for keeping stuff.

-Give every page -something- as a starting summary. Either use Laconic, and then auto-generate if no Laconic exists, or auto-generate everything.

-Tag every page as needs_summary. Because it's going to be a long job, and it might actually get done if people can tell what still needs doing.

From the way you said "the majority will be autogenerated in any case", I'm assuming the auto-generate fallback is already part of A2 through A4.

So, that gives me a final reply of A1 or A4, with a moderate preference for A1.

-Morgan.

[nylor]

I usually don't even look at the laconic pages, so I don't really have any strong feelings about them one way or the other. I would lean towards A3, I think - if there's a particularly good joke worth saving from the laconic page, it can be the first line of the series description, you know?

[Labster]

Sorry about being away from this thread for a while, but we're still working out the last few bugs in the software and the test import.  I was told yesterday that we'd have a site open for alpha testing today... and obviously that didn't happen.   You know, software development.  But we're in the home stretch.
I just noticed this over on TV Tropes:

TV Tropes by TV Tropes Foundation, LLC is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.
Permissions beyond the scope of this license may be available from thestaff@tvtropes.org.
Privacy Policy

Now IANAL, but:

3b. You may Distribute or Publicly Perform an Adaptation only under: (i) the terms of this License; (ii) a later version of this License with the same License Elements as this License; (iii) a Creative Commons jurisdiction license (either this or a later license version) that contains the same License Elements as this License (e.g., Attribution-NonCommercial-ShareAlike 3.0 US) ("Applicable License").
...
7.  Distribution
This License and the rights granted hereunder will
terminate automatically upon any breach by You of the
terms of this License. Individuals or entities who have
received Adaptations or Collections from You under this
License, however, will not have their licenses terminated
provided such individuals or entities remain in full
compliance with those licenses. Sections 1, 2, 5, 6, 7,
and 8 will survive any termination of this License.

Does that mean that they terminated the right to host their own content?  The use of that "may" word is tricky, but it still worries me, if only because it looks like they're paving the way to make a profit from their content.
-- ∇×V

[Sweno]

From my dealing with other open source software it is fairly common to have a Creative Commons license, and a separate non-CC license if you want to use the work in ways that wouldn't be allowable under the CC.

One of the big hooplas about CC 3.0 is how strongly it leans towards copy-left (the use of only as you picked out), cutting off the dual licensing that used to be common under CC v2

IANAL, but my understanding of the situation is that once they put it under CC 3.0 they gave up the ability to make any commercial licensing deals.

If it can be proven that they commercially licensed the content somehow, then it would take an actual lawyer to determine if they lost the ability to distribute the work at all.
-Terry
-----
"so listen up boy, or pornography starring your mother will be the second worst thing to happen to you today"
TF2: Spy

[robkelk]

Does that mean that they terminated the right to host their own content?  The use of that "may" word is tricky, but it still worries me, if only because it looks like they're paving the way to make a profit from their content.

No – what that license says is that nobody else is allowed to make a profit from their content. They can do whatever they want with their own stuff... assuming that all the contributers surrendered the rights to it when they contributed it.

(Fenspace is available under a similar license, except that we don't require "ShareAlike". That doesn't restrict my ability to make money off of my character "Noah Scott" if I so desire despite that character being quite fully integrated into Fenspace.)
--
Rob Kelk
"Governments have no right to question the loyalty of those who oppose
them. Adversaries remain citizens of the same state, common subjects of
the same sovereign, servants of the same law."

- Michael Ignatieff, addressing Stanford University in 2012

[Labster]

Yeah, but no one has surrendered their rights here.  Here's the entire barrier to entry for editing:

Being known on the wiki means that the software is aware of your wiki handle (user name). It reserves the use of your handle for you, and provides some services to known people that are not available to completely anonymous folks. Posting and editing are reserved for people 13 years of age or older. By posting or editing here, you are stating that you are at least 13 years old.
Handle:    
(spaces will be removed)
Password:    
(passwords are case sensitive)
Create Know-age!
You can also use this form to just set/reset your cookie on this machine, if you are already known. Put your handle and password in the boxes above and press:
Already Known. Just bake a new cookie.
Quick link to You know, that thing where ...
"Well, duh," you are thinking. "It's a user account." Not exactly. We don't bug you for an email address. Using this form is the entire process. We do need to put a cookie on your system, so you'll have to allow that. You will be logged in on the machine you are using now automatically, unless something happens to the cookie.

New works automatically come into copyright by their creators under the Berne Convention, under the control of their creators.  "As soon as a work is "fixed", that is, written or recorded on some physical medium, its author is automatically entitled to all copyrights in the work, and to any derivative works unless and until the author explicitly disclaims them, or until the copyright expires." (Wikipedia, CC-BY-SA).
TV Tropes is not the creator of the content on their site, nor have I granted them ownership of the content I personally wrote, so they have no legal authority to waive the rights under section 8 of the CC-BY-NC-SA license.  They cannot possibly offer more rights than the creator grants, without having first acquired the copyright from the creator.  One could assert other rights, like fair use, but that's obviously not coming into play here.
Honestly, I prefer the CC-By license -- here's the content, cite me, do whatever you want with it.  But since I can't just change the license from TV Tropes (and neither can they) without the written consent of everyone who ever edited, there's really only one possibility for the license on the new wiki.
-- ∇×V

[Labster]

Hi everyone,
First off, I'm an idiot.  They actually relicensed the TV Tropes wiki to CC-BY-NC-SA from CC-BY-SA.  So any earlier content, naturally, would be still under a commercial-okay license.  This actually presents a major fork in the road for us, too.  Our data set is still commercial-okay.  Should we set the non-commercial bit on our license as well?  I'm not a fan of restrictive licenses, as someone writing and publishing a book wouldn't be able to include content from one of the analysis pages, say.  However, this would also mean that we couldn't migrate any new content to our site, so it would be a hard fork.  (They'd actually be able to grab content from us, but I'm not sure that will be an issue.)  Anyone have thoughts on this?
Second, the site has moved up to top priority in our software developer's schedule, as Evan needs to finish the software for his work and for his social life (a LARP wiki), as well as our project.  Which is good, because it's not like I can afford to pay him or anything.
Third, I've changed my mind about importing the images, if only because I didn't want to leave a bunch of broken image markup around.  I downloaded the entire image set some time back, which is only ~2.5 gigabytes.  Of course, no one at TV Tropes actually noticed me doing so, so there goes the logic for banning me, lol.  The copyright on the images is more obscure, being a mix of fair use, the older creative commons license, and God knows what else.  Looks like I'll have to set up a system for DMCA takedown requests, and hope for the best.
-- ∇×V

[robkelk]

Anyone have thoughts on this?

SA is already restrictive - you're telling everybody down the line that their license has to be the same as yours. (Okay, TVTropes is saying that.)
--
Rob Kelk
"Governments have no right to question the loyalty of those who oppose
them. Adversaries remain citizens of the same state, common subjects of
the same sovereign, servants of the same law."

- Michael Ignatieff, addressing Stanford University in 2012

[nylor]

If you're an idiot, then I'm .. well, less well-off than that. I'm not fully conversant with the Creative Commons license options.  If I understand correctly..
Right now we're in the Commercial Okay option, which means ... we can only use TV Tropes content from before they changed licenses? Because we can't impose a less restrictive license than they did, yes? Or do I have that exactly backwards?

[Bob Schroeck]

If you keep the image set, you'll need to drop my avatar from it. The rights issue on it is ... murky. In short, Dave didn't give me full rights to it; my use here and on my website are okay; elsewhere is... probably not. Though Dave's no longer around on the Net to object, we should play it safe.

(They'd actually be able to grab content from us, but I'm not sure that will be an issue.)

They probably will, too -- which will simply be the most profound complement they could give a competitor.

I like the idea of keeping the copyright commercial-okay, partly so that academics who publish research which includes the wiki won't need to remove it, and partly because I like the idea of someone eventually doing a book form of the wiki.

Looks like I'll have to set up a system for DMCA takedown requests, and hope for the best.

In the interests of being far more creator-friendly than TVT, I was going to urge the new wiki have comprehensive contact information for this kind of thing (unlike the old TVT system of hiding the one contact address three page-levels down and not actually listing any policies anywhere). One thing that should differentiate ATT from TVT is its willingness to engage creators who contact ATT on both a friendly and an unfriendly basis.
-- Bob
---------
Then the horns kicked in...
...and my shoes began to squeak.

[Labster]

We're probably going to have a couple of phone numbers that people can call if they have concerns.  It doesn't mean we'll always agree with them, but having a real conversation with another human voice always calms issues on wikis in my experience.
Another policy that will change:  Explicitly listing copyrights, and asking editors to submit under our particular license.  Seriously, they don't have you agree to any statement, terms of service, or anything.  The grant of copyright is only implicit in the fact that you edited the site.   A mature website should never have these kind of huge legal holes.
Okay, a quick and unofficial summary of Creative Commons Licenses:

• CC0 (Zero): Essentially create a public domain resource, even in countries that don't have public domain.
• CC-By: Attribution only.  You can add any license on top of this, so long as the attribution condition remains.
• CC-By-NC: Attribution/Non-Commercial: You can relicense the content, but you cannot use the content or its derivatives for private inurement.
• CC-By-SA: Attrib./Share Alike.  You can only relicense the content under a similar or more restrictive license.  You can make money off the content, so long as you also make it available under the CC-By-SA license.
• GNU Free Documentation License (GFDL): Roughly the same terms as CC-By-SA.
  
• CC-By-NC-SA: The most restrictive license, content can only be relicensed under the same.  All derivative works must be noncommercial and must be made available under the same license.

You can always increase the restriction level of CC license on derivative works, but not the other way around until the copyright term expires, unless the content's creator otherwise grants a new permission on the content.
On an unrelated note, I found an old archive of Troper Tales in a usable format.  Should those be added back into the mix?  I don't really want to bring up old debates, but apparently I'm doing it anyway.  Does anyone have a strong opinion on Troper Tales?  Because I sure don't have much of an opinion.
-- ∇×V

[robkelk]

On an unrelated note, I found an old archive of Troper Tales in a usable format. Should those be added back into the mix? I don't really want to bring up old debates, but apparently I'm doing it anyway. Does anyone have a strong opinion on Troper Tales? Because I sure don't have much of an opinion.

Get the meat in first, worry about the fancies later.
--
Rob Kelk
"Governments have no right to question the loyalty of those who oppose
them. Adversaries remain citizens of the same state, common subjects of
the same sovereign, servants of the same law."

- Michael Ignatieff, addressing Stanford University in 2012

[Bob Schroeck]

What Rob said.
-- Bob
---------
Then the horns kicked in...
...and my shoes began to squeak.