TV Tropes 2.0: Database level redesign (Not in active development yet)

"Also, I think letting people put comments under the trope pages is an awful idea."

Seconded. This is a wiki encyclopedia, not a blog or a news site. Commenting will clutter up the trope pages, greatly lengthening their load time. It'll also encourage drive-by trolls and Single-Issue Wonk types (who tend not to sign up for forums) and thus require a lot of extra modding. And let's not even get into spam...

If people want to comment on tropes, that's what the forums are for.

Please. No commenting feature.

"Way off in the future" being this Monday at noon, I guess.

Or whatever the countdown on the main page is for. What is it for, anyhow?

edited 4th Feb '16 7:57:09 PM by Aldo930

That comments section is going to be such a nightmare. And there's still no bulletin board for news.

I only saw the commenting on the main page, not on any of the tropes. Can we get some more detail on what's up with that? It seemed to be some sort of social media thing, not something that's going to interfere with the rest of the site.

Honestly, as long as the core of the trope pages remains the same, I really don't mind the interface that much. I do think that adding the comments in general is a bit much, but other than that I figure it'll annoy me for a week or two and then I'll get distracted by reading trope lists or whatever again.

Decided to check a random page I've edited recently (Kingdom Hearts II), and the version on the new site is from over a month ago.

I'm assuming it'll be fixed in the transition?

edited 4th Feb '16 8:50:37 PM by Karxrida

I have to agree with others, in that I don't like the idea of letting people comment on trope (and possibly work) pages. Everything else looks okay so far.

Okay, my first reaction is a kneejerk Uncanny Valley sort of a thing. So ignore that.

When I look at it a bit longer the design actually pretty smooth and easy on the eyes. I like the blue v2 version a lot and I'm sad to see it go, but this one's not half bad and the Darth and Sugar Wikis are a vast improvement.

• The round buttons for the subpages look more cluttered than the current, squareish ones - it's okay when there's only a few, but in big works with twenty-odd subpages it's even harder to find the one you want. (Squarer buttons also take up less room, which again is noticeable on the big pages.) I like the icons, though, very pretty and finally coherent, and thank the gods the page you're on is actually selected.
• The sidebar is a bit too wide for me. Again, very subjective - but on my smallish screen it shunts the content either far to the right or far to the left. Could be one of those things you have to get used to. The main problem is that once you scroll down and run out of sidebar, you just have this huge blank space on one side (about 1/4, even in wide mode) and it feels pretty unbalanced. [1]◊
  Like sitting in a waiting room where all the paintings are hung just ever so slightly crooked.
• Basically what AliceMacher said. Comments under the trope depictions on the main page is asking for trouble, given how well-known TV Tropes has become. Maybe keep comments hidden by default, possible to toggle in the profile preferences?
  • On the other hand having trope depictions on the main page looks pretty nice. Might actually induce me to stay on the main page for more than the milliseconds it takes me to find the search bar.

Other notes:

• No preview button on the edit page? Admittedly I'm being hypocritical here, since I usually forget to use it.
  • holy shit the help option is beautiful
• The search bar on the search page (not the one in the header) shears the bottom off p/q/etc.
• Wide load leaves no whitespace to the left of the tvtropes logo and the right of your user icon, so both are jammed against the edge of the page.

edited 4th Feb '16 8:57:48 PM by LongLiveHumour

The database is old on Dragon, but when it goes live the code will be applied to the current database.

edited 4th Feb '16 8:51:27 PM by Fighteer

Looks good, then.

Only complaint is the aforementioned comments and the inability to left-align the toolbar.

You can log in and left-align from your profile, just as in the current version.

edited 4th Feb '16 8:54:45 PM by LongLiveHumour

Oh, sweet.

No real complaints, then.

Compact is good.

That's the main issue I have with site design in general. Things taking up needless space is annoying. As much of the space as possible should be used for the actual content of the pages. Interface doesn't really matter that much, as long as you can clearly read it.

If there's a search bar or menu bar or anything like that that follows the scrolling, it needs to be closable (if that's what the Sticky Header option is, good work). There are some cases where even a large one fits well (such as Dictionary.com), but TV Tropes isn't one of those pages.

Anything other than maximising the actual content is secondary as far as a reader is concerned.

On a smaller note, I like that lighter blue colour for links. The current darker colour sometimes blends in with black text if light conditions in the room are bad.

It looks pretty good to me so far. Nothing's glaringly wrong that has me thinking it should be changed right off the bat.

Good Stuff

The sticky header option. Having the Index category be a major category along with Work and Trope. It doesn't look like comments are enabled on work/trope pages. Please let it stay this way.

Bugs

In night-view mode, (my default browsing mode) the folder icons on folder headings look like the icon. In night-view mode, the section headers (that use the "!" markup) background color doesn't change even though the text changes color so it looks white text on very light grey background.

That's all a cursory glance gave me at the moment...

edited 4th Feb '16 11:11:59 PM by ObsidianFire

Will the older site design still be available, somehow? It'd be a nice option.

edited 5th Feb '16 12:22:46 AM by YasminPerry

This thread is for discussing the database overhaul, which is not the same thing as the DRAGON overhaul. Please start a new thread to discuss the DRAGON overhaul or use the DRAGON forum (the login form is tropes+kickstarter105 on http://dragon.tvtropes.org/pmwiki/topics.php for this, thanks!

edited 5th Feb '16 1:12:15 AM by drewski

Done.

So could anyone explain to me what exactly the countdown on the main page is for? This is really confusing.

I am guessing that they plan to deploy 3.0 (Wow. Talk about a misleading name choice there...) in that time.

Finally, I can use the site again. For some reason, I couldn't access it at all.

Database hiccup. The countdown is when Dragon (aka TV Tropes 1.3) will go live.

Are there any new functions coming to the forums?

I don't believe so. This thread is a better place to discuss all things Dragon - https://tvtropes.org/pmwiki/posts.php?discussion=14546627510A75066000&page=1#11

Just read over the whole thing. I have 1 long item re: item hierarchy and 7 other data-base related items. I also came up with 5 sidebars, but this probably isn't the place.^note For those wondering: Restrict Page-voting to those who have been Known for a set period, How TRS tickets are dispensed, an "Image Launcher" idea for new possible pictures, How "set" committees are a bad idea for content-related decisions such as images, and that the Forums should also be hierarchical by sub-topic for these referencing ideas to work.

1. I'd put in intermediate levels between Overall Franchise and Work (to cover subfranchises) and between Work and Episode:

• Franchise: The overall brand of a group of works
• Subfranchise: An intermediate level containing works having a more common connection than merely being part of the overall franchise. EDIT: Continuities would be classed at this level, though a single continuity could also be a group of Subfranchises, the two Mega Man timelines for example.
• Work: An individual Series/Volume/Game/Movie/etc. i.e. the Work itself. Medium would be defined at this level.
• Arc: A grouping of episodes within a work. Usually time-based (i.e. "Seasons") but may also be a long-term storyline.
• Episode: An individual entry within a work.

In the example Fighteer gave in post #147, the Subfranchise would be Batman 1966 TV series and would contain the Series proper (with its 3 Seasons) and the 1966 movie. i.e.:

Batman
...

• Batman 1960's TV
  • Batman
    • Season 1 (individual episodes would be the next level down)
    • Season 2^note Though to me there's little difference between Seasons 1 & 2.
    • Season 3
  • Batman: The Movie
• [Next Subfranchise]
  ...

Harry Potter

• Philosopher's Stone
  • Book
  • Movie
• Chamber of Secrets
  ...

Mega Man

• Mega Man (Classic)
  • Each of the mainline Mega Man Games (11+ hits)
  • Each of the Rockman World Games (5+ hits)
  • Power Fighters
  • Power Battle
  • Battle & Chase
  • Wily & Light's Boardgame
  • Mega Man the Cartoon
• Mega Man X
  ...

2. Folders may be considered embryonic sub-pages by default. The data restructure should automate the folderization of pages, allowing both medium- and franchise-based sub-pages to branch off at certain thresholds. This will also allow different options for low speed (shows top-quote, description, folder links, and stinger only, folders loaded only when opened), medium speed (as low plus page-image), and high speed (folders pre-loaded but still closed) users, as well as the ability for Known Tropers to set their default "speed". (Speed for Unknowns will be defaulted to "Low".)

On a work page, the folders could be set by "Plot", "Settings", "Characters", and "Other" and sub-classed as "Early" (advertizing materials and first act) and "Later" (2nd act and later, i.e. Spoiler Territory). "Characters" would also be subclassed as the individual characters and groups; the Early vs Late distinction could allow for soft pages for Walking Spoilers and would also allow the distinguishing of early character traits vs ones showing up later. If part of a (Sub)Franchise, "Recurring" Tropes can also be corralled as per the Super Sentai pages.

3. I think that, for simplicity, examples of the form Text-Link-Text should be flagged for manual reentry. After all, if the work/trope is not the first Wiki Word, then it could be directed to the wrong entry.

4. With each Primary Guess requiring its own header, a nesting system would be... well... it'd be possible to implement it in WMG. Headscratchers is a bit more questionable; the time may have come for a revamp of that page type. On that note, the Trope Pantheons do have somewhat of a structure, so they can be converted to the database format as well. Only the Titles for Pantheon elements (e.g. Slayer of the Cinnamon Challenge) should be posted on that element's particular page, if that.

Also, would we have several "alternate texts" for examples, then? One for Main Work, one for the Trope in question, one for Character Third-Person, and one for Self-Demo?

On that note, some of the JFF pages are Pages for unpublished compliations and Fictional (work-based) entries in various Useful Notes categories. Perhaps for the latter, we should group them in on their parent pages with a "Fictional Versions" header, effectively treating them like Fanfic versions of characters/settings.

5. Square Peg Round Trope entries would be quite the storage memory hog under the proposed system, then.

6. What about Tropes that are too omnipresent to list examples yet blocking it from the Works page would leave the works page lacking? Conversely, could examples for Spoiler Tropes be blocked (or toggled) on a Works page while still being allowed on the Trope page?

7. Wouldn't the watchlist be maintained as present? The pages would still exist; they'd just be redefined as a description and a list of entries bound by a common tag.

8. Yeah, translated pages should be at the same point within the respective hierarchy. Although, now I'm thinking about works that themselves are translated, or even rewritten for a different culture with the same language. Maybe the "Arc" level should also be used to cover different-language versions that have notable discrepancies from the original publication.

edited 13th Feb '16 10:31:43 PM by DonaldthePotholer

1. I suspect that we'll allow multiple layers of hierarchy, so that you can indicate that a work is both a member of a series and a spinoff/adaptation of another work/series.
2. It is my impression that, in the current system, folders consume similar bandwidth and browser resources whether they are opened or not. Their use is mainly for visual organization. That said, yes, the idea is to make them automatic rather than manual.
3. Most text-link-text examples do indeed link to the correct place, I suspect. That said, the conversion parser should probably apply some heuristics to see if they need to be called out for manual attention. Those heuristics could include whether there is a crosswick.
4. Frankly, WMG and Headscratchers disgust me on general principles; they have so very little use and would be much better integrated as part of a forum topic. That said, a lot of people apparently like them, and so we'd either implement them as threaded discussion pages much like the current Discussions, or as free-form text elements hooked onto an article and editable from them.
5. Not sure I get you here. If there are tropes with a lot of shoehorned examples, they would occupy a similar amount of storage space as the articles that now hold them. The conversion will be linear at worst in terms of the delta in storage: [n] lines of example text = [n] example elements. In practice, it'll be <[n] because many examples will be hooked up automatically with their crosswicks, so two examples may become only one in the final database.
6. We have in mind that articles can be set to suppress the display of examples on themselves, on reciprocal articles, or both. So an Omnipresent Trope could be set to hide all examples when viewed directly, but you'd still see them on work pages that have examples of the trope. Of course, one reason why we say that those tropes shouldn't collect examples is the ridiculous size of such a collection, but that will no longer be an issue with the new system and auto-pagination/categorization.
7. Yes, the Watchlist would be much the same as at present; rather than be notified of gross changes to the text of the article, you'd be notified of any change to any element. It would even get smarter, as the automatic nature of crosswicking would let you know when anyone added an example of the trope or work anywhere on the wiki.
8. For works with multiple language versions, meta-tags would handle that. "Work X is available in [English], [Japanese] (original), [German]." We could even design it so that title elements link to a translation of the work. "In English, this work is titled X; in Spanish this work is titled Y," but at a database level. I see no need for distinct articles, however, unless the work is extremely different between translations. There's nothing stopping that though. One possible type of relational link between articles would be "translation of".

edited 13th Feb '16 10:52:29 PM by Fighteer