Join 3,438 readers in helping fund MetaFilter (Hide)


The Official API of the Marvel Universe
February 20, 2014 5:52 PM   Subscribe

You mean Prince JSON of Spartoi? Marvel Comics has released in beta an API for access to part of its character and comic book database. Currently, it covers 30,000 comic books across 7,000 series and 8,000 characters in work by 5,000 creators, and provides results to queries in standard JSON format.

The Marvel API and developer platform, currently limited to 1,000 API calls per day per key holder, is designed to allow fan sites in particular to create data visualisations and queries based around Marvel history, and as a resource for partners and retailers, but has an obvious application in trivia-offs also - although it has yet to be demonstrated whether it can judge who would win a fight between $hero and $villain.

A guide to setting up queries by Ricky Yin.

Marvel.com's interactive API tester. (Account or social login required.)

A demonstration by Adobe's Raymond Camden, which pulls and displays a random comic book cover from the Marvel archive. Code.
posted by running order squabble fest (31 comments total) 40 users marked this as a favorite

 
Meanwhile, over in the music industry the intermediaries have extensive back-catalogues of material which is out-of-print and no mechanism for research... (Heck, even Gracenote/CDDB/etc. was essentially crowd-sourced to begin with - I know in the early days of sticking CD's into computers, I spent quite a bit of time inputting data)

Kudos to Marvel !
posted by jkaczor at 6:04 PM on February 20 [5 favorites]


Some academic needs to do this for the characters in Russian and Chinese novels and other ones where there are dozens of them it's difficult to keep straight. I hear that GOT is similar?

Is this perhaps the product of a synergy with writing/storyboarding software tools? I wonder if visualization apps and other things like this might become so developed and polished and standardized that a dataset is released with the launch of a book or comic or film and it becomes a scaffolding that allows the storyteller to range beyond what someone would normally be able to keep track of... a thousand characters, ten thousand, or something like that. The unassisted human mind is left in the dust, straight outta Accelerando.

This season's datasphere serial seems boring and tedious? You probably need to upgrade to the nightly release of the TVTropes cognitive bookmarklet.
posted by XMLicious at 6:16 PM on February 20 [4 favorites]


Does DC have anything like this?
posted by Brandon Blatcher at 6:41 PM on February 20


jkaczor, don't go there. Just. Don't. Go. There.

OK, fine! You remember that video, posted awhile back, about timezones? Whom gods destroy, they make program timezones?

Luxury!

Music catalogs are horrific. List by titles? List by composers? Maybe by artist? Oh, we used to dream about such a neat and tidy database! Opus numbers? Wha---? Why would anyone, oh, very well. Wait? The "Moonlight Sonata" is, for some damn reason, Opus 27, Number 2, but everybloke calls it "Sonata Number 14," because it was, like, his 14th attempt at doing something pretty for the masses. Very well.

No, no, no, no. It's just music, it has a score. It's easily reproducible. Who cares if the bloody "Funeral March" was performed by the Boston Symphony Orchestra (conducted by the thrice-damned Leinsdorf) on that particular day. Oh, well. The Kennedy assassination. The first one, I mean. Sad that. Well, perhaps, we could add a subquery, here or there, for the odd special cases.

Wait... are you seriously asking me who did the recording? And the pressing? Why don't just ask me what pants Leinsdorf was wearing?!

Well of course, we had it tough.

We had this old comic book database that we were pretty proud of. Every issue cataloged, back to the thirties, every story noted, every author, artist and inker annotated, when we could find them. Links to image covers. Oh, we were happy in those days!

But then came the variants and your bloody foil covers! And the ridiculous issues that weren't "numbered" in a sensible 1, 2 and 3, like you'd think, but "A," "B" and "C!" The lads in the mill, they worked hard to pull that off, I can tell you.

And the "reprints," the knock-off editions! No, you fool, it says "Famous First Editions" right on the cover. Go on, get off! Stupid git.

And then the reboots! "Action Comics #1!" What again?! It's like you are asking us to walk through time in the snow, uphill, both ways.

But you trying telling the young people of today that, and they won't believe you.
posted by SPrintF at 7:14 PM on February 20 [14 favorites]


$5 to whoever hacks in and replaces all the author names with Jack Kirby.
posted by benzenedream at 7:27 PM on February 20 [14 favorites]


They keep rebooting the server.
posted by srboisvert at 7:35 PM on February 20 [8 favorites]


Dude, I have real work I need to be doing.
posted by signal at 7:38 PM on February 20 [2 favorites]


$5 to whoever hacks in and deletes anything credited to Rob Liefeld.
posted by entropicamericana at 7:51 PM on February 20 [4 favorites]


$5 to whoever hacks in and replaces all the author names with Jack Kirby.

I was going to chip in a No-Prize for anyone who put Kirby before Lee in every database record (but not deleting, because I don't believe you can deny that Stan Lee played a role [albeit self-serving] in Marvel's success.) I will however offer $5 for whoever runs 's/Bob Kane/Bill Finger/g' on the the eventual DC version of this, because you just know they'll expose shell access and run the thing off badly-permissioned text files.
posted by sysinfo at 7:53 PM on February 20 [1 favorite]


+1 for reasonable documentation
posted by tilde at 7:54 PM on February 20


WHAT
posted by middleclasstool at 8:10 PM on February 20


Does DC have anything like this?

Not to my knowledge - I mean, they hopefully have a database, but they haven't released an API. I suspect one issue is that these databases need a fair bit of cleaning - they have been good enough to work out what a new writer on a title ought to read, but not necessarily good enough to open to the public...
posted by running order squabble fest at 8:29 PM on February 20


SPrintF - I see I hit a nerve... another sympathetic one... yes, I have looked into this madness before... (and timezones too, one of the other toughest challenges I had back when I was a programmer dealt with timezones, and considering it was a railway application, it was technically "life and death"... yikes...)
posted by jkaczor at 8:47 PM on February 20


Can someone pithily explain to me what this is, as if I were someone who doesn't know a single thing about comic books or web development (THEORETICALLY SPEAKING)?
posted by threeants at 9:11 PM on February 20


Can someone pithily explain to me what this is, as if I were someone who doesn't know a single thing about comic books or web development (THEORETICALLY SPEAKING)?

Think of all the information in the marvel universe. All the superheros, bit characters, comic book titles, issue numbers, writers on each issue. Now imagine you would like to use that information in a program. Maybe you want to create a program that displays random comic book covers with Wolverine on them.

Now that there's this API, you can ask Marvel questions about their data. You can say "can I get a list of comics that have Wolverine in them", and Marvel will give you the data in a format that is easy for your program to work with. You can then say "give me a list of all the writers involved in those comics", or "give me a list of other characters in those comics" or "give me the date, price, and ISSN for each of those comics", and Marvel will give you that data too.

This opens up a lot of doors for people who like comics and want to create new and interesting applications using comic book data. You could chart how often Wolverine and Hulk were in the same issue vs Wolverine and Cyclops, year over year. Or which writers used the least number of characters in their comics.

Possibilities!
posted by Jairus at 9:37 PM on February 20 [3 favorites]


So basically someone logged a ton of data into a database that's optimized for versatile use?
posted by threeants at 9:47 PM on February 20


I'm imagining Jessamyn just covering her mouth and laughing. Oh, you boys think you had it tough? We had to live in an old master collection, with a load of old paperbacks dumped on us every morning. Library? Hah!

Well, when I say "library," I just mean a hole in human culture, covered by a sheet of tarpaulin. But it was a library to us!
posted by SPrintF at 9:49 PM on February 20 [3 favorites]


So basically someone logged a ton of data into a database that's optimized for versatile use?

Sort of - the data is in the database, sitting on Marvel's servers. The API - the Application Programming Interface - is basically a set of instructions about how that data should respond to requests from the outside world. And, when it gets a request from the outside world, it outputs the information in JSON format - essentially a standard format for exchanging data, and thus one that can be represented easily in different forms by different applications*.

So, you could, as Jairus says, make varied queries about contents of the database - like finding out which Fantastic Four Comics involved Uatu the Watcher _and_ Eternity _and_ Reed Richards _and_ Galactus**. Our you could visualise how many comics featured Wolverine, year by year.

But you can also mix that information up with other data sets, potentially. For example, if a comic collecting database releases an API, you could combine the two and query which authors' comics appreciate most or least in value. Or, for that matter, if you had access to the data, which states produced the most prolific writers of stories about Star-Lord...


*Like, comma-separated values files can be open in spreadsheet programs, where they appear as spreadsheets, or in speaker support programs, where they appear as charts and graphs.

** Issue 262 - "The Trial of Reed Richards pt 2"
posted by running order squabble fest at 10:03 PM on February 20 [1 favorite]


Great, a way to be more pedantic about continuity. Just what comics need.
posted by The Hamms Bear at 10:50 PM on February 20 [1 favorite]


DC only needs to go back to Flashpoint at the New 52.
Sure, it might require more work than they put into actually thinking about how it might all hang together, but it would be modest. And then people can stop wondering about The Robin Problem because there will be proof that Stephanie Brown never existed!
posted by Mezentian at 11:24 PM on February 20


Does DC have anything like this?

They are trying, but a meta-cognitive war between Earth-1, Earth-2, Earth-S, and Earth-♪ keep rewriting the database.
posted by GenjiandProust at 1:58 AM on February 21 [4 favorites]


Sadly a war with Earth-2 is their next major crossover, following the war with Earth-3.
posted by Mezentian at 2:51 AM on February 21


Know enough to realize how cool this is. Don't have the skills to do anything with it. :(

I'm in the middle like a bird without a beak.
posted by hot_monster at 4:10 AM on February 21 [2 favorites]


Okay... there are a fair number of open APIs now. Someone's been keeping a list, right? I know the NYTimes has one.

Who's doing the fun experiments mashing them up?

Apparently a bunch of folks.
posted by DigDoug at 4:22 AM on February 21 [1 favorite]


Stephanie Brown is back :)
posted by sevenyearlurk at 4:52 AM on February 21


Sadly a war with Earth-2 is their next major crossover, following the war with Earth-3

Earth-1 had always been at war with Earth-n+1!

Report to the Ministry of Truth for Reeducation, Citizen!

posted by GenjiandProust at 6:03 AM on February 21 [2 favorites]


Great, a way to be more pedantic about continuity. Just what comics need

This has nothing to do with continuity, just brand control. Marvel has zero interest in continuity between stories, but they care massively who is looking at what. "Someone just noticed that the Dazzler and Squirrel Girl combo always makes sales go up. Very interesting."

Personally I would LOVE a continuity database: i.e. "Which changes will be referenced in the future?"

Such a task would be much easier than it sounds. Having followed the Fantastic Four in some detail over the years, a "five year continuity" database would have very few entries after 1968, and probably only one entry after 1988 (Valeria's return seems to have stuck).
posted by EnterTheStory at 6:23 AM on February 21


Though I respect this a lot, I'd be fired if that were my job, after killing JSON off and countless screaming RGONOTS.
posted by maryr at 9:03 AM on February 21 [6 favorites]


Some academic needs to do this for the characters in Russian and Chinese novels and other ones where there are dozens of them it's difficult to keep straight. I hear that GOT is similar?

Hah! "Dozens"
posted by ChurchHatesTucker at 10:41 AM on February 21 [1 favorite]


> "Hah! 'Dozens'"

Game of Thrones has greater than a thousand named characters.
posted by wires at 10:14 AM on February 22


What I meant is that for example Dream of the Red Chamber contains about forty major characters according to redologists, though yes the total number of named individuals counting every relative and passerby and the equivalent of Star Trek redshirts is an order of magnitude greater or more.
posted by XMLicious at 12:24 PM on February 22


« Older Awesome choir (SLYT)...  |  Brian Williams performs "Rappe... Newer »


This thread has been archived and is closed to new comments