Unfortunately, and I say this from excruciating experience, this isn't a complete/accurate representation of all M:TG cards.
I spent the summer of 2011 scraping Gather to build out a product catalog and API for a start up that didn't start. I really wish that I had taken the time to put the data out as a torrent.
M:TG has 20 years of strangeness to take into account when creating a data model... And Gatherer has its own issues. For example, the way symbols are represented in card text is inconsistent. (e.g. The tap symbol can be represented as both "ocT" and "{T}", and the representations for mana symbols have evolved inconsistently as well.)
Wizards of the Coast are the only ones who can build this API legally. They hold the copyright on all of the data related to M:TG and I truly wish that they would open up it up so that we, as a people, wouldn't have to spend hundreds of hours recreating the wheel. Poorly. [0]
Fortunately, I stopped what I was doing and started on a board gaming "thingy" with friends instead: http://www.shutupandsitdown.com/
[0]: That isn't to say that the mtgapi.com guys/gals did a shoddy job. It's just a big task that WotC should be forward looking enough to do themselves.
While I have WotC on the line... Why wasn't the D&D 5th Edition SRD hosted on GitHub during development?
Even without counting the two Collector's Edition print runs of the original core set, it's also missing the Anthologies printing, the FNM promo printing and the judge promo printing along with the version that appeared in the Coldsnap theme deck "Kjeldoran Cunning" (which kept the Ice Age art and expansion symbol but put it in the modern card frame).
Meanwhile, this has all of those printings listed:
While I have WotC on the line... Why wasn't the D&D 5th Edition SRD hosted on GitHub during development?
Hell yeah. I'm tempted to contact some of the creators of some of my favorite RPG systems (Jason Bulmahn, Jonathan Tweet, Monte Cook), and beg them to agree to make a completely free RPG. Or at least design the core mechanics. I'm thinking maybe some Creative Commons license, which would allow commercial use. Get them to make a bid, as to how much it would cost to have their work-for-hire for however long they estimate it would take to make the Core Rules of this game...
...and then have a Kickstarter, to raise the funds to do it.
And make implementations of the Core Rules in several programming languages, probably JavaScript first (good for the Browser, or for Node.JS), with Reach Goals for other languages...
Let me tell my personal story: I was a DM for D&D 3rd Edition, and I chose to run through Return to the Temple of Elemental Evil. I had a hard time running it, though, so I made an electronic version of the campaign book. I scanned every single page, I made a wiki on my laptop, one for each NPC. I made a wiki page, with the scanned content for every single monster that showed up. I made wiki pages for every spell (with the scanned content of every spell, right there.) I didn't have to settle for the OGL content, I had the ACTUAL CONTENT FROM THE BOOKS, so my rule-lawyering players couldn't complain that the DM's rules were slightly different from their rules. I made hot-clickable maps, to take me to rooms. I made spreadsheets with every NPC, and their Initiatives, and Hit Points, and any effects on them...
...and now that I've made this thing, for personal use, it would be ILLEGAL for me to share this with anyone.
That sucks.
I've also played around with making computer role playing games, based off of some fantasy role playing game, but most of the "FREE" ones out there, are not available for commercial use, or I'd have to pay a licensing fee.
That sucks.
Ah well, I'm done rambling for now...
One more quick point: I don't begrudge people for making copyrighted, commercial art like games and campaigns. But I think the time for the "Linux" of Fantasy Role Playing Games has come.
I've also dealt with the data and can confirm it's a stunning headache. Try building a search UI around the inconsistencies and you'll go insane. Even Gatherer has issues!
From what I can tell, there's only one or two guys at WotC that come close to understanding the data. With the rest of the organization in mind, I highly doubt WotC will open up their data - the business guys wouldn't understand it. It's a closed-source kind of company.
Wow, just looking at the example card on that makes me realize how ridiculous MTG has gotten in terms of powerful cards. Each edition seems to get closer to a "tap to win game" card.
The dev team learned their lesson from the earlier days[1] and do strict evaluations of cards and interactions before releasing them. The most popular format, "Standard", is played with the last 2 "blocks" (3 sets of card "sets"), and 1 "Core Set". So basically the last 1.5—2 years worth of cards. Those interactions are studied carefully before being released. Very infrequently do you run into something truly overpowered and needing to be banned from the competitive scene.
Not to mention, in the most basic sense, cards like Counterspell[2] will probably never be printed in a Standard playable set, but cards like Cancel[3] will. One being too cheap to cast and being almost game breaking, while the other has its power balanced just right for the cost.
Yeah, after the hellish Tempest/Urza block they really cut back, Mercadian Masques was almost critically under-powered, and they've gotten so much better since that time at providing balance.
If you haven't played in a while, you should check it out again. There's a game[1] now for Steam/Xbox/PS3/Android/iOS to get you back into the swing of things and show you many of the current cards.
The API was just a feature. The actual product was a multi-merchant market place w/ a shipping estimator that was a 1-1 implementation of EasyPost (YC S13).
I spent the summer of 2011 scraping Gather to build out a product catalog and API for a start up that didn't start. I really wish that I had taken the time to put the data out as a torrent.
"Magic: The API" doesn't appear to support:
* Multi-faceted cards such as "Fire // Ice"(http://gatherer.wizards.com/Pages/Card/Details.aspx?multiver...)
Both return the data for the "Fire" facet, but with different artists: http://mtgapi.com/api/v1/fetch/id/27165 & http://mtgapi.com/api/v1/fetch/id/27166
* Double sided cards such as "Student of Elements" (http://gatherer.wizards.com/Pages/Card/Details.aspx?multiver...)
Only shows data for "Tobita, Master of Winds": http://mtgapi.com/api/v1/fetch/id/78691
* Double cards such as B.F.M (http://gatherer.wizards.com/Pages/Card/Details.aspx?multiver...)
Only shows 1/2 of the cards: http://mtgapi.com/api/v1/fetch/id/9844
* Multi-sided cards such as "Cloistered Youth // Unholy Fiend" (http://gatherer.wizards.com/Pages/Card/Details.aspx?multiver...)
Only shows the "Unholy Fiend" data: http://mtgapi.com/api/v1/fetch/id/221212
M:TG has 20 years of strangeness to take into account when creating a data model... And Gatherer has its own issues. For example, the way symbols are represented in card text is inconsistent. (e.g. The tap symbol can be represented as both "ocT" and "{T}", and the representations for mana symbols have evolved inconsistently as well.)
Wizards of the Coast are the only ones who can build this API legally. They hold the copyright on all of the data related to M:TG and I truly wish that they would open up it up so that we, as a people, wouldn't have to spend hundreds of hours recreating the wheel. Poorly. [0]
Fortunately, I stopped what I was doing and started on a board gaming "thingy" with friends instead: http://www.shutupandsitdown.com/
[0]: That isn't to say that the mtgapi.com guys/gals did a shoddy job. It's just a big task that WotC should be forward looking enough to do themselves.
While I have WotC on the line... Why wasn't the D&D 5th Edition SRD hosted on GitHub during development?