Awesome idea. You completed my embedded Japanese game idea that I have wanted for a while now. I have had the concept of a dungeon crawler RPG that required Japanese to play and basically had a simple SRS. It was started, but died after I lost interest in building yet another SRS and trying to get Japanese to render sanely in SDL with truetype font madness.

What you have done is given me the missing link. Use anki and push card answering activity to the server.

For now, I have gutted your plug-in and am using a XMLRPC call to send experience to a dungeon crawl server. The idea is young, but experience is translated to energy and the crawler is much akin to the fish aquarium model wherein the player doesn't directly interact with the game. The interaction is in the form of energy that is required to crawl a dungeon, hunt for gold, loot and fight creatures. The idea is based off of a crawler concept on the 300 game mechanics site: http://www.squidi.net/three/entry.php?id=22

Again, thanks and great job!

brianobush Wrote:Awesome idea. You completed my embedded Japanese game idea that I have wanted for a while now. I have had the concept of a dungeon crawler RPG that required Japanese to play and basically had a simple SRS. It was started, but died after I lost interest in building yet another SRS and trying to get Japanese to render sanely in SDL with truetype font madness.

I actually had a similar idea ages ago, but python ended up being to slow to render the map so I lost interest. You could probably find the thread somewhere. Though in my game you were supposed to answer the cards in game during fights, this energy idea sounds much better.

brianobush Wrote:For now, I have gutted your plug-in and am using a XMLRPC call to send experience to a dungeon crawl server. The idea is young, but experience is translated to energy and the crawler is much akin to the fish aquarium model wherein the player doesn't directly interact with the game. The interaction is in the form of energy that is required to crawl a dungeon, hunt for gold, loot and fight creatures. The idea is based off of a crawler concept on the 300 game mechanics site: http://www.squidi.net/three/entry.php?id=22

So how do you play the game? Does it play itself if you have enough energy?

Quote:So how do you play the game? Does it play itself if you have enough energy?

The general idea, is that it is basically idle until you start shoveling energy in. I also detect start/stop event (anki startup, shutdown) which can be used to start quests, e.g., 'Get 30 experience points in the next twenty minutes.' After which, it plays itself. Playing itself is basically dungeon crawling, hunting for loot, battling with evil mummies, etc. Mini-games could make it interesting, or crafting (ala minecraft) to help your crawler battle better. In the end, it will be a glorified progress quest with counters for gold, experience points, rooms searched, goblins killed, etc. Most content will be procedurally generated.

So far, no code for the actual game component, just scribbles. Just the client implementation in Anki and a simple server. Python should be fine for this and I will use pygame for the graphics, which will be minimal. If I have issues, I can whip out C/SDL to polish it, but pygame is easier to build out ideas with.

Keep the ideas rolling out, since I have only a small set of actual plans worked out.

bombpersons Wrote:I actually had a similar idea ages ago, but python ended up being to slow to render the map so I lost interest. You could probably find the thread somewhere. Though in my game you were supposed to answer the cards in game during fights, this energy idea sounds much better.

What all did you have on the map at once? Were you doing lots of transformations on the sprites each frame? I've found that pygame (backed by an array library for transformations) is actually pretty reasonable for speed; it's just the moment you start modifying individual pixels without an array library that the speed becomes horrid.

That said, you may want to check out panda3d. It's a python 3d game engine by Disney/CMU with performance critical parts in C. On par with pygame for ease of use, but far more comprehensive and is a fully featured, modern, 3d engine thus you automatically get nice features like scenegraphs.

overture2112 Wrote:

bombpersons Wrote:I actually had a similar idea ages ago, but python ended up being to slow to render the map so I lost interest. You could probably find the thread somewhere. Though in my game you were supposed to answer the cards in game during fights, this energy idea sounds much better.

What all did you have on the map at once? Were you doing lots of transformations on the sprites each frame? I've found that pygame (backed by an array library for transformations) is actually pretty reasonable for speed; it's just the moment you start modifying individual pixels without an array library that the speed becomes horrid.

I was only drawing the map tiles that were visible on the screen. No characters or anything. Maybe my map drawing algorithm was just really unoptimized...

Though, if I were to do it again, I would probably use a hardware accelerated library like sfml or panda3d like you said. I'm sure its possible to do in pygame with enough optimizations though.

Hi,

thank you for all the ideas, recommendations and links on further information!
I can't comment all of them here because it would take to long, but I read each one of them and am really grateful for your suggestions.
I am currently working on a simplified version of the city conquering and think I can publish it this week. Little by little I then can add more features. But I want to get a working version out as soon as possible. Still the first version will be quite "fun to play", in my opinion.

Also glad I could help other people like overture and brianobush having ideas for their own project. Seems like there are a lot of people actually thinking about the connection of learning and playing. Looking forward to see the results :-)

Any idea how to efficiently grab a (nice) picture for each of 250+ Japanese cities and convert them to maximum 200px width?

I could use the wikipedia api which returns a list of all the images used in an article, but I still would need to choose an appropriate picture manually, which is a lot of work.

Also I am not sure if I want to make a 3 Megabyte plugin. Maybe I should offer the images as an additional download package? On the other side, 3 megabyte is not that large, is it?

3 megabytes isn't too big I don't think. As far as collecting the images is concerned... Can't think of any other way to do it other than manually if you want to make sure the images are correct.. I'm sure people wouldn't mind helping with the images though.

Proxx Wrote:Any idea how to efficiently grab a (nice) picture for each of 250+ Japanese cities and convert them to maximum 200px width?

Fetching and selecting the pictures may be a pain, but it should be relatively easy to mass convert them to the correct dimensions using ImageMagik's convert.

Proxx Wrote:Also I am not sure if I want to make a 3 Megabyte plugin. Maybe I should offer the images as an additional download package? On the other side, 3 megabyte is not that large, is it?

That's smaller than a full sync of most decks and doesn't have to be downloaded often, so I'd think that's fine.

Manually downloading is one problem, the other one is copyright. Although most of the images are gnu or cc licence, I would still need to mention the photographer somewhere. Which means I would also need to get the authors name for every image.

For now I downloaded google maps pictures with an marker on the relevant city. That's not really exciting, but the easiest solution I found so far. (Now I have 250+ images of Japan where only the markers position changes...)

Would be great to replace those with images taken by people here on the forum. But I doubt those would cover every city.

I took every city in Japan with more than 100.000 inhabitants and now have 267 cities. That gives a great opportunity to level up, but maybe is a little bit to boring in noone know the cities?
Well I guess I'll go with 267 cities, and if I get to much negative feedback I'll remove some.

Proxx Wrote:I took every city in Japan with more than 100.000 inhabitants and now have 267 cities. That gives a great opportunity to level up, but maybe is a little bit to boring in noone know the cities?
Well I guess I'll go with 267 cities, and if I get to much negative feedback I'll remove some.

Too many cities/levels is only a problem if all the more famous places are saved til the end.

Also, what about creating a public google spreadsheet with City Name, Link to Picture, Link to Wikipedia page, and maybe a Did You Know blurb (which could be especially neat for less famous places); split the work?

Hi,

here is the link to the 267 largest Japanese cities including number of inhabitants and wikipedia link: https://spreadsheets.google.com/ccc?key=...y=CP-FuNYO

They are sorted by size, so there will probably be a lot of not famous places on the first levels. If I'd take for example only 100 cities, either the highest level would be reached faster, or the time span between level-ups would increase...

Hello, i'm happy that you have this idea, i thought about it myself, but not as well as you did you're taking it seriously! One question, do you use a MVC framework for your php development? Well if you use one such as cakephp or zend framework ill be happy to help under your direction

I've been hoping for something like this to come out for a long time. I'll have to try it once I get home from work.

One suggestion though, and this one is a doozy.

Add competition to the game. Not the against the computer competition, but competition against other players. Maybe a scoreboard or something. Nothing breed necessity to learn something as the thought of possibly failing against someone. Fear is a powerful motivator. ;D

Probably not a great idea for everybody, but the people who thrive on competition will love it.

Thank you for your suggestions!
@lastchance: I started developing the server application in Yii framework. I stopped it after two days because it was getting to complicated and I wanted to focus on the core feature of the plugin. Probably I am going to connect Anki with an online server in a future release. Would be great to get some help then :-)

@KMDES: Yes, I am planning something like this for a future version of AnkiWarrior. The problem in Anki is, that cheating is relatively easy as you rate yourself. With my plugin, I want to encourage Anki users to study more, not to change their study and rating behavior. Maybe I'll make competing in a small groups possible, where people know each other, or competing against oneself.

I just uploaded the first "official" release of AnkiWarrior to the shared plugin folder of Anki. It should be downloadable right now. If there are any problems, please let me know.

Here's a brief description:

In the current version of AnkiWarrior, you have to "conquer" Japanese cities.
You start with a small city and for every flashcard you answer in Anki, you'll "convince" more and more of the citys population (i.e. with your fantastic language skills). As soon as you've convinced every inhabitant, the city is yours; you'll reach a new level and move on to the next bigger city. Currently, only Japanese cities are implemented.

More information here: http://ankiwarrior.hallojapan.de/

I am still looking for good pictures. If you have any suggestions, please contact me.

I'll try to implement more features in the future. The current version should work fine and is a good basis for future development.

Hey there,

I found your plugin today in the plugin list and I think it's a really great idea!

I would suggest you try to keep it as open as possible and make it easy for users (especially non programmers who just add content) to make contributions. The whole Japan scenario is nice, but users are studying many different things with Anki and it would be interesting if users could create their own settings and choose from a lot of user generated settings. Just different city lists that everybody can create and which are automatically imported would be enough for the beginning.
I myself use Anki to study Chinese and would be happy to contribute to a chinese city list.

Later when you have even more elements you can also allow user generated items or attractions or whatever. I think as long as you make it as easy as possible to add things there will be lots and lots of user generated content and yourself and all future programmers contributing can concentrate on creating the gameplay itself while the content is generated by the community.

Keep up the good work!
Marcel

This combined with the unreal anki plugin is ridiculously awesome =)
I love the conquering cities idea, brilliant!

What would be really awesome, would be if you added more cities from all over the world and eventually you can take over the world, then the solar system, then the galaxy, then the universe, then the multiverse and then whatever comes after the multiverse (multimultiverse?) =)

Not sure if this is a bug or not (as far as I can tell nothing actually happens), but I get this error message whenever I open a Wikipedia link:

Traceback (most recent call last):
File "C:\Users\XXXXX\AppData\Roaming\.anki\plugins\JxPlugin\hack.py", line 38, in JxAnchorClicked
QDesktopServices.openUrl(QUrl(url))
NameError: global name 'QDesktopServices' is not defined

Any idea what the JxPlugin is objecting to? Both plugins still work fine fwiw.

Also, great job!

Thanks for you comments guys :-)
I definitely plan to allow user generated and also additional conent (different cities or complete scenarios). For now I've started with Japanese cities as a basis. But the plugin will certainly grow over time. I cannot gurantee how fast, but I am trying to implement the most common requests and suggestions.

harhol Wrote:Any idea what the JxPlugin is objecting to? Both plugins still work fine fwiw.

It seems like JxPlugin is overwriting the handling for links in the help-bar and, as far as I can see, this is a bug in JxPlugin as it doesn't import the QDesktopServices. When JxPlugin is active, this problem should occur for every link to a webpage from the help bar. Maybe you should try and contact the author if JxPlugin.

Hoho, this is fun, but at this rate I'll never conquer anything. Gotta stop slacking and add more cards.

Just a bug I noticed; if you undo a review, your convinced population counter remains as if it hadn't been undone.

@Proxx, in fact I work as a web developper (php/mysql/jquery etc) but I don't know much about server config and all these things unhappily :/

About the competition and the issue of people who might over-rate their knowledge, what if people are competing against people who are studying exactly the same thing (ie: heisig, KO2000 etc), and everyone is tested on cards he has already done from time to time, for example people are asked for the definition of a kanji, or for keywords, and he must give the answer (type it). It could be 10 questions in a limited set of time, of course answers are too complex to be checked by the computer, but the community could see the answers of everyone and judge if the guy is faking or not.

Not sure if my idea is clear lol !

Also, since pressing "again" means you'll see the card again soon, but still convinces people, it leads to the odd situation that pressing "again" and then "hard" convinces more people than just pressing "hard".

Come to think of it, the fact that "easy" and "very easy" increase the card's interval by quite a bit may mean that in the long run they get you less points than pressing "hard" a lot, depending on settings. I can't think of an easy solution for this, though, or even if it's something worth caring about.

Reviewed Wrote:I can't think of an easy solution for this, though, or even if it's something worth caring about.

If you want to protect against people changing their responses to optimize for the game, the easiest solution is to scale the reward by time until the card is due again (which should be 0 for "again", despite the fact that some people have failed cards set to delay for 10min). Then there's no effective advantage to purposely failing, stating the card was harder, or stating that the card was easier than it was.

Now I feel silly for not having thought of that.