Plans for new OHR Gamelist

[shakeyair]

There should obviously be a report button if anyone uploads a game with a virus in its package or illegal content or something but automatically deleting games just for being unpopular is a terrible, terrible idea.

This. Also, I am against having a page with no download/no link to an external download. Games in development are abandoned far too often, it would wind up just being clutter.

[Mogri]

But lets say we list every OHR game on CP, SS, RPG Online, etc. What would be fair to rehost without asking, and what wouldn't? (But ask IM and Moogle first anyway.) Screenshots and game descriptions? As Surlaw pointed out, that would be bad, as then those screenshots and descriptions would need updating when they change elsewhere. So if practical I would favour only caching that stuff from other sites if at all, and periodically reimporting. But that's not to say you couldn't add content to a game listing manually instead... that's the whole point of a gamelist after all.

http://www.slimesalad.com/forum/gamedump.php has a list of every game on the site along with its detail page URL and associated files, plus a timestamp for the most recent update. External integration with the SS gamelist should be trivial. James has already done it.

Unrelated: I don't understand why this is in General Discussion.

[Spoonweaver]

Unrelated: I don't understand why this is in General Discussion.

where should it be?

[Meowskivich]

The Twilight Zone, obviously.

[Chronoboy]

Although a python function to check a game's password is perfectly possible (and nohrio might already do it) passwords are *very* easy to crack, so knowing it is not sufficient proof of game-authorship.

Shhh. -- I thought the password used a one way hash of some sort.

This is true! ... but even as a one way hash, it is still super easy to crack!

Okay... Crack this MD5 one-way hash password: 4397f32366e0acf4913d1c9986134762 :p

If you are unable to reverse it into a clear-text password, then it works the way it should.
To confirm the user typed in the correct password, the password to be verified is hashed, and both hashes are compared.

http://www.slimesalad.com/forum/gamedump.php has a list of every game on the site along with its detail page URL and associated files, plus a timestamp for the most recent update. External integration with the SS gamelist should be trivial. James has already done it.

Thank you, when it comes time to perform the import, I will look into this and use it as a source.

[Bob the Hamster]

Okay... Crack this MD5 one-way hash password: 4397f32366e0acf4913d1c9986134762 :p

If you are unable to reverse it into a clear-text password, then it works the way it should.
To confirm the user typed in the correct password, the password to be verified is hashed, and both hashes are compared.

The OHRRPGCE doesn't use an MD5 hash. It uses a shorter simpler hash.

A brute force attack on a hashed password does not need to recover the exact password, it just needs to find one collision. So no, it would not be possible to recover the exact password the author intended, but it is indeed easy to find a hash collision that works to unlock the game-- and from the perspective of anyone who is not the author, there is no way to tell the difference between the original password and a hash collision password that works.

And, no, we will not be changing the OHRRPGCE's password storage system. :)

[Chronoboy]

The game list is still being developed, and for those with at least some Python experience and want to contribute, here's a quick Django tutorial I wrote to get you started:

http://www.pythondiary.com/tutorials/si ... jango.html

It explains how to build a CRUD app, which the game list consists of mainly, if you read the tutorial and then visit the source code repo on BitBucket, you will begin to understand how the code works.

EDIT: I'd also like to mention that if you see my blog, you will notice that everything on it is completely cross referencing. This is how I'd love to make the game list. For example, click on Django under the title, and magically, you will see everything available for Django, more tutorials, and reviews. Links in various posts are also references as well, click South for example, and you will see page that tells you more about it and any tutorials that might be available. This is why I usually create custom solutions as I have the freedom to do anything, and why my blog was custom made and isn't an off the shelve blogging software.

[TMC]

Please excuse the late reply.

It sounds like there's fairly general agreement about what the goals of this project should be. I think in-development games aren't worth listing unless there's something to say about them, such as a preview submitted to a magazine (which would make them not very different to the scores of games with a review but no working download link, which I would like to see listed as well). But clutter shouldn't be a problem because the point is to index everything, so intelligently sorting the most relevant games to the top will be vital anyway. Just sorting by number of likes sounds too simple to me.

I would love to be involved in this project. But engine development and my studies are unlimited timesinks, so I wouldn't want anything else to depend on me. Lately I've been spending time on the port and the wiki. I don't know when I could start. I'll have a closer look at that tutorial and your code later.

Checking the password would still be a good way to provide evidence of authorship. The certainty threshold doesn't have to be set too high if previous versions of a page are kept and mods can undo any damage and monitor what's going on. nohrio does have a password checking implementation, though it is for the previous password version. The current version is trivial to implement, but older versions are missing. We even already have a function to generate collisions, but most people don't know that...

You can try out a parser and formatter for HS I wrote in Python 4 years ago (slightly out of date) here: http://tmc.castleparadox.com/forms.php. The code is here. It's a mess.

These might also be interesting: http://tmc.castleparadox.com/gamelist.py, http://castleparadox.com/search-gamelist.php?mirror=true (the later is used by the motherhamster mirror). The former is obviously meant to imitate the SS game dump (I can't remember why I did that). It doesn't access the actual database: consider it unfinished. Also, are you aware of rpgbatch.py? Here are some examples of output: gamelist.txt, scanlog.txt,scrlog2.txt

[Chronoboy]

Please excuse the late reply.

It sounds like there's fairly general agreement about what the goals of this project should be. I think in-development games aren't worth listing unless there's something to say about them, such as a preview submitted to a magazine (which would make them not very different to the scores of games with a review but no working download link, which I would like to see listed as well). But clutter shouldn't be a problem because the point is to index everything, so intelligently sorting the most relevant games to the top will be vital anyway. Just sorting by number of likes sounds too simple to me.

I would love to be involved in this project. But engine development and my studies are unlimited timesinks, so I wouldn't want anything else to depend on me. Lately I've been spending time on the port and the wiki. I don't know when I could start. I'll have a closer look at that tutorial and your code later.

Checking the password would still be a good way to provide evidence of authorship. The certainty threshold doesn't have to be set too high if previous versions of a page are kept and mods can undo any damage and monitor what's going on. nohrio does have a password checking implementation, though it is for the previous password version. The current version is trivial to implement, but older versions are missing. We even already have a function to generate collisions, but most people don't know that...

You can try out a parser and formatter for HS I wrote in Python 4 years ago (slightly out of date) here: http://tmc.castleparadox.com/forms.php. The code is here. It's a mess.

These might also be interesting: http://tmc.castleparadox.com/gamelist.py, http://castleparadox.com/search-gamelist.php?mirror=true (the later is used by the motherhamster mirror). The former is obviously meant to imitate the SS game dump (I can't remember why I did that). It doesn't access the actual database: consider it unfinished. Also, are you aware of rpgbatch.py? Here are some examples of output: gamelist.txt, scanlog.txt,scrlog2.txt

Thank you, these are all great resources. I wasn't aware of rpgbatch.py, I figure it goes through an entire directory and scans ZIP files for proper RPG files within? I also saw the size of all those games, a whopping 4.3GB, wow! I think creating links to working gamelists makes sense for now, as that is a lot of data to transfer over. The game list will act more of an indexer/search engine for OHR games when released. I need to find the best way to grab the metadata from the gamelists, and will need to personally go through them to determine the genre. So when first released, don't expect it to work perfectly, as most games will be Unfiled, but will still act as a central list with titles and descriptions fully searchable.

[TMC]

Yes, basically. You feed rpgbatch a list of rpg files, rpgdir directories, zip files containing those, or directories containing files/zips. Each can be marked as a different source.
I've run rpgbatch on the CP, Op:OHR, and ohrhits gamelists, plus a small collection of "extra" games which were not on any of the gamelists, such as Motrya and unreleased games, and James ran it on his mirror of SS. I have been meaning to download all the games off the Bahumut RPG Community site as well.

rpgbatch is a small module used by the scripts which actually process each encountered game. They live in tools/rpgbatch in the OHR SVN repository. Here's some more examples (from analyze.py): All unique characters in fonts. All unique master palettes. (These, and most other, runs don't include SS, as only James can do that.) Fun stuff. Mostly we use them to help us make decisions about fixing bugs in a breaking way.

Edit: Genres and so on would be best labelled by community members anyway. But I think the Game type, Genre and Style categorising need rethinking and agreement first. Using tags instead would be more flexible. For example "fan game" is completely orthogonal to everything else, and so is "tech demo". Tech demo could be a game completeness level instead. So my first impression is: have game type, genre and completeness categories, and make everything else tag/label-based.

Edit2: Hmm, actually even labels within each category aren't exclusive. A game can be both Simulation and Strategy, and so on. And the style category isn't a bad idea, I just don't like the options available.

[Chronoboy]

Yes, basically. You feed rpgbatch a list of rpg files, rpgdir directories, zip files containing those, or directories containing files/zips. Each can be marked as a different source.
I've run rpgbatch on the CP, Op:OHR, and ohrhits gamelists, plus a small collection of "extra" games which were not on any of the gamelists, such as Motrya and unreleased games, and James ran it on his mirror of SS. I have been meaning to download all the games off the Bahumut RPG Community site as well.

rpgbatch is a small module used by the scripts which actually process each encountered game. They live in tools/rpgbatch in the OHR SVN repository. Here's some more examples (from analyze.py): All unique characters in fonts. All unique master palettes. (These, and most other, runs don't include SS, as only James can do that.) Fun stuff. Mostly we use them to help us make decisions about fixing bugs in a breaking way.

Edit: Genres and so on would be best labelled by community members anyway. But I think the Game type, Genre and Style categorising need rethinking and agreement first. Using tags instead would be more flexible. For example "fan game" is completely orthogonal to everything else, and so is "tech demo". Tech demo could be a game completeness level instead. So my first impression is: have game type, genre and completeness categories, and make everything else tag/label-based.

Edit2: Hmm, actually even labels within each category aren't exclusive. A game can be both Simulation and Strategy, and so on. And the style category isn't a bad idea, I just don't like the options available.

The current site is only a prototype, nothing there is set in stone. I made to show what a new game list could be capable of. In fact, using Django means I can easily mold it whichever way, it's a very flexible framework.

[Spoonweaver]

So, things seem to be moving along in this project. I have to give you props Chronoboy, putting together that sample site actually brought this project from brainstorming to In Progress. Well done!

Anyways, the 2 goals of the site that I believe should be the most focused on again are Creating a Complete OHR Game Database, and Making a Site Focused more on Game Players than Game Developers.

I believe you have good plans for the database. I'll go ahead and second TMC's idea of not listing in progress games. I do like the idea of having a page for lost games. If no one has a download for the game but other things exist for it, I think we should list it. And maybe just have a Download Lost image where the download button should be. At the bottom of the description we could urge players to email us if they in fact do have a copy of the game.
I also, think that ultimately it would be good if the site actually hosts the files, but I understand the road blocks that something like that will cause ( memory and developers not wanting an extra copy of their game hosted ) I think that if the site gets set up and becomes viable that things will change though, so for now it should just go ahead and just link to download pages.


As for the player friendly part of the goals that I believe the site should focus on. I think a lot of things need to be thought out that haven't been touched on in this thread yet. How is this site going to actually be a better game host then slimesalad? If it's only equal to slimesalad in presentation then the site lacks most of it's purpose. And, seeing as how Slimesalad has a half way decent gamelist of it's own, with filters and ordered gamelists, I think this new site has it's work cut out for it. It doesn't just have to work it has to do something that most game hosting sites don't do so well. It has to present games in a pleasing way that makes players want to play them. The site itself needs to make players excited about the games it's hosting. Basically, this site has to somehow be better than just a complete list of OHR games on the OHR wiki.

I think the site needs to have a Great title, a Logo, it's own graphical style, Great Design, Outside Marketing and a person to write reviews for games that would rotate in and out of being hosted on the main page.

[Chronoboy]

So, things seem to be moving along in this project. I have to give you props Chronoboy, putting together that sample site actually brought this project from brainstorming to In Progress. Well done!

Thanks!

Anyways, the 2 goals of the site that I believe should be the most focused on again are Creating a Complete OHR Game Database, and Making a Site Focused more on Game Players than Game Developers.

Yes, this is definitely something I am planning.

I believe you have good plans for the database. I'll go ahead and second TMC's idea of not listing in progress games. I do like the idea of having a page for lost games. If no one has a download for the game but other things exist for it, I think we should list it. And maybe just have a Download Lost image where the download button should be. At the bottom of the description we could urge players to email us if they in fact do have a copy of the game.
I also, think that ultimately it would be good if the site actually hosts the files, but I understand the road blocks that something like that will cause ( memory and developers not wanting an extra copy of their game hosted ) I think that if the site gets set up and becomes viable that things will change though, so for now it should just go ahead and just link to download pages.

Download Lost does sound like a good idea. The current version of the site only supports self-hosted games ironically. I will be adding the ability to link to another gamelist's download page to create the central database idea. I will be leaving in the hosted component for developers who choose to use it.

As for the player friendly part of the goals that I believe the site should focus on. I think a lot of things need to be thought out that haven't been touched on in this thread yet. How is this site going to actually be a better game host then slimesalad? If it's only equal to slimesalad in presentation then the site lacks most of it's purpose. And, seeing as how Slimesalad has a half way decent gamelist of it's own, with filters and ordered gamelists, I think this new site has it's work cut out for it. It doesn't just have to work it has to do something that most game hosting sites don't do so well. It has to present games in a pleasing way that makes players want to play them. The site itself needs to make players excited about the games it's hosting. Basically, this site has to somehow be better than just a complete list of OHR games on the OHR wiki.

The advantage of this site will be that it is like a clay we can mold and can support anything our imaginations can ponder.

I think the site needs to have a Great title, a Logo, it's own graphical style, Great Design, Outside Marketing and a person to write reviews for games that would rotate in and out of being hosted on the main page.

Yes, as seen from the games I attempted to develop, I'm not an awesome graphics artist. I specialize more in programming and logic(I'm like Spock). I hope someone from the community will be-able to contribute a website template and some spiffy graphics later down the development line.

Update:
Just had a new idea which lots of people in this community would like, a Let's Play Insert OHR game here section, this would be-able to centralize all of the Let's Play's from YouTube allowing potential players to see the game in action super easily.

[Sparoku]

Although a python function to check a game's password is perfectly possible (and nohrio might already do it) passwords are *very* easy to crack, so knowing it is not sufficient proof of game-authorship.

Shhh. -- I thought the password used a one way hash of some sort.

This is true! ... but even as a one way hash, it is still super easy to crack!

I must be retarded then, cuz I've been trying to find a way to crack a certain game just to update it's God-awful graphics (Not even to re-upload it or anything) and haven't been able to.

Also, sorry if I necro'd this.

[Bob the Hamster]

I must be retarded then, cuz I've been trying to find a way to crack a certain game just to update it's God-awful graphics (Not even to re-upload it or anything) and haven't been able to.

If it is your own game, just mail it to me, and I can crack it for you.