Request Programs

[nevon]

I can't help but feel that this is a problem that has already been solved. I mean, we're hardly the first people to think about this problem, and there are already achievement systems out there. So how have other people solved it?

[EMB]

I don't know what language you're thinking about building the backend in

PHP, very powerful and is installed by default on almost every web server.

Now, my thoughts on stopping fake achievements is that we may aswell not bother making it secure. Look at the biggest platforms out there, their achievement systems are broken routinely and the world doesn't stop. Securing it would also make it alot harder for the developer to create achievements.
Nevon, your post popped up just as I was typing this and to answer your question, they haven't.

[tentus]

I've thought about that. Although, I admit, I was only thinking of a technical solution.

A scenario which it doesn't fix, is something like:

Johnny tries to win an achievement for winning a level with 100% accuracy. He finds it hard, but still wants to brag to his friends, so after five attempts, he just starts Notepad and types:

Code: Select all

ass.achieve('perfect_aim')
ass.send()

He brags about his skills and his friends envy him, so decides will do it more often, only using appropriate spacing so they don't get suspicious.

The question is then: how much of a problem is things like this (imagine loads of people doing this, it isn't hard to come up with)?

Fair enough. As an asocial personality, I'm not really bothered by that situation, but I can see how people would feel that it devalues their achievements. I'm strapped for ideas that address that which don't have truly massive work requirements though.

[nevon]

I don't know what language you're thinking about building the backend in

PHP, very powerful and is installed by default on almost every web server.

Well then it's easy as shit to switch between databases (unless you're planning to do fancy stuff, in which case I'd say just target MySQL for now).

[Robin]

As an asocial personality, I'm not really bothered by that situation, but I can see how people would feel that it devalues their achievements.

Never underestimate the power of social thinking in humans.

(unless you're planning to do fancy stuff, in which case I'd say just target MySQL for now).

If openness is an issue, I'd like to suggest a recent fork of MySQL rather than MySQL itself. I don't know much about this, but there seems to be something going on in database land and it looks like MySQL proper is going the way of OpenOffice.org.

[EMB]

Right, I'll start coding the backend.
Hopefully it will simply be a single HTTP POST request, containing all of the relevant data. (Anyone oppose this method?)

(unless you're planning to do fancy stuff, in which case I'd say just target MySQL for now).

If openness is an issue, I'd like to suggest a recent fork of MySQL rather than MySQL itself. I don't know much about this, but there seems to be something going on in database land and it looks like MySQL proper is going the way of OpenOffice.org.

What do you mean by this?

[Robin]

What do you mean by this?

MySQL is currently owned by Oracle. Oracle... doesn't like open source very much. While they can't stop it, they are still trying and this has led to the forking of OpenOffice.org into LibreOffice, and MySQL into a myriad of forks. I don't know which is the most promising, but I do know they all freely exchange source code, so I doubt it will matter very much. I doubt MySQL itself will be able to keep any kind of momentum. If you're interested in the subject, Henrik Ingo's blog might be able to provide more and better information.

[EMB]

MySql is stable ATM. PHP natively supports it. If it does eventually collapse, I'm sure PHP will make it easy for people to change.
Besides, phpBB runs off of MySql so if it is hosted on the LÖVE server (As I would expect.) then it means less work.

[BlackBulletIV]

Here's my opinion on binary merging and achievements. While binary merging doesn't hide your code, it will appear to hide your code to most users. Only a select few users would know to run it through an unarchiver, either because they know what the game is built with, or they're experimenting with different things to get at the code. To most users, it would appear that the source code is locked away (and anyway, many people in the public wouldn't know how to read code in the first place).

[EMB]

Just found a problem: passwords.
Personally I don't wish to transfer them in plaintext, but Lua has no hashing methods as standard...
Anyone got any ideas?