Steven thinks Oceangrams are doomed to failure, because sooner or later the spammers will infest the thing and then 99% of the messages will be for
C*H*E*A*P drug5 n0 perscr1pshin neaded! I’m sort of amazed this hasn’t happened already. It does seem sort of obvious. Maybe the community is just too small to go after. They just now hit mesage #500,000 delivered last night. Since the site has been running for a while now, this means the message volume is probably too low to interest spammers.
But this thing presents a few other interesting hurdles that spammers would need to overcome:
- If you wanted to spam the Ocean, you would have to write some new software. The interface is a flash game, so you can’t use existing bulk mail programs. You have to reverse engineer whatever protocol the Oceangrams use. I’m sure this wouldn’t be a big job, but it would still need to be done so that your spambot had an interface to the Ocean.
- URLs do not become hotlinks in bottle messages, so users would have to cut & paste the URL you sent them into their web browser. This is going to lower the spam usefulness even more. If people respond to 1% of all spams when they can click on the link, how much of that 1% will be willing to take the extra step of cut & paste? If the response is lower, then spammers need to make up for it with an even LARGER message volume.
- Unlike email, you don’t know what happens to your message. A spammer can experiment with spam filters and see what messages make it through, but if there is a spam filter on Oceangram, how do you know? The system can silently eat spam messages and the spammer will have a hard time discovering which ones, and why.
- The reduced delivery rate will exacerbate all of the above: For each connection, you can only get a message every 5 mins or so. So the spammer will need a LOT of connections if they want to see if their messages are making it through.
- Unlike emails or USENET, the server can easily throttle sending speed so that you can’t send messages faster than once every 10 or 15 seconds. That means the spammer is going to need a LOT of connections. The server can combat this by limiting the number of messages sent from a single IP. This means a user can open 20 windows to GET grams if they want, but they can still only send at the rate of once every 10 seconds or so. Better yet, ALLOW the messages. If the spammer sets up 100 instances and is pounding away at the server with several messages a second, it should be easy to spot. Instead of rejecting, just accept them and then silently discard 99% of the messages. It will take the spammer a long, long time to figure out what the message threshold is and how fast they can safely send.
- The system has a fixed delivery rate based on number of users, not the number of messages. If a spammer emails 100 spams, then you will get all 100 next time you check your email. If they add 100 spams to the Oceangram pool, it increases the odds that you will get spam, but you’ll still only get messages once every five minutes or so.
None of this will keep spammers out forever. These are hurdles, not walls. But this may discourage them until Oceangrams gets a good deal bigger than it is now. They probably won’t bother until the community is large enough that it will be worth the initial investment of coding time to get it rolling.
I’ve been signing my ‘grams “- SY”. I haven’t started putting the link to my site into my messages yet, though. Heh.
In Defense of Crunch
Crunch-mode game development isn't good, but sometimes it happens for good reasons.
What did web browsers look like 20 years ago, and what kind of crazy features did they have?
Programming Language for Games
Game developer Jon Blow is making a programming language just for games. Why is he doing this, and what will it mean for game development?
There are two major schools of thought about how you should write software. Here's what they are and why people argue about it.
Silent Hill 2 Plot Analysis
A long-form analysis on one of the greatest horror games ever made.