I’m seriously asking how you think your password will get leaked. is there some flaw in the socket code I posted?
Yes, that’s exactly what I meant. There’s a huge gaping hole in your code that nobody has pointed out yet, and I think it’s going to leak all our passwords.
:
well it’s not like you told me what WOULD leak it!
ffs … :-\
That was the point of my sarcasm.
If I knew what would leak it, I wouldn’t think it would be leak it, as I would tell you about it.
Call it a case of dont-trust-hacks-with-your-password-itis.
well you kept saying your password would leak, so I was seriously curious. half of my decisions for java unlimited is by asking for opinions and advice. I believe this has made the site more solid and enjoyable.
but, if you’re using a password that you would protect with your life on a forum … that’s a mistake right there already. if you trust this site, you should trust a site that is doing the same thing you do every day to login.
more ideas welcome.
You’re posting the passwords through HTTP rather than HTTPS. The result of this transfer is that anyone who’s standing in the middle of the path of your packets (or on the same subnets) could intercept the packets, decode them, and have your password. For safety reasons you should really be using an HTTPS post.
What I find laughable about this is that JGO doesn’t use HTTPS to log in its users. Passwords are sent in the clear, period. Which begs the question why Markus is using a highly sensitive password that he feels he is unable to lose on a site that is extremely low security.
Remember Markus, security is like an Onion. Keep those layers separate. 
(you probably mean “raises the question”, not “begs the question”)
Thanks for agreeing with me, jbanes.
That’s another reason not to send the passwords to woogley. Instead of one network to worry about, there suddenly are two networks to worry about.
If both JGO and Woogleys site used https, I would trust this idea more.
The “highly sensitive password” is the password for my login on this site, and a bunch of other web forums I don’t consider important enough to give unique passwords. The worst a hacker could do is to post a bunch of posts in my name.
But that doesn’t mean I’m going to give you or anyone else my password.
Allright, here’s my honest opinion.
Obviously, the more popular games will get noted in one way or another (if not due to download counts and high votes, then because some site will link to them, and generate more attention). This extra exposure will surely let them have more viewers, and thus more points. Popularity will be exponential, so that while less popular games might end up with nothing more than a few points, the popular ones might get hundreds of points. This, in my opinion, has to be thought over.
First of all, since I don’t believe cheating would be such an issue (and, rather easily detected at that, unless they’re changing IP once a second, and you don’t do that), I’d say public voting would be best. Second, I don’t agree with the “72 hour poll idea”, since I believe the “three favorite games” will be the same ones for each and every user (or, almost), giving perhaps 5 or 10 games hundreds of points, and leaving the rest with none. Absolutely no points at all. A weighted average based on the 1/2/3/4/5 for every game would be better. Now the problem with this is the timing issue - newer games would have been exposed for shorter periods of time, and the average would mean less with fewer votes, etc, etc. You could solve this using two techniques:
- Make sure the system works, check it, double-check it, but do not put it on JU until the contest is over. That way, every game can have the same exposure time-wise.
- Use a weighted average of some sort, something like that the IMDB uses for the top 250 movies:
EDIT: Minor changes.
I think that this algorithm would work fine, as it would balance the voting so that all games would get about the same amount of votes:
`
Let MaxDifference = 20
Let M = GetGameWithLeastAmountOfVotes(Games)
For each game G in Games
Let G.youMayVoteOnThisGame = false
Let G.showThisGameInTheListOfGamesYouCanVoteOn = false
If G.votes < (M.votes+MaxDifference)
Let G.youMayVoteOnThisGame = true
End If
If G.votes < (M.votes+(MaxDifference/2))
Let G.showThisGameInTheListOfGamesYouCanVoteOn = true
End If
End For
`
Then you have a special page for voting on games where their information is shown, only games with the property showThisGameInTheListOfGamesYouCanVoteOn == true would be shown. The games are shown in ascending order of votes. New votes won’t be accepted on games with a property value of youMayVoteOnThisGame == false , they are not the same property to handle lots of voters in a nice way.
This will make sure that all games get reviews and thus have a fair chance in this competition. Any “non-vote balancing” system who presents the games in a certain way will be more or less unfair, perhaps it still will be “fair enough” but I would prefer this method.
[edit] Corrected the algorithm
Agree!
A bit over the top, but sure why not. Would be a shame if a game gets 500 votes and an average of 4.8 and gets beaten by a game that gets 3 votes and average of 5.0…
Reagrding pwd:
Woogley, would it be possible to change it so that instead of “checking login at JGF” you could automatically send a message to a user (from your user account maybe)? If that could work, then each judge would only have to give you its login name. You could then send a “voting key” (possibly generated by the user name, so it would be the same if requested several times) to the users message box… The judges would then only supply the voting key when they give grades. That way no JGF pwds would be exposed, and only JGO users can vote to prevent cheating (might have to check so that users with < 10 posts can’t vote, but I doubt it). Probably a bit more complex for you to create, but a good challenge right?
If that doesn’t work, then we might want a poll on who would take part with the “insecure methode”. It might be the best option anyway, if it’s anly a few who refuse to take part. There is not discussion that it is just “wrong way of doing it”, but I’d just change my pwd before voting and change it back when done, no need to make a big deal out of it I think.
[quote]A bit over the top, but sure why not. Would be a shame if a game gets 500 votes and an average of 4.8 and gets beaten by a game that gets 3 votes and average of 5.0…
[/quote]
I would say that my simple method (see my post above) is better as it will make sure that all games get ratings, something a weighted vote algorithm won’t make sure of. If my method is used we won’t even need a weighted vote solution, as all games will more or less have the same amount of votes. And the least significant argument, but still an argument - my method is very simple to implement.
the “insecure method,” as you call it, is the exact same method you are using to login to this very forum every day. since this is true, I’m not really considering taking that longer route at this time (because that’s almost as annoying as registering to java unlimited anyway)
as far as the forumula goes… it’s a nice idea and everything, but for 30 games? I highly doubt any of these games would get 500 votes. and as far as some games not getting enough attention, well, that’s part of the challenge. Look at how jbanes generates attention - storyline, website, etc. If you’re going to post some puzzle game without a working screenshot or whatever, then of course you won’t get a high download count.
[quote]highly doubt any of these games would get 500 votes. and as far as some games not getting enough attention, well, that’s part of the challenge.
[/quote]
What about my suggestion then? It would be dead simple to implement, and it would result in a more fair competition.
I am still considering yours, Donald, but are you saying we forbid voting on a game that clearly deserve more votes …?
No, you’d have to use an average instead of possibly unlimited scores - remember, games don’t deserve more votes, they deserve more points. Donald’s idea works just as well as mine, although I can see problems with that one too. Nevertheless, I much prefer it over the system you proposed. Besides, I don’t think storylike, website, etc should count for much, since they are not fit within the 4k, and they aren’t truly part of the game (the game is what’s important here, not how good you are at generating attention.). Granted, attention-grabbing effects and additions such as these should clearly be allowed, and they’ll help you generate attention, not high votes. Making more people look at and enjoy your game is one thing, getting votes as a free addition is an entirely different matter.
EDIT: Corrected grammer a bit. Changed text slightly.
EDIT2: Oh, one more thing. You said 30 games - let’s hope that number increases. I certainly believe it will, since there may be quite a few new entries the last couple of days before the deadline
[quote]I am still considering yours, Donald, but are you saying we forbid voting on a game that clearly deserve more votes …?
[/quote]
What I am saying is that we should never let the difference in votes between the game which has got most responses and the one with the least number of responses should be larger than MaxDifference.
How much points a vote would contribue with is important here. For example on a scale from 1-10 then 1 isn’t that much and will lower the mean value, while 10 is indeed very positive. The algorithm simply forces all games to get votes, if that’s a good or a bad thing depends on if the game is good or bad. The important thing is that all games would get feedback from the users, so that people just doesn’t vote on a few games and don’t bother trying the others.
When the number of votes on all games are about the same all games will be available for voting. Also a single user should only be allowed to vote on a specific game once, but I took that one for granted. Games who “deserve votes” will get their votes as soon as the game with the least amount of votes “catches up” (and it will), and as those games are of good quality the votes will be of high value.
The focus on using this way is that all games should get responses, not to alter the results in any way. Great games will still get great scores and the game most people give high scores will win.
edit: I corrected my faulty way of thinking.
Oh, by the way - I’d love to see that comment feature implemented. If you have the time, woogley, please do, because that way non-JGO users could comment on the games so that we could improve our games according to their ideas
I sort of see the concept, but I still think the voters would be a bit dumbfounded to find that their favorite isnt on the list because it was voted too many times
don’t forget I said you’d be voting your top 3 games, what are the chances of everyone having the same top 3? maybe top 5?
take my opinion with a grain of salt because I havent fully grapsed your idea yet, Donald. but it’ll probably click after I get home from work 
Morre, yes, I admit I have put that off too long
I know it’s top 3, but I’d say it have to be top 10 in order for it to work well. Now, top 10 wouldn’t work anyway, because that takes too much time. Not for me, not for most developers I think, but for the casual browser (and we’ll have to hope and count on that they find the site - that’s what the new contest and site is all about, isn’t it?)
With top 3, I still think 5-10 games would get all the votes, and the rest of the games would score zero. Literally.
[quote]don’t forget I said you’d be voting your top 3 games, what are the chances of everyone having the same top 3? maybe top 5?
[/quote]
Well in that case the method I am pushing for can’t be used, it can only be used if you are allowed to vote once on all the games in the contest. If that is not the case, then the situation you describe can indeed occur; where a game who I want to give my vote can’t be found in the “currently available for votes” page.
Personally I would like to vote on all the games, as I would take the time to give them all a chance and rate them as soon as I played them. In case you’ll change it into that then the method I’ve talked about can be used (if it’s not that clear when I talk about it, check the psuedo-code as I think it speaks for itself). I would prefer it this way, but I guess you have some reason to only allow three votes.
The important thing is that Markus Persson won’t win! Just kidding, he will get what’s coming to him from me whatever voting system is used.
[quote]With top 3, I still think 5-10 games would get all the votes, and the rest of the games would score zero. Literally.
[/quote]
True, if voting for your own game isn’t allowed.