... If someone wish, he can cheat the on-line voting.

It's not like we were having an election for PM's President, or some such... :)  I always thought of these polls as being for entertainment only, and to provide some context for funny, wise or witty remarks, which all in all just helps to get to know each other a little better, and gives some rough feedback what "the community" thinks.

I've personally never felt the urge to vote more than once, except maybe if I couldn't decide on which option to vote. And even then, I just picked one and went on with life.  I'd suppose the majority of Monks are handling this in a similar way (?)

How does that help any? Everyone can now vote more often than before! The majority of people currently have one vote, but you just gave them a second vote.

If I'd make a change, it would be to give the following results:

+--------------+-----------+------+--------------------------------+
| IP Address   | User      | Vote | Comments                       |
+--------------+-----------+------+--------------------------------+
| 192.168.1.10 | vcTheGuru |  Y   | First instance of IP and user. |
| 192.168.1.10 | Anonymous |  N   | IP already voted.              |
| 192.168.1.10 | tye       |  N   | IP already voted.              |
| 192.168.1.10 | Anonymous |  N   | IP already voted.              |
| 192.168.1.11 | vcTheGuru |  N * | User already voted.            |
| 192.168.1.11 | Anonymous |  Y   | First instance of IP and user. |
| 192.168.1.11 | tye       |  N   | IP already voted.              |
+--------------+-----------+------+--------------------------------+
[download]

The * marks the only change from the current setup.

However, the gains from this change are *very* minor: It would force the "cheat" to logout to place any vote after the first. It won't stop him from placing as many votes as he has IP addresses. It's not worth the effort.

Congratulations, you two have come up with two of the four "solutions" I came up with long ago. One makes things a bit fairer for people behind a firewall with a shared public IP address while doubling the chances for repeated voting for the most typical situations. The other does the best job of reducing repeat voting (but is still trivially worked around and so doesn't improve the current situationn all that much).

So neither is much if any of an over-all net improvement. Hence, I don't see much point in spending resources to implement either.

For the record, describing all five options in terms of how the code would work rather than via an example chronology:

Current method. You can't vote if:

• Someone has already voted from your IP address

First proposal above (by vcTheGuru). You can't vote if any of the following are true:

• You aren't anonymous and have already voted under this user ID
• You are anonymous and someone has already voted anonymously from your IP address.

Second proposal above (by ikegami). You can't vote if any of the following are true:

• You aren't anonymous and have already voted under this user ID
• Someone has already voted from your IP address

The proposal I first seriously considered long ago. You can't vote if any of the following are true:

• You aren't anonymous and have already voted under this user ID
• You are anonymous and somebody (anonymous or otherwise) has already voted from your IP address

The second proposal I considered long ago. You can't vote if any of the following are true:

• You aren't anonymous and have already voted under this user ID
• You are anonymous and somebody (anonymous or otherwise) has already voted from your IP address
• You aren't anonymous and somebody has already voted anonymously from your IP address

I personally think that the last one is the best of the bunch. But it still doesn't reduce the number of duplicate votes that anyone can cast. It is a bit "fairer" for people sharing a public IP address, but then most people have another public IP address that they don't share (at least not with the same people), especially if they go slightly out of their way. So, when you consider that this is just about rather silly polls and people can still reply to them, it never became a priority to me to thrash the DB design and rewrite the "deal with poll votes" code (which isn't just "one place", FYI).

Oh, vcTheGuru wrote:

Hi, thanks for the reply and considering this seriously.

Well, I hope this reply does show that I had already considered this seriously (if, in fact, only briefly). For the record (just in case you didn't notice), my previous reply was a bit sarcastic. In particular "this is our highest priority" was irony. Other parts contained sarcasm as well but I was sincere in noting that you hadn't demonstrated that you had considered all of the possibilities. And I am glad that you did follow through on that point.

The other part nobody has yet mentioned is voting on previous polls. Something that is also not worth "fixing", IMHO (it isn't, strictly speaking, "broken" and you can't really "fix" it completely either, but there could certainly be improvements made on some fronts but not without trade-offs that are too expensive on other fronts).

- tye        

"I feel like none of the method is fool-proof someone wish to screw up any voting :(" ding ding ding we have a winner! ;)

The only way to "fix" this is to disallow anony votes which wouldn't be very nice so, oh well.