Voting - Ticket System

[boecko 0]

Hi,

i've switched on the sponsorship-mechanism of mantis and modified it a bit:

* money -> votes

* you can give a ticket votes from 1-10

Benefit?

BIS would see, which tickets are troubling most of the users or which features should be implemented.

This doesn't affect priorities.

Feel free to comment. Bye

Edit:Voting is online .. no need for testing

[kronzky 5]

Perhaps I should get some glasses, but where do I actually place my vote? I couldn't find a voting button anywhere on the form...

And how many votes do I have, and what happens when I run out of them?

Update:In case anybody else might run into this issue, and you're running Firefox, check your userContent.css.

Most likely you're having an ad suppression definition in there that is looking for a layer called "sponsor". Remove that and you're all set.

[boecko 0]

Perhaps I should get some glasses, but where do I actually place my vote? I couldn't find a voting button anywhere on the form...

And how many votes do I have, and what happens when I run out of them?

Click on the link:

http://bugs.armed-assault.net/view.php?id=1623

There must be the text "Users voting for this issue" in english.

[suma 8]

BIS would see, which tickets are troubling most of the users or which features should be implemented.

This doesn't affect priorities.

I like this a lot. Such feedback is very important for us. I will be using the view sorted by the number of votes a lot.

[INNOCENT&CLUELESS 0]

a dream becomes true! The restriction is so that each mantis user can only give one vote, and you can even change it if you changed you mind. Excellent! Stupid Question: 10 equals highest support?

[boecko 0]

a dream becomes true! The restriction is so that each mantis user can only give one vote, and you can even change it if you changed you mind. Excellent! Stupid Question: 10 equals highest support?

More votes, more support.

but I would make lower range ...

e.g.

1-5

regards

[Monkwarrior 0]

I like this a lot. Such feedback is very important for us. I will be using the view sorted by the number of votes a lot.

Allrighty, that's all we needed to hear.

Let's all vote now gents.

Monk.

[boecko 0]

Ok .. ladies and gentlemen ..

Voting is online ...

you can give 0-5 votes to each bug/issue

If you see some problem, report it in the 'mantis bts'-project and vote for it.

Side effect: I disabled all languages besides english. I don't want to translate into polish, russian or czech .

[messiahua 0]

Very good feature, I should say... Now everyone can show how important is the issue for them

Quote[/b] ]Side effect: I disabled all languages besides english. I don't want to translate into polish, russian or czech .

If there are not much strings to translate, then I can help with russian

[sbsmac 0]

I like the concept but I'm not convinced that allowing a range of values for each user (1..5) is terribly useful. As I understand the system at the moment, users can vote for as many issues as they like. There's therefore no incentive not to add 5 votes for every issue a user is interested in.

I'd suggest that there is just as much information about user interest available to BIS if you allow users to only post one vote for each issue and continue to allow them to post votes for an arbitrary number of issues. Even this suffers from the flaw that users are not really being asked to express a preference since they can say they are interested in everything.

Presumably the original 'money' concept allowed users to 'spend' a limited number of votes across some arbitrary collection of issues. My own opinion is that this is more complicated than it really needs to be. I suspect most forum users here rarely visit the BTS and certainly aren't going to want to juggle their virtual bank-balance when they do.

The simplest solution would be just to allow users to choose their top 5 or 10 open issues so that at any time BIS could see the '10 most wanted bug-fixes/features'. OTOH, I appreciate that Mantis probably doesn't support a straight-forward concept!

[boecko 0]

I know,

but that's the only way it can work right now.

I've already implemented a max value. because this was unlimited in the original underlying sponsorhip model.

see

http://www.mantisbt.org/bugs/view.php?id=668

But I think.. we should try it first. I'm confident, that we will see a representative top-10orwhatever list.

Bye

[kronzky 5]

Yeah, I don't really see the need for a "rated" vote (1-5).

Then it just comes down to who "screams" the loudest...

One person, one vote.

Keeps it easy, and then the total is a true representation of how popular that particular issue really is.

[boecko 0]

Ok ...

changed it to one vote.

This eradicates the need for a vote contingent for each user. This would be a problem with the unknown amount of ArmA Bugs .

Other changes:

* disabled email notification after voting

* disabled automatic monitoring

* reduced the Issue History-Activities for each vote

regards

[sbsmac 0]

>I've already implemented a max value

Boeko, why not just set the max-votes to 1 rather than 5 ?  At least then you have the situation where users can either vote for an issue or not.

It would also be really interesting to see some stats eventually on how many active users of the system there are and how many are voting.  I suspect that there might need to be a bit more promotion of the system before it gets enough use to start showing real preferences.  I've taken to pointing people to the BTS whenever possible but there are still a lot of forum users who probably never visit it.  Maybe you could also make the default filter sort by votes so that users are immediately exposed to the new functionality ?

*Edit* Ah - you did it whilst I was posting

[kronzky 5]

Would it be possible to disable people voting on their own submissions?

Otherwise *every* report would end up with that mandatory self-vote to avoid being flushed into the abyss...

[boecko 0]

Na ...

it's not the reporters bug, it's a bug of ArmA.

So the reporter was the unfortunate to report it first.

I don't think this distorts the statistic.

[raedor 8]

I like this system as well. Now go ahead, make "Spread BTS voting" signatures and get some people voting, not only whining about bugs.

[suma 8]

I like the concept but I'm not convinced that allowing a range of values for each user (1..5) is terribly useful. As I understand the system at the moment, users can vote for as many issues as they like. There's therefore no incentive not to add 5 votes for every issue a user is interested in.

This was my initial though as well. However, thinking about it somewhat more, actually adding 5 to all bugs is the same as adding nothing to any of them - it has no influence on sort ordering.

Still, it is true one dedicated person with a lot of spare time could influence a lot, but such think can be detected by the admin anyway (if it really happens).

[sbsmac 0]

Quote[/b] ]Still, it is true one dedicated person with a lot of spare time could influence a lot, but such think can be detected by the admin anyway (if it really happens).

Working on the weekend ?

My point was really that there was really very little information to be obtained from the fact that someone has given an issue the maximum number of votes so you might as well make the maximum number 1 as Boeko has done. As you say, the ability to vote on an unlimited number of issues means that 'gaming' of the system is currently possible in theory but the 'cost' (ie, effort) of doing so probably means that it won't happen.

The more interesting question is how (if) the community can do something to help with processing the large number of open issues (over 440). Even reading the details of each of these reports list probably takes the best part of a couple of days ! Clearly, addressing them all is unlikely. At our company we tend to assume an average of 1 resolved issue per man-day is pretty good going! Maybe some of us could move stuff into the 'confirmed' state after reproducing problems on the latest build and verifying sufficient information is present in the report ? This would at least cut out some of the initial investigation that would otherwise be required by BIS. I'd be happy to help with this as time permits.

*Edit* Ok so the discussion here

goes some way to addressing this by temporarily rejecting issues with not enough information. It doesn't handle issues which _have_ been confimed by others though.

*Edit again* Hmm - I should read the entire forum before posting Guidelines discussion talks about the 'acknowledged' state which I previously thought meant that BIS had acknowledged the issue.

[suma 8]

I like this a lot. Such feedback is very important for us. I will be using the view sorted by the number of votes a lot.

Having sad that, it seems such sorting is not working now at all. I tried sorting both ascending and descending by Votes, but voted issues are never listed at the top.

[boecko 0]

@Suma

it works ...

but 5 sticky issues are always on top .. because they are sticky

[suma 8]

Error on my side: I had some filter set from before, which I have forgotten about, but the server did not. There were no voted issues satisfying the filter.

[INNOCENT&CLUELESS 0]

About possible "abuse" of voting:

Assuming that some BTS users would abuse the voting by

organizing voting to push a certain issue and ask/force users to give highest vote to a certain TT I guess that can be avoided by:

- selecting the votes you are interested in which means (if boecko can make that filtering on user name lists) for editing&scripting TTs BI should be able to assemble a list/several lists of users also engaged in mission editing/addon making/island making which have a name already. If BI for example gets the opinion that I vote for every bullshit I can excluded from that list. It is a kind of racism, but it would help a lot.

So BI should be able to assemple several usergroups(lists) which are not related to any query as such. It should be possible to give this list a name. If BI makes a query for a certain bug/category/all bugs, they should be able to see the scoring based on all users and based on the selected usergroups.

Hence BI could have a telling view how different users would think about a certain issue.

@boecko: feasible?

@BI: would this help?

[boecko 0]

@I&C

no too much rocket science. .

[suma 8]

I think we can wait for I while to see how much abuse there will be and not waste effort now on fixing problems which perhaps never come real.