Hi *,
just a short question:
If there is a page, and User A comes along and flaggs it, after which User B comes along and flags it another way, what do we do?
a) User B overrides User A, we only store the latest flag b) User Bs flag is used, because it was the latest (keeping both flags) c) Apply some method to determine an average d) Show all flags
What do you think?
Cheers,
Joerg
On 10/03/07, Joerg Baach lists@baach.de wrote:
Hi *,
just a short question:
If there is a page, and User A comes along and flaggs it, after which User B comes along and flags it another way, what do we do?
a) User B overrides User A, we only store the latest flag b) User Bs flag is used, because it was the latest (keeping both flags) c) Apply some method to determine an average d) Show all flags
What do you think?
I would think that we should store all flags (for blame purposes, if nothing else ;-)), but we should use the most recent state of each flag for each revision - if A sets flag X true (or whatever), but ignores flag Y, and then B sets flag Y true and flag X false, we should show "X false, Y true, Z unset" (probably skipping unset flags)...
Individual authors' actions are equal ("idempotent", as it were) - this isn't meant as a voting system, nor AIUI is it meant to have a concept of one users' flag outranking another's based on the /user/ (so that one flagger being a sysop and the other not should have no effect on the weight given by the code to their actions) - this seems to be the way that we've developed the idea of wikis for MW and Wikimedia's wikis especially.
Thoughts?
Yrs,
I don't think it is useful to have a feature for setting a version that is flagged sighted as unsighted. The point is: if the version was flagged as sighted and shouldn't be, then it contains vandalism. This vandalism should be removed and thus a new version should be created, which automatically gets the sighted-flag, but this time for the correct reason. The the last editor should then approach the editor who set the flag wrong.
For other flags this is indeed more complicated. In the case of reviewed, the possibility of storing all flags seems to be reasonable, as well as the possibility of removing flags. If a flag is removed, it should still be possible to see this in some way, for the reason James mentioned (blame purposes).
As for the rest, I agree with James, flasg from different users shouldn't be weighted.
Bye,
Philipp
On Saturday 10 March 2007 15:22:28 Joerg Baach wrote:
If there is a page, and User A comes along and flaggs it, after which User B comes along and flags it another way, what do we do?
a) User B overrides User A, we only store the latest flag b) User Bs flag is used, because it was the latest (keeping both flags) c) Apply some method to determine an average d) Show all flags
What do you think?
I agree with P. Birken. It is not useful having a feature for unsetting a flag. If there is a problem with the currently flagged version (for example vandalism, or the half done article example with an accidential flag) then a newer version has to be created and/or flagged. Someone who complains about a specific version has some specific critics on content. If it is really such a problem for him he should solve the problem himself. Everthing else would quickly cause lame vote wars (you know voting is attractive because you're not forced to think nor need to write anything ;-).
Furthermore I am against multiple flags of the same kind for one version (and even making an average out of it). This renders flagging into voting. This is something we strictly have to avoid. Voting or voting like procedures instead of flagging render creation of sock puppets much more attractive. Something we definitely have to avoid.
Beside that the history page would look much more complicated and crowed with information if it is possible to change flags afterwards.
That said I am not against flagging an article as "sighted" and "checked" as these are two different things (so for example a version could have the "minor", "sighted" and "checked" flag). However it would be useless to allow people flagging an already "checked" version as "sighted".
Cheers, Arnomane / Daniel Arnold
wikiquality-l@lists.wikimedia.org