[SciPy-Dev] Voting functionality on scipy central
Sun Sep 22 13:56:20 CDT 2013
On 22 September 2013 18:51, Surya Kasturi <email@example.com> wrote:
> On Sun, Sep 22, 2013 at 4:09 AM, Sebastián Ventura <firstname.lastname@example.org>wrote:
>> Personally, I believe the best solution would be to have reputation be a
>> field on revisions that gets copied (from the last revision) when a new
>> revision is created for the same submission. But after the new revision is
>> created, it is no longer linked to the past revision (so both of them can
>> be voted separately).
>> This idea is quite different. Could you please tell me why is that only
> last revision should carry on reputation? If there are 3 revisions, should
> 3rd rev carry reputation of 1 and 2 or only 2?
All revisions would have reputation, not just the last one... I don't know
if that answers your question. Here's a simple example: Laura makes a new
submission, and it gets 3 upvote; those 3 upvotes are stored on the first
revision; Laura then creates a new revision for the reputation, both
revisions now have 3 points. Someone downvotes the first revision and
upvote the last one: now the first revision has 2 points, and the last one
has 4 points. If someone now were to create a new revision for the same
submission, it will have the same reputation as the last revision, 4
points. Someone again could downvote that revision, and it would have 3
points, while the first revision would still have 2 points, and the second
revision 4 points. Is it any clearer?
> I think we need to be a bit clear as to when a new revision is usually
> 1. Does new revisions are made making some minor changes? or
> 2. Does revisions are made only if there is change in functionality of
> More generally, what kind of changes are made usually in revisions?
> If the changes are very minor and latest revision is almost same as
> previous revision, its totally fine to carry on the reputation.
> But if there were any major, significant changes in latest revision, we
> never know which one is better until another user votes.
Maybe we can just add an option to copy reputation or not? I'm not even
sure if that would work, just throwing it out here.
> SciPy-Dev mailing list
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the SciPy-Dev