On Mon, Mar 28, 2011 at 7:55 PM, Fred Bauder <fredbaud(a)fairpoint.net> wrote: Where do you suggest to store it? --vvv

On Mon, 28 Mar 2011, Victor Vasiliev wrote: There's no reason an ordinary comment on the talk page can't be used for this purpose. While it's true that talk pages can be edited by others, the history cannot be edited and will always show what text actually came from the subject. It will still have to be verified that it really is from the subject, but that's true for all such notes whether on the talk page or not.

On 28 March 2011 16:55, Fred Bauder <fredbaud(a)fairpoint.net> wrote: Mmm. But those authors and editors usually have the personal note sitting in their files, and are entirely confident that it's verifiable - because it is, for them. But on a collaborative project, this doesn't scale; we can't have the "personal communication" copied to a thousand editors. So we fall back on services like this "iCorrect", or encouraging people to post things on their own websites, etc etc. The issue of whether or not we'd accept a personal note is less critical - I think most reasonable editors are willing to accept first-hand commentary as valid for uncontroversial points, and are capable of weighing it accordingly (or ignoring it, where appropriate) for controversial ones. There's no danger that we'd feel obligated to change articles purely on the subject's say-so if it didn't seem a good idea from an editorial perspective - we're good at handling that sort of concern!. Mainly, we need to worry about the best way of getting first-hand commentary out there for us to use it, and of doing so in a manner appropriate to a collaborative project. So something which can validate ID, and do so with appropriate reliability and privacy, in a manner that makes it trustworthy, but then makes the correction (or rebuttal, or what have you) public. iConnect seems to do this, but has the undesirable aspect of charging a high fee to do so - thus severely limiting its practicality. So, how can we square the circle of making the correction public, whilst keeping the information which verifies the identity of the author private, in a trustable fashion - and do this all in a way that works for *us*. Are there existing services out there which *do* undertake to verify the identity of their users which we could work off, or are we going to have to create one - at suitable arms length from WMF? -- - Andrew Gray   andrew.gray(a)dunelm.org.uk