Email discussion between myself and Tom about the Wikipedia 1.0 idea and [[Wikipedia:Version 1.0 Editorial Team]], forwarded with his permission. I've removed email addresses and some names (not secret, just that the discussion in question is at the named page, not here).
Note that I've also created [[Category:Wikipedia 1.0]], a project category which is pretty much for use as a working set of 1.0 documents, rants and so forth.
- d.
---------- Forwarded message ---------- From: David Gerard Date: Mon, 28 Feb 2005 23:23:37 +0000 Subject: Re: Wikipedia reputation To: Tom Haws
On Mon, 28 Feb 2005 14:21:56 -0700, Tom Haws wrote:
I noted with that same sinking feeling [XXX]'s response to your note at [[Wikipedia talk:Version 1.0 Editorial Team]]. It seems the conventional way of talking about 1.0 is to add an impossible burden of review work on top of the Wiki miracle. But what you and I are thinking of is not much more than a programming effort to harness the miracle of the Wikipedia dilettantism mill to automatically tell us which articles (or article versions) are suspect/trusted, adult/family fare, fledgling/mature, or any of a selected group of qualifiers that could be used as selectors for various purposes. As you, I am confident the wiki editing process would rate the entire encyclopedia on even multiple aspects within days or weeks. This doesn't have to go on for months.
I've answered [XXX] trying to explain why editorial committees are unlikely to scale. I can't see them working - I know how damned hard it is to get *one* article past WP:FAC, in areas I'm a subject matter expert in!
I think editorial committees are an exercise in futility, but the people fond of such an approach can't be told otherwise. As such, I'd probably leave them to it and see if they come up with anything useful.
Of course the first item of business is to discover the relevant forum to spread and build consensus on such a manner of thinking. I would think Jimbo (having experienced it all) would be thinking in the same direction as us, but from all I have seen, he has been pretty tight-lipped. I hopefully imagine that his silence (assuming I haven't simply missed the "right" places like the mailing list) is indicative of dismay with the conventional proposals. I wonder if there perhaps is a way to get his honest ideas, which of course we would value highly without quoting him as an "authority".
I think Jimbo is waiting for others to get on with it ;-)
Presumably when it happens, it'll be simple and elegant and obvious in retrospect.
Asking him really does work ;-) wikien-l is a good place to sound out ideas.
Other than getting into Jimbo's mind, and dedicating my own front page to the issue (which I have done some time ago, and am willing to revise per your thoughts), I don't know where to go with this. Who and what is relevant to this issue?
I've created [[Category:Wikipedia 1.0]]. Possibly others will add stuff to it. e.g. their own userspace thoughts pages.
One thing: I really don't think our technical tools (MediaWiki features) are up to the job.
URGENT: - rating system. May scale, editorial committees don't. VERY USEFUL: - references syntax (many mooted, none implemented).
I should start a page on meta to this effect ...
I see this issue as urgent because of the fund-raising implications it has.
I don't actually see it as urgent in that sense. I see it as something to get *right*, because if we don't do that it's going to splutter anyway.
I hope we can discuss specifics in a wider forum soon.
I'll be taking it to wikien-l myself. Mind if I send this mail there? Or you can :-)
p.s. Your response will be bounced by my server, but excuse and ignore the mistreatment. I will let your mail through.
Ooooh, whitelists. Evil! I use Bayesian, of course, because Thunderbird and gmail do ;-)
- d.
I notice Jamesday http://en.wikipedia.org/wiki/User:Jamesday proposed "...tagging based on quality metrics with no specific print or CD target. Anyone else could then use those metrics to select articles for any arbitrary target they desire, with only those selecting for non-online targets being liable." His context was legal concerns about an Editorial committee producing a CD or Paper Wikipedia.
I think this is exactly the direction we need to go. We need to develop a Mediawiki mechanism for wiki-tagging metrics on all versions of all articles. Not weighted voting; that's anti-wiki. Harness the dilettantism (thanks David Gerard), and do straight Wiki-tagging of all attributes except Trust. Set trust according to known trust of saving editor (anon, newbie, seasoned, admin, etc.).
Tom Haws "And [the angel] said unto me: Knowest thou the condescension of God? And I said unto him: I know that he loveth his children; nevertheless, I do not know the meaning of all things."
David Gerard wrote:
Email discussion between myself and Tom about the Wikipedia 1.0 idea and [[Wikipedia:Version 1.0 Editorial Team]], forwarded with his permission. I've removed email addresses and some names (not secret, just that the discussion in question is at the named page, not here).
Note that I've also created [[Category:Wikipedia 1.0]], a project category which is pretty much for use as a working set of 1.0 documents, rants and so forth.
- d.
---------- Forwarded message ---------- From: David Gerard Date: Mon, 28 Feb 2005 23:23:37 +0000 Subject: Re: Wikipedia reputation To: Tom Haws
On Mon, 28 Feb 2005 14:21:56 -0700, Tom Haws wrote:
I noted with that same sinking feeling [XXX]'s response to your note at [[Wikipedia talk:Version 1.0 Editorial Team]]. It seems the conventional way of talking about 1.0 is to add an impossible burden of review work on top of the Wiki miracle. But what you and I are thinking of is not much more than a programming effort to harness the miracle of the Wikipedia dilettantism mill to automatically tell us which articles (or article versions) are suspect/trusted, adult/family fare, fledgling/mature, or any of a selected group of qualifiers that could be used as selectors for various purposes. As you, I am confident the wiki editing process would rate the entire encyclopedia on even multiple aspects within days or weeks. This doesn't have to go on for months.
I've answered [XXX] trying to explain why editorial committees are unlikely to scale. I can't see them working - I know how damned hard it is to get *one* article past WP:FAC, in areas I'm a subject matter expert in!
I think editorial committees are an exercise in futility, but the people fond of such an approach can't be told otherwise. As such, I'd probably leave them to it and see if they come up with anything useful.
Of course the first item of business is to discover the relevant forum to spread and build consensus on such a manner of thinking. I would think Jimbo (having experienced it all) would be thinking in the same direction as us, but from all I have seen, he has been pretty tight-lipped. I hopefully imagine that his silence (assuming I haven't simply missed the "right" places like the mailing list) is indicative of dismay with the conventional proposals. I wonder if there perhaps is a way to get his honest ideas, which of course we would value highly without quoting him as an "authority".
I think Jimbo is waiting for others to get on with it ;-)
Presumably when it happens, it'll be simple and elegant and obvious in retrospect.
Asking him really does work ;-) wikien-l is a good place to sound out ideas.
Other than getting into Jimbo's mind, and dedicating my own front page to the issue (which I have done some time ago, and am willing to revise per your thoughts), I don't know where to go with this. Who and what is relevant to this issue?
I've created [[Category:Wikipedia 1.0]]. Possibly others will add stuff to it. e.g. their own userspace thoughts pages.
One thing: I really don't think our technical tools (MediaWiki features) are up to the job.
URGENT:
- rating system. May scale, editorial committees don't.
VERY USEFUL:
- references syntax (many mooted, none implemented).
I should start a page on meta to this effect ...
I see this issue as urgent because of the fund-raising implications it has.
I don't actually see it as urgent in that sense. I see it as something to get *right*, because if we don't do that it's going to splutter anyway.
I hope we can discuss specifics in a wider forum soon.
I'll be taking it to wikien-l myself. Mind if I send this mail there? Or you can :-)
p.s. Your response will be bounced by my server, but excuse and ignore the mistreatment. I will let your mail through.
Ooooh, whitelists. Evil! I use Bayesian, of course, because Thunderbird and gmail do ;-)
- d.
WikiEN-l mailing list WikiEN-l@Wikipedia.org http://mail.wikipedia.org/mailman/listinfo/wikien-l
--- Tom Haws hawstom@sprintmail.com wrote:
I notice Jamesday http://en.wikipedia.org/wiki/User:Jamesday proposed "...tagging based on quality metrics with no specific print or CD target. Anyone else could then use those metrics to select articles for any arbitrary target they desire, with only those selecting for non-online targets being liable." His context was legal concerns about an Editorial committee producing a CD or Paper Wikipedia.
Ack! No more blasted meta tages (meaning templates). A smarter system is needed that does not require an editor to edit or in any way change the content of an article by using said system.
-- mav
__________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Daniel Mayer wrote:
Ack! No more blasted meta tages (meaning templates). A smarter system is needed that does not require an editor to edit or in any way change the content of an article by using said system.
Well, mav, we can flag for Confidence without any editor intervention by simply attaching our confidence in the saving editor to the article version. Anon saves get low confidence, newbies get a little higher, etc. But any other quality or other metric is going to have to be done by humans.
I personally would be delighted to see nothing more than a Confidence flag on articles so people could select for distribution or presentation based on Confidence. That solves at least the problem of burning a CD with some percentage of vandalized articles, or articles that do not represent "the best we can do at the moment". We could even show anons high-confidence articles only, and tell them when they click "Edit", "Hey, you weren't really looking at the latest version. Here's the latest and the diff. Edit it. If it's been trashed, go into History to edit and save as the latest the version you were viewing."
Tom Haws "And [the angel] said unto me: Knowest thou the condescension of God? And I said unto him: I know that he loveth his children; nevertheless, I do not know the meaning of all things."
Daniel Mayer wrote:
--- Tom Haws hawstom@sprintmail.com wrote:
I notice Jamesday http://en.wikipedia.org/wiki/User:Jamesday proposed "...tagging based on quality metrics with no specific print or CD target. Anyone else could then use those metrics to select articles for any arbitrary target they desire, with only those selecting for non-online targets being liable." His context was legal concerns about an Editorial committee producing a CD or Paper Wikipedia.
Ack! No more blasted meta tages (meaning templates). A smarter system is needed that does not require an editor to edit or in any way change the content of an article by using said system.
Magnus Manske's rating system appears to be along those lines, i.e. it's a separate thing entirely to the article space. We should assume we're talking about something like that ;-)
- d.