Wikipedia-l

wikipedia-l@lists.wikimedia.org

8289 discussions

Slashdot on MIT free courseware
by lsanger＠nupedia.com 11 Oct '01

11 Oct '01

http://slashdot.org/articles/01/10/10/1454231.shtml Doesn't look like anyone has yet mentioned Wikipedia...Jimbo? Somebody? Larry

1 0

Re: [Wikipedia-l] Search database not updated? (fwd)
by Jimmy Wales 10 Oct '01

10 Oct '01

I temporarily disabled the cron during our traffic crisis. I will re-instate it now. Also, I'll run the job by hand now, which will take a couple of hours to update the engine. lsanger(a)nupedia.com wrote: > How frequently is the Wikipedia search engine updated now? Someone else > expressed some concern about this recently, I think...they seemed to think > it was updated less often than once per day. > > ---------- Forwarded message ---------- > Date: Fri, 21 Sep 2001 22:54:50 -0500 > From: Axel Boldt <Axel.Boldt(a)metrostate.edu> > Reply-To: wikipedia-l(a)nupedia.com > To: wikipedia-l(a)nupedia.com > Subject: [Wikipedia-l] Search database not updated? > > >From discussion at http://www.wikipedia.com/wiki/Wikipedia_commentary/Search_Engine > it appears that the search database has not been updated in a while. Could someone > with access to the server please check if there's a working cron script in place to index > the site daily? > > Axel > > [Wikipedia-l] > To manage your subscription to this list, please go here: > http://www.nupedia.com/mailman/listinfo/wikipedia-l > -- ************************************************* * http://www.wikipedia.com/ * * You can edit this page right now! * *************************************************

1 0

Re: PHP wikipedia
by Jimmy Wales 09 Oct '01

09 Oct '01

Yes, I agree! It will be fine to use the German wiki. It is large enough to have lots of links to test, but small enough to be quick and easy. Magnus Manske wrote: > Larry, that's a misunderstanding: We only want to copy the German data to my > test wiki! > > Don't panic, > Magnus > > > -----Original Message----- > > From: lsanger(a)nupedia.com [mailto:lsanger@nupedia.com] > > Sent: Tuesday, October 09, 2001 10:13 PM > > To: Michael Dill > > Cc: Magnus Manske; jwales(a)bomis.com; lsanger(a)nupedia.com > > Subject: RE: PHP wikipedia > > > > > > What we need is a test wiki. I don't much like the idea of using the > > German wiki as a test wiki--they've done a huge amount of work already, > > and we don't want to mess it up. Maybe we can make a copy of it and put > > it at http://test.wikipedia.com/ . > > > > Larry > > > > On Tue, 9 Oct 2001, Michael Dill wrote: > > > > > Magnus, > > > > > > Go ahead and try the German version for the first > > > tests. We can probably get Clifford Adams or someone > > > to help sort out the current perl/CGI data formats. I > > > beleve that Jimbo might have some information there... > > > > > > I don't know enough to help with the data conversion, > > > but if we cant get any support from Clifford or > > > someone at BOMIS, I will see if I can find a > > > programmer or two to look at the tarball and help you > > > with the structure. > > > > > > Jimbo, > > > > > > Would you help point us in the right direction here. > > > > > > Thanks to you all for great work! > > > > > > Mike Dill > > > Phone 408-545-3675 > > > > > > --- Magnus Manske <Magnus.Manske(a)epost.de> wrote: > > > > --- Mike Dill <michael_dill(a)yahoo.com> wrote: > > > > > Yes, call the current PHP script at > > > > > wikipedia.sourceforge.net a Beta. > > > > > > > > OK! > > > > > > > > > I think we need to clear out everything in the > > > > test > > > > > database and see how to install a copy of the > > > > current > > > > > data. This will allow us to test/debug the script > > > > for > > > > > the data conversion. It will probably take a dozen > > > > > trys at porting it before we get the conversion > > > > script > > > > > into a reasonable shape. Then we can ask all the > > > > > wikipedians to look at all the pages to check for > > > > the > > > > > remaining script inconsistancies. > > > > > > > > We not necessarily need to clean the DB. And, should > > > > we try another > > > > wikipedia (e.g., the German one) first? It's much > > > > smaller, so conversion > > > > won't take that long for the first tries. If that > > > > works, we can purge the DB > > > > and install the "real thing". What do you think? > > > > > > > > > If we can get those bugs out, then asking Jimbo to > > > > > make the transition would be a reasonable > > > > proposition. > > > > > Until we can port the current data with some > > > > accuracy, > > > > > I feel that we should not push it on the world. > > > > > > > > Agreed. > > > > > > > > > Lets try a data convertion from the tarball and > > > > see > > > > > what comes out. When it seems reasonable to us > > > > then > > > > > Larry and Jimbo can decide when it should be > > > > > implemented. > > > > > > > > I looked at the German tarball, which seems to be > > > > quite old, and I suppose > > > > the "real" data is in the "page" sub-sub-...-subdir. > > > > Do you know if that's > > > > it? Are just the old versions stored as diffs, or > > > > the current version as > > > > well? I'm not that good at perl/CGI, so... > > > > > > > > Magnus > > > > > > > > > > > > > __________________________________________________ > > > Do You Yahoo!? > > > Make a great connection at Yahoo! Personals. > > > http://personals.yahoo.com > > > > > > -- ************************************************* * http://www.wikipedia.com/ * * You can edit this page right now! * *************************************************

2 1

[Nupedia-l] Discussion and a proposal (fwd)
by lsanger＠nupedia.com 09 Oct '01

09 Oct '01

This might be of some interest to you. This is a discussion, posted on Nupedia-l, of some problems Nupedia is facing, together with a proposed solution. One part of the solution is to add a link to Wikipedia article pages saying (something like) "Submit this article to Nupedia." This would take you to a page where further information would be solicited. Successful submission of an article to Nupedia would be noted on the Recent Changes page, as would final approval of any Wikipedia-originated Nupedia article. This *isn't* approved yet. It's just FYI. If you want to discuss it, it would be better that you do so on Nupedia-l rather than Wikipedia-l. You can subscribe to Nupedia-l here: http://www.nupedia.com/mailman/listinfo/nupedia-l --Larry ---------- Forwarded message ---------- Date: Mon, 8 Oct 2001 16:42:38 -0700 (PDT) From: lsanger(a)nupedia.com Reply-To: nupedia-l(a)nupedia.com To: nupedia-l(a)nupedia.com Subject: [Nupedia-l] Discussion and a proposal Please, read through this whole thing before responding to any part of it. I know it's long, but it sums up my thinking on the issues pretty well. I. First problem: bottlenecks I've been looking over all the many comments about the future of Nupedia with great interest. One strand of the discussion, started by Andreas Flack, is that Nupedia's system can be improved in various ways. One of these ways was to create reminder e-mails. As Magnus said, we already have many such e-mails (it took Toan and I some weeks to design them properly, last January and February, I suppose it was). It's worth noting that most people simply delete these reminder e-mails as soon as they receive them. Some of the other proposed changes are fairly cosmetic changes that would very likely speed things up somewhat, but overall, not very much, I suspect. I designed the process with the help of people from advisory-l and nupedia-l, and Toan and I designed the system to implement the process. I've also addressed problems at every step of the way. As I said (in great detail, which I am not going to repeat) last spring, there are bottlenecks at *every* step--dozens of potential bottlenecks, really. Getting rid of a few of them will speed things up, yes, but not significantly over the long haul. This is why I'm opposed to confining our changes to the system to relatively cosmetic changes. Another point discussed by Andreas and others and replied to by G. B. Lane is that we should do what we can in order to find replacements for delinquent reviewers and writers. As G. B. well said, this is a heck of a lot harder than it sounds. Finding a qualified reviewer for a single article might be easy for the most distinguished leaders of a field, but it can be quite difficult for people who are not so high on the academic/research totem pole, particularly when one must sell the *project* at the same time. On this point, I think we should all defer to the experience of G. B.; if you're not an editor, you don't understand the problem as well as he does. In sum, even our most dedicated, motivated members have complained of endless bottlenecks in the process. I don't think that removing just one or two of the bottlenecks is going to speed things up very much. The problem is not cosmetic, to be solved by small tweaks to the system. II. Second problem: lack of adequate credibility Many people, both potential and former participants, have been turned off by the fact that participation in the Nupedia project requires that people do for free what they would do elsewhere for pay. Brad Eden's comments are typical in this regard and totally understandable. The fact that Nupedia is an open content project--that our articles are freely distributable to anyone by anyone, and not just by the website Nupedia.com--does not seem to matter to them. And why should we expect it to matter to everyone? Just because *we're* convinced, that doesn't mean they will be. In response, we can sing the praises of open content, that it involves contributing free information to the world and that this cause is worthy. But, of course, there are many people who won't be convinced, and we shouldn't wring our hands if they're not convinced; we should instead regard their skepticism as a constraint on how we design our projects. It's difficult enough to organize academics to do difficult academic work, for which they'd ordinarily be paid, for free. It's even more difficult to get them to do that sort of work if they feel, as Brad feels, that their time is being wasted because the system forces them to work with people that they feel are underqualified. I'm *sure* that's one of the weaknesses of the current system that runs Nupedia--let alone Wikipedia. (This is not to say, however, that I agree with Brad's dim assessment of the state of Nupedia's music category at present.) I don't think we should simply dismiss these concerns. We can pontificate all we like about the virtues of open content, but that's not going to change the *facts* about the people we would like to have involved. I know that Jimbo and others are very concerned about making Nupedia more open, like Wikipedia, and I *completely* sympathize; but I think we should resign ourselves to the fact that the people whom we want to write for Nupedia don't want to be part of a wide-open project. Instead, we should make it as easy as possible for them to get to work in their *own* element. I don't think we can *expect* many academics to work closely with nonexperts--and that's what we ask academics to do if we ask them to join a fully bazaar-like model, like Wikipedia. It seems to me it would be best to have as efficient (but still high-quality) article-creation model as possible for them to use, and if they become impressed Wikipedia along the way, grand--we can welcome to Wikipedia with open arms. If we make it easier for Wikipedians to submit good Wikipedia articles to Nupedia, we will thereby expose Nupedian academics to the virtues of Wikipedia. III. The range of options open to us Therefore, the way I see it, if we want to redesign Nupedia's editorial process, we can choose revisions from a continuum with (at least) two extremes. At one extreme, Bomis would try to make Nupedia an ueber-elitist project, not unlike the Stanford Encyclopedia of Philosophy, accepting submissions from and making reviewers of only the leading researchers in any field. Another option in this direction, but less extreme, would be for us to revert to a traditional peer review system, completely redesigning (but radically simplifying) the system. At the other end of the continuum--and this is actually *not* seriously proposed by anyone--Nupedia would become another version of Wikipedia. Since Wikipedia already exists, this is silly; but in this direction, Nupedia could become *no more than* an approval process for Wikipedia. The latter would cause a mass exodus of people like Brad--most of our most qualified, distinguished participants on Nupedia. I know some of you on Nupedia-l just don't care about having the participation of academics and bona fide experts, but I think this is a huge mistake. If our open content encyclopedia projects are going to achieve their full potential, we *must* have the participation of these people. It would be a mistake not to let our *most qualified* potential participants stay in the project--on their own terms. So, I think Brad has an excellent point. It's one that Michael Witbrock emphasized to me very forcefully and persuasively when I visited him in Boston: to succeed, the project has to have even *more* credibility among its most qualified potential participants. I think Andreas' suggestion, that we feel free to replace reviewers and authors as necessary, is fine, but it's implausible that it would help much to adopt this policy explicitly, for the reason G. B. Lane gave. As I'd put it, for any given topic, on Nupedia, there is bound to be only a very few people who are *both* interested in writing on that topic *and* willing to revise someone else's work. Similarly with reviewers--it's hard enough finding just *one* qualified reviewer for an article. In other words, Nupedia's problem isn't that the project is closed to our finding replacement authors and reviewers when necessary--it's that we couldn't find enough *qualified* people even if it *were* generally open. I think you'd find most Nupedia editors would agree with this. IV. The proposal It's with these sorts of thoughts in mind that I have come around to the following proposal. This is something Jimbo suggested many months ago and that Claus Wilke has articulated recently. In fact, it was probably suggested by some wise person when we were designing a review system in the first place. I. Replace the current Nupedia system with a simple, traditional review process, to be designed by the members of advisory-l. This would at a minimum involve (1) writers either being assigned an article or simply submitting articles to the editor, (2) the editor deciding whether to assign the article to a reviewer or reviewers or reject it outright, (3) a reviewer or reviewers either saying yea, nay, or conditional pass (maybe) on articles (in one simple step), and (4) the article being passed by the editor. After that, I would strongly recommend having a public feedback system that would be something like the open review step we now have in place. In other words, I think we should basically eliminate steps four through six, open review and copyediting, and leave quality in the hands of editors and expert reviewers--then *supplemented* by public feedback. Copyediting would be the responsibility of authors, who would have to solicit help, if needed, on pain of having articles simply rejected. Copyediting comments would be solicited in the public feedback process. (By the way, I would like to say at this juncture that Ruth Ifcher has been probably the hardest-working member of Nupedia, and one of the most widely appreciated, and this latter proposal should by no means be construed as (absurdly) implying that her contributions as Nupedia's chief copyeditor have been without value.) If an editor were to approve very much substandard work, I would be willing to field complaints and, if necessary, take whatever action was appropriate. II. Create a link on Wikipedia article pages labelled: "Submit this article to Nupedia." This would link to a page that would solicit further information from the submitter and then place the Wikipedia article in the queue. Submission information would be reported on the Wikipedia [[Recent Changes]] page, as would final approvals of articles in the Nupedia system. III. Delete the Chalkboard (move its contents to Wikipedia, with the permission of authors--Magnus, mainly :-) ). IV. Put off a decision whether to make a separate article approval process for Wikipedia, on the hope that Nupedia might turn into a reasonably efficient approval process for Wikipedia articles. (This latter, after all, is how many of us have envisioned Wikipedia and Nupedia working together.) So, that's it. I'd like to solicit comments here and then, sometime soon, unless fatal flaws are discovered, take this proposal to advisory-l. Larry [Nupedia-l] -- a Nupedia.com mailing list Help build the largest and finest encyclopedia in the world! To manage your subscription to this list, please go here: http://www.nupedia.com/mailman/listinfo/nupedia-l

1 0

Of interest, re control of content
by lsanger＠nupedia.com 08 Oct '01

08 Oct '01

Owning the Future: Content Discontent http://www.techreview.com/magazine/oct01/shulman.asp I'm not sure if anything follows for the prospects of open content projects, but it's very interesting, I think. Larry

1 0

New Scientist interview
by lsanger＠nupedia.com 08 Oct '01

08 Oct '01

I was interviewed about Nupedia and Wikipedia (about equally) this morning by New Scientist: http://www.newscientist.com/ They describe themselves as "The World's No.1 Science & Technology News Service," and I imagine most of you have heard of them. The article is going to be about the open source movement's influence on other kinds of projects. Actually, it wasn't perfectly clear to me what the subject was, but the reporter seemed to imply that open content projects would be highlighted. --Larry

1 0

Wikipedia back to "normal" :-)
by lsanger＠nupedia.com 08 Oct '01

08 Oct '01

It looks to me like Wikipedia's recent traffic problems have been fixed, and it's back to its formerly fast self (and links to new articles appear as regular links!). It's nice to note that, despite being a Saturday, and despite being plagued by traffic-related problems, yesterday we *still* had as many visitors and pageviews as the average pre-Sept. 11 weekday. :-) Larry

1 0

sorry to report, but
by Michel Clasquin 07 Oct '01

07 Oct '01

Nope, it's still unusable. I've been trying to update my Wolf page for twenty minutes now and it just won't lock. Larry, Jimbo this might scare off newbies who stumble onto the site while this is going one. Maybe disable editing while you are working on it? --

2 1

Re: [Wikipedia-l] who?
by Gareth Owen 06 Oct '01

06 Oct '01

Michel Clasquin <clasqm(a)mweb.co.za> writes: > WHO isnt worried about bandwith, then? <g> > > Good luck getting everything back to normal Amen. So whats going on? Has the load uncovered actual bugs, or does the present-and-basically-correct code just not scale well? Having said that, I have done a fair bit of actual work today :) -- Gareth Owen

3 2

Wikipedia development
by Anders Törlind 06 Oct '01

06 Oct '01

Hi all! As you might have been noticing, the 'pedia is feeling a bit under the weather, so i thought i could help out magnus Manske a bit with the new code. So here comes a couple of questions for Magnus, the reply of which might prove interesting for more people than me i hope :-) 1. How does one go about with the CVS download from Sourceforge? 2. Is there a design specification, so one might get an overview of how the code hangs together? 3. How do you get it up and running on your home installation of Apache? 4. Is there any special area that needs attension? and last but not least 5. Where do I get hold of the best PHP tutorial/documentation on the net :-) Cheers! /Anders Törlind ps. If this topic has already been discussed today, i apologize: I just set my subscription off digest... ds

3 6

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

Wikipedia-l