Suppose I'm editing an article and my browser crashes.
Were my <Preview>s stored anywhere? If so, how can I find them?
-- Toby Bartels
toby(a)math.ucr.edu
PS: Is this the correct forum for this, or should I ask on <wikitech-l>?
On Friday 21 June 2002 08:50 pm, Karen wrote:
> I would think that the best policy is consistency... no, Australian
> placenames do not repeat from state to state (much). Not to the extent
> that it would interfere with entries... but at the same time I think
> that if the policy for other countries is city,state then we should
> stick to that. I know that whenever I refer to an Australian city in
> writing an entry I refer to it with that format because I've seen it
> elsewhere. But then I generally link to them as [[Sydney]], [[New South
> Wales]] rather than as a single phrase because there IS only one Sydney
> (and one Melbourne, one Darwin etc)!
Unfortunately, there is a Sydney in Canada, Florida, and Virginia; a
Melbourne in Florida, Arkansas, and Iowa; and a Darwin in Minnesota,
California, Nevada and Illinois (there are probably several more of each). In
all these examples there was only one non-American city and in my search I
couldn't find a single additional Australian city by any of these names.
The reuse of city names in the US is just too big an ambiguity issue to
ignore and therefore city names in the US should be treated differently than
city names for other nations. The general trend for US cities in wikipedia
is to name them in the format of [[city, state]] (which is what they are
actually specifically referred to within the US and abroad). I agree with and
encourage this trend because of the need to differentiate between US cities
and to have predictable names for them.
Consistency in city naming should be <i>within</i> a nation and not for all
nations in my opinion. But since most other nations don't have internal
ambiguity issues that even begin to compare with the United States, I propose
that those cities should be in the form [[city, nation]] (as most non-US
cities are specifically referred to -- at least according to a sampling on
Google and my own experience) or just [[city]] for noted exceptions (there
are important consistency and minor ambiguity issues with the [[city]]
format though....). There might be some nations other than the US that are
exceptions to this proposed naming convention, but these can be explored on a
case-by-case basis.
BTW, before we revisit it, [[Melbourne, Florida, United States of America]]
and [[Melbourne, Arkansas, United States of America]] are both WAY too
tedious and would hardly ever be directly linked to because the long form
(even with USA or US) is not really used much at all ("United States of
America", USA or US not being needed to disambiguate). Likewise, [[Melbourne,
Victoria, Australia]] is probably just as bad ("Victoria" not being needed to
disambiguate). In short, "simplify, simplify" - but be internally consistant.
Technical fixes such as redirects are just that -- technical fixes that skirt
the main issue (in my experience, users who prefer needlessly long article
names are notorious for not providing redirects and if we encourage long
names these users will soon outnumber more sane individuals wanting to
preserve easy and natural linking).
maveric149
> The trusted hand requirement sounds quite reasonable to me. The
> fact is that there are people using the upload facility as
> storage space for their own personal junk and there just isn't
> enough room for it. People like that do NOT sign up and stick
> around for a month to load their stuff up - they see the
> opportunity and go for it on the spot.
What opportunity? TO spend time uploading files that someone else
will just delete immediately? I think the only reason it's a problem
now is that the current software makes it hard to clean up the mess.
The new software makes it much simpler to find large images, images
that have no links to them, etc.; so I'm not planning on any
restricted rights there just yet. If it's still a problem we can
revisit the issue again, but I'm not sure it will be.
0
>> So then, why don't we just take the path of least resistance
>> and use whatever format is the dominant one? Oh drat, I just
>> checked and found out that that path is to not disambiguate
>> at all.....
> I would think that the best policy is consistency... no,
> Australian placenames do not repeat from state to state (much).
> Not to the extent that it would interfere with entries... but at
> the same time I think that if the policy for other countries is
> city,state then we should stick to that.
Consistecy in itself is not that important. The most important thing
is maintaining ease of editing. When people are editing an article
and put brackets around something on a whim, there should be a good
chance that there's a useful article at the other end of the link.
The most important thing helping those odds is simplicity. _Some_
consistency helps that too--once users are generally accustomed to the
fact that U.S. city articles look like "Paris, Texas", the he's likely
to put "Fremont, California" in the brackets and (rightly we hope)
assume that it works.
Outside the U.S., I think it's natural for people to use "Sydney,
Australia" and "Lyon, France". I don't think most English speakers
will have any trouble with the minor inconsistency that American
cities have states and other cities have countries. Canadian ones
might us provinces as well--I think Americans and Canadians both tend
to think of them that way ("Toronto, Ontario"). And that way we won't
have to change the title of "Montreal, Quebec" when they seceed :-)
0
---------- Forwarded Message ----------
Subject: Junk uploads, bogus user accounts and "trusted hand" status
Date: Sat, 15 Jun 2002 15:48:28 -0700
From: Daniel Mayer <maveric149(a)yahoo.com>
To: wikipedia-l(a)nupedia.com
I'm sure many of you have noticed that recently there has been a growing
problem of people creating user accounts for the sole purpose of using
wikipedia's upload utility to place copyrighted and otherwise inappropriate
material on the server.
We might want to discuss further limiting the use of this utility
because it's abuse is beginning to be a maintenance issue for the syops and
the presence of these bogus user accounts is also over-reporting the number of
true (as in contributing) users we have.
I know the "trusted hand" status is in the database already so it might be
easy to limit upload ability to "trusted hand" or greater status users. It
would also be nice to have this status granted automatically if say two
conditions are met: 1) a user account is at least 30 days old and 2) this
user has edited a certain number of articles (I would be liberal and set this
at 30, but I wouldn't mind having it set at up to 100).
And while we are at it we might also want to grant trusted hands the ability
to do things that currently only sysops can do -- like move pages or edit
certain protected pages (those set to "is_trusted" perhaps? "is_sysop" would
then only be used to protect certain established policy pages and for
emergencies -- if a "trusted hand" went on a rampage it would be simple to
just change the user's status). I wonder if a script could be written to
change the status of all current users that have edited a certain number of
articles? The sysops could always do this by hand if need be.
Oh, and if something like this were established we should encourage sysops to
manually promote new users that are able to understand the wikipedia concept
and its policies early to "trusted hand" status before the 30day&30article
requirement is met.
Just a thought....
maveric149
-------------------------------------------------------
On Friday 21 June 2002 08:29 pm, I wrote:
> Oh, and if something like this were established we should encourage sysops
> to manually promote new users that are able to understand the wikipedia
> concept and its policies early to "trusted hand" status before the
> 30day&30article requirement is met.
I should have said "30day&30edit".
maveric149
On Friday 21 June 2002 09:02 am, Pierre wrote:
> Could someone put on the upload page a note that this is not a file-sharing
> system or a file storage area? I think people look for "upload files", find
> our upload page, and upload things without realizing that this is an
> encyclopedia.
Here is an excellent candidate to <i>not</i> allow Google to index.
BTW, now that we have Jimbo back, can we revisit the idea to limit use of the
upload utility? There must be several dozen user accounts that were created
for the sole purpose of using our upload utility to store files. I like
Koyaanis Qatsi's idea:
On Sunday 16 June 2002 12:01 pm, Koyaanis Qatsi wrote:
> Speaking of which, the old software used to have a function where you
> could hover over a username and the IP address would show up (.xxx at
> the end for most, but I imagine sysops would see the full thing).
> That's bring that back, mmmmkay?
That way if a logged-in user where consistently uploading illegal stuff then
that person's IP could be blocked. BTW, I recently had a nightmare about a
group of vandals creating accounts to perform a systemic assault on
wikipedia. And of course, since it is not possible for mere sysops to block
logged-in users I was pretty much helpless -- protecting pages isn't
practical because it is not possible to predict which pages will be hit. I'm
glad that was just a dream (yeah I know - I should get out more often :-).
I will repost my previous (slightly edited) email about another idea to
combat upload abuse for further comment in a different post.
maveric149
On Friday 21 June 2002 09:02 am, tarquin wrote:
> might Google also be following "pages that link here", "random" and
> "history"?
> Wikipedia freezes almost every day around 3pm UK time, which is odd as
> it's during the pediod with the fewest edits. Could that be when google
> comes a-crawling?
> we had problems on UnrealWiki with downloading robots calling too many
> diff links & oveerloading the server.
>
> tarquin
Is this fixed in the new code? I thought we decided not to allow Google to
index edit links (along with older versions of articles in History, log pages
and special pages other than maybe RecentChanges)? I don't see a problem with
allowing Google to follow links through "pages that link here" though -- so
long as this doesn't cause undue hardship on the server (if anything, this
should decrease the amount of time it take Googlebot to do its thing).
maveric149
It Takes a Village to Save a Site
By Paul Boutin
2:00 a.m. June 21, 2002 PDT
Begging for funds to keep your bankrupt site afloat rarely works. Unless
you're Rusty Foster, whose tech community site Kuro5hin just raised
$35,000 -- and a few eyebrows -- in less than a week.
"I gotta say, I'm quite impressed at what they've pulled off," said
Slashdot.org founder Rob "CmdrTaco" Malda.
Slashdot's site, along with its free source code, inspired Foster to
found Kuro5hin in December 1999 as a similar forum for "technology and
culture from the trenches."
But a string of disappointing ad models led Foster to post a long
missive on Monday titled "We're Broke: The Economics of a Web
Community." Enumerating his income and expenses, Foster asked readers to
help him meet the site's $70,000 annual operating budget.
"This has been coming for a while. We didn't have any money coming in,"
Foster said by phone from Peaks Island, Maine, where he and his wife
relocated from San Francisco last year.
"I spent a lot of time asking myself: Are we a magazine, are we a media
company?" Foster said. "We're a community; that's why the whole
advertising thing doesn't work for us. Advertising works in media
because the company showing you advertising is providing you value by
writing the articles. On a site like mine, the transaction breaks down
because the visitors are writing the articles themselves."
Such introspection and analysis are standard fare on Kuro5hin, whose
tech-savvy members discuss everything from software license details to
the U.S. government's failure points in design, specification, and
implementation. There are few places you can have an intelligent
discussion like that.
"Rusty doesn't interfere with the way the site evolves," said Robin
Bandy, one of Kuro5hin's most prolific participants who contributes
under the name "Arkady."
But Foster's announcement shook up members, who realized their reluctant
leader might have to give up the site for a paying job.
"Kuro5hin is about Rusty Foster the same way Slashdot is about Rob
Malda," said writer Doc Searls, an early booster of the site. "He's a
wise young man in the literal sense of the word. He's a programmer, but
from a different variety -- more intellectual, and less focused on
technology."
"I asked the community what I should do, because that's always worked
before," Foster said.
Bandy agreed, "Rusty honestly wants the site to be a community. A year
ago he tried to give the site over to the readers. The overwhelming
reaction was, 'Nah, you're doing a good job, keep it up.'"
This time, though, Foster's financial bind helped him find a way to
bring others on board without selling the site. "We're talking to
accountants and lawyers," he said about converting his one-man company
to not-for-profit status under Section 501c of the Federal Internal
Revenue Code.
Besides the tax breaks, 501c status would enable Foster to bring in
other Kuro5hin members as trustees to oversee the site.
There should be no lack of volunteers: Foster said the largest donation
received by mid-Wednesday was $240, as the turnout came from hundreds of
small donors rather than a single sugar daddy.
Leslie Nakajima, a vice president at PR agency Bite Communications in
San Francisco, explained the pledge drive's surprise success as proof
that, unlike many sites, Kuro5hin's worth is apparent to its fans. "They
have an obvious value proposition," she said.
Less so for the site's name. "'Corrosion.' It's a pun on my name,"
Foster said. "Rusty ... corrosion ... get it? Don't worry, nobody else
ever got it, either."
>
>
>From: Pierre Abbat <phma(a)webjockey.net>
>Date: Thu, 20 Jun 2002 16:40:39 -0400
>Subject: [Wikipedia-l] Upload page
>
>Could someone put on the upload page a note that this is not a file-sharing
>system or a file storage area? I think people look for "upload files", find
>our upload page, and upload things without realizing that this is an
>encyclopedia.
>
When i look whit google for "upload files" the wikipedia upload-page is
hit number 2.
http://www.google.be/search?hl=nl&ie=UTF-8&oe=UTF8&q=upload+files&btnG=Goog…
That is not good.
Mayby include this on that page;
<META NAME="ROBOTS" CONTENT="NOINDEX, NOFOLLOW">
and why not also do this whit all non-current pages ? -- giskart,WikiNL