geni wrote:
On 12/1/05, Daniel Mayer <maveric149 at yahoo.com> wrote:
(better RC patrol features, trust networks to filter RC and watchlists, article validation, etc).
RC patrol is reaching it's pratical limit. I'm seeing more vandalism on my watchlist and no matter how good the people doing it there is simply no way for them to know if a large percentage of it is true or not
Would it be an idea to make RC patrol organised, i.e. people (admins or not) sign on for shifts, checking all edits in a given block of time? Two patrollers for each time period should be enough, three or four for evenings US time. You do RC patrol - would something like this make sense?
(I don't know how long the time periods would be - an hour? 30 minutes? 10 minutes?)
- d.
On Thu, 2005-12-01 at 17:10 +0000, David Gerard wrote:
geni wrote:
On 12/1/05, Daniel Mayer <maveric149 at yahoo.com> wrote:
(better RC patrol features, trust networks to filter RC and watchlists, article validation, etc).
RC patrol is reaching it's pratical limit. I'm seeing more vandalism on my watchlist and no matter how good the people doing it there is simply no way for them to know if a large percentage of it is true or not
Would it be an idea to make RC patrol organised, i.e. people (admins or not) sign on for shifts, checking all edits in a given block of time? Two patrollers for each time period should be enough, three or four for evenings US time. You do RC patrol - would something like this make sense?
(I don't know how long the time periods would be - an hour? 30 minutes? 10 minutes?)
Unworkable - who knows when they can do it.
Just queue new articles so they have to go via an admin before joining the main namespace (they can sit in New: before that).
Not much code.
On 12/1/05, Justin Cormack justin@specialbusservice.com wrote:
Just queue new articles so they have to go via an admin before joining the main namespace
Why? Isn't that missing the point of having a wiki in the first place?
On Thu, 2005-12-01 at 17:38 +0000, Tony Sidaway wrote:
On 12/1/05, Justin Cormack justin@specialbusservice.com wrote:
Just queue new articles so they have to go via an admin before joining the main namespace
Why? Isn't that missing the point of having a wiki in the first place?
The suggestion was to have a constant patrol, and process everything very quickly. If you want to do this it should be enforced.
Any mechanism where things have to go through some form of category change say is better than just hoping things dont get missed, and lets you avoid the situation of not knowing what someone else has reviewed.
Doesnt have to be done by admins, although vandals might work it out pretty quickly if they could sign off their own new articles.
On 12/1/05, Justin Cormack justin@specialbusservice.com wrote:
The suggestion was to have a constant patrol, and process everything very quickly. If you want to do this it should be enforced.
Any mechanism where things have to go through some form of category change say is better than just hoping things dont get missed, and lets you avoid the situation of not knowing what someone else has reviewed.
Doesnt have to be done by admins, although vandals might work it out pretty quickly if they could sign off their own new articles.
We tried this. We switched it off. Pretty much all new articles are cheacked. With edit rates of at least once a second there is no way we can chech every edit.
-- geni
geni wrote:
On 12/1/05, Justin Cormack justin@specialbusservice.com wrote:
The suggestion was to have a constant patrol, and process everything very quickly. If you want to do this it should be enforced.
Any mechanism where things have to go through some form of category change say is better than just hoping things dont get missed, and lets you avoid the situation of not knowing what someone else has reviewed.
Doesnt have to be done by admins, although vandals might work it out pretty quickly if they could sign off their own new articles.
We tried this. We switched it off. Pretty much all new articles are cheacked. With edit rates of at least once a second there is no way we can chech every edit.
-- geni
Ya, your best bet is going to be some sort of heurustical (sp?) or other automated content-analysis and flagging system....
On 12/1/05, Justin Cormack justin@specialbusservice.com wrote:
On Thu, 2005-12-01 at 17:10 +0000, David Gerard wrote:
geni wrote:
On 12/1/05, Daniel Mayer <maveric149 at yahoo.com> wrote:
(better RC patrol features, trust networks to filter RC and watchlists, article validation, etc).
RC patrol is reaching it's pratical limit. I'm seeing more vandalism on my watchlist and no matter how good the people doing it there is simply no way for them to know if a large percentage of it is true or not
Would it be an idea to make RC patrol organised, i.e. people (admins or not) sign on for shifts, checking all edits in a given block of time? Two patrollers for each time period should be enough, three or four for evenings US time. You do RC patrol - would something like this make sense?
(I don't know how long the time periods would be - an hour? 30 minutes? 10 minutes?)
Unworkable - who knows when they can do it.
Obsesives who have wikipedia nights and the like.
Just queue new articles so they have to go via an admin before joining the main namespace (they can sit in New: before that).
Not much code.
Due to the level of new page patrol that is one area that is pretty much covered and to be honest all the admin viewing in the world is not going to be able to detect false info in a boarderline notable person.
vandalism that hits already existing articles is more of a problem at this point.
-- geni
Justin Cormack wrote:
Just queue new articles so they have to go via an admin before joining the main namespace (they can sit in New: before that).
Just to clarify, are you talking about New Pages or Recent Changes? (I think the discussion started as RC, and moved to NP).
The practical problem I see with this scheme (which is worse for RC) is that any edits started between one user's submission and admin approval will amount to edit conflicts.
There are a number of ways such conflicts could be managed, but then the system stops being so simple.
On 12/1/05, David Gerard dgerard@gmail.com wrote:
geni wrote:
On 12/1/05, Daniel Mayer <maveric149 at yahoo.com> wrote:
(better RC patrol features, trust networks to filter RC and watchlists, article validation, etc).
RC patrol is reaching it's pratical limit. I'm seeing more vandalism on my watchlist and no matter how good the people doing it there is simply no way for them to know if a large percentage of it is true or not
Would it be an idea to make RC patrol organised, i.e. people (admins or not) sign on for shifts, checking all edits in a given block of time? Two patrollers for each time period should be enough, three or four for evenings US time. You do RC patrol - would something like this make sense?
If it were posible it would probably have some utility if only to give us some idea of the amount of rescoures we have availible.
(I don't know how long the time periods would be - an hour? 30 minutes? 10 minutes?)
10 minutes. Those who can go longer are free to sign up back to back. those of us who tend to get disstracted by seeing things we can fix can still make a contribution
-- geni
geni wrote:
On 12/1/05, David Gerard dgerard@gmail.com wrote:
Would it be an idea to make RC patrol organised, i.e. people (admins or not) sign on for shifts, checking all edits in a given block of time? Two patrollers for each time period should be enough, three or four for evenings US time. You do RC patrol - would something like this make sense?
If it were posible it would probably have some utility if only to give us some idea of the amount of rescoures we have availible.
(I don't know how long the time periods would be - an hour? 30 minutes? 10 minutes?)
10 minutes. Those who can go longer are free to sign up back to back. those of us who tend to get disstracted by seeing things we can fix can still make a contribution
If I had time this weekend (I'm going to the other end of the country tomorrow morning and not returning till Sunday, which will be way busy), I'd float a roster idea on WP:AN and WP:ANI myself. But I don't, so I suggest someone else boldly do so :-)
- d.
--- David Gerard dgerard@gmail.com wrote:
Would it be an idea to make RC patrol organised, i.e. people (admins or not) sign on for shifts, checking all edits in a given block of time? Two patrollers for each time period should be enough, three or four for evenings US time. You do RC patrol - would something like this make sense?
What we need are teams of people working together in a way that minimizes duplicated effort. As is, there is no way to know if a diff has been looked at once before, never before, or 200 times. There is also no way to know what was being looked at (simple vandalism, subtle vandalism, accuracy, etc). Knowing at least the first part (how many times a diff has been looked at) will help a great deal. Sorting RC based on that would be even better.
Better still would be to sort RC and similar lists based on trust networks. For example, I trust you and many other people. So edits either made by you and diffs checked by you should be OK in my eyes. It would be nice if my RC and similar lists de-emphasized your edits and edits checked by you. This could go one step farther: I trust your ability to judge whether or not other people are trustworthy. So it would be nice if I could trust by proxy all the users you trust. This creates a trust network.
This could also be done P2P via an offline editor and thus minimize server load. The offline editor would do all the sorting after downloading raw data from Wikipedia.
I think it is obvious that we need bigger and better guns to fight vandals.
-- mav
__________________________________ Yahoo! Music Unlimited Access over 1 million songs. Try it free. http://music.yahoo.com/unlimited/
On 12/1/05, Daniel Mayer maveric149@yahoo.com wrote:
What we need are teams of people working together in a way that minimizes duplicated effort. As is, there is no way to know if a diff has been looked at once before, never before, or 200 times. There is also no way to know what was being looked at (simple vandalism, subtle vandalism, accuracy, etc). Knowing at least the first part (how many times a diff has been looked at) will help a great deal. Sorting RC based on that would be even better.
Better still would be to sort RC and similar lists based on trust networks. For example, I trust you and many other people. So edits either made by you and diffs checked by you should be OK in my eyes. It would be nice if my RC and similar lists de-emphasized your edits and edits checked by you. This could go one step farther: I trust your ability to judge whether or not other people are trustworthy. So it would be nice if I could trust by proxy all the users you trust. This creates a trust network.
This could also be done P2P via an offline editor and thus minimize server load. The offline editor would do all the sorting after downloading raw data from Wikipedia.
I think it is obvious that we need bigger and better guns to fight vandals.
This "trust network" sounds not entirely different to the whitelist on CDVF. It shouldn't be too difficult to create a shared whitelist that members of a specified group can all use and add to. Multiple lists could be created and you could subscribe to as many as you like and that the members accept you.
This sounds eminently possible to me, if a fair bit of work. Do people with some technical knowledge have an idea if it is possible?
-- Sam
On 12/1/05, Sam Korn smoddy@gmail.com wrote:
On 12/1/05, Daniel Mayer maveric149@yahoo.com wrote:
What we need are teams of people working together in a way that minimizes duplicated effort. As is, there is no way to know if a diff has been looked at once before, never before, or 200 times. There is also no way to know what was being looked at (simple vandalism, subtle vandalism, accuracy, etc). Knowing at least the first part (how many times a diff has been looked at) will help a great deal. Sorting RC based on that would be even better.
Better still would be to sort RC and similar lists based on trust networks. For example, I trust you and many other people. So edits either made by you and diffs checked by you should be OK in my eyes. It would be nice if my RC and similar lists de-emphasized your edits and edits checked by you. This could go one step farther: I trust your ability to judge whether or not other people are trustworthy. So it would be nice if I could trust by proxy all the users you trust. This creates a trust network.
This could also be done P2P via an offline editor and thus minimize server load. The offline editor would do all the sorting after downloading raw data from Wikipedia.
I think it is obvious that we need bigger and better guns to fight vandals.
This "trust network" sounds not entirely different to the whitelist on CDVF. It shouldn't be too difficult to create a shared whitelist that members of a specified group can all use and add to. Multiple lists could be created and you could subscribe to as many as you like and that the members accept you.
This sounds eminently possible to me, if a fair bit of work. Do people with some technical knowledge have an idea if it is possible?
-- Sam
It's posible but it hits a brick wall pretty fast. A lot of the vandalism on my watchlist is from anons.
Optimiseing is good but it hits problems:
You can only go so far. If you sart ignoreing too many edits because you think they are safe vandles will start to exploit that.
-- geni
On 12/1/05, geni geniice@gmail.com wrote:
It's posible but it hits a brick wall pretty fast. A lot of the vandalism on my watchlist is from anons.
Optimiseing is good but it hits problems:
You can only go so far. If you sart ignoreing too many edits because you think they are safe vandles will start to exploit that.
Yep, a lot of the vandalism is from anons. I personally would never whitelist anons, as that just isn't reliable. But with a whitelist you cut out all the users who do make good edits, meaning you don't have to check them. All the sharing does beyond what already exists is optimises it for use by many people.
Of course, not everyone should use this system. Any universally-used system has loop-holes. Having numerous systems means the loopholes disappear, or at least become insignificant.
I'm sure there *are* loopholes, but I haven't seen them yet. Care to expand?
-- Sam
On 12/1/05, Sam Korn smoddy@gmail.com wrote:
On 12/1/05, geni geniice@gmail.com wrote:
It's posible but it hits a brick wall pretty fast. A lot of the vandalism on my watchlist is from anons.
Optimiseing is good but it hits problems:
You can only go so far. If you sart ignoreing too many edits because you think they are safe vandles will start to exploit that.
Yep, a lot of the vandalism is from anons. I personally would never whitelist anons, as that just isn't reliable. But with a whitelist you cut out all the users who do make good edits, meaning you don't have to check them. All the sharing does beyond what already exists is optimises it for use by many people.
Of course, not everyone should use this system. Any universally-used system has loop-holes. Having numerous systems means the loopholes disappear, or at least become insignificant.
I'm sure there *are* loopholes, but I haven't seen them yet. Care to expand?
When I do RC patrol I do it the old fashioned way and whitelist everyone who isn't an anon. Other than say saturday morning and the like that still means there are far to many edits for me to cheack all of them.
-- geni
On 12/1/05, geni geniice@gmail.com wrote:
When I do RC patrol I do it the old fashioned way and whitelist everyone who isn't an anon. Other than say saturday morning and the like that still means there are far to many edits for me to cheack all of them.
That's a problem inherent in allowing anonymous edits, I suppose. My suggestion about common whitelists and also some kind of edit validation would make this work in general. Add semi-protection to most-vandalised articles and I think the strain of RC patrol would be considerably cut.
-- Sam
On 12/1/05, Sam Korn smoddy@gmail.com wrote:
On 12/1/05, geni geniice@gmail.com wrote:
When I do RC patrol I do it the old fashioned way and whitelist everyone who isn't an anon. Other than say saturday morning and the like that still means there are far to many edits for me to cheack all of them.
That's a problem inherent in allowing anonymous edits, I suppose. My suggestion about common whitelists and also some kind of edit validation would make this work in general. Add semi-protection to most-vandalised articles and I think the strain of RC patrol would be considerably cut.
semic protection makes things worse because you lose the fly paper effect. Consider:
Vandalotices edit tag for the first time. Hits whatever low profile article they happen to be on.
Thinks cools what else shall I hit. I know GWB
Vandal hits GWB gets noticed and previous vandalism gets picked up at the same time.
-- geni
On 12/1/05, geni geniice@gmail.com wrote:
semic protection makes things worse because you lose the fly paper effect. Consider:
Vandalotices edit tag for the first time. Hits whatever low profile article they happen to be on.
Thinks cools what else shall I hit. I know GWB
Vandal hits GWB gets noticed and previous vandalism gets picked up at the same time.
If every edit is getting checked (which is by no means a pipe dream) then that won't be a problem.
By the way, could you run spell-check on your mailing-list posts? It's somewhat distracting when you're a pedant and this isn't a wiki...
-- Sam
On 12/1/05, Sam Korn smoddy@gmail.com wrote:
On 12/1/05, geni geniice@gmail.com wrote:
semic protection makes things worse because you lose the fly paper effect. Consider:
Vandalotices edit tag for the first time. Hits whatever low profile article they happen to be on.
Thinks cools what else shall I hit. I know GWB
Vandal hits GWB gets noticed and previous vandalism gets picked up at the same time.
If every edit is getting checked (which is by no means a pipe dream) then that won't be a problem.
It isn't happening at the moment. Until it does you can have my antivandalism tools when you pry them from my cold dead hands
By the way, could you run spell-check on your mailing-list posts? It's somewhat distracting when you're a pedant and this isn't a wiki...
-- Sam
Someone has taken to emailing with random headers attacking my spelling and me. You may rememer them they surfaced on this list trying to inforce a slightly odd version of english grammer. I'm waiting to see how long it takes for them to explode.
-- geni
On 12/1/05, geni geniice@gmail.com wrote:
Someone has taken to emailing with random headers attacking my spelling and me. You may rememer them they surfaced on this list trying to inforce a slightly odd version of english grammer. I'm waiting to see how long it takes for them to explode.
Thanks for the amusing mental picture. I'm sure they are having a fit.
-Matt
Daniel Mayer wrote: {snip}
What we need are teams of people working together in a way that minimizes duplicated effort. As is, there is no way to know if a diff has been looked at once before, never before, or 200 times. There is also no way to know what was being looked at (simple vandalism, subtle vandalism, accuracy, etc). Knowing at least the first part (how many times a diff has been looked at) will help a great deal. Sorting RC based on that would be even better.
Better still would be to sort RC and similar lists based on trust networks. For example, I trust you and many other people. So edits either made by you and diffs checked by you should be OK in my eyes. It would be nice if my RC and similar lists de-emphasized your edits and edits checked by you. This could go one step farther: I trust your ability to judge whether or not other people are trustworthy. So it would be nice if I could trust by proxy all the users you trust. This creates a trust network.
This could also be done P2P via an offline editor and thus minimize server load. The offline editor would do all the sorting after downloading raw data from Wikipedia.
I think it is obvious that we need bigger and better guns to fight vandals.
-- mav
{snip} Something like that would be great! For every vandalism that gets reverted, I can almost gaurantee 3 or 4 of us checked, and tried to rollback. I'd also love to have more active admins in the #wikipedia-en-vandalism channel (at least today I would have) The last few days, it's been getting crazy. Lotta serial vandals.... IMHO, almost anything that could reduce the duplication in effort would be great. Maybe changing the bot, so that... a user could 'check in' with the bot, and it would rotate vandal warnings to the users that are 'checked in', via privmsg, and the users could in turn, respond to the bot {notvandal, vandal, warnvandal1(2,3,4,5), etc} another nice function to this 'dream bot', would be for it to notify admins of users as they pass thier 4th or 5th warning that week/day/hour/whatever.... (on the same rotating scheme, as to prevent duplication of effort there, too)... I've been debating on writing such a bot in PHP:PEAR/SmartIRC, but work keeps getting in the way... Anyhow just my thoughts on the subject, as a relatively new vandal whacker
[[User:Vilerage]]
Daniel Mayer wrote:
What we need are teams of people working together in a way that minimizes duplicated effort. As is, there is no way to know if a diff has been looked at once before, never before, or 200 times. There is also no way to know what was being looked at (simple vandalism, subtle vandalism, accuracy, etc). Knowing at least the first part (how many times a diff has been looked at) will help a great deal. Sorting RC based on that would be even better.
I don't know if "Mark as patrolled" is currently switched on ... it does help doing RC on Uncyclopedia.
- d.
On 12/1/05, David Gerard fun@thingy.apana.org.au wrote:
Daniel Mayer wrote:
What we need are teams of people working together in a way that minimizes duplicated effort. As is, there is no way to know if a diff has been looked at once before, never before, or 200 times. There is also no way to know what was being looked at (simple vandalism, subtle vandalism, accuracy, etc). Knowing at least the first part (how many times a diff has been looked at) will help a great deal. Sorting RC based on that would be even better.
I don't know if "Mark as patrolled" is currently switched on ... it does help doing RC on Uncyclopedia.
- d.
No. We tried it and it didn't work.
-- geni
On 12/1/05, geni geniice@gmail.com wrote:
No. We tried it and it didn't work.
It could, however, be made to work if it were integrated into the IRC feeds and made to work with CDVF. I think there would have to be some bueaucracy with it, though. There would have to be a level of user (I know this is already programmed in) that can mark edits as patrolled, and this would have to be separately granted. It would just need a little evidence of vandal fighting, just so the simple vandals can't get in.
This doesn't seem unreasonable if it were to be programmed.
-- Sam
On 12/1/05, Sam Korn smoddy@gmail.com wrote:
On 12/1/05, geni geniice@gmail.com wrote:
No. We tried it and it didn't work.
It could, however, be made to work if it were integrated into the IRC feeds and made to work with CDVF. I think there would have to be some bueaucracy with it, though. There would have to be a level of user (I know this is already programmed in) that can mark edits as patrolled, and this would have to be separately granted. It would just need a little evidence of vandal fighting, just so the simple vandals can't get in.
This doesn't seem unreasonable if it were to be programmed.
-- Sam
Have you any idea how fast wikipedia gets edited? the system breaks down because by time someone could be looking at what edits have been missed they are 50 edits ago and off the first page.
-- geni
The answer is pretty obvious in that more diverse ways of sorting RC for different uses. People can get organized all they want, but its pointless to do without the tools to handle the massive numbers. Watchlists only work for hardcore users, and even then it can be a pain, when you find you inflexible it is. (Multiple watchlists would be nice... [1],[2],[3],[4]...).
The sorting by all available namespaces, doesnt really help sorting out the "too much data" problem because 98 percent of the "vandalism" is in the Main namespace -- ie. articles. The sort RC by anon or logged-in users is useful, but its only a one-trick pony (in two different ways), because its only useful for RC patrolling, and does so focusing only on IPs (who happen to make up 20 percent of WP's contributors).
The answer seems like it will be some kind of fine-tuned context-subject sorting, based on the Category scheme. So people can sort RC by SUBJECT OF INTEREST, rather than by an abstract data quantity like login/IP/namespace etc. It's got to be harder to do, as it draws from a user-defined list of categories. There will no doubt be overlap, etc. (Has noone tried to graphically map wikipedia by the way?)
I wonder if the Wikipedia:TOC couldnt just be replicated with Related Changes links instead of Cat links. Someone with sysop privs could even stick that at the top of the main RC....
Stevertigo
--- geni geniice@gmail.com wrote:
On 12/1/05, Sam Korn smoddy@gmail.com wrote:
On 12/1/05, geni geniice@gmail.com wrote:
No. We tried it and it didn't work.
It could, however, be made to work if it were
integrated into the IRC
feeds and made to work with CDVF. I think there
would have to be some
bueaucracy with it, though. There would have to
be a level of user (I
know this is already programmed in) that can mark
edits as patrolled,
and this would have to be separately granted. It
would just need a
little evidence of vandal fighting, just so the
simple vandals can't
get in.
This doesn't seem unreasonable if it were to be
programmed.
-- Sam
Have you any idea how fast wikipedia gets edited? the system breaks down because by time someone could be looking at what edits have been missed they are 50 edits ago and off the first page.
-- geni _______________________________________________ WikiEN-l mailing list WikiEN-l@Wikipedia.org To unsubscribe from this mailing list, visit: http://mail.wikipedia.org/mailman/listinfo/wikien-l
__________________________________________ Yahoo! DSL Something to write home about. Just $16.99/mo. or less. dsl.yahoo.com
Copied WikipediaTOC to WikipediaTOCRC, appending Special:Recentchangeslinked/ to each. Should be identical to the TOC, and useful for general cats.
Stevertigo
--- stevertigo vertigosteve@yahoo.com wrote:
The answer is pretty obvious in that more diverse ways of sorting RC for different uses. People can get organized all they want, but its pointless to do without the tools to handle the massive numbers. Watchlists only work for hardcore users, and even then it can be a pain, when you find you inflexible it is. (Multiple watchlists would be nice... [1],[2],[3],[4]...).
The sorting by all available namespaces, doesnt really help sorting out the "too much data" problem because 98 percent of the "vandalism" is in the Main namespace -- ie. articles. The sort RC by anon or logged-in users is useful, but its only a one-trick pony (in two different ways), because its only useful for RC patrolling, and does so focusing only on IPs (who happen to make up 20 percent of WP's contributors).
The answer seems like it will be some kind of fine-tuned context-subject sorting, based on the Category scheme. So people can sort RC by SUBJECT OF INTEREST, rather than by an abstract data quantity like login/IP/namespace etc. It's got to be harder to do, as it draws from a user-defined list of categories. There will no doubt be overlap, etc. (Has noone tried to graphically map wikipedia by the way?)
I wonder if the Wikipedia:TOC couldnt just be replicated with Related Changes links instead of Cat links. Someone with sysop privs could even stick that at the top of the main RC....
Stevertigo
--- geni geniice@gmail.com wrote:
On 12/1/05, Sam Korn smoddy@gmail.com wrote:
On 12/1/05, geni geniice@gmail.com wrote:
No. We tried it and it didn't work.
It could, however, be made to work if it were
integrated into the IRC
feeds and made to work with CDVF. I think there
would have to be some
bueaucracy with it, though. There would have to
be a level of user (I
know this is already programmed in) that can
mark
edits as patrolled,
and this would have to be separately granted.
It
would just need a
little evidence of vandal fighting, just so the
simple vandals can't
get in.
This doesn't seem unreasonable if it were to be
programmed.
-- Sam
Have you any idea how fast wikipedia gets edited? the system breaks down because by time someone could be looking at what edits have been missed they are 50 edits ago and off the first
page.
-- geni _______________________________________________ WikiEN-l mailing list WikiEN-l@Wikipedia.org To unsubscribe from this mailing list, visit:
http://mail.wikipedia.org/mailman/listinfo/wikien-l
__________________________________________ Yahoo! DSL Something to write home about. Just $16.99/mo. or less. dsl.yahoo.com
WikiEN-l mailing list WikiEN-l@Wikipedia.org To unsubscribe from this mailing list, visit: http://mail.wikipedia.org/mailman/listinfo/wikien-l
__________________________________ Start your day with Yahoo! - Make it your home page! http://www.yahoo.com/r/hs