Forty Three Thousand untagged posts DONE. March 20, 2008 11:54 PM   Subscribe

Holy crap, the back tagging project (recent update) is complete! 43,000 old posts have been keyworded by a team of hundreds of volunteers here. A huge thanks to all, I never thought we'd finish up all those old MeFi posts.

It's funny, I put a "keyword" table in the database around 2001 and was thinking of optionally letting people assign keywords to their posts but I didn't implement it here until January of 2005 (after seeing it take off so well with del.icio.us and flickr). After a couple years of using them, we realized having 33,000+ posts on mefi and 9,000+ early ask mefi questions without tags would be a huge undertaking down the road to catch up and tag those old posts.

interrobang floated the idea in late 2006 and in May of 2007 Jessamyn launched the distributed project of back tagging. pb built a cool ajaxy screen that would display ten posts at a time and let you drop in a string of keywords that all became tags, with a refresh button to give you ten more. Ten at a time, page after page, for almost a year until today when it was finally complete.

A big thanks to everyone that helped out (there's a small message on your userpage) because it was a real slog. I would do a dozen here and there but after a few months I still only had a few hundred complete. An extra special thanks goes out to our top super taggers -- here's the top ten and how many posts they tagged:
  1. CKmtl (7,825)
  2. nkknkk (4,999)
  3. mumkin (2,923)
  4. donnagirl (1,997)
  5. dilettante (1,545)
  6. scratch (1,520)
  7. blue_beetle (1,456)
  8. jessamyn (1,419)
  9. ThePinkSuperhero (1,004)
  10. Rumple (995)
We'll expose the tagging info in the infodump soon and the next project will be tag normalization and grouping and all sorts of stuff we couldn't do until we had everything tagged. So thanks again!
posted by mathowie (staff) to MetaFilter-Related at 11:54 PM (161 comments total) 14 users marked this as a favorite

Amazing! Great work, super heroic back-taggers! I'm sorry I petered out...
posted by carsonb at 11:59 PM on March 20, 2008


Great job, guys. Good team effort. Nice work. But now that that's out of the way, could you guys tag all the comments too?

That would be just super.
posted by Rhaomi at 12:02 AM on March 21, 2008 [1 favorite]


Question for you troopers, did you use the initial tagging screen more, or the one with the smart suggestions?
posted by carsonb at 12:03 AM on March 21, 2008


I forgot to mention that I was trying to buy CKmtl a gift to thank him for the hard work, but he's in Canada and Amazon refuses to let me send him a xbox360 as a gift, and apparently amazon.ca doesn't even sell xboxes. WTF Microsoft/Amazon?!

If anyone knows of an easy way to get a xbox360 sent to someone up north without it being trapped in customs or having warranty issues (that's what Amazon claimed), I'm all ears.
posted by mathowie (staff) at 12:09 AM on March 21, 2008


Yeah, just send the xbox to me. I'll make sure CKmtl gets it.
posted by dhammond at 12:12 AM on March 21, 2008 [1 favorite]


That is AWESOME. Huzzahs all around!
posted by LobsterMitten at 12:25 AM on March 21, 2008


I believe you should be able to buy through the Canadian electronics store Futureshop. That would be the first place I'd call, if they can process an order with a US card.
posted by LobsterMitten at 12:28 AM on March 21, 2008


WOW. Thank you back taggers! I appreciate the insane amount of effort this must have taken.
posted by hurdy gurdy girl at 12:35 AM on March 21, 2008


Awesome, I just went to the tagging page to pass some time and saw it's all done! Feeling vaguely disappointed, but mostly chuffed.
Yippee!
posted by goshling at 12:43 AM on March 21, 2008


Wow, that is pretty amazing.

If only there were a way to harness... oh wait, there was a way.
posted by -t at 12:48 AM on March 21, 2008 [1 favorite]


carsonb: I didn't use the smart suggestion version. Forgot it existed in fact, since I'd just bookmarked the other.
posted by mumkin at 12:55 AM on March 21, 2008


Matt: If you're shipping over the border, make sure it's not sent UPS or Purolator. Those guys charge a customs handling fee that's a hassle. Sending USPS with insurance does not incur the same fee. There are sometimes short delays due to customs, but it's not too bad in my opinion. Maybe it's different if things don't go smoothly.

No idea about warranty though. Maybe check Best Buy or Future Shop's online stores?
posted by ODiV at 12:56 AM on March 21, 2008 [1 favorite]


nice badges, bitches!
/so jealous
posted by Ambrosia Voyeur at 1:21 AM on March 21, 2008 [1 favorite]


Huzzah!

*goes to check that someone tagged #19 as theshimmeringveil*

*comes back, mildly disappointed*

But it's good that 19, 1142, and 9622 all got metafilterhistory tags, so it all evens out! Well done, folks.
posted by stavrosthewonderchicken at 1:31 AM on March 21, 2008


Well done, and thanks! I'm proud to be part of community where people care enough to do things like this.
posted by tomcooke at 2:16 AM on March 21, 2008


I like the little "I was a back tagging superstar!" in those profiles.

I can't even look at my profile anymore.
posted by Corduroy at 3:07 AM on March 21, 2008


That metafilter history tag page is pure gold.
posted by shelleycat at 3:27 AM on March 21, 2008


I miss the old Metafilter without tagging.
posted by terrapin at 3:52 AM on March 21, 2008 [4 favorites]


Does getting onboard, tagging three or four pages, and then forgetting to do it again make me a superstar? Slacking really works when we all pull together as a team, we did a great job guys!! I think Matt should buy us all a beer.
posted by Meatbomb at 3:56 AM on March 21, 2008


Awesome. Big round of applause.
posted by Dreama at 4:01 AM on March 21, 2008


I'm really glad the back-tagging superstar congratulations didn't remove ThePinkSuperhero's magnificent OLD SKOOL shock color page. That page has provided an ocean of pink comfort to me in trying times, and I hope it never, ever changes.
posted by cgc373 at 4:26 AM on March 21, 2008


So I guess this would be a good time to volunteer?

Seriously, thanks for the work, guys and gals!
posted by TedW at 4:34 AM on March 21, 2008


Awesome! Three cheers for all the contributors... hip-hip--hooray! Hip-hip--hooray! Hip-hip--hooray!!!

Keep up the good work:)
posted by hadjiboy at 4:39 AM on March 21, 2008


Great job! I always meant to do more, and I always didn't. I can't believe the big numbers those folks put up.
posted by OmieWise at 5:10 AM on March 21, 2008


Let the front-tagging begin!!!!!
posted by Fuzzy Skinner at 5:17 AM on March 21, 2008 [1 favorite]


Good job everyone. Is it possible for all of us superstars to find out how many we each did? I'm sure I'm an order of magnitude behind the leaderboard, but I'd still like to know.
posted by Rock Steady at 5:18 AM on March 21, 2008


Oh, well done! I wish I was as awesome as all the people that worked on this.
posted by dg at 5:24 AM on March 21, 2008


Yay!
posted by Blazecock Pileon at 5:25 AM on March 21, 2008


2. nkknkk (4,999)

The Obsessive/Compulsive Quasi-significant Number Alliance begs to know -- you couldn't force yourself to do just one more??

Great work everyone. Future generations will thank you.
posted by Dave Faris at 5:28 AM on March 21, 2008


Gotta drink to that! *swig*
posted by Melismata at 5:43 AM on March 21, 2008


Dave Faris, that was my first thought too! We had no idea what our numbers were as we went, or TRUST ME I would have done at LEAST one more.
posted by nkknkk at 6:04 AM on March 21, 2008


Goodness, and I thought I'd done my share when I tagged my own FPPs. You people really put in a lot of effort.
posted by orange swan at 6:10 AM on March 21, 2008


Y'all did a tremendous job. We'll update the tag data in the Infodump; it wouldn't be too hard to put together a little report on the backtagging saga based on that data, probably.
posted by cortex (staff) at 6:30 AM on March 21, 2008


Fucking taggers!

j/k
posted by briank at 6:36 AM on March 21, 2008


Great job, thanks.
posted by bru at 6:37 AM on March 21, 2008


Sweet!
posted by cashman at 6:38 AM on March 21, 2008


69296. yeti (0)

Thanks to everyone that did this for lazy guys like myself.
posted by yeti at 6:50 AM on March 21, 2008


Yay, good job back taggers!
posted by ThePinkSuperhero at 6:55 AM on March 21, 2008 [1 favorite]


It just seems so appropriate that the last tagged post was an I/P single newspaper link post that is now a brokenlink.
posted by Razzle Bathbone at 7:00 AM on March 21, 2008


I, for one, welcome our tagging overlords!
posted by DreamerFi at 7:00 AM on March 21, 2008


I am eternally grateful.
posted by grateful at 7:03 AM on March 21, 2008


TheonlyproblemisthatIforgettoputinspaceswhenIwritenowfromdoingthefinalpushlastnight.
posted by Razzle Bathbone at 7:03 AM on March 21, 2008 [1 favorite]


Can I nominate this thread for the metafilterhistory tag? And what about this deleted thread? (Ok, maybe that one isn't as pretty as Sketchzilla, but it's historic, no?)
posted by yeti at 7:06 AM on March 21, 2008


Let the front-tagging begin!!!!!

Actually, it's funny you should say that. This project got me thinking about whether the older posts might now have better tags than the newer ones, since the backtagging team seemed to be very diligent and professional about what tags to add. I wonder if the pre-tag era posts will be easy to find now while the newer ones will be all, you know, butts lol.
posted by danb at 7:11 AM on March 21, 2008


Woohoo!
posted by klangklangston at 7:22 AM on March 21, 2008


The back-tagging project is not done. It is only done as far as it is currently possible. I can point at 15,992 untagged threads.
posted by Plutor at 7:25 AM on March 21, 2008


Yay. We were talking about giving everyone some little userpage notification about this but we never thought we'd finish so quickly so last night I was a little "er.. thanks for all your hard work everyone" and tossed in a few animated GIFs. CKmtl has been really doing great things with helping ferret out badly typoed tags so we can fix them and we'll be building something to help us clean up the set of tags we are working with. I think it's going to be neat to see them all in the infodump.
posted by jessamyn (staff) at 7:26 AM on March 21, 2008


Not to second guess anyones choice of tags, but after these tags are cleaned up maybe somesort of front-tagging project may be a good idea.

Maybe a script that pulls 10 posts that have only 1 or 2 tags and then MefiMails a tagging posse member. They could then go in and add more tags if necessary or tick a box saying the tags are fine.
posted by Razzle Bathbone at 7:27 AM on March 21, 2008


The back-tagging project is not done. It is only done as far as it is currently possible. I can point at 15,992 untagged threads.

Zoinks!

Whoever tagged that last post, don't ask for that Wii just yet!
posted by cashman at 7:29 AM on March 21, 2008


Plutor - if you can send the details of those threads to mathowie/pb we can build them in to a clean-up brigade. Not knowing the mechanism of how the tagging machine chose untagged posts [and we skip untagged music, project, meta and jobs posts] it's possible that there are some hiding there. But almost 16,000? Really?
posted by jessamyn (staff) at 7:39 AM on March 21, 2008


The back-tagging project is not done. It is only done as far as it is currently possible. I can point at 15,992 untagged threads.

Well, let's not overstate the case; a few thousand of those 15,992 are in fact shockingly green and well-tagged, owing to the strange hybridine nature of late 2003, early 2004.
posted by cortex (staff) at 7:39 AM on March 21, 2008


(In otherwords: metatalk has no tagging capacity.)
posted by cortex (staff) at 7:40 AM on March 21, 2008


Well at least tagging metatalk threads would be simple: lamecallout flameout whinging
posted by Razzle Bathbone at 7:42 AM on March 21, 2008 [2 favorites]


I'm curious how many threads I tagged. Is there a way to find out?
posted by Blazecock Pileon at 7:43 AM on March 21, 2008


Ha! Cute profile bug/thinggie. Superstar! [poses triumphantly, promptly falls over backwards]

carsonb: I just used the plain tagging screen. I saw the tag suggester mentioned once, but forgot about it since I was already in the habit of using the plain one.
posted by CKmtl at 7:47 AM on March 21, 2008


(In otherwords: metatalk has no tagging capacity.)

Which, if you think about it, is probably a good thing. I mean, the heat that some of the callouts here begin with could lead to real interesting ugliness:

Tags:

callout
douchebag
jackassery
threadshitting
hitlerlike
babyeater
selflinker



...actually, could we get tags in MeTa? This would be like jet fuel for flameouts.
posted by quin at 7:49 AM on March 21, 2008


I miss the image threads...
posted by jouke at 7:53 AM on March 21, 2008


And I'm thrilled that the project is finally (more or less) completed! And I'm amazed at the numbers the top 10 put in; you guys must have ocean-like depths of patience.
posted by quin at 7:54 AM on March 21, 2008


you guys must have ocean-like depths of patience

Well, at least 3 of us are librarians - that may explain a thing or two.
posted by donnagirl at 8:00 AM on March 21, 2008


It explains the bizarre proliferation of "LOCrulesDEWEYdrools" tags, at least.
posted by cortex (staff) at 8:01 AM on March 21, 2008


Watch it cortex, you are getting dangerously close to a big library kerfufle between the pro-subject heading people and the pro-tagging people. Believe you me you do not want to get caught in that.
posted by Razzle Bathbone at 8:06 AM on March 21, 2008


Good stuff!
posted by Artw at 8:09 AM on March 21, 2008


Those 30 or so threads I tagged must've really put us over the top. Hooray for me!
posted by loiseau at 8:11 AM on March 21, 2008


Jessamyn came up with the idea of a distributed project of back tagging,

Hooray to Jessamyn for having this fantastic idea!
posted by interrobang at 8:13 AM on March 21, 2008 [1 favorite]


**Does happy little bee dance, now much more accurately wiggling his butt, thus communicating the precise location of the juicy posts and their nectar filled links**
posted by Toekneesan at 8:14 AM on March 21, 2008 [1 favorite]




I'd be all for front-tagging or tag-review (mostly because now I'm selfishly desperate for another huge mind-numbing but strangely fascinating project to do while I watch Tivo) although it does have some implications - ie overwriting FPPs tags for "better" ones is different than adding tags to an untagged post.

That said, there are some crappy recent tags out there... it's like an itch... must... scratch...

btw, yay CMKmtl! I thought I was giving you a run for your money there, but you left me in the dust!
posted by nkknkk at 8:16 AM on March 21, 2008


cortex: "(In otherwords: metatalk has no tagging capacity.)"

Sorry for being so oblique. Yes, the lack of tagging in MetaTalk is shocking, especially considering how useful metafilterhistory could be.

quin: "Which, if you think about it, is probably a good thing. I mean, the heat that some of the callouts here begin with could lead to real interesting ugliness"

How is this a bad thing? People can use those "ugly" words in their comments or their posts. Are tags somehow stronger? Admins can edit them just as easily.

The only dangers I can see are that tags are unflaggable, and even non-oligarchists can tag threads by mutual-contacts no matter how old they are. Do admins have an easy "latest tags added" view?
posted by Plutor at 8:18 AM on March 21, 2008


Yeah, but how many posts were tagged in a prescriptive manner, and how many descriptively?
posted by blue_beetle at 8:23 AM on March 21, 2008


Not everyone can be a soopastah.
posted by Blazecock Pileon at 8:31 AM on March 21, 2008


Great, now I can go back and check to see how my old posts have been handled, and to remove tags where necessary.

(Kidding! Thanks, volunteers.)
posted by Eideteker at 8:32 AM on March 21, 2008


I was a back tagging superstar!

Was, was?! AM!!

Well done all and sundry. I AM sundry.
posted by peacay at 8:34 AM on March 21, 2008


*hands peacay some tomatoes*

Get to work on these, then.
posted by Eideteker at 8:36 AM on March 21, 2008


Okay, so the Wii then. Who won?! Or was that a jokey thing.
posted by cashman at 8:38 AM on March 21, 2008


*splats cashman*
RTFP.
posted by peacay at 8:43 AM on March 21, 2008


..or this.
posted by peacay at 8:47 AM on March 21, 2008


A real slog?!?

*Flagged as sloggerist*
posted by slogger at 8:49 AM on March 21, 2008


the more you know, CKmtl inadvertently helped me fix my disastrous taggage, it may still be ugly, but at least I no longer have 30 character tags that look as if they'd sound like quaint Nordic towns.
posted by dawson at 9:55 AM on March 21, 2008


or rather, the names of said towns
posted by dawson at 9:56 AM on March 21, 2008


Amazing. Thanks, taggers!
posted by rtha at 10:06 AM on March 21, 2008


That's outstanding. Thanks to the community-service-minded folks who helped with tagging. It's just as awesome as people cleaning up the vacant lots and planting gardens in town. It makes the place better, and it's really cool that people were willing to volunteer. I hereby offer you the Thousand Pixels of Light award.
posted by Miko at 10:11 AM on March 21, 2008


That's just staggering. CKMtl tagged almost eight thousand posts, nkknkk did 5, and mumkin did 3. Almost a third of the entire backlog was done by just those three people, and Ck did damn near 20% alone.

What a gigantic time investment. Thanks so much!
posted by Malor at 10:13 AM on March 21, 2008


Wow! That's awesome!
posted by vacapinta at 10:16 AM on March 21, 2008


Malor: "Almost a third of the entire backlog was done by just those three people, and Ck did damn near 20% alone."

If by "almost a third" you mean 36%.
posted by Plutor at 10:19 AM on March 21, 2008


Oh, leave him alone. He's an economist, not a mathematician.
posted by vacapinta at 10:24 AM on March 21, 2008 [4 favorites]


If by "almost a third" you mean 36%.

The pedantic nature of MeFites generally assured this project's success.

I don't mean that in a BAD way, but as someone who was driven to tag over a thousand of these myself, there is really something compelling about missing metadata.
posted by jessamyn (staff) at 10:25 AM on March 21, 2008


Aw, I wanted to help out with this. Too bad it's over, I still would love to help.
posted by flatluigi at 10:26 AM on March 21, 2008


He was saying it was almost a third coming from the other direction.
posted by cortex (staff) at 10:26 AM on March 21, 2008 [1 favorite]


EPIC WIN.
posted by spiderskull at 10:34 AM on March 21, 2008


Pecay, I thought there was a Wii for the person who tagged the last post. Which would be Bazzle Rathbone. Or Razzle Bathbone.

So kindly unsplat me.
posted by cashman at 10:40 AM on March 21, 2008


Tomorrows Post "Whoops I accidentally deleted them."
posted by jeblis at 10:53 AM on March 21, 2008


you guys must have ocean-like depths of patience.

When the alternative is combing through buckets of gravel looking for vole teeth....
posted by Rumple at 10:56 AM on March 21, 2008


Sheesh, I was just eyeballing it, not running a calculator. :)
posted by Malor at 11:56 AM on March 21, 2008


I feel like I didn't do enough to deserve my badge. Good job, everyone!
posted by Alvy Ampersand at 12:01 PM on March 21, 2008


vole teeth, you say...?
posted by jessamyn (staff) at 12:11 PM on March 21, 2008


Is that what the kids are calling it nowadays? Back tagging?

[NOT TAGGIST]
posted by found missing at 12:27 PM on March 21, 2008


My God, I did like, ten before I got distracted. The idea that people did over 1,000--at CKmtl did almost 8,000 just blows my mind. Dude, that was an eighth of the workload! Congratulations all, and thank you!
posted by Anonymous at 12:34 PM on March 21, 2008


Congrats people. Not sure why Plutor seems to have a hate on. This is truly impressive.
posted by purephase at 12:37 PM on March 21, 2008


I forgot to mention that I was trying to buy CKmtl a gift to thank him for the hard work, but he's in Canada and Amazon refuses to let me send him a xbox360 as a gift, and apparently amazon.ca doesn't even sell xboxes. WTF Microsoft/Amazon?!

Yeah, Amazon doesn't ship any electronics to Canada and amazon.ca doesn't sell any electronics. It sucks.

Futureshop.ca or bestbuy.ca should do it.
posted by dobbs at 12:42 PM on March 21, 2008


I don't think Plutor has a hate on; I think he has a welling-desire-for-related-but-metatalk-specific love on.
posted by cortex (staff) at 12:46 PM on March 21, 2008


purephase: "Not sure why Plutor seems to have a hate on."

No, screw you!

I'm sorry. I'll take the weekend off from the Internets, now.
posted by Plutor at 12:47 PM on March 21, 2008


According to the tagging page, there's 1 untagged Mefi post:

MeFi: 1 untagged posts. | 33,562 tagged so far.
AskMeFi: 0 untagged posts. | 9,426 tagged so far

I wonder which one?
posted by ThePinkSuperhero at 12:59 PM on March 21, 2008 [1 favorite]


I wonder which one?

Is that like in historic preservation, where they intentionally leave a tiny strip of wall unpainted so we can see what an original surface was like?
posted by Miko at 1:01 PM on March 21, 2008


Yes, it's our very own truth window.
posted by donnagirl at 1:24 PM on March 21, 2008


Aw, Plutor. *hug* You shoulda taken me up on the cookies. You'd be tasting MeTa win right now. Oh well, I'd give you a greasemonkey merit badge if I could.
posted by Ambrosia Voyeur at 1:31 PM on March 21, 2008


Well done, everyone!

And for those who complain that MeFi is too America-centric, we now have tagible proof to the contrary: only 1,091 of 43,000 posts were related to the US.
posted by UbuRoivas at 4:14 PM on March 21, 2008


It was amazing to me how many broken links I came across when I was tagging. Speaks to the temporary nature of the net, I guess. Most of the dead ends I found were links to news sites. On the other hand, I found some gems like this that hadn't been updated for quite some time, yet remained live.

A tip of my hat to the folks who did the bulk of the work. Wish I could have done more.
posted by SteveInMaine at 4:18 PM on March 21, 2008


UbuRoivas, only 1,091 posts were TAGGED "usa." "America" "unitedstates" were also probably used, and lots of American-centric politics posts weren't tagged with any country identifier, but instead with "recount" "georgebush" etc. It's still better than nothing, but don't draw too many definitive conclusions yet.

FWIW, in my (limited, random, "give me 10 more") experience, ancient MeFi history was VERY America-centric.
posted by nkknkk at 4:35 PM on March 21, 2008


only 1,091 of 43,000

Out of ~33,500 actually. ~9,500 of that total number were untagged AskMe questions.

But yeah, there was a lot of US stuff in there. Which led to a new talent: being able to type "USA politics election 2000 (or 2004) GeorgeBush GWB Bush AlGore Gore (or JohnKerry Kerry) brokenlink" with my eyes closed and without a keyboard in front of me.
posted by CKmtl at 4:47 PM on March 21, 2008 [2 favorites]


One thing about broken links, which I almost made a MeTa post about when I started tagging, is that a lot of posts with broken links were news stories on Yahoo News or Google News or other aggregator sites. Those links expire more quickly than most proper news sites do these days. So, my recommendation is that if you are posting newsfilter, it is best to link straight to the Globe & Mail or the Washington Post or whatever, because those links will last longer, as lots of newspaper sites are archiving things decently.
posted by loiseau at 4:51 PM on March 21, 2008 [2 favorites]


It was funny, I think probably atleast half of the last 50-60 that I did were broken links - single link posts. I know I counted on one of the last batches of 10 that 7 posts were broken.

When it came to those 2000 or 2004 election posts, thank god for auto-complete. I would just start typing algore or johnkerry and hit enter.
posted by Razzle Bathbone at 4:53 PM on March 21, 2008


*goes back through own posts, changes all tags to "quonsar"*
posted by mr_crash_davis at 8:02 PM on March 21, 2008


OK, now, you're it!
posted by dhartung at 10:50 PM on March 21, 2008


Very cool, thanks!

I regret missing this. It sounds like fun.
posted by owhydididoit at 11:03 PM on March 21, 2008


What meatbomb said. I'm embarrassed to get the profile page shout, even though I didn't stick with it for long. I should have made that login a page that opened on booting FF, but I was always, like, "oh, I need to go find that url for tagging," ... but mostly didn't, after the first few days. :(

You other guys are the best, though, and I agree that this project is a concrete example of metafilter as an involved community. Extra-sparkly kudos to Jess for the idea.
posted by taz at 7:50 AM on March 22, 2008 [1 favorite]


I suppose stats could be run on which news sites are the least likely to expire, that'd be an interesting piece of data nerdery.

From what I recall, the Washington Post was sort of mid-range in terms of brokenness. The Globe&Mail was too. CNN, The Guardian, The NYTimes, and BBC were almost always still viable links.
posted by CKmtl at 7:52 AM on March 22, 2008


Huh. Such stats could be run. Just grab every post that has a brokenlink tag, strip out the links, and hash 'em up by domain, I guess.
posted by cortex (staff) at 8:18 AM on March 22, 2008


Hmm, that'd give a rough look at it. But there are cases where brokenlink (or deadlink, or whatever other variants were used) were used for a non-news link in a post containing a news item. So there'd be false positives for brokenness.

Since there's also the posts' dates as data, it might be possible to even rank news sites in terms of how long they last too. "X% of Washington Post links die after 3 years, Y% of BBC ones after 7 years" etc.

Yay, stats!
posted by CKmtl at 8:52 AM on March 22, 2008


Yeah, it'd be rough, but it wouldn't require human intervention.

The dates thing is a good idea, yeah.
posted by cortex (staff) at 9:00 AM on March 22, 2008


Brokenlink is a weird tag, in that it's never going to be assigned by the OP at the time of posting. The most recent brokenlink is from Feb 10, 2007, and I assume it's the result of backtagging, since I backtagged this one from Jan 29, 2007. When did tags become mandatory instead of optional, anyway? Has there been any discussion about automated link-checking, with a MeFiBot that appends a brokenlink tag to posts that it gets an error from?

I tried to reliably tag news organization names, as did others, so perhaps the intersection of name with the brokenlink tag is somewhat meaningful. For example, the organ's tag (# brokenlinks), with the date of the most recent broken link that appears to be the org's fault (as opposed to the poster's):

[ WashingtonPost (95), WaPo (79) ] April 3, 2004
Yahoo (67) Oct 13, 2005
CNN (48) Jan 2, 2005
BBC (30) Nov 20, 2003
[ NYTimes (28), NewYorkTimes (12), NYT (6) ] ???
[ LATimes (18), LosAngelesTimes (1) ] Feb 16, 2005
[ Guardian (18), TheGuardian (3)] July 16, 2004
Reuters (8) Dec 8, 2004
GlobeAndMail (5) Nov 8, 2002
TimesOfIndia (3) Feb 16, 2002

The New York Times is a bit unusual, in that its dead links all appear to be related to feed partnerships. So, stuff with a /reuters or /aponline path may be dead, just like at Yahoo, but otherwise the NY Times is quite healthy. There are some old urls for which the domain is partners.nytimes.com or the 14 posts linking to channel.nytimes.com that 404, but if a mod would care to edit them to be at www.nytimes.com, the rest of the path is valid and resolves to the intended article. (Oh, and there's one that uses www.nyt.com that breaks.) Matt was automatically rewriting NYTimes urls at one time, but only for a few weeks.

Are there any plans to allow more complicated tag searching, by the way? Being limited to finding the intersection of two tags is...well, somewhat limiting.
posted by mumkin at 1:45 PM on March 22, 2008


Brokenlink is a weird tag, in that it's never going to be assigned by the OP at the time of posting.

Of course, now I must dedicate myself to figuring out a way to prove this wrong without creating a post that will immediately get deleted for being useless.

Hmmm.
posted by quin at 1:51 PM on March 22, 2008


It's worth noting that while there are only 263 posts tagged NewYorkTimes, for example, at least 2,222 posts link to stories there. Tags can present a somewhat distorted view of the site if you put too much trust in them, and the guidelines that OPs have had about how to tag their FPPs are less detailed than the ones the backtagging project ostensibly followed.
posted by mumkin at 2:16 PM on March 22, 2008


Also, weird: I just pulled up the tagging page to review its instructions and was presented with this post to tag. Did it just get deleted and re-added? Seems curiously long after the fact for it to be killed and resurrected.
posted by mumkin at 2:18 PM on March 22, 2008


Weird, mumkin; no admin action on that post recently, no. It's not impossible that someone (the poster, maybe) could have stripped the tags out of it after the fact just recently, though.
posted by cortex (staff) at 2:35 PM on March 22, 2008


Wow, good work guys!
posted by dejah420 at 3:15 PM on March 22, 2008


Brokenlink is a weird tag, in that it's never going to be assigned by the OP at the time of posting.

Of course, now I must dedicate myself to figuring out a way to prove this wrong without creating a post that will immediately get deleted for being useless.

It'd have to be some really cool 404 page, right?
posted by Rock Steady at 4:33 PM on March 22, 2008


Cool 404 pages ready for your FPP
posted by Rumple at 5:09 PM on March 22, 2008


double
triple
quadruple
posted by cortex (staff) at 5:16 PM on March 22, 2008


Yay! Well done everyone! I am proud of myself - an unusual feeling for a sunday morning.
posted by Jofus at 3:05 AM on March 23, 2008


Also, can we change our profile pages to say 'I Was A Teenage Backtagging Superstar'?
posted by Jofus at 3:10 AM on March 23, 2008


now if only we had a team of volunteers that hunted down no longer valid old links and replaced them with working links to archived versions or ... ok, sometimes this webbernet thing sucks.
posted by krautland at 9:53 PM on March 23, 2008


now if only we had a team of volunteers that hunted down no longer valid old links and replaced them with working links to archived versions

Or even just replaced them with working links to other random stuff.
posted by Rock Steady at 4:19 AM on March 24, 2008


I'm back. I'm sorry for my previous outbursts. I love you all. Tag team ftw.
posted by Plutor at 7:11 AM on March 24, 2008


aww, thanks, Plutor. If we could tag metatalk posts, we could add "apology" and "redemption" to this one.
posted by donnagirl at 9:11 PM on March 24, 2008


I see we have numbers in our profile now.
posted by tellurian at 5:23 PM on March 25, 2008


Hmm. There are some discrepencies between the numbers reported by mathowie above and the numbers in our profiles. CKmtl's profile has short-changed him by 2060 posts, for example, if the top 10 list is accurate. I'm just 638 short, myself. Anyone server-side know which "score" is more accurate?
posted by mumkin at 5:42 PM on March 25, 2008


Not that it was a contest or anything. Just, you know, wondering at the different values.
posted by mumkin at 5:48 PM on March 25, 2008


Yeah we noticed that too. I think because the count was basically supposed to be a pretty rough estimate, it was counted if someone tagged the same post twice. basically if you loaded tags into a post and then reloaded them, that counted as two. I'm sure pb or mathowie can come in and explain more specifically, but basically for really high-number taggers who are also tagger-reloaders, this discrepancy shows up a little more starkly
posted by jessamyn (staff) at 5:51 PM on March 25, 2008


Gosh, in that case I'm surprised that I only reloaded tagged posts 638 times. That was definitely part of my standard workflow.
posted by mumkin at 6:04 PM on March 25, 2008


Yep, as jessamyn mentioned, the original backtagging counts that I put up were a quick indication of how things were going. I didn't know it would be used as an official tally. When I went to add the numbers to profile pages today, I realized that my quick stats were counting log entries instead of unique post IDs. (oops.) So if someone tagged the same post multiple times (which was easy to do with backtagging system), the post was counted multiple times.

The counts on the profile pages are more accurate because they're counting the number of unique posts tagged. Sorry about the discrepancy, I should have done it that way from the beginning.
posted by pb (staff) at 7:44 PM on March 25, 2008


Ah, that's what it is.

Yeah, I did that a lot. Add tags, click them to check if they were used before or what other stuff they linked to. Delete/change some dead-end ones, correct typos, add new-found related tags were appropriate, etc.
posted by CKmtl at 7:45 PM on March 25, 2008


I cringe a little when I see "I was a back-tagging superstar!" with only 200 posts tagged. (I did mean to get back to it again, but lost the bookmark in a computer wipe, and never resumed.... sorry.)

If you MUST put a note in there, I'd prefer something along the lines of "I helped with backtagging! (200 posts tagged)". Save the real praise for the people who really did the work. "I did a lot of backtagging!" for 250+, "I was a backtagging star!" for 500+, "I backtagged until my fingers fell off!" for 1000+.

Save the superstar tag for the top three, who so, so earned it.
posted by Malor at 12:20 AM on March 26, 2008


Agreed. My paltry 60 posts (a good 40 more than I thought I'd done) should really only warrant a "The back-tagging project just happened to occur at the same time as i was off work with gastric flu and, against the run of probability, was able to hijack someone nearby's wireless internet connection."
posted by Jofus at 1:50 AM on March 26, 2008


Actually, after seeing Plutor's comment, "I was a member of the tag team!" could also work for low-count wimps like me. :)
posted by Malor at 2:00 AM on March 26, 2008


I cringe a little when I see "I was a back-tagging superstar!" with only 200 posts tagged. (I did mean to get back to it again, but lost the bookmark in a computer wipe, and never resumed.... sorry.)

I agree with almost every word of this. "superstar" makes me feel really guilty for free-riding.
posted by LobsterMitten at 7:46 AM on March 26, 2008


Most people on MetaFilter did nothing.

Anyone who helped out in the slightest bit is a superstar in my mind and my mind was the one that put the note on the user pages [with pb and mathowie's assistance].
posted by jessamyn (staff) at 7:49 AM on March 26, 2008


Jessamyn is able to modify others' profiles with the power of her mind. Don't get her angry, or her mind might change your listed gender to "yes plz lol".
posted by Plutor at 8:04 AM on March 26, 2008 [6 favorites]


Most people on MetaFilter did nothing.

Well, I made some of the posts in the first place. That has to count for something.
posted by Dave Faris at 9:33 AM on March 26, 2008 [1 favorite]


Most people on MetaFilter did nothing. FOR THE BACKTAGGING PROJECT WHICH IS WHAT THIS THREAD IS ABOUT.

FTFM.
posted by jessamyn (staff) at 9:45 AM on March 26, 2008 [1 favorite]


FTFM = Florida Thoroughbred Farm Managers

Now THERE's a dream job.
posted by SteveInMaine at 10:13 AM on March 26, 2008


I love you jessamyn.
posted by Meatbomb at 12:35 PM on March 26, 2008 [1 favorite]


Bless you, Plutor, for making that joke so I didn't have to.

Also, if we're all superstars, do we get sparkly leotards, as I once imagined I might wear if I ever became a superstar?
posted by LobsterMitten at 1:33 PM on March 26, 2008


I think I'd rather Matt bought us some sparkly new features instead of a truckload of Danskins. But you should certainly feel entitled to wear a sparkly leotard, LobsterMitten. Just print out your user profile and carry it with you, in case anyone questions your right to be shiny.

Incidentally, I notice that this unfortunate member is identified as a superstar, but without a parenthetical total. A corner case of some kind?
posted by mumkin at 5:25 PM on March 26, 2008


mumkin, you only get a parenthetical total if you tagged over 10 posts. If someone only tagged 1-9 posts, they're a superstar, but not a parenthetical-worthy superstar.
posted by pb (staff) at 5:48 PM on March 26, 2008


I only just now noticed that I'm a parenthetical-worthy superstar.

I rock so hard!

Thanks for the merit badge!
posted by trip and a half at 5:27 AM on March 29, 2008


Hah! Just snagged another one [proof]. What is that elusive blue post?
posted by tellurian at 6:42 PM on April 2, 2008


Looks like that single MeFi post is just a glitch in the system. The query that does the counting thinks post ID 24 is untagged, but that post doesn't exist. We get some new untagged posts from time to time, which just means we need to tighten up our form validation a bit.
posted by pb (staff) at 7:34 AM on April 3, 2008


Post 24 is one of the great mysteries of the early days. Every time I go poking around in the origins of the site I trip across that sucker. It exists but it doesn't. It's missing, but it isn't. If you look at 19 and 25, the first two really real posts, you can see it referenced in the "Newer" and "Older" nav links (respectively):

"Download Abiword..."

Very strange.
posted by cortex (staff) at 9:54 AM on April 3, 2008


Change it so that 24 is a real post. Make it something goofy and recursive, like a link pointing to itself with the claim of being the most meta link ever posted on the site.

Never mention it to anyone, and see how many people eventually stumble across it using the Random link.
posted by quin at 10:05 AM on April 3, 2008


"Make it something goofy and recursive, like a link pointing to itself with the claim of being the most meta link ever posted on the site."

Eh, it's been done.
posted by Eideteker at 10:48 AM on April 3, 2008


I feel bad for #11
posted by thetenthstory at 1:02 PM on April 8, 2008 [1 favorite]


« Older Meetup! What time? Where?   |   Big love for community heroism Newer »

You are not logged in, either login or create an account to post comments