Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
Wikitravel: Bazaar-style Travel Guide Development

By evanp in Internet
Thu Nov 06, 2003 at 07:56:00 PM EST
Tags: Internet (all tags)
Internet

In July 2003, my partner Maj and I founded a project called Wikitravel. Our goal is to create a free (as in freedom), complete, up-to-date and reliable world-wide travel guide. The project was inspired by the success of Wikipedia, the free encyclopedia, but also by our total frustration with inaccurate and incomplete commercial travel guides.


The Cathedral

Try this exercise: next time you're in a bookstore, skim through one of the commercial guidebooks available for your hometown. Do you even recognize the city they're talking about? Where the hell are those restaurants they review? Can't you think of at least a dozen better ones without breaking a sweat? Are there stores, museums, parks and theaters you know that are given little or no coverage? How deep is the discussion of the history, politics and culture of your city or region? How many fine points and details are glossed over? How many important points are omitted? How many just plain factual errors are there?

Once you realize how superficial and inaccurate guidebook coverage of your own home town is, it's hard to believe that coverage of other places is any better. Unless Yourtown USA is the black hole of travel writing -- and what are the odds of that? -- it follows that there's a whole lot of the world that travel guides just don't cover.

The situation with travel guidebooks today is a perfect example of the weaknesses of Cathedral-style content development. Any given guidebook is composed and updated by one or two writers, with maybe a handful of overworked fact checkers and editors providing backup and QA. With tight deadlines and small budgets, these tiny teams have to create and check hundreds or thousands of listings. Is it any wonder that so much falls between the cracks?

I don't mean to put down travel guide writers and publishers. Most of them are dedicated professionals with a real love of adventure and the places they cover. They do the best they can within the framework they have. But that framework is broken, and it shows.

The Bazaar

The crazy part is that there's hundreds of thousands of travelers out beta-testing these guidebooks every day. Each time a traveler shows up at the door of a cafe that went out of business six months ago, or takes a wrong turn on the way to the museum and winds up wandering around lost for hours, that's a bug that should be fixed. And every time that traveler discovers something unlisted -- a rockin' nightclub, a kooky roadside attraction, a cheap and charming B&B -- that's a feature that should be added.

But travelers don't have the ability to update the guides themselves (except in the margins of their own copy). Sure, travel guide publishers usually accept updates, suggestions, complaints and ideas from travelers. But this outside input is treated the same way bug reports and feature requests are treated in your average Cathedral programming shop: haphazardly or not at all.

So travelers end up trying to accumulate and share knowledge outside the bounds of the guidebooks. They talk to other travelers in train stations and cafes, swapping info on where they've just come from for info on where they're going to. They make travel blogs and post to travel bulletin boards, trying to share their info.

These are all great things, and they're part of what makes travel fun. But the problem with all these workarounds is that they ignore the central issue: nobody's fixing the guides.

Wikitravel

What we want to do with Wikitravel is create an environment for Bazaar-style travel guide development. We use a Wiki model, inspired by Wikipedia, so that any reader with a Web browser can edit any page. We want travel guides written by and for travelers (or stay-at-homes!), instantly updateable with the latest information. Because everyone's working on the same document and revising it collaboratively, readers only need to go one place to find the most authoritative travel data.

Maj and I and about a hundred other Wikitravellers have spent the last few months working together to figure out how to make this project a success. We decided what we want to have in Wikitravel and what we don't. We've designed templates and procedures for creating city, country, and regional guides, so each new article gets easier to write. And we've made plans for creating mini-phrasebooks and maps as well as recommended travel itineraries and essays about travel topics.

As I write this, we have over 500 articles in Wikitravel right now, with about 10 new ones starting every day. Lots of them are still seedlings, hardly deserving the name "guide", but bit-by-bit, with contributions from anonymous users drifting in off the Internet, we can see them growing to full-bloom before our eyes.

And every piece of knowledge on Wikitravel is 100% free. We use a copyleft license, the Creative Commons Attribution-ShareAlike 1.0, which lets anyone use the guides we create for pretty much any purpose they can think of. We hope that travelers can find hotels and restaurants online, whether they're already on a trip or just planning at home. We want them to be able to print out guides and stick them in their back pocket for walking around town. We want hotels, tourist information centers, travel agents and other travel industry companies to be able to print our guides directly and provide them to customers, or download them and create private house-brand travel brochures and what have you.

In the future, we hope to have travel guides downloadable to portable computers or PDAs. And we also want to set up a system so that readers can use print-on-demand technology to create ad-hoc travel guides: that is bound books tailor-made for each traveler's interests and itinerary.

Last but definitely not least, we want commercial travel guide publishers to take the knowledge we as a community are creating and include it in their commercial publications. We think everybody, both information producers and information consumers, can benefit from the Bazaar.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Related Links
o Wikitravel
o Wikipedia
o Cathedral- style
o Wiki
o Wikipedia [2]
o Attributio n-ShareAlike 1.0
o Also by evanp


Display: Sort:
Wikitravel: Bazaar-style Travel Guide Development | 74 comments (54 topical, 20 editorial, 0 hidden)
ESR's model (1.54 / 11) (#3)
by spamfish on Wed Nov 05, 2003 at 08:51:30 PM EST

Fails to mention the fact that it is controlled by a "benevolent dictator." I think FreeBSD personally is better, Linux has been stagnant at 2.4.X for quite some time, while FreeBSD keeps moving on up. Though I have some disdain for the whole open source movement (stupid zealots) this is a good idea. I enjoy travelling, perhaps I will contribute.

Cows taste alright.

Linux is moving, too. (2.20 / 5) (#16)
by Highlander on Thu Nov 06, 2003 at 05:12:03 AM EST

Latest Linux version is 2.6.0. That version just isn't ready for the public yet, but it shows linux is not stagnant.

I doubt it is a good idea to measure the success of an OS by the number of its releases.

OTOH, with linux you need the new releases for the drivers that come with it. Drivers are easier to come by for Windoze, or what do you think ?

Moderation in moderation is a good thing.
[ Parent ]

LInux is moving (1.33 / 9) (#40)
by Replicabot on Thu Nov 06, 2003 at 02:53:13 PM EST

Yes, I've heard that ever since IBM bought Linux, great things have been happening. I guess that's what happens when you get experienced corporate leadership with lots of resources behind a project. I suppose it also means that Linux will die when IBM gets another one of its phases when the leadership is poor and it performs behind the market.

K5 is my anti-drug! What's yours?
[ Parent ]
Damn (none / 1) (#47)
by evanp on Thu Nov 06, 2003 at 06:33:39 PM EST

You're good.

[ Parent ]
Awesome (none / 3) (#5)
by dennis on Wed Nov 05, 2003 at 09:21:53 PM EST

I've wanted something like this ever since I read Hitchhiker's Guide to the Galaxy.

Last time I traveled extensively, Let's Go Europe was just reliable enough so you started to trust it just before it screwed you.

-1, Probably Illegal (1.64 / 17) (#6)
by SIGNOR SPAGHETTI on Wed Nov 05, 2003 at 11:00:37 PM EST

This is just a way for unscrupulous criminals, pirates and gypsies -- and in the Middle East, terrorists -- to post bogus travelogues and drive tourists to "charming", "out-of-the-way" locations where they will be beaten, their monies stolen, and titillating worse. If that's what you want in a holiday then fine: visit wikitravel and loiter in train stations. But I do not think AOL subscribers wearing loud Hawaiian shirts, Florida Marlins baseball caps and size 12 bowling shoes want that! I did not understand your theory of Cathedrals and Bazaars, or why you thought this story was an appropriate place to mention it.

--
Stop dreaming and finish your spaghetti.

Shhh! (2.70 / 10) (#9)
by Tatarigami on Wed Nov 05, 2003 at 11:59:37 PM EST

Luring AOL users to their deaths in isolated corners of the world is the whole point!

[ Parent ]
I'm on AOL. (1.50 / 8) (#10)
by SIGNOR SPAGHETTI on Thu Nov 06, 2003 at 12:15:47 AM EST

=:-o

--
Stop dreaming and finish your spaghetti.
[ Parent ]

What's your point? [nt] (2.40 / 5) (#13)
by Tatarigami on Thu Nov 06, 2003 at 12:34:37 AM EST

;)

[ Parent ]
Suspected Troll (2.83 / 6) (#23)
by evanp on Thu Nov 06, 2003 at 10:18:29 AM EST

Well, I have a feeling I'm falling into a troll trap, but you've got a good point about Wikis: since anyone can update an article, what keeps the information reliable? There's a lot of theory on that on the Web. Maybe I'll try to dig some up. I didn't really want this article to turn into an apology for the Wiki Way, though. The whole idea is so counter-intuitive, it's hard to explain.

[ Parent ]
As a regular contributor to Wikipedia... (2.75 / 4) (#44)
by Russell Dovey on Thu Nov 06, 2003 at 04:30:02 PM EST

I agree that the concept of an "electronic graffitti wall" would seem to be useless for reliable information. However, the strengths of Wiki are manifold.

Like all good systems, Wiki exploits human nature to operate smoothly. There will always be train-spotter nerds who watch their favourite topics like a hawk, alert to any ignorant additions or changes. They will generally correct any mistakes within hours, if not minutes.

Areas with more general interest will have more nerds, so will be more reliable, but fringe topics usually have at least two people who have wasted their brainspace on this one topic. Also, Wiki's ability to instantly regress an article to a previous changepoint ensures that vandalism is easily fixed.

A travel Wiki is a stroke of inspired genius. While Paris and London might be well-served by travel books, a person living in, say, Australia is often forced to correct the knowledge of tourists with bad information. This will allow locals to correct bad information easily and without complication.

"Blessed are the cracked, for they let in the light." - Spike Milligan
[ Parent ]

I like it! (2.85 / 7) (#7)
by simul on Wed Nov 05, 2003 at 11:35:21 PM EST

When I travel, I'll write things down (favorite hotel, restarant, etc.) and post them. If everyone just wrote things down and posted them, then the guide would grow organically.

It's smart and simple.

Idea: encourage the use of keywords like "vegetarian" and "adventure" and "relaxing" for people searching for particlar types of travel experience. Maybe enhance with a checkbox of keyowrds on each category of post.

Read this book - first 24 pages are free to browse - it rocks

Oh... (none / 3) (#8)
by simul on Wed Nov 05, 2003 at 11:36:14 PM EST

If we all just posted stuff about our own city... then the gude woudl grow really fast.

In my mind I assumed you had to travel first.

Dumb!

Read this book - first 24 pages are free to browse - it rocks
[ Parent ]

And once again... (none / 2) (#15)
by dasunt on Thu Nov 06, 2003 at 02:28:32 AM EST

simul writes
Idea: encourage the use of keywords like "vegetarian" and "adventure" and "relaxing" for people searching for particlar types of travel experience.

And once again, the vegans are left out...



[ Parent ]
Too true (none / 2) (#32)
by killmepleez on Thu Nov 06, 2003 at 12:19:21 PM EST

There is absolutely nothing "adventurous" or "relaxing" about vegans.

__
"...if there isn't a delivery mechanism, it's not a weapon of mass destruction, it's a paperweight of mass destruction."
--Parent ]
Thanks (none / 1) (#22)
by evanp on Thu Nov 06, 2003 at 10:15:09 AM EST

We like the idea, too.

[ Parent ]
-1, Kuro5hin President is a Travel Agent now (nt) (2.00 / 6) (#17)
by A Proud American on Thu Nov 06, 2003 at 08:47:59 AM EST

{Conflict of interest}

____________________________
The weak are killed and eaten...


Great Point (none / 2) (#19)
by evanp on Thu Nov 06, 2003 at 09:57:24 AM EST

Shoot, I forgot to mention travel agents! After all, they could definitely benefit from having tons of useful, Free information about destinations that they can print out directly, or download and use in house-brand brochures... seems like a goldmine to me.

[ Parent ]
wiki == suck -nt- (1.07 / 13) (#24)
by Suppafly on Thu Nov 06, 2003 at 10:32:28 AM EST


---
Playstation Sucks.
I like the idea (2.75 / 4) (#25)
by l3nz on Thu Nov 06, 2003 at 10:56:47 AM EST

Still we'll have to see how it will be developed. There is probably room for abuse - what protects contributions to abuse? if, say, somebody has a restaurant or a hotel and they keep putting it first on the list and add bad comments to other resorts in order to attract visitors.

Popk ToDo lists - yet another web-based ToDo list manager. 100% AJAX free :-)

I kinda didn't want to get into this stuff (3.00 / 6) (#26)
by evanp on Thu Nov 06, 2003 at 11:04:10 AM EST

I really wanted to kind of gloss over the mechanics of Wiki editing. But, to address your question, here's the theory: someone comes in and tries to edit articles in an abusive way -- either with blatant advertisements, or to undercut other companies, or to give bad information, or just vandalism and trolling ("HLUAGLALGL LIK MY BALLZ!!@!"). For each person who does that, there's another 5 people who actually care about the integrity and quality of Wikitravel, and go overwrite those bad changes with real info. If the Abuser comes back and does it again, the Fixers just fix it again. And so on. The theory is that Fixers care more than Abusers. Eventually, the troll or vandal is just going to give up, and the "real" information will remain. Yeah, it sounds crazy, but for some reason this is actually how it works. Wikipedia gets vandalized on a daily basis, yet it remains an excellent source of information. There's lots of ways this concept can be subverted -- automated vandalism, subtle vandalism that runs under the radar -- but more or less it works. Yeah, I don't get it, either. But there you have it.

[ Parent ]
I wonder (3.00 / 4) (#31)
by l3nz on Thu Nov 06, 2003 at 12:19:02 PM EST

This is true - and this is the reason why Wikis exist, otherwise the'd be long time dead. :-)

What I was asking is that there is quite a difference between writing "Julius Caesar was a Finnish leader" and "I did not find that hotel to be so good"; it's maybe harder to spot trolls and abusers. Don't you think this could - say - water down the quality or the usefulness of the system?

Anyway I like the idea! :-)

Popk ToDo lists - yet another web-based ToDo list manager. 100% AJAX free :-)
[ Parent ]

Subtle subversion (3.00 / 4) (#34)
by evanp on Thu Nov 06, 2003 at 01:13:41 PM EST

Yes, there's actually a great possibility of subtle subversion. Remember, though, that we don't have a posting for "Restaurant X", with 300 comments tagged at the end ("I like it", "I didn't like it", "I like it", "I didn't like it", etc.). There's just one listing for a restaurant, and that's definitive. If it's controversial, we try to settle down on one description. We haven't had a problem with this yet, though. But, as you said, some of this stuff is subjective. A _lot_ of stuff isn't -- what hotels are in a city, what the address of the museum is, or the phone number for the bus company.

[ Parent ]
There's a similar (but non-wiki) site (2.50 / 6) (#29)
by fraise on Thu Nov 06, 2003 at 11:20:58 AM EST

IgoUgo

Been around for about three years now. I used and contributed regularly during its first two years, but stopped once I ran out of info and a reason to consult it. Nowadays I get better information directly from locals. But based on how it was back then, my opinion is that it worked well because A. submissions were reviewed by the site's editors, and B. users could rate submissions. I'm not sure how well a wiki-style travelogue will work, because it will be wide open to spam from hotels and such. I used to be responsible for France:Travel and Tourism on the ODP, so I have experience with this!! People with commercial interest in tourism just never give up when they have a free outlet. (Drove me to leave the ODP, in fact, because I got so sick of deleting spam.)

Abuse (3.00 / 6) (#30)
by evanp on Thu Nov 06, 2003 at 11:35:22 AM EST

So, two comments: first, I just commented on abuse. The basic Wiki idea is that people with dedication to the integrity of the Wiki put in the time to correct abuse, while abusers eventually get bored and go home.

The other comment: man, I hope we get to the point where hotels and restaurants are trying to spam the site. I figure once they're clamoring for Wikitravel endorsement, that means we've made it.

Problems like these I wanna have.

[ Parent ]

Travel guide sites (2.50 / 4) (#35)
by evanp on Thu Nov 06, 2003 at 01:17:40 PM EST

So, there are a number of these kind of travel guide sites. I like them, mostly. Most are good.

The problem I find with them, though, is that there's not really a <u>single</u> guide for any particular destination. So if I want to know about Florida, I have to read 8 different people's guide to Florida, and then mentally merge them into a single guide.

That's not so bad, and in fact it has some great benefits -- you can find people you identify with, or who share your interests, and follow their guide more closely.

The big problem, for me, is that there's no quality control by readers. You've just got all these little Cathedrals instead of one big Cathedral. It's nice when people can just go change those spelling mistakes themselves.

[ Parent ]

this is a cool idea, +1 FP (3.00 / 4) (#37)
by urdine on Thu Nov 06, 2003 at 01:38:15 PM EST

I hope this works. Travel guides are the perfect topic for a wiki-style web structure. Nice presentation in the article, too.

Thanks (2.75 / 4) (#38)
by evanp on Thu Nov 06, 2003 at 01:41:36 PM EST

I hope it works too. I had hoped that the discussion in the article gave a good idea of our motivations and what the problem was that we were trying to solve.

It's nice to get positive feedback. Thanks.

[ Parent ]

Excellent (3.00 / 4) (#39)
by CaptainSuperBoy on Thu Nov 06, 2003 at 02:35:33 PM EST

Excellent application of a wiki. I'm going to Europe for three weeks next month, I'd love to have the use of a guide that reflected reality for a change. Unfortunately there are a lot of holes in the Wikitravel project but it still seems like a good resource. I'll try and contribute once I get back.

As guidebooks go, I've heard the Lonely Planet series is pretty good. Not perfect, but I think they are a little more accurate than some of the other ones.

Sort of offtopic, we are going to start in Turin, Italy and go by train through Southern France and into the Barcelona area of Spain. Any suggestions on what places we should go or what places we should skip?

--
jimmysquid.com - I take pictures.

Wiki + Reality = ? (none / 2) (#62)
by tkatchev on Fri Nov 07, 2003 at 05:23:46 PM EST

= TROLL.

   -- Signed, Lev Andropoff, cosmonaut.
[ Parent ]

There are great guides out there. (2.50 / 4) (#45)
by Tezcatlipoca on Thu Nov 06, 2003 at 05:32:05 PM EST

I take issue with your assertion that the commercial guides are somehow lacking.

You have to diferentiate between the normal tourist that is going to visit a place for one or two weeks and the long term tourist, like trouly yours, that gets stranded in a foreign land longer than intended.

As things, stand most commercial travel guides fullfil perfectly the need to inform tourists about a place they are visiting. Thanks to those guides I have made wonderful walking tours of places as diverse as Bangkok, Cape Town or Vienna and have managed to survive the horrific traffic in Rome by riding the public transport system as suggested by most guides.

Now, as a Mexican, I surely know that the "Mercado de Sonora" is extremely interesting to check the exotic fresh produce that Mexico City consumes every day and that there is a little known art market in the Monumento a la madre where painters gather every Sunday to sell their wares. Well, this may be interesting, but a tourist that will be in Mexico City once in a lifetime better goes and check the Teotihuacan pyramids, which will be fully explained in any guide you care to mention.

I think your project is interesting but there is no need to knock down stablished resources that are extremely useful.

Might is right
Freedom? Which freedom?

What about the more exotic things (none / 2) (#46)
by spamfish on Thu Nov 06, 2003 at 06:00:41 PM EST

besides landmarks and the like? I personally like to walk around the city and find whatever interests me instead of going to see some landmark since most landmarks are boring anyway. Seeing some old building on a guided tour does not give you as much sense of the culture and society of these people than roaming around.

Cows taste alright.
[ Parent ]

Bagging on commercial travel guides (none / 2) (#48)
by evanp on Thu Nov 06, 2003 at 06:38:28 PM EST

You know, it's funny: one of our admins made his first contributions because he was so disgusted with coverage of Mexico City by commercial guides.

But, hey, I know: some people like the commercial guides. I know that having something is better than having nothing. And a lot of the people who work on the guides get a lot of great stuff done.

I think the point is: guidebooks could be a lot better. We can improve them by letting travelers, and residents, and just regular people update the books.

That's all!

[ Parent ]

A problem with this model (2.87 / 8) (#52)
by theantix on Thu Nov 06, 2003 at 08:34:13 PM EST

I like what you are doing with it, I really do.  But looking over the wiki for my home city, Vancouver BC, I notice a problem that could be an indicator of bad things to come in the future.  When discussing how to get to/from Vancouver by train, It says " VIA rail service to points East leaves Vancouver at 17:30 on Tuesdays, Fridays and Sundays. This means you will be travelling through the mountains at night, so if you want to see the scenery, look into the Rocky Mountaineer as an alternative." and links to this page. If you read that linked page, it seems to be basically an advertisement for the Rocky Mountaineer train service.  

The statements in the original page are misleading, because they don't mention that in the summer you have five hours of daylight after 5:30PM, and you get plenty of scenery on that trip.  And when you wake up in the morning you are still going through the mountains -- I know because I've taken it and it's amazing scenery.  Combining the misleading statement on the first page and the advertisement on the linked page, the result is alarming to me.

My problem is not with that directly -- duh, it's a wiki, I can clean it up.  But if someone from Rocky Mountaineer wants to keep advertising his site on your wiki it would be pretty hard to stop.  It seems like a structural problem to your idea, because there are many other buisinesses out there that will try to get free advertising on your website and have good reason to be more dedicated to maintaining updates to the wiki than the travelers who you intend to do the updates.

I'm just curious if you have anticipated this type of problem, and if you have come up with a way to handle these activities.

--
You sir, are worse than Hitler!

I recommend random killings (none / 2) (#53)
by CaptainSuperBoy on Thu Nov 06, 2003 at 09:11:01 PM EST

No but seriously, for a topic to be wiki-able it has to generate enough enthusiasm so that the regulars are more vigilant than the advertisers and the vandals. I don't know if travel guides are a sustainable topic for a wiki but it's worth a shot. C2's design patterns wiki is an extraordinary resource and it's much more specialized than a general travel site, but still keeps an active enough user base to stop any problems. Of course design patterns are a lot less susceptible to stealth advertising than travel guides, so who knows.

--
jimmysquid.com - I take pictures.
[ Parent ]
Rocky Mountaineer (2.50 / 4) (#57)
by evanp on Fri Nov 07, 2003 at 09:56:09 AM EST

The person who wrote the article about the Mountaineer is not an advertiser. She's just an enthusiast of the Rocky Mountaineer.

Advertising is just another form of abuse. At this point, I wish I had written about abuse in the main article, even though that wasn't the point of the article, because if you look over these comments, there's like 18 about abuse. So, heh.

Anyways: abusers come in and abuse once. People who care about the site fix it. The abuser may come back once or twice, but eventually they give up. That's the magic of Wikis: those who care more win.

[ Parent ]

Right (none / 1) (#68)
by theantix on Wed Nov 12, 2003 at 07:14:17 PM EST

Look, I understand the wiki concept and how it works generally for sites like wikipedia.  But judging from your flippant response I can see that you really haven't thought this through all that far.  Who is going to update this site, who is your main source of people inputting information.  I would venture that the "best" people you want to input information on, say, Barcelona are the people who have visited Barcelona recently and want to share good and bad information.  To me that is why the wiki concept is good in this application.  

Of course obvious abuse could be corrected..... but I'm not talking about defacing a page or something that one could easily spot.  I'm talking about a subtle sort of free advertising that would be difficult to spot, like the page I linked to earlier -- even though it may not have been advertising after all like you claim.  

However, while I could see myself adding information on the various places I've been to in order to help fellow travellors, I'm not going to devote enough time to it to prevent a restaurateur from replacing my recommendations with his -- hell, I probably wouldn't even notice it as abuse.  The restaurateur has sufficient motivation to constantly update the site to recommend his establishment, but I for example would not.

I really hope that you take my concerns into consideration because I strongly feel that problems will arise if you don't have a strategy to combat this sort of problem.  Which is a shame, since I also feel that this kind of site would be great for a wiki, if managed properly.  Either way good luck.  =)

--
You sir, are worse than Hitler!
[ Parent ]

Abuse and Commercial Incentive (none / 3) (#70)
by evanp on Sun Nov 16, 2003 at 12:29:07 PM EST

Well, first of all, let me say that I didn't mean to be flip. I really was kicking myself for not answering the abuse question in the article itself, since it's been the main question I've gotten on K5 so far.

Now, to the meat of the question: Wikitravel has a few means of countering abuse. First, let me point out that rampant abuse is a problem of success. I only hope we have to deal with abuse problems. It means lots of people are looking at the site.

The first, and most effective, is simply have a large number of dedicated Wikitravellers watching for changes to the site, and reversing any changes that look abusive. The software we use has a lot of tools for this: lists of recent changes, "diff" utilities to show what's changed on an article, page histories, etc. Administrators can also protect a page -- keep it from being edited from anyone except an admin -- if it's the subject of an edit war, with people reverting each other's changes right and left.

I can't overstate how effective this is. We have a handful of abuse cases per week, and they're usually reverted within minutes.

Second, all changes are logged either by user name (for logged-in users) or IP address (for anonymous users). If a particular user or IP address continues to be abusive, we can simply block that user or address from editing Wikitravel pages. Yes, it's a crude tool -- abusers can simply create a new user ID, or go to a different computer, or whatever -- but it's the second line of defense.

And that's it. But what if a concerted attack comes against the site, such as a DDOS attack, from thousands of computers, automated so editors can't keep up with the changes? The answer is simply that we haven't seen it yet, and neither has Wikipedia. They get thousands of changes a day, and have 160K+ articles. If this happened, we could always lock down the database so no new changes can happen, until we figure out how to deal with the attack. Damage can be kept to a minimum.

Now, what I'm getting from your concern is that you think that this is not enough, for two reasons:

  • Commercial advertising is a subtle enough abuse that it will fly under the radar of a cadre of dedicated editors.
  • People who want to do commercial advertising are motivated enough to launch continuous abuse attacks that go around our two levels of protection.

My response to the first is that if advertising attacks occur, they'll either be a) not subtle or b) harmless. If they're not subtle, they'll get zarked. Other editors will keep them honest and fair, and within the bounds of our manual of style. Super-huge blink-tagged stuff with inappropriate claims of excellent deals will be fixed just as well as "HULAGHULAGHL LIK MY BALLZ!!!1!".

If the advertisements are subtle, that is, well-written listings that fit in with our MoS, well, then, who cares? Truthful, accurate information that is helpful to travellers... what's the problem? Frankly, I'd love to have businesses keep their own listings on Wikitravel up-to-date. I'm not particularly interested in the motivations of contributors, but rather with the quality of their contributions. It's not our job to look into the hearts of whoever's on the other side of the screen. If a single destination gets overloaded with listings by umpteen restaurants or hotels, an editor will come in and trim them.

To your second argument -- that businesses are more motivated -- I agree in theory, but not in practice. As I said, I only hope someone gets motivated enough to want to spend big bucks to put their listing on our site. When being listed in Wikitravel is worth mounting a huge attack against the site... I mean, isn't that pretty much a sign that we've made it? I also wonder what businesses would bother with this -- either paying someone to sit and edit, edit, edit a single page, or develop an automated attack.

I also want to point out that any talk of travel and destinations is necessarily going to name names of actual businesses, with addresses, phone numbers, and reviews. There's no way we're going to have pure content without any businesses listed -- that's just not worthwhile for travelers. Unless you're a hobo, you have to pay someone to get you from point A to point B, to feed you, to give you space to sleep. It's better to have names for these people than not. That's part of what a travel guide is for.

Your expectation that we provide some other kind of "protection" seems a little strange. What do you expect us to do? Get certified affidavits from each and every contributor that they have no commercial interest in the subjects they're writing about? Have an elaborate submission process, where certified editors review each and every spelling correction to make sure it doesn't have a subliminal advertisement in it?

The power of Wiki is that anyone can edit a page. By opening the floodgates, you allow anything, the good along with the bad. Experience with Wikis such as Wikipedia shows that the good far outweighs, and overpowers, the bad. Until we see otherwise, I don't think we're going to hobble Wikitravel with obstacles to contributing, and lose the power of open, anonymous editing, until we really see that the model doesn't work.

[ Parent ]

awesome (none / 1) (#71)
by theantix on Tue Nov 18, 2003 at 06:30:40 PM EST

Okay, I think you've addressed my concerns pretty well.  I see your point about subtle commercial tags not being really bad, and while it might not be the best way of getting the best results I would say that it also might not be the terrible fiasco that I implied it would be.

Now that you've convinced me in the idea behind th site, all I have to do is contribute to it... ;-)  Thanks for your time in (a) designing the site, and (b) addressing my concerns.  Cheers and good luck again.

--
You sir, are worse than Hitler!
[ Parent ]

Oh, man (none / 1) (#72)
by evanp on Wed Nov 19, 2003 at 09:45:44 PM EST

I love it when that happens. Thanks for making me think all that through -- I've got a couple more articles about Wikitravel on the burner, and I really need to hone these arguments.

[ Parent ]
The Proprietary Competition (2.50 / 4) (#55)
by pde on Fri Nov 07, 2003 at 12:07:58 AM EST

There is pre-existing proprietary competition:

Lonely Planet runs a set of discussion forums called The Thorn Tree which attempt to do (some) of the same things.

Apparently, LP have recognised that their paid writers have no hope of collecting as much information as a horde of 'net using travellers.

I know a few people who've picked up very valuable recommendations there.

Visit Computerbank, a GNU/Linux based charity

Travel Forums (none / 1) (#58)
by evanp on Fri Nov 07, 2003 at 11:26:13 AM EST

I think there's a lot of value to travel forums. The hard part, though, is trying to find particular information about a place you're going within a forum. You can either try searching the forum for some keywords -- "Bangkok" "Youth Hostel Munich" -- or post a question to the forum, and then check back over and over for answers.

The idea behind Wikitravel is that you don't have to do that. People go ahead and just write up their information about a place, and if something changes, someone else comes in and corrects it. Readers only check one place for the latest information.

[ Parent ]

Wiki Problems (2.50 / 4) (#56)
by Graymalkin on Fri Nov 07, 2003 at 02:10:12 AM EST

I really like this idea. I think a decent addition to the Wiki model would be some form of article rating. I would describe it similarly to Amazon's comment rating. If I write an article on getting around in LA and people find it helpful they can say so. I would get a reward for writing a helpful entry albeit a virtual reward. Someone spamming would get an unhelpful or spam rating and it would be easier for people to avoid such entries.

With a Wiki model you have to build up some trust between the article writer and the reader. If I give directions to Canter's that send someone into Ventura they should have the ability to tell other people not to listen to my directions. Optimally someone is going to spot my mistake and correct it. If I however am constantly giving bad advice or directions users should have a way to flag my entries. If I constantly give good directions and advice I could also be flagged as someone the user trusts.

Other than that I think the idea is pretty cool. The opinion and advice of locals can be invaluable when you're traveling or planning a trip.

Some problems with ratings (none / 2) (#59)
by evanp on Fri Nov 07, 2003 at 11:30:06 AM EST

Well, the big problem with ratings is that articles in a Wiki don't have any single author. Some of our articles on Wikitravel have 20 or more! Everyone throws their little bit of info into the mix, and it all gets mashed up together. So it'd be hard to rate an article based on one contributor's reputation.

If some contributor puts in wrong or abusive information on one article, it will be changed to something more useful or correct. If a contributor is continuously abusive, disruptive, or continues to vandalize articles or give purposefully incorrect advice in order to mess with people, we can ban them from contributing to the Wiki. We haven't had to do this yet, and I think it's only happened a few times on Wikipedia, but we can if we must.

[ Parent ]

Ok (none / 2) (#63)
by Graymalkin on Fri Nov 07, 2003 at 09:53:35 PM EST

I see what you're saying. I wasn't really thinking of the Wiki as a mishmash of different writers. Would there be a way to keep track of the editions made by particular users for use in a rating system? At least some way for people to see that you or I have made good changes or editions and maybe some other user has made a bunch of inaccurate ones.

[ Parent ]
Keeping track of changes (none / 1) (#64)
by cep on Sat Nov 08, 2003 at 02:33:43 PM EST

In a way, there is. For each article, there is a version history, so each (partial) author can see the changes made by others. (This is in fact my experience with Wikipedia, but Wikitravel seems to use the same software.)

[ Parent ]
Cool! (none / 2) (#60)
by bigbtommy on Fri Nov 07, 2003 at 03:12:55 PM EST

As a wiki owner (LawWiki), this is a pretty neat use of it. I didn't quite understand why until reading this article. The Bazaar model rules, and I'm just seeing the benefit of it on LawWiki as we get the first few contributors. Good luck with Wikitravel.
-- bbCity.co.uk - When I see kids, I speed up
Thanks (none / 1) (#61)
by evanp on Fri Nov 07, 2003 at 05:08:07 PM EST

I figure the next few months are going to be the hardest -- not only keeping up momentum, but dealing with success. We'll see what happens; keep your fingers crossed. And good luck with LawWiki. I just posted there; I think I need to write an article soon about why the GFDL is so inappropriate for short publications.

[ Parent ]
This is going to FUCK UP rusty's hotel site. nt (none / 2) (#65)
by Nigga on Mon Nov 10, 2003 at 11:44:27 AM EST


--------
The fuck happened to Nigga?

CIA factbook info - what for? (none / 2) (#66)
by frijolito on Tue Nov 11, 2003 at 06:39:24 PM EST

OK, when I first read about this I thought to myself: "hey, good idea! I'll go ahead and enter some travel tips on my small country". It was only until today that I had a free moment, so I went and hit the Guatemalan page.

Then I find there a whole crapload of information: an import from the CIA World Factbook 2002, and a disclaimer saying it was only a starting point. Fine, I think, so I hit the FAQ. Turns out there's also a rather large manual of style (which I barely skimmed), and a country template which all country information should follow. After some more reading, I find that the CIA information should belong in one of the categories of the country template. And after a while I figured that if I wish to use the template to edit my country, I have to edit the template's page, copy the "wikitravel markup code", and paste it in my country's page.

What a pain. All the while I'm playing along, but thinking that in reality all that CIA info shouldn't be there. What a new user should see when she hits an unedited country is an empty country template, and that's it. So whenever someone new hits the page, they would spend less time on figuring out how to input stuff, and more time entering information.

This is my suggestion: Instead of having the factbook import as the "default" information on a country, the country template should be there. This way you won't deter users from "plunging in". Actually, I was just about ready to give up on the whole deal, but then I thought you guys would probably appreciate the feedback... and now, after typing all this, I feel kinda guilty and will try to implement the suggested changes in my country's page.

Still, I think new users should be welcomed by an empty "country template", rather than a crapload of information collected by, of all people, the CIA.

CIA Factbook Import (none / 2) (#67)
by evanp on Wed Nov 12, 2003 at 04:13:26 PM EST

So, I completely agree with you about working with the CIA Factbook import pages. Most of the information that the CIA wants on a country is not really cogent to travelers.

The main reason we did the import in the first place was to lay out a framework for all the countries in the world. We also figured that at least some of the information would be useful for writing real Wikitravel country articles.

Also, it's a common situation for Wiki systems that people feel more comfortable adding to an existing page than they do starting with a blank page. So having pages for all the countries we could find was, we thought, a step in the right direction.

The sad fact is that the CIA factbook pages are really hard for people to work with. The tabular format for information is really ugly, and it keeps people from really diving. Dealing with changing over from the CIA factbook pages to real pages is a big problem right now; you're not the only one to have it.

Thanks for the feedback, though.

[ Parent ]

Update on this (none / 1) (#69)
by evanp on Thu Nov 13, 2003 at 05:46:48 PM EST

After some work, there's a new Country article template which is a lot more usable, and which doesn't include the CIA factbook info. Thanks for your input; it was a real kick in the ass to get this going.

[ Parent ]
The same, yet different. (none / 0) (#73)
by Fricka on Sat Nov 22, 2003 at 10:18:20 PM EST

http://www.virtualtourist.com

It's not the same as a wiki but it appears to be driven by individuals adding content. I found it while looking up information for Chicago... which turned out to be right on the money. I thought the entries were engaging and adding them in seemed straight forwarde, I was tempted to add information in about my own town (but I had to finish packing instead).

As other posters have noted, I think Wikitravel will need an obvious and easy way for folks to add information in -- and one that doesn't look like one will have to commit their entire day to fill out. I was able to figure out how to join and add information to virtual touris in a few seconds, with wiki I had to drill down to my town then was confronted with a large blank square.

Just some feedback, not knocking the idea, as I'd love to see it grow.

----------- Support my Internet and Gaming Hobbies: http://www.OffLineTshirts.com

Some differences (none / 0) (#74)
by evanp on Wed Nov 26, 2003 at 03:00:44 PM EST

Yes, there are some similarities with Virtual Tourist. There are some differences too, though.
  • Multiple guides. Each user on VT can put up their own guide to a destination. There are usually a handful of guides to each one. It's up to readers to combine the information on the different guides into a coherent whole. On Wikitravel, we have one guide for a destination. Contributors are responsible for combining their own knowledge into a single guide. The reader just has to print it out and put it in their pocket.
  • You can't correct someone's mistakes. If someone puts up a guide that's really good, but with a few factual errors, you can't correct it. You can post your own guide with the corrections, or tips, or whatever, but it's up to the user to update their own guide. The reliability benefit doesn't come through unless everyone can edit.
  • Thousands of tips. Virtual Tourist's user contributions consist of thousands of tips. For Chicago, for example, there's 2750 tips (today). Each one has to be navigated to separately. For Wikitravel, we'd like to produce one set of related documents called the Chicago guide, which combines all that info.
  • It's not Open Content. That's probably the most important one. The content on Virtual Tourist isn't open; unless specifically stated by contributors, you have no right to re-publish any VT content.

I think ultimately services like VT or Lonely Planet's Thorntree are useful, if you're willing to put in the time and effort to go searching around for each little tip. Hopefully, you find the tip that says that the hotel you're planning to stay at burned down... If not, tough luck.

I kind of think of these user-contribution sites as a fake bazaar. Users aren't really trusted to work on content directly; the services simply allow users to post their own stuff. Users can't edit or update other users' work, which is where Wikitravel shines.

As to ease-of-use for wikis: yeah, that's tough. We are kind of in a bind: some people want more structure for new articles, some people want less. You got a plain white box; another person had problems with our template. When people ask, I usually just tell them to put in what they know, however they want, and someone else will come in and reformat it later. That's how wiki works, after all.

[ Parent ]

Wikitravel: Bazaar-style Travel Guide Development | 74 comments (54 topical, 20 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!