Anybody want to work on Wikipedia?

[Stephane Dumas]

Didn't AARoads.com tried to start a Wikipedia fork who focus on roads a while ago? Maybe we could revive this project.

On a off-topic sidenote, an ex-founder of Wikipedia will try to create a Wikipedia alternative titled "Encyclosphere".
https://www.youtube.com/watch?v=8PrWGMyJgpI

[Scott5114]

Every so often I get a desire to go work on Wikipedia's highway articles in the inland west region, but I find the environment over there tends to be pretty nit-picky on small details which can be annoying sometimes. The lack of allowance for "original research" can also be frustrating especially when it's a logical conclusion I can draw myself from available sources -- like I probably can't put "This route most likely ran along X Road".

That said, I will fix any errors on sight if I happen to notice them.

The attention to small details is usually a combination of desire for consistency across the several thousand road articles, along with compliance with policies that are forced upon the road projects from the rest of Wikipedia. After working on them for a while, you get the sense of how everything works and it isn't much of a bother.

The way around the "original research" policy is to simply state what the sources say–"Y map says the route ran in A area, and Z map shows the route running in B area" and so on, allowing the reader to decide that the route most likely ran along X road.

It's important to remember that early on, it was a fight to get Wikipedia to even accept roadgeek content. This meant that the road project editors had to dot every i and cross every t to demonstrate that the articles were worth keeping around. This mindset is still part of the project.

Didn't AARoads.com tried to start a Wikipedia fork who focus on roads a while ago? Maybe we could revive this project.

I don't remember hearing of any plans to do so. The thought of forking had been raised by Wikipedia project contributors a few times whenever it looked like there was a possibility that the editing environment was getting too toxic to road contributors. In every case, it was decided that the Wikipedia infrastructure (servers, hosting, development environment) was too valuable to split away from, and that doing so would sacrifice the audience, since most people would end up on Wikipedia first anyway.

That last part (the audience) is crucial, because one of the things that the project has always struggled with is attracting editors with knowledge and enthusiasm from lower-population states. Much of the work that remains to be done can only be done through a moderate amount of research. Road research isn't particularly hard, but it does require becoming knowledgeable on where the sources are. This differs from state to state; an editor who is proficient in editing Oklahoma roads will have to learn a whole new constellation of sources to be able to edit Texas roads. An editor also needs an interest level deep enough to sustain the effort across hundreds of pages. In general, roadgeeks tend to be enthusiastic about areas they have a close personal connection to, through living there or visiting frequently for some other reason. This means there is a fair bit of collaboration in high-population states like the coasts, but the interior states tend to be more of a crapshoot as to whether a roadgeek who just happens to live there stumbled onto Wikipedia at some point and was adamant about continuing the work solo.

With that in mind, none of the forks were proposed to have the AARoads name attached, which naturally carries a fair bit of weight with the roadgeek community. This could make it easier to attract editors than a completely independent wiki. That being said, I am not entirely sure how one could relax the policies (other than throwing out the ones that entirely don't apply) without sacrificing quality.

On a off-topic sidenote, an ex-founder of Wikipedia will try to create a Wikipedia alternative titled "Encyclosphere".
https://www.youtube.com/watch?v=8PrWGMyJgpI

I knew without looking that this had to be another Larry Sanger project. This is about his third or fourth bite at the apple when it comes to establishing a Wikipedia alternative. Each time it has failed because his problem with the Wikipedia model is that experts (read: academics) do not carry enough weight. On Wikipedia, an expert has just the same amount of a "vote" (keeping in mind that Wikipedia as a rule does not conduct first-past-the-post voting) as anyone else. Which is just as well and good, since otherwise you'd have randoms claiming to have degrees and not being able to verify that, as has happened already on Wikipedia without it mattering. Then you have a whole other set of things to argue about.

In the past, Sanger's projects have failed because either his experts are too overbearing (and people don't want to essentially deal with a boss when doing a project for fun), or the experts aren't enthusiastic about the project, and lack of participation means attempts at articles never get vetted, or a combination of both.

That being said, this attempt at a Wikipedia replacement discards the wiki model in favor of a content network model. Unless I'm misunderstanding, anyone on here could agree to write a post on their blog to Encyclosphere standards and have it be reviewed and picked up by the network. This is interesting, because it eliminates the inherent conflict that arises when multiple authors edit the same article. However, it also removes the collaborative process that makes Wikipedia a useful sort of information, and opens the door to bias, since each article would be a single-author piece.

[Alex]

Didn't AARoads.com tried to start a Wikipedia fork who focus on roads a while ago? Maybe we could revive this project.

Back in 2009 contributor Jessie Bender suggested adding a wiki to the site that he wanted to help organize. I looked into it and eventually installed Mediawiki successfully. But for the effort, he disappeared and after it sat idle for a several months, I ultimately just removed it.

[rschen7754]

When it came to forking, originally it was concerns like "they're deleting all the road articles" and "they're forcing standards on us that we don't want". Eventually after we got almost 1000 good articles and 60 featured articles that went away for several years.

Lately the Wikimedia Foundation itself has been the main concern with their questionable priorities in software development and their interfering in community affairs like https://en.wikipedia.org/wiki/Wikipedia:Community_response_to_the_Wikimedia_Foundation%27s_ban_of_Fram. But even if things went completely south I don't know if we would have the momentum to start something new (editing is down on the U.S. road articles because many of the primary contributors including myself are busy with outside matters, and there would have to be a massive investment of labor at the beginning just to get the site going).

Not to mention all of the antispam measures that would have to take place (I was a Wikimedia steward for a year and there's entire stewards where all they do is remove spam).

[Scott5114]

I'm not sure of the "massive investment of labor" part–I just migrated a wiki not too long ago. It's a small wiki (~200 pages), but the majority of labor was converting from the old wiki engine's markup to MediaWiki. Going MediaWiki to MediaWiki you can use Special:Export and Special:Import to make things easier (and preserve contrib history), and you don't even have to worry about transferring files if you're exporting from Wikipedia, since the base MediaWiki install still links to Commons.

On a specialist AARoads wiki, I assume we could avoid spammers by doing a similar "user must be approved by an admin before editing" scheme as we do on the forum and requiring registration to edit, since there would be no Foundation mandate to allow editing by anons.

[rschen7754]

It's not just the page imports (10000+ for the US), it's that we would have to import every single template that we use. We would also have to set up all sorts of policies since I don't think "just do it like it was done on Wikipedia" would fly with newer editors. That and having to keep up with all the MediaWiki patches.

[hotdogPi]

I would rather keep everything on Wikipedia.

[vdeane]

I don't think "just do it like it was done on Wikipedia" would fly with newer editors.

I would think "just do it like it was done on Wikipedia, except for cases where we decide otherwise" (in other words, start from Wikipedia's policies and making changes as needed, rather than from scratch) could work.  If someone was willing to import all the stuff and do all the admin maintenance, at least.

[hotdogPi]

A state route that is 3 miles long in a rural area is at least as notable as a village with a single digit population, and the latter is definitely going to be kept in any deletion discussion.

Also, many more people will read about road articles if it's on Wikipedia itself compared to if it's in its own wiki.

[Scott5114]

It's not just the page imports (10000+ for the US), it's that we would have to import every single template that we use. We would also have to set up all sorts of policies since I don't think "just do it like it was done on Wikipedia" would fly with newer editors. That and having to keep up with all the MediaWiki patches.

Other than the biggies (infobox road, the RJL templates) I think that would be doable. Most of the more complex templates, like the cite templates, could be re-implemented in a simpler form to ignore edge cases that don't apply to road articles (would we really need coding to handle scholarly journals?), and for templates that are only used once or twice across the project, judicious substing would be possible.

I wouldn't worry about policy too much. Wikipedia has long had so many policies that they're a turn-off to newbies. No need to replicate that; streamlining would be better where possible.

I do think in a roadgeek-focused wiki, though, you would have to make allowance for sources that Wikipedia deems non-reliable. By which I mean, if some other roadgeek already did the research on their personal site, I don't think the new community would be okay with disallowing that as a cite. I would argue for a tier-based system, where personal sites are citeable (Tier A), but should be replaced by news articles (Tier B) or DOT sources (Tier C) whenever possible. The FA equivalent would still have to be free of Tier A sources.

A state route that is 3 miles long in a rural area is at least as notable as a village with a single digit population, and the latter is definitely going to be kept in any deletion discussion.

Neither of these were a given back when the roads project first started. Someone initially automatically generated all of the US city articles from US Census data (this is why they all read the same for the smaller cities). At the time it was very controversial ("do we really need all of these articles?" "if we allow this then we will have to allow articles on every village in Ghana too"). At one point it was necessary for me to have a road notability FAQ to refer people to.

Going through this shaped some of the mindset of the road editors, and partially explains why we usually don't have full articles on things like FM roads and county routes.

[hotdogPi]

I just added "retroreflectivity" to Wiktionary.

[rschen7754]

California State Route 76 is the featured article today (June 20).

[D-Dey65]

When it came to forking, originally it was concerns like "they're deleting all the road articles" and "they're forcing standards on us that we don't want". Eventually after we got almost 1000 good articles and 60 featured articles that went away for several years.

On a lot of non-road articles, I can relate to that quite easily. I've been at it for a good 15 years or so, and far too often the powers that be on Wikipedia, as well as the Wikimedia Commons just piss me off.