How to fork a wiki Options

[emesee]

Step 1

Sign up for hosting somewhere. You need to have PHP and MySQL installed, at least.

Dreamhost will work nicely, and quite often, you can signup, with a unique domain name for just 10 dollars, for the first year.

There are many other hosts to signup with. It will often run about 5-15+ dollars a month. If you are not getting mega ammounts of traffic, a host somewhere within this range should suffice.

Step 2

Install MediaWiki. This can have a bit of a learning curve, but once you get the hang of it, it is not super super difficult. See:
http://www.mediawiki.org/wiki/Manual:Installing_MediaWiki

Step 3

Get the XML dump of the wiki that you wish to fork.

Many can be found at:
http://download.wikimedia.org/backup-index.html

If you want to fork Commons, then see:
http://yousefourabi.com/blog/semantic-web/...ages-with-wikix

If the wiki you wish to fork does not have an XML dump available, then you still could fork the wiki, but it could be a pain. You would have to get the page list at [[Special:Allpages]] and then either export all pages manually at [[Special:Export]] (with a large wiki, this could take a long, long time) or else setup some sort script to do it (find it using Google, if it exists, or write it yourself).

Step 4

Import the dump. See:
http://www.mediawiki.org/wiki/Manual:Importing_XML_dumps

Appendix

Configure the Wiki

Then of course, configure MediaWiki at sometime in here. You may want to Import [[MediaWiki:common.css]] from the big wiki.

You can enable the use of images from Commons. See:
http://www.mediawiki.org/wiki/Manual:$...oreignFileRepos

A pretty decent overview of other good to know configurations are at:
http://wiki.dreamhost.com/MediaWiki

You may need to tweak the configuration over time. See:
http://www.mediawiki.org/wiki/MediaWiki

The MediaWiki wiki has a lot of information on it, and is organized perhaps pretty reasonably.

(IMG:smilys0b23ax56/default/smile.gif)

[CharlotteWebb]

You left out the all-important detail of how to get other people to contribute to your site rather than to Wikipedia, and how to keep it from degenerating into a grudge site being colonized primarily by unpersons banned from TOW, etc.

If two trees diverge in a yellow wood, but not even Robert Frost visits your wiki, is it still a fork? (IMG:smilys0b23ax56/default/tongue.gif)

[Eva Destruction]

You left out the all-important detail of how to get other people to contribute to your site rather than to Wikipedia, and how to keep it from degenerating into a grudge site being colonized primarily by unpersons banned from TOW, etc.

If two trees diverge in a yellow wood, but not even Robert Frost visits your wiki, is it still a fork? (IMG:smilys0b23ax56/default/tongue.gif)

You also forgot step 5: realise that the moment you start hosting it (and consequently become the "publisher") you become liable for any libels, errors etc and for taking steps to fix them (Jimbo & pals can at least point to the fact that they try to fix errors), and also become subject to the laws of your country as opposed to Florida's comparatively lax libel laws. And while IANAL, I suspect that because you'd be actively importing any offending content – as opposed to just being the host on which other people post libels – you'd be on very shaky ground if you tried to claim s230, should anyone take exception to any of the offending material.

[CharlotteWebb]

You also forgot step 5: realise that the moment you start hosting it (and consequently become the "publisher") you become liable for any libels, errors etc and for taking steps to fix them (Jimbo & pals can at least point to the fact that they try to fix errors), and also become subject to the laws of your country as opposed to Florida's comparatively lax libel laws. And while IANAL, I suspect that because you'd be actively importing any offending content – as opposed to just being the host on which other people post libels – you'd be on very shaky ground if you tried to claim s230, should anyone take exception to any of the offending material.

Hmm, do the hundreds of known mirror sites claim (or need to claim) any kind of immunity? That's probably news to them.

How about the "scraper" sites which just load the current version of a Wikipedia page and pass it along to you (substituting their own site's name and logo, etc.)?

[Eva Destruction]

You also forgot step 5: realise that the moment you start hosting it (and consequently become the "publisher") you become liable for any libels, errors etc and for taking steps to fix them (Jimbo & pals can at least point to the fact that they try to fix errors), and also become subject to the laws of your country as opposed to Florida's comparatively lax libel laws. And while IANAL, I suspect that because you'd be actively importing any offending content – as opposed to just being the host on which other people post libels – you'd be on very shaky ground if you tried to claim s230, should anyone take exception to any of the offending material.

Hmm, do the hundreds of known mirror sites claim (or need to claim) any kind of immunity? That's probably news to them.

How about the "scraper" sites which just load the current version of a Wikipedia page and pass it along to you (substituting their own site's name and logo, etc.)?

Those listed at Wikipedia:Mirrors and forks/All are mostly specialist projects that have just forked a few pages and have checked them before hosting their own versions. The mirror sites should, at any rate, not be editable and purely mirror the Wikipedia pages, and are essentially just an alternative means of accessing the main site's content. Major scrapers like BBC Music that mirror Wikipedia pages (as opposed to using Wikipedia articles as a base on which to write their own articles) are festooned with legal disclaimers and tend to have a policy of suspending particular articles on request if issues are raised regarding accuracy. As I understand it, what Emesee is talking about (correct me if I'm wrong) is the mechanics of duplicating the whole of Wikipedia and using it as a basis – it would be virtually impossible to even attempt to check the 2million+ articles to ensure an accurate starting point, without Wikipedia's high user base and equally high (but often ignored) casual-browser-who-quietly-fixes-an-error count. As I understand it, this was one of the main reasons Larry ditched the original plan to import-and-clean-up the whole of Wikipedia as a starting point for Citizendium.

Forks can work if there's a substantial amount of people who edit in a non-malicious manner and keep pages up to date. This happens if the wiki is associated with a topic which can interest a number of people from a highly populated forum (WoW wiki, Star Wars wiki and Star Trek wiki are certainly successful, and more accurate than the wikiprojects on Wikipedia related to those topics).

So would it be possible to start a wiki based on a narrow topic (to lure the Right Kind Of People onto the ground floor) then gradually broaden its scope to include, eh... potentially everything? Has anyone ever taken this approach?

Conservapedia – which started out as a host for teaching materials for Eagle Forum University – is one that springs to mind.