User:Ned Scott/transwiki

Rough ideas for guidelines/advice on exporting articles outside of Wikimedia Foundation projects.

Things to do

• Create/propose an "export article" link in the AFD template.

• Create/propose a undeletion-for-transwiki request page and find interested admins to help out. Noting that articles deleted for BLP or copyright reasons cannot be undeleted, but others are normally not a problem.

• Work with people who are good with pywikipediabot to create/improve scripts for the Wikipedia-to-external wiki process.

• On this same idea, set up a request for bot-help kind of page for transwiking.

• Seek non-python scripts to lessen the learning curve on at least some tasks.

• Find out if programs and scripts like AWB can be adapted to work on external wikis. When you have a smaller group of editors, script-powered editing becomes pretty essential.

General advice

• Categorize your images by the work of fiction it is in. Images are often forgotten about in the transwiki process, and besides individual article usage, are not often grouped together in a way that makes moving them easy.

• Categorize your redirects by work of fiction. This is important for keeping article history of merged articles, as well as showing you what common alternative titles should be redirected on the new wiki.

• Look to other similar external wikis for advice and ideas. You don't have to re-invent the wheel, and you shouldn't get too hung up on trying to make everything specific to the new wiki when it doesn't need to be.

Content advice

• Don't give everything it's own article. While it might be tempting to shake off the restrictions of Wikipedia by giving everything its own article, it doesn't always make for a good move from an organizational standpoint. Something having its own article or not should not be a sign of importance, but rather simply about information organization. Preemptive redirects are a good way to guide new users to this idea, without actually banning any article from being independent. Take some time and discuss article structure, maybe even make a graphical tree to give other users an idea of what you have in mind. This not only helps you understand what articles should be made, but helps you keep articles linked when they should be. Of course, such a tree doesn't have to be exact with each page in the wiki, but even a general topical tree can help keep things organized for both editors and readers.

Example

• You might also want to keep "main" articles similar to Wikipedia articles, as general over-views, and have a companion article for the details, to make finding the basic information faster (one positive side effect of this is for wikis that wish to keep unexpected spoilers to a minimal). For example, you might have a main character article, but then have "relationships" articles (for teh omg romance), which would also mean you wouldn't have to update two character articles each time something small changed in their relationship. One possible idea, and there are many more to consider.

Taking from Wikipedia's example

• When you're on an external wiki, you're almost always starting from scratch. Even basic templates we've come to use as fundamental functions have to be copied over, such as a simple Template:unsigned. It would be great to round up some of Wikipedia's most useful and most common operations templates under a category or page. A toolbox for new wikis, of sorts.

• On a new wiki you also find yourself with the need for rules, guidelines, and policies. This is another way Wikipedia is a great resource, you not only can re-use articles, you can re-use our policies and guidelines under the GFDL as well. To save time and effort, I've told people on Digimon Wiki to use Wikipedia's same guidelines and policies, but with a list of exceptions (wikia:digimon:DigimonWiki:General guidelines). Eventually, when time permits, a simplified guideline set can be copied over. Wikia's own basic guidelines are often derived from Wikipedia's. (On that same note, for those who are wondering, Wikia only sets some very basic rules, and the rest is up to the consensus of the local community of editors for each wiki.)

• Use Wikipedia's MOS guidelines. Wikipedia's WP:MOS guidelines emphasize on how to format the text, titles, and so on, and can normally be applied with ease to other wikis, since they normally don't contain content advice. Even when they do, it's still easy to just link to the Wikipedia MOS, and have a write up of exceptions on the external wiki. Alternatively, you can selectively use the MOS, pointing to specific issues such as naming conventions, etc.

Importing/Exporting

• Special:Export's instructions are outdated, and some of the functions don't actually work as advertised. However, using this url directly will export the entire history of an article:

• http://en.wikipedia.org/w/index.php?title=Special:Export&pages=NAME_OF_ARTICLE&history=1&action=submit

• Save the resulting file as something.xml

• Use a simple find/replace function in a text editor. Find all occurances of "</username>" and replace it with "@en.wikipedia.org</username>". This will properly attribute the contributions, as well as avoiding wiki-identity fraud or innocent confusion.

• A user with admin access on the external wiki can then import the file using Special:Import. Remember that if you are importing to Wikia, there is a size limit just below 2MB.

Notes

• There are other ways to "transwiki" an article, such as a simple copy/paste, and then linking back to Wikipedia for the full page history (using templates such as centralwikia:Template:Wikipedia). I would only use this route when the article being linked won't likely be deleted. Even then, it doesn't hurt to have a second copy of the article history. However, this is an ideal method when the article on Wikipedia is still active, since that way you don't have to re-import every time one copy is updated.

• The other method is simply cut and pasting the article itself, then cut and paste the list of contributors from the history page of the article to the talk page. From what I'm told, this is fully GFDL compliant, but less desired since you can't see other revisions, or see who did what on an article.

• Wikia seems to have a 1.9MB file size limit for importing XML files (the format articles are in when exported). Articles larger than this have to be chopped up and imported. For myself, I've been trying to find/create a script to do this, as I currently have several large xml files to import (some are up to 60MB). Smaller files can be combined into larger documents to make importing faster. I've created an Apple Automator workflow for this (with my newbie skills), but I'd like to make something a little more flexible/easy/efficient/etc. My own experience with scripts is limited, so finding people to help with this process will be very important.

• One of the problems people seem to face on Wikia is that they feel alone and unable to get help. Wikipedia has a very strong and active community, and I think many people would be glad to help, but don't know where help is needed, or even that someone was looking for help at all. There should be a better way to connect everyone.

• Due to something called Tidy, many templates that work on Wikipedia will stop working on other wikis, such as those hosted by Wikia. These are templates that mix HTML table code with wikitable code. Helping people convert these templates, or trying to get Wikipedia to use more pure-wikitable templates, would greatly help this issue.

• Since many external wikis have less people to check for vandalism, an external-wiki anti-vandalism bot network would be ideal. Anti-vandal bot code is not publicly available, and it's release is limited to only those who are trusted and need to run such bots (this is done to prevent vandals from finding weaknesses and workarounds). However, it might be possible to extend anti-vandal bot services to multiple external wikis. The workload demand would be far less than on Wikipedia, so it comes down to finding volunteers who already have access to anti-vandal bots and would be willing to expand their use, or finding one or two people who might be trusted with access.

• There is also the Wikia:Annex, which was set up as a staging area for articles to be transwikied. The instructions on the main page tell a user how to transwiki when still keeping the article history on Wikipedia.

• It's more than just Wikia. While Wikia is likely to be the most comment host / target for external-trasnwiking, ideally guidelines should be set up for any GFDL compatible/Mediawiki site (eventually non-mediawiki as well, but things are hard enough as it is that it would be more efficient to just focus on Mediawiki wikis for the time being)

• It's more than just fiction. Entertainment and fiction are easy examples of articles that are good candidates to be transwikied, but there are also wikis for many other things, such as garagebands and minor organizations, and just about everything else.