Moving a wiki to Miraheze

From Miraheze Meta, Miraheze's central coordination wiki
This page contains changes which are not marked for translation. Please make a request at Meta:Administrators' noticeboard or contact any translation administrator or administrator via their user talk page to mark the updated page for translation.
Moving a wiki to Miraheze
Welcome to Miraheze, we're glad to have you here!

Moving a wiki to Miraheze isn't a difficult chore. Whether the wiki is hosted by yourself, Fandom, or any other MediaWiki host, moving a wiki to Miraheze is possible. As long as you have the database dump ('XML' file), the task is easy!

Exporting a wiki[edit | edit source]

The first step in the process is to generate an XML dump file of your wiki (also known as a database dump). This file will contain all pages on your wiki (including user pages and image descriptions but not the images themselves which are separate). The method will vary on the wiki.

Moving from Fandom[edit | edit source]

The button to request dump generation from Fandom

Fandom (formerly known as Wikia) provides an XML dump of each wiki which can be accessed via [[Special:Statistics]] on your wiki. Please note that this file is not generated every day and may be outdated. If it is outdated, wiki administrators can click the "Send request" button there to request a newer dump be generated. If your wiki has no active administrators, you'll have to ask Fandom support to generate one for you. Once that is done, skip to the "Importing a wiki" section.

 Note: Fandom does not allow for wiki closures (except in certain circumstances such as small, niche wikis), mass deletions of pages or linking to external, non-Fandom wikis on the Main Page. Their policy only allows for creating a public post regarding the move which may be linked to on the Main Page. After 2 weeks, the link to the new wiki must be removed from the post.[1]

Images[edit | edit source]

If you wish, you may provide Miraheze with image dumps from your Fandom for us to import. Note that under Fandom's Terms of Use, automated scraping of images is a violation of Fandom's Terms of Use and note too, that images are not provided in XML dumps. Our understanding is that downloading them manually is still legal, but Miraheze is not able to assist you in that process. You may have additional rights under Article 20 of the GDPR in the EU in relation to images that you personally uploaded, due to any metadata linking it to your account.

Moving from WikiDot[edit | edit source]

Moving from WikiDot may pose a more unique challenge. WikiDot does not run MediaWiki, they use a proprietary software solution which is incompatible with MediaWiki.

Some scripts exist to convert WikiDot backups to MediaWiki-compatible XML files. One which has had reported success in the past is this script. Note that it requires an outdated version of Python (2.9) to run and that it is a shell script which may not be intuitive for less technically knowledgeable users.

Exporting via maintenance scripts[edit | edit source]

If you have server access and if you are moving an entire wiki, the easiest way to do so is using a maintenance script to export your entire wiki. For more information, checkout the documentation on dumpBackup.php on MediaWiki.org. Once you have a database dump, skip to "Importing a wiki."

Exporting images via maintenance scripts[edit | edit source]

If you have server access, you can use maintenance script dumpImages.php. Once you have an image dump, skip to "Importing a wiki."

Creating a dump via a script[edit | edit source]

If these methods don't work, you can use WikiTeam's wiki archiving tool to make an XML dump and download all available images at the same time. While it is a useful tool, do note that it does rely on an outdated version of Python to run. We can't guarantee it will work on all wikis, nor we will run it for you. You'll have to run it on your own and prepare the data yourself.

Exporting from Special:Export[edit | edit source]

You can export a wiki manually using Special:Export. That page is self-explanatory, just make sure to uncheck "Include only the current revision, not the full history" as most licenses (such as Creative Commons) require attribution to be rendered so failing to do that will cause legal headaches. While this is useful for smaller wikis or if you're exporting small amounts of pages, it can be tedious to use this interface if exporting an entire wiki.

Caveats[edit | edit source]

  • Apart from the WikiTeam tool method, none of the methods here export image files. Methods system administrators for processing (via a Phabricator ticket).
  • None of these tools move user accounts.
  • As always, any data transferred must have the proper copyright, and you must have a licence or ownership that allows you to host the content at Miraheze (e.g. Creative Commons). If you need to change the licence of the wiki as a whole from CC-SA to some other licence, please note this in your upload request.

Importing a wiki[edit | edit source]

There are a few methods for importing a wiki. Generally, you should always have us do large imports as MediaWiki will display an error if the import is over a few kilobytes big. If your import is small enough, however, you can do it yourself without needing us.

Do it yourself[edit | edit source]

If your dump is really small (less than few kilobytes), you can try Special:Import on your wiki (not this wiki, Miraheze Meta). Images cannot be restored imported using that page. If the import is small enough, your import should finish, and you’ll be able to use your wiki.

If the import fails, please see below section.

 Note: If you choose to do it yourself, you’ll have to manually upload photos yourself. You can use the normal upload interface to do this, or you can use the MsUpload extension to drag and drop and upload groups of photos through the wiki editor.

Have us do it[edit | edit source]

XML (Content) Imports: If your XML dump file is big enough that MediaWiki will likely fail to process it in an appropriate amount of time (files over a few kilobytes), you’ll have to request that a system administrator import it for you. To request that we import the file, please use Special:RequestImportDump and fill in the form.

For image imports, please use Phabricator (Phabricator account required). You can send us your images file in a few different ways:

  • On the Phabricator task, paste a link to where it can be downloaded from (i.e. Google Drive, Dropbox)
  • Send it to sre-mediawiki(at)miraheze.org
  • If it’s small enough, upload it directly to Phabricator

In the meanwhile, while you wait for the import to be done, you should avoid creating new pages that share the same name as pages on your old wiki. That will cause a conflict and the imported version of the page will not override your newly created page. If this happens, you yourself need to manually restore an older revision of the page through the history tab to get an older version of the page to display.

 Note: We disable any editing (including account creation) on the wiki during the import process: Your wiki users need to create an account on Meta (this wiki!) and use that username and password to log in on the wiki until the import is complete.

See also[edit | edit source]

References[edit | edit source]