- ThurWiki
- Migration
Moving across without losing pages.
ThurWiki imports from Confluence, MediaWiki, DokuWiki, XWiki, BookStack, Notion and Markdown archives, step by step with a preview. Nothing is written before the last step. And whatever the wiki cannot convert, it tells you about, instead of quietly leaving it out.
ThurWiki beside what you already know.
An honest comparison. Confluence is a good tool, and so are MediaWiki and DokuWiki. The differences lie where a company works every day: in the writing, in the content from other systems that ThurWiki fetches through macros (ready-made building blocks for the page), and in where the data sits.
| ThurWiki | Confluence Cloud | MediaWiki | DokuWiki | |
|---|---|---|---|---|
| An editor without markup (control characters such as == Title ==), a ribbon like Word's | yes | yes | partly, VisualEditor as an extension | partly, through extensions |
| Tables: sort, filter and number while reading | yes | sorting | sorting | through extensions |
| Charts from a table, without an add-on | yes | through apps | through extensions | through extensions |
| Content from bexio, WHMCS, Abacus on the page, fetched instead of typed out | yes, 13 macros | no | no | no |
| Jira, Git, CRM, calendar, e-mail on the page | yes, 17 macros | Jira yes, the rest through apps | through extensions | through extensions |
| An AI assistant with the rights of the user, a model in your own house possible | yes | Rovo, only in the provider's cloud | no | no |
| Simultaneous editing | yes | yes | no | no |
| Four-eyes approvals | yes | through apps | through extensions | through extensions |
| Full-text search inside PDF, Word, Excel | yes | yes | through extensions | through extensions |
| Image search across nine archives with licence attribution | yes | no | Wikimedia Commons | no |
| draw.io in our own house | yes | through an app, at the app maker's | through an extension | through an extension |
| Data in Switzerland, on the provider's infrastructure | yes, ThurData | data residency per region, provider in the USA | if self-hosted | if self-hosted |
| Operation, backup and updates by the provider | yes | yes | yourself | yourself |
| Help inside the interface, written for end users | 60 topics at the question mark | the maker's documentation | the community | the community |
As of September 2026, based on the makers' public statements and without warranty. What can be achieved with extensions depends on the particular module and how well it is maintained.
The import in five steps.
You find the import in the administration under "Import pages". An import wizard takes you through five steps. Use the preview: it shows the page tree as it would come into being, with titles and attachments.
Choose the source
Confluence (Cloud, Server or Data Center), MediaWiki, DokuWiki, XWiki, BookStack, a Notion export or an archive of Markdown files. For the wikis through their API, with an address and access; for exports as a file.
Check
The wiki looks at what is inside and reports whatever it cannot read.
Choose the space
Where the pages should go. We recommend an empty space: that way you see what has arrived.
Preview
On the left the page tree of the source, on the right the chosen page as it would look after the import – with tables, attachments and a note about everything the wizard had to guess.
Run it
Pages, attachments, images, links. A run can be repeated without pages appearing twice: the wiki knows what it has already fetched.
What is carried over
Pages with title, text, tables, images and attachments. The page tree. Links between the pages, rewritten to the new addresses. draw.io drawings, still editable. The common Confluence macros, converted.
What the wiki does not know
A macro ThurWiki does not know leaves a visible line with its name and an entry in the import report. If it fell away without a trace, something would be missing from the page and nobody would notice. That is the worst thing an import can do.
What you do afterwards
The import brings content, not rules. It is best to create the users before the import, so that mentions find their target. Rights and groups you set afterwards, and deliberately: a move is the best opportunity not to carry the tangle of old permissions across with you.
A good way to run a move.
These tips come from a real customer's content, against which we tested the import: 917 Confluence pages with some 2900 macro occurrences.
A small piece first
Twenty pages into an empty space. Look at what arrived. Only then the rest.
Not everything has to come
Whatever nobody has opened in three years needs no new home. Sort it out before the import.
The titles come across, including the bad ones
"New page (2)" stays "New page (2)". Reckon with follow-up work; an import does the bulk, not all of it.
Run in parallel, then switch off
Let the old wiki run on read-only until the new one has settled in. The import can be repeated to pick up changes made in the meantime.
Three starting points, three routes.
From Confluence
Cloud, Server or Data Center, through the API or from the export. Spaces become spaces, the page tree stays, the common macros are converted. Your team finds the editor in the same place: a ribbon, tables, macros. The mapping of the macros is on the macro page.
Confluence macros and their counterparts →From MediaWiki, DokuWiki, XWiki, BookStack
Through the wiki's API, with an address and access. Out of wiki markup come pages anybody can edit. From BookStack you choose a shelf; every book becomes a top-level page in the target space, chapters become pages with sub-pages. The greatest gain is rarely the import itself: it is the day somebody outside IT writes a page for the first time.
The editor →From Notion, Markdown and a drive full of files
A Notion export or an archive of Markdown files becomes a page tree. Word files are not imported but attached and shown on the page; the search finds their content. A drive full of instructions becomes a place with search, rights and history.
The search →Questions about migrating
Which systems can ThurWiki import from?
Confluence (Cloud, Server and Data Center), MediaWiki, DokuWiki, XWiki, BookStack, a Notion export and an archive of Markdown files. The import runs in five steps with a preview; nothing is written before the last step. A run can be repeated without pages appearing twice.
What is carried over by the import?
Pages with title, text, tables, images and attachments; the page tree; links between the pages, which the wiki rewrites to the new addresses; the common Confluence macros, converted into ThurWiki macros. Not carried over are rights and users: the import brings content, you set the rules afterwards. It is best to create the users before the import, so that mentions find their target.
How long does a move take?
That depends on the size and on how tidy the old wiki is. An import does the bulk of the work, not all of it; reckon with follow-up work, and use it to sort out old material. We recommend bringing twenty pages into an empty space first and looking at the result. After that, the effort for the rest can be judged well.
Can we run both wikis in parallel for a while?
Yes. The import leaves the source system untouched. What has proved itself is to keep the old wiki running read-only until the new one has settled in, and then switch it off. You can start a further import run at any time; it picks up what has changed in the meantime, without creating pages twice.
What do we gain over Confluence?
An editor that needs no getting used to, content from bexio, WHMCS, Abacus, Git and CRM directly on the page without add-ons, an AI assistant with the rights of the user, four-eyes approvals, image search with licence attribution, draw.io in our own house. And: your data sits in Switzerland on ThurData's own infrastructure, a Swiss company with no US parent; the US CLOUD Act (an American law giving US authorities access to data held by American providers, including abroad) does not apply.
Show us your wiki. We will show you how it looks in ThurWiki.
With twenty of your pages in a demo wiki, imported with a preview, before you decide.