These are some common dilemmas. If you have a question, don't hesitate to add it.
- How can I mark a change as "trivial" or "minor", so that it doesn't clutter the recent changes? If you make several edits in a row, Hatta will automatically make all but the first one as trivial, provided you don't change the comment. Other than that, you can't.
- Can I run this wiki as CGI? It is possible, using a CGI wrapper for WSGI applications, but it wasn't tested and may be slow. See Plain CGI on shared hosting for details.
- What license is this released under? It's free software, but since I'm not good with legalese, I used the MU License: If you are asking what license this software is released under, you are asking the wrong question. On the other hand, Mercurial itself is GPLv2 and it's being used as a library, so it has to be GPLv2 or compatible.
- How to translate Home, Changes etc. page titles into another language? The Home page is unfortunately hard coded at the moment (maybe it would be better to use Start or even an empty link?), the rest is defined on the Menu page, where you can use
[[link|label]]
to have different text for the Changes link. Proper internationalization support is upcoming. - All right, I can edit a given page, but how do I add one of my own? As with most wikis, you create a link on some existing page, and then follow it – you will see an editor for the new page. The links to non-existing pages are even displayed differently.
- Secondly, how fast is it? I run it as a single FastCGI process with flup on my server, it takes about 13MB of RAM and can serve about 18 requests per second for a wiki page (front page in this case), and about 25 requests per second for a binary file (the logo in this case). This is comparable to other wiki engines, like MoinMoin. Of course this depends highly on the server you use and its load. Hatta also makes use of HTTP caching mechanisms, HEAD requests, conditional requests, etags, etc., so its performance should be better than that of more dynamic wikis with macros and detailed permission systems.
- How much stuff can it keep? I.e. does it scale? Each page is a separate file, so the number of them is only limited by your filesystem and perhaps Mercurial's limits, if it has any. The number of pages doesn't affect browsing speed – each page is just read and rendered independently of any others, so the only thing that affects their display speed is their content. Same with editing, except that saving involves updating backlinks and committing to the repository – which depend to a some degree on the number of pages, linearly. Search is indexed, based on a hash map, so it's $\log n$, where $n$ is the number of pages. Results of some tests can be seen at Performance page.
- It too could use Git as its backend, as another option. Unfortunately the library that provides python bindings is far from being mature and has to be installed separately from GIT, which makes it rather unattractive for this kind of utility. Also, I wrote this wiki mainly to fit my own needs, and I don't use GIT
Maybe have a look at Piranhia or Gazest. I'm sure there are more of them. - Are there any plans for making it possible to use a different language than English? Yes, there are Translations.
- Does hg track all binary files? I heard that hg will eat up a lot of memory when deal with large files. (Not a good news for my low end server.) Yes, all Hatta pages – binaries and text – are treated the same and stored in the repository. You probably shouldn't use it to keep movie files and ISO images.
- How/Where can i ask questions and get help before creating bugs that are just misundertanding? The developer is available at the channel #wiki on irc.freenode.net, and can be also contacted via e-mail and jabber, see Radomir Dopieralski.