1 # A few bits about the RCS backends
5 ``web-edit'' means that a page is edited by using the web (CGI) interface
6 as opposed to using a editor and the RCS interface.
11 Subversion was the first RCS to be supported by ikiwiki.
13 ### How does it work internally?
17 RCS commits from the outside are installed into M.
19 There is a working copy of M (a checkout of M): W.
21 HTML is generated from W. rcs_update() will update from M to W.
23 CGI operates on W. rcs_commit() will commit from W to M.
25 For all the gory details of how ikiwiki handles this behind the scenes,
26 see [[commit-internals]].
28 You browse and web-edit the wiki on W.
31 ## [darcs](http://darcs.net/) (not yet included)
33 Support for using darcs as a backend is being worked on by [Thomas
34 Schwinge](mailto:tschwinge@gnu.org).
36 ### How will it work internally?
38 ``Master'' repository R1.
40 RCS commits from the outside are installed into R1.
42 HTML is generated from R1. HTML is automatically generated (by using a
43 ``post-hook'') each time a new change is installed into R1. It follows
44 that rcs_update() is not needed.
46 There is a working copy of R1: R2.
48 CGI operates on R2. rcs_commit() will push from R2 to R1.
50 You browse the wiki on R1 and web-edit it on R2. This means for example
51 that R2 needs to be updated from R1 if you are going to web-edit a page,
52 as the user otherwise might be irritated otherwise...
54 How do changes get from R1 to R2? Currently only internally in
55 rcs\_commit(). Is rcs\_prepedit() suitable?
57 It follows that the HTML rendering and the CGI handling can be completely
58 separated parts in ikiwiki.
60 What repository should [[RecentChanges]] and [[History]] work on? R1?
62 #### Rationale for doing it differently than in the Subversion case
64 darcs is a distributed RCS, which means that every checkout of a
65 repository is equal to the repository it was checked-out from. There is
68 R1 is nevertheless called the master repository. It's used for
69 collecting all the changes and publishing them: on the one hand via the
70 rendered HTML and on the other via the standard darcs RCS interface.
72 R2, the repository the CGI operates on, is just a checkout of R1 and
73 doesn't really differ from the other checkouts that people will branch
78 #### Another possible approach
80 Here's what I (tuomov) think, would be a “cleaner” approach:
82 1. Upon starting to edit, Ikiwiki gets a copy of the page, and `darcs changes --context`.
83 This context _and_ the present version of the page are stored in as the “version” of the
84 page in a hidden control of the HTML.
85 Thus the HTML includes all that is needed to generate a patch wrt. to the state of the
86 repository at the time the edit was started. This is of course all that darcs needs.
87 2. Once the user is done with editing, _Ikiwiki generates a patch bundle_ for darcs.
88 This should be easy with existing `Text::Diff` or somesuch modules, as the Web edits
89 only concern single files. The reason why the old version of the page is stored in
90 the HTML (possibly compressed) is that the diff can be generated.
91 3. Now this patch bundle is applied with `darcs apply`, or sent by email for moderation…
92 there are many possibilities.
94 This approach avoids some of the problems of concurrent edits that the previous one may have,
95 although there may be conflicts, which may or may not propagate to the displayed web page.
96 (Unfortunately there is not an option to `darcs apply` to generate some sort of ‘confliction resolution
97 bundle’.) Also, only one repository is needed, as it is never directly modified
100 This approach might be applicable to other distributed VCSs as well, although they're not as oriented
101 towards transmitting changes with standalone patch bundles (often by email) as darcs is.
105 Regarding the Git support, Recai says:
107 I have been testing it for the past few days and it seems satisfactory. I
108 haven't observed any race condition regarding the concurrent blog commits
109 and it handles merge conflicts gracefully as far as I can see.
111 As you may notice from the patch size, GIT support is not so trivial to
112 implement (for me, at least). Being a fairly fresh code base it has some
113 bugs. It also has some drawbacks (especially wrt merge which was the hard
114 part). GIT doesn't have a similar functionality like 'svn merge -rOLD:NEW
115 FILE' (please see the relevant comment in mergepast for more details), so I
116 had to invent an ugly hack just for the purpose.
118 ## [mercurial](http://www.selenic.com/mercurial/)
120 Being worked on by Emanuele Aina.
122 <http://techn.ocracy.org/ikiwiki>