Can we please get email support a place in the roadmap please? -- Antony Stubbs 25-Jun-2008 19:48 PDT
Under 0.1.0, the roadmap discusses reverting to prior version from the history tab. Was that feature released with 0.1.0? -- scroco 08-Aug-2006 13:20 PDT
Thought I'd move this discussion to the comments section ...
Since the topic came up, I took a closer look DB persistency, and in PersistencyHandler.writeTopic(...), lies this code (0.3.4 beta6):
if (topic.getTopicId() <= 0) {
addTopic(topic, params);
} else {
updateTopic(topic, params);
}
if (userVisible) {
if (topicVersion.getPreviousTopicVersionId() == null) {
TopicVersion tmp = lookupLastTopicVersion(topic.getVirtualWiki(), topic.getName(), params);
if (tmp != null) topicVersion.setPreviousTopicVersionId(new Integer(tmp.getTopicVersionId()));
}
topicVersion.setTopicId(topic.getTopicId());
if (Environment.getBooleanValue(Environment.PROP_TOPIC_VERSIONING_ON)) {
// write version
addTopicVersion(topic.getVirtualWiki(), topic.getName(), topicVersion, params);
}
String authorName = topicVersion.getAuthorIpAddress();
Integer authorId = topicVersion.getAuthorId();
if (authorId != null) {
WikiUser user = lookupWikiUser(topicVersion.getAuthorId().intValue(), params);
authorName = user.getDisplayName();
}
RecentChange change = new RecentChange(topic, topicVersion, authorName);
addRecentChange(change, params);
}
A successful addition of a new topic would call addTopic(), addTopicVersion(), and addRecentChange(). But those activities are not contained in a single transaction. If addTopic is successful but addTopicVersion fails, the work done in addTopic will not be rolled back. -- scroco 29-Sep-2006 10:26 PDT
Right. Now I see it. I missed the fact that writeTopic is overloaded. The writeTopic(...) I was looking at is always called by writeTopic(Topic topic, TopicVersion topicVersion, ParserOutput parserOutput) which handles the transaction properly. Thanks for the clarification. -- scroco 29-Sep-2006 11:09 PDT
Instead of getProperty() the method must be renamed to getValue(); otherwise setup.jsp won't compile.
... <option value="<%= DatabaseHandler.DB_TYPE_MSSQL %>"<%= Environment.getValue(Environment.PROP_DB_TYPE). ...
Archived from the Feedback page:
Due to my database needs, I want to set up the database and JAMwiki settings by hand, as opposed to running through the automated setup. Can you make the SQL schema available for download? Or even better, make the source available for download other than through subversion (I don't run subversion and have no desire to do so). -- scroco 27-Jul-2006 12:50 PDT
Thanks for posting the source. My situation involves the way we have set up our database security, we have an admin user that owns tables (e.g. "admin_user"), and various other users which access the tables, each with different access rights. So I don't want to create tables using the same user that will be accessing them. And further, when querying them, I'll be prefacing the table names with the owner name (e.g. admin_user.wiki_table). I'll probably need to modify the source to accomplish that. Do you have any plans to externalize all the SQL into a properties perhaps? Or do you plan to support EJB architectures? -- scroco 27-Jul-2006 14:10 PDT
This page looks much better after the cleanup, and the removal of the Changelog items. Thanks! -- Ryan 25-Jul-2007 21:46 PDT
Archived from the Feedback page:
In the road map it is mentioned that you keep an eye on jsr170.
The first time I saw this specification I thought it would be ideal for a wiki. It has import/export, versioning, access control, xpath queries etc... All you need for the data-side of a wiki. Any change of being the first wiki to have this jsr as a base? -- [Kees]
Not sure whether this should be in the list for 0.6.x since it wasn't approved, but it seems urgent: captcha needs to be an option in the permissions system since "require captcha to edit" is a permission level
Credits for authoring also need not to be hardwired but instead to be part of BottomArea and easily removable or customizable (more or fewer names, etc.); You don't want silly pseudonyms to have to be displayed, or lots of IP numbers; It's also not correct to call all non-real-names "anonymous" since often someone responsible knows who they are and this might be legally thorny. If one associates all bald IPs with the tag "anonymous", fine, but it is just as important to be able to put another tag on specific ranges or lists of IPs.
Also, block IP and ban user should not be confused. One implication of this is that users should be able to claim specific IP numbers as themselves post-facto, and that a much softer, easy to challenge and undo, way of associating IP numbers with an identity or "user" might be required (though it's probably better to associate them with a faction that might only have one person in it).
For some reason edits to this page are especially likely to result in this error:
"An unknown system error has occurred. The error message is: java.lang.Exception: Failure while executing insert into jam_topic_version ( topic_version_id, topic_id, edit_comment, version_content, wiki_user_id, edit_type, wiki_user_ip_address, edit_date, previous_topic_version_id ) values ( ?, ?, ?, ?, ?, ?, ?, ?, ? )."
It's been seen at least a dozen times on this page alone.
Not just yet! You write that you want to make jamwiki a preferred solution among commercial users. The most important features to reach this goal are, based on my experience, LDAP-integration, user-rights based on LDAP-groups and a wysiwyg editor. I would like to see theses three points in your roadmap for upcoming releases. I work for a large german car manufacturer and we are searching for a new wiki-solution.