Just a quick reminder that on Wed, 12 Nov at 9am PST/17:00GMT is our Voice of the Customer webinar with Nate Nash and Jay Hariani of BearingPoint.
As founding members of the Enterprise 2.0 Capabilities Group, they have led large implementations of social media technologies for commercial and public organizations. Notably, they were responsible for implementing the BearingPoint Enterprise Wiki, powered by Confluence, for over 17,000 global consultants.
Currently, their wiki boasts 11,720 pages, 6,661 comments, 1,503 news (blog) items, with 62 originating countries!
* Event: ~45 min Webinar
* Date: Wednesday, November 12, 2008
* Time: 9am PST/17:00GMT
Updated: Video now available



5 Comment(s)
Confluence can now support increasingly large sites, thanks to the first beginnings of performance improvements implemented over the last few months. There's more work to do, however, in making Confluence a practical enterprise solution.
First, comprehensive information security is sorely lacking, with metadata leakage still evident as explicit page links betray confidental information despite underlying page content access restrictions. Viewing and editing comments cannot be restricted to designated groups, so private managerial or otherwise confidential discussions can't be conducted. While communities on the internet may think of themselves as 'democracies', corporations are not, and can not be, due to competitive, compliance, and legal realities.
While Confluence is well suited to "internet" collaboration, the realities of today's enterprise also demand support for secure, controlled information access. Engineering, Marketing, Legal, Finance, HR and other communities may choose to publish general content, however there are many occasions where information must be tightly controlled variously for legal, compliance, and competitive reasons.
The only practical way to publish such confidential content today is to implement each in a separate workgroup instance. Better design and control of the search engine would address many of these concerns, and reduce the excessive lifecycle costs associated with secure departmental instances.
Second, backup and restoration functionality does not meet what many of our clients would consider basic market requirements for architectural maturity and stability.
Backup and restoration should be abstracted, or filters for each additional version added, so that content can be seamlessly rolled forward regardless (within reason) of version. Future-proofing content investments is critical to enabling scaled enterprise deployments over the long term. Enabling higher performance and larger sites just elevates the importance of backup and restoration seamlessly supporting technology upgrades that occur the larger lifecycle context.
By Jim Barrick at November 12, 2008 3:46 PM
Are you guys going to post the video from this webinar?
By Slawomir Fryska at November 16, 2008 5:07 AM
@Jim
On Restrictions. We walk on a tightrope when designing the product; on one side we risk falling into being too restrictive and killing what makes a wiki successful, ease of use and openness, on the other we are too free and make Confluence useless in commercial environments. To take one of your examples, we have no plans to allow for restrictions on comments (of course we have very useful restrictions on the entire page). We think that it makes the product far more complex without offering a significant value to a broad enough range of our customers. For the needs we cannot satisfy, or don't want to, we think people will find other lightweight solutions (such as creating a child page with permissions so that managers can talk on the issue, or they can use email in this limited case).
As for controlled content, we at Atlassian hold a lot of our private data on the wiki, including HR, Finance etc. Many of our 6000+ enterprise customers (in defense, in government, in finance) do too. Our page and space restrictions whilst far simpler and easier to use than archaic, lumbering Enterprise Content Managements systems, is more than adequate for the task. There is no problem with search finding any of this confidential data.
On backup and restore. Your criticisms are valid here. All though we do a good job at small instances, larger instances can have trouble with backup. We have however helped many large (Fortune 1000) customers through these issues and have good documentation and support for backing up the DB itself, which is a preferable, safer method of backing up large instances. We are concentrating far more resources (at least half the team) to looking at architectural, performance, scalability, stability, and maintenance issues, the results of which can be seen in the recent releases. We of course need to do more in the future.
Please don't read my comments as negative or defensive ones. The points you raise are important and we think about them every day. It's just that our answers may not always be the one's you'd like. The other factor is that as with everyone we have limited resources and with the large customer base we have we cannot please everyone, all the time. We are making good progress however in supporting and making a product that works just as well for a large customer as a smaller one and I hope you continue to use Confluence so that you can see these improvements being delivered over the next year.
Confluence Product Manager
Atlassian
By Adnan Chowdhury at November 17, 2008 8:03 AM
@Slawomir Yes we are going to post a video of the webinar. We are currently working through the proper approval channels at BearingPoint and will publish the video to Atlassian TV as soon as we get that approval (hopefully this week).
By Bill Arconati at November 18, 2008 3:27 AM
You can catch the BearingPoint webinar here:
http://blogs.atlassian.com/news/2008/11/going_global_wi.html
By Morgan Friberg at November 18, 2008 4:20 AM