I don't know much about Cenqua or their products, but BileBlog's entry on their purchase by Atlassian (WARNING: explicit language, NSFW) bashes Atlassian, which I do have recent experience with. In a nutshell, his statement:
[Atlassian is a] company whose informal motto is ‘make it pretty, sell it, then maybe ponder making it work if some turds pay enough for it’.
is not at all our experience. In fact, I've exchanged email with the president of Atlassian and offered to pay to have some refactoring of the user schema completed sooner. Rather than bilking us for cash- as BileBlog suggests would be their character- their director of engineering has been in touch with our engineering team and all agreed it was an important change that they will take on themselves in the (tentative) 2.7 release. This has also been true for many other bugs and RFEs we've reported and/or voted on, which have already been fixed in the 2.5.5 and 2.5.6 releases- all of which were functionality related, not eye candy.
Of course YMMV, but thought I should share our positive experience with Atlassian, and am looking forward to what will come of the Cenqua acquisition.
Posted by Mikael Gueck on August 06, 2007 at 03:18 PM PDT #
Thanks, Rama, for the post! We've grown really fast as a company and have doubled our customer count year over year; with all those customers there are bound to be negative experiences. On the whole, it seems like we have a lot more supporters (of course, we'd love for everyone to have an awesome experience with our products). I'm happy to respond to constructive criticism, but BileBlog is so far outfield that we ignore it.
Posted by Jon Silvers on August 06, 2007 at 05:07 PM PDT #
Mikael: We made a reasonably bad, or at least an expedient decision early on in Confluence's development with regards to user management (*cough*OSUser*cough*). There's been a work-around for the problem you describe for a year or so now, it just takes a bit of manual configuration tweaking.
We've been holding off on making that configuration the default for backwards compatibility reasons—most people _don't_ have 50,000 users in Confluence, and would be more disrupted by upgrading to a new architecture, whereas most people who do have 50,000 users have them in LDAP and will already be using the new code already.
As for the rest, hyperbole is the BileBlog's stock in trade. Every company, product, person or codebase has its flaws, perceived or otherwise, and Hani's pretty good at poking sticks into people to see how they react. Now we're not exactly the plucky little guy any more, we should expect to take the lumps that go with it. :)
Posted by Charles Miller on August 07, 2007 at 01:44 AM PDT #
Charles: I'm only critical of the quality because I really like your products and would like to keep using them (as well as buying Clover and Crucible now that you made the prices right.)
I'm genuinely interested in Rama's opinion on general Confluence source quality, because I really haven't had the opportunity to discuss the subject with others who have read through it in the context of a major deployment, and would like to reflect on whether I'm being overly critical or if there's something I've missed.
Posted by Mikael Gueck on August 07, 2007 at 11:18 AM PDT #
Mikael- I did most of my Confluence evaluation prior to having access to the source code- Igor (paging Mr. Minar!) would be able to more accurately comment as he did spend time in the source.
Posted by rama on August 07, 2007 at 01:43 PM PDT #
Message received :)
Here is my opinion:
http://net3x.blogspot.com/2007/08/is-atlassian-and-confluence-code-base.html
Posted by Igor on August 08, 2007 at 12:25 PM PDT #