On Mon, 29 Sep 2003 11:34:41 -0400, Mario Cormier wrote: > Could a kind committer have a look at the problem and the patch, and > hopefully apply it to the source code? This one is a major issue for our > application, even under moderate load, and it has been sitting in Bugzilla > for a month and a half already. I couldn't submit a test case reproducing > this, as it is dependent on load and concurrency, but the stack trace > makes it obvious what the problem is. The patch is very straightforward, > so I'm very confident that it works. Of course, the fact that I've been > running it for all this time and never have seen the problem again, nor > any deadlocks due to the additional synchronization, makes me all the more > confident. > > I want to be able to use Xindice in production applications, but seeing > how patches to make it suitable for such use are left to (bit)rot makes me > wonder if this would be a good idea. In any case, it doesn't encourage me > to investigate more problems and fix them, and I'm probably not the only > one. Of course I could maintain my own version, as I've been forced to do > so far, but I have neither the time nor the energy. I need to be able to > work with the rest of you guys, or find another solution.
I have a system that supports many users. The biggest problem I have is the indexing and caching, I've never tun into you problem before :S. As a side note you may find Sysunit (sysunit.codehaus.org) a great tool for creating multithreaded tests. > Sorry for the rant, it's Monday morning so take it with a grain of salt. > I just want this project to work because it could be SO useful in SO many > situations. Thanks for listening. I don't take it personally and I'm sure most of the other developers don't either. I'll take a look at it tonight and try and come up with a unit test that exposes it (it might be a matter of creating a unit test at a lower level). -k.
