Welcome to Community Server Sign in | Join | Help

Feedback Halftime

Well, it needs some major refactoring, but the background queue for processing outbound autodiscovery, trackback and ping back looks like it's working against some of the known examples. I replaced a lot of System.Diagnostics calls with log4net earlier, so that should make identifying exception cases easier.

Autodiscovery and the outbound pings themselves are fairly latent, I didn't want them blocking post completion, the background approach appears to work okay and allows for failover retrys. I really just try to go back to the same place X times at Y delay, so it's not that smart as far as failover goes really.

Since inbound trackback and pingback won't really gain much advantage for queuing (and doing so might delay a blocking autodiscovery process on other folks' clients), they should be fairly simple. Although I say this and then remember I want to do some of the auto-excerpting for pingback that Sam's host has and some decent back-validating. So who knows.

It strikes me a little funny that I've gotten as much functionality as I have in there and I still don't have a serious management interface beyond the xml-rpc apis. Oh well, once I modified things to support w.bloggar in MT mode, it really hasn't popped up. So that's probably next up after inbound feedback and then back to the ASP.NET client app itself. Good fun.

Published Wednesday, April 23, 2003 11:28 AM by grant
Filed Under:

Comments

No Comments
Anonymous comments are disabled