1) We’ve discontinued our unpopular outsourced telephone answering service. Too many people told me that they didn’t like talking to nimrods reading scripts who knew nothing about the actual TalkShoe service. So now when you call support at 724-935-8265, you may have to leave a message if you don’t reach us live, but you’ll know that if nobody else picks it up, I will be notified within minutes.

Keep in mind that if it’s not a system-down issue, you should use our People-Powered Customer Support at www.talkshoe.com/support (or click the “Help” link from any webpage).

2) We’ve equipped our support staff with a new audio streaming restoration capability to fix streaming issues on the fly. We plan to put this capability directly into hosts’ hands in an upcoming release. Separately, we’re working to improve our audio streaming system so that it will be much more reliable.

3) We’ve fixed a server deadlock problem that’s caused two significant outages in the last month. Unfortunately, we’ve subsequently discovered a new one, which has already caused one outage (Wednesday afternoon, 6/25/08). That one will be fixed in the next week.

4) By popular demand, we’ve restored the monthly reports that were disabled when we paused the TalkCash revenue sharing program. This allows you to view your stats on a monthly basis, even though earnings are zero.

5) Even though it’s a new feature from last release, don’t forget to tell your callers to use *8 (*T) to request to talk.

TalkShoe Community Calling is up from May to June to new record levels. Against a backdrop of a few hosts being unhappy with the pausing of our TalkCash program, this is very encouraging. I personally thank you.

1. We are moving all of the recordings to an eternal direct attached storage (DAS) unit. This unit has built in fault tolerance, is designed for heavy read/write access, and can be attached to any computer on our system. If the computer hosting the DAS unit fails, we can move it to another computer within minutes.
2. While we do backup all recorded content in triplicate, we’ve discovered that our restoration process is very slow. We are in the process of implementing a new backup procedure. One of the three copies will be copied to a “hot” back up unit. If the original unit fails for any reason and cannot be restored, we can move the hot backup unit into production within an hour.
3. We are changing the way we process our recorded content. Currently, we process all content serially. In other words, if 20 files are waiting to be processed, all 20 files must be processed before a single one gets posted to the internet. Early next week, we will have a new process in place that will process files in parallel. Smaller files will be processed very quickly without having to wait behind large files that take a long time to process.

With these changes, if we experience the same system failure, we will be back online again within minutes, with ALL content restored.