A little more than two months ago, I wrote about Google Reader’s awful ability to ‘lose’ older, unread items from my queue. Since then, I have made it a daily effort to trim at least one day of items from the older side of the heap each day, so that this loss would not happen to me again.
At the start of July, I went away for a week of vacation. Our destination had no internet (apart from 19K AOL dial-up – the horror). Before we left, I made a concerted effort to trim enough old items from my heap so that they would not be subject to this issue. When I came back, I was pleased to see that everything old was still there, with abut 3,000 new items stacked on top. I’ve spent my web time for the last ten days working those 3,200 unread items down, each day carefully trimming the oldest stuff to ensure no surprises there. Last night, I’d winnowed the pile down to 1,032 items, the oldest being from July 3. That’s only two weeks ago – well under the 3-4 week threshold I’ve seen in the past.
This morning, Google Reader reports an unread queue of about 400 items, the oldest from July 8. So five days worth of about 600 posts – posts which I’d thinned out from the chaff so that they were actually worth reading – are all gone! Poof!
My wife says to look at it as a blessing, and I see her point, but data loss is a cardinal sin in computer systems, and that’s what this is. It’s unannounced, arbitrary, and variable.