The Second Life™Market Place (MP) has been a source of pain and frustration for many. For myself I rate it as an annoyance to my Second Life experience. I continually hope they will get it fixed and working well. But, keeping track of what is going on with the MP is enough of a challenge I tend to ignore it and follow Darrius Gothly’s coverage over at dgp4sl.com.
Today I see he has a new article out: Wanton and Willful Fraud by Linden Lab. Darrius tends to rant. But, his rants are well explained and backed up by his experience.
He does say that the new Direct Delivery (DD) service of the MP is a good thing and generally works better than the Magic Boxes ever did.
Darrius’ problem with the MP now is in the area of payments. He explains what is being attempted and he and I see it as a good thing. His problem is the merchants are NOT getting paid. Customers get billed and amounts deducted from their accounts. The majority of items are delivered to the customer. The Linden sales commissions are deducted and credited to the Lab. But, the merchant gets nothing and does not even see that a sale is supposed to have occurred. This Darrius calls fraud. It certainly fits the legal definition.
Darrius notes that other web based market places work flawlessly… I would not expect any web sales site to be perfect. But, the percentage of transaction failures is typically super small. So, with some salt I’ll take the use of ‘flawlessly’ as reasonable here.
What has him going is; over the years when the MP fails to work correctly it is always a fail that takes money from the merchants. Darrius sees this as a pattern. In the four years I’ve been around I would say that is probably true. But, I don’t track all the problems so I would have a hard time proving it.
However Darrius is not relying on a pattern to make his accusation of deliberate fraud. His claim is that there is no way to program a transaction that completes for 2 steps of a three step transaction process. So, the Lab must have deliberately done something.
Well… computer databases that handle transactions, they are used all over the web, have transaction features. The idea of these features is that if a transaction fails, all parts of it are rolled back. So, if you buy something from Amazon and the server cannot complete the transaction, the charge to your credit card is back out automatically. The whole thing is unwound and the as far as the database is concerned the sale never took place. There is an error report and somewhere some programmers are suddenly looking to see what went wrong and make sure that does not happen again.
In a well built transaction system, Darrius is right. This type of failure would not be possible.
But, there is nothing to require a programmer to use those database transaction services/function built into the systems when they build their web app. I know lots of small site builders that skip the complexity and avoid setting up SQL transaction sequences. They simply throw a record at the data base. I skip that in simple it payment only sites or donation buttons.
So, I can’t go along with Darrius on the idea it is impossible to accidently program a sit that fails part way through a series of entries needed for a multi-part transaction. Therefore, it is not de facto proof of deliberate fraud. But, the pattern of an apparent trend in the type of failure looks bad.
Again that pattern of failures is incomplete for me. I don’t know how anyone from outside the Lab could ever see the entire pattern of failures without a court case and legal depositions of Lindens working on the project.
So, make your own call. Just remember we are operating on belief and opinion based on incomplete facts. There is evidence the Lab is trying to make it right (paying merchants for missed sales), which defeats the point of a deliberate fraud.