Saturday, April 7, 2012

the things we laid do not amount to much

From Josh Susanto:
In a way, it's better that this happened to me than that it didn't. I finally bought 3 listing enhancements, all for the borked item, which I am giving away for free in order to publicize my own belief about why this is really happening. Thanks, Lindens! Since you're willing to sacrifice the whole company to avoid firing one person, let's just get it over with, OK?
Damn. So I went and checked. This is his new item, renamed "BORKED ITEM 1455293". The description as it stood today is well worth quoting in full:
The recent Marketplace system failures are not a bug but deliberate sabotage as evident in the chronology of the 14xxxxx cluster of borked listings.

These listings correspond to a calendar period of items listed in Sept/Oct 2010, notably affecting the highest concentration of items with seasonal demand and high rates of paid promotion; paid promotion to be wasted if the items cannot be sold during the promoted period, due to their seasonality.

The affected seasonal items pertain to a peak season for user birthdays (around 16 September), Halloween, Thanksgiving, Christmas and New Year's.

The same calendar period of Sept/Oct begins at essentially the same time as would have begun the 13 Sept Direct Delivery code deployments, had they been allowed to continue beyond the attempted sneak-deployment on 13 September (the attempted sneak-deployment is not conjecture, it's fully documented)

That the period of borked 2010 items and the timing of attempted sneak-deployment in 2011 indisputably correspond in terms of statistical product types could mere coincidence, but I have compelling reason to believe otherwise...

1) System failures appearing to be magic box problems immediately following the deployment of mesh at the beginning of Aug 2011 were timed to disrupt a rush in sales of mesh items. LL maintain that the problem was not related to mesh code, but has been less clear about the timing being mere coincidence. That the magic boxes continued to work just fine when attached to an avatar standing on Linden land suggests no code problem with the magic boxes themselves. Further experimentation revealed that the problem was also not lag, as boxes also worked well in various heavily lagged sims, when worn as an attachment. The pattern of observable information was that sims where boxes were advertised as rezzed all seem to have been borked, specifically.

After considering what this incident has in common with what followed, my understanding is that an escalation of apparent box malfunction was manufactured in order to produce support for the Direct Delivery program; the next thing on the list after mesh deployment; thus the weird "coincidental" immediacy of apparent box malfunction after mesh deployment.

The idea that these malfunctions can be explained by an increase in sales after the mesh release is not supported by sales figures. In fact there was no increase in sales due to the mesh release; there was a decrease in sales due to the malfunction.

Moreover, a drop in sales can't be indirectly caused by an increase in sales unless the causative increase happens first.

And it just didn't happen.

2) The attempted sneak deployment, I've already mentioned was not only unannounced, but merchants had been deliberately led not to expect it (see Merchants Forum archives). The rest of the DD deployments were announced as not to follow as originally planned, but gift orders for the annual birthday peak on 16 September were nonetheless disrupted, and there was some continued impact on merchant utility of listing enhancements, on merchant confidence in listing enhancements, and on the market more generally.

Notably, the quarterly report for this period has been withheld by LL, which can only mean it's worse then the previous one, which is already REALLY BAD.

3) Subsequent code deployments also appear calculated to produce the maximum loss of merchant utility per merchant dollar invested, and by essentially the same means as the chronology targeting of the 14xxxxx borked cluster.

A) February 14th 2012 ("The Valentines Day Massacre") was same-day borked, as consistent with an effort to disrupt a preponderance of same-day gift-function orders, also in stealth-release mode, as on 13 September, more or less in a simple replay of that incident.

B) Easter was borked on a longer, almost perfect time frame by a "more complete" DD deployment on 21 March, but items from the 14xxxxx cluster have been reported as borked HOURS BEFORE THE DIRECT DELIVERY RELEASE.

Look at this again if you need to. It's a LOT of coincidence.

It's also a lot of very similar, repeated failure without anyone getting sh##canned for some mysterious reason.

So what's the real plan behind this?

My guess is that it's an attempt to drive commerce back to the grid by gradually destroying the practicability of other options, while also trying to seem to provide better options... by offering continuous "improvement efforts" which amount to little or nothing more than madatory downgrades and "bugs" which, when a transaction is involved, inevitably favor LL over merchants or their customers.

And WHY?

The grid is too big and land revenues are not high enough to allow the company to survive unless something changes, especially as demand for land continues to decline. Driving commerce back in-world may not work, but it's the kind of thing we should expect desperate people to try.
I don't know how much of this is accurate, I really don't. But it sounds utterly damning even if only half-right.

And bet me--we're not going to hear any Linden response to these accusations. I'm virtually guaranteed that's going to be what happens.

*hears crickets*

Yep. Not surprised.

No comments: