Prok, by your definitions I’m a FIC techie-wikki. You’re free to ignore everything under this point. You won’t, but you are more than welcome to.

The problems with your post, however, should be addressed.

I suspect SL is a complex enough system that it shows some emergent properties. That means laying the blame at a specific change may not be possible, it could, literally, be impossible to predict the outcome of a given change. However, that doesn’t mean all it’s parts are impossible to see and correct and predict.

This update saw several significant changes to the client, even if they’re not all significant to you:
The map drawing was changed so hidden islands are no longer hidden (apparently driven by an internal pressure for scalability).
Mapping friends (which you characterise as “Flipper’s change”) was also changed.
A whole host of new lsl functions about parcels were introduced (I’m not aware that Flip asked for this, although I do know there was a feature vote on it).
A new tab for web displays was added to the profile (this was a “fun project” of Kelly’s according to his posts on it).

Also, almost certainly inadvertently, as part of that change, classifieds were stuffed.

Classifieds, as I’m sure you know, but I will spell out anyway, are created in profiles, affect the map drawing and affect parcels. Any of the four big changes above *could* have be responsible, as could a combination of them. Given it took a week to fix, and I do have a little faith that they had people working hard on it for that period, I’d suspect it was a combination of two or more of those things. The fact that search was also hosed, which is a parcel level thing, makes me suspect that at least SOME of the blame is in the coding of the new lsl functions, since they’re the only thing working at a parcel level.

Saying that the developers should have tested to see if such a change occured is entirely justified. It is, however, also _possible_ that the load on the preview grid never got high enough to affect it: many if not all sql queries have a time out period in their host software. It is possible the reason the searches failed is because they were timing out when there were 10k+ people around on a regular basis, looking at maps, searching classifieds etc. in loads far greater than the preview grid ever gets.

Your comments after “Why make it personal about Flipper?” doesn’t add up. Or, rather, they add up to a personal attack on him rather than any reasonable commentary on the release. You are proud of your holistic view, your lack of attachment to concepts such as cause and effect and apparently of your lack of technical understanding. There are definitely times when the first two can be useful and you should be proud of them. The last most probably reflects choices you’ve made in life about education, there are definitely things in life that can be useful you know more about than me. Sadly, when a little thought and technical know-how shows there are simpler explanations perhaps it’s time to think before you dive in splashing the blame around.

Flip doesn’t implement features you twit, Linden Labs does. See heres how the chain works for the Dim.

1. Residents want features.
2. LL Decides hey thats a good idea.
3. LL codes in the feature.
4. Feature fails and fucks everything up.
5. LL fails to rollback during the early phases of failure.
6. LL can no longer rollback due to the ammount of time and integration of the new patch.

At no point did Flip fail you or the community. The blame for this problem resides in its entirity with Linden Labs. As a systems administrator managing quite frankly a server farm amost certainly twice as large as LL’s I find their disaster recovery plans to be downright gross and negligent.

Never EVER do you put a patch in place without a secondary rollback procedure in place that can quickly throw you back to a previous version.

Do note for a server farm the size of the grid a day (24 hours) would be an acceptable downtime for a catastrophic event, which this patch HAS been.

I again point to this absolutely retarded Tao of Linden happy dappy dumbshit way of running things. Phil claims to have almost NO turnover at his office. That tells me where his problem is right there.

Tech sector jobs, even good ones have an average turnover “job life” of two to two and a half years. If a place is so awesome to work that they aren’t seeing any turnover ODDS ARE NO ONE IS BEING HELD ACCOUNTABLE FOR FAILURES.

When the grid goes down, and stays down for 6+ hours, I’d be hiring someone new and firing someone old. If some patch I applied destroyed some major functionality of my windows or linux network and I had no ability to roll back, ODDS ARE I would be looking for a new job. Tommorow.

Quit blaming the wrong people by throwing a wall of bullshit out. People are tired of you pushing your little agendas.

What I’d like FlipperPA to do is to analyze your lost sales, my lost sales, my tenants’ lost sales, with this patch, which contains a feature he lobbied so long and hard for (being able to hide his online status from his very own friends!) and weigh whether all the lost sales have been worth it.

Because that’s what it’s about. It’s not about a causal relationship — of course tekkie wikinistas can only see it in that light. It’s about the drive for some people who only care about the platform and its features to push through coding changes no matter what, even at the expense of the world that has grown up in SL. There are Platformers and there are Worlders. Flipper and all his little friends like Aimee and the gang at SC are Platformers.

Why make it personal about Flipper? Well, because Flipper purports to be a Leader of the Community and Organizer and EmCee of SLCC. That’s why. He imagines *he* is the community. OK. Then justify why new features, resume-polishing exercises for Lindens and Lindesidents,has to take priority.

Why 1.13 ever got off the boards is beyond me. Why it would be rolled out after ostensible testing when the results of rolling it out would be to lose 80 percent of sales for many people, is beyond me.

It comes from the sheer cynical indifference of Linden coders to peoples’ livlihoods, which in no way, shape or form do they tie to *their own*.

They have now grown cocky and confident enough that they can say, well there’s always another guy to buy not one island, but 50, *and* get it reported on the front page of magazines and newspapers, so who cares about these server load-tests who make micropayments and want to play store.

I guess we all have to conclude that playing store with these nasty kids is no longer fun. Either we start to sell to all their big new friends, if we have something they want, or we fold our tents.

Within moments of this being posted there’s an annoucement to the effect of what I said appearing on the Linden Blog. I wish I could claim insider knowledge, perhaps the influence of the Herald is that great!