Josh Susanto

>Being unprofessional is always a fail for a business. Well said. Even I would probably STFU about at least certain technical failures if they were simply handled professionally. But the reason that there are so many fails to handle unprofessionally must simply be that the decision process that leads to them is, itself, not a professional decision process.

>I did the survey. In that case, please note that I do, in fact, apologize for the comment rather than suddenly accusing you of being Malefactor Linden. I still think you're encouraging people to get ahead of themselves, and I still think it's a bad idea. But I do not find your motives to be suspect after all your previous contributions here. I AM SORRY.

Yeah. If a pyromaniac has been burning down 2 buildings a day and then suddenly cuts back to 1 building a day, I guess I should try to sound encouraging, sure. "I appreciate your effort." How sincere did that sound?

>You have attributed conditions to villainy that simply result from stupidity. Stupidity isn't consistent. Villainy is consistent. So far LL has consistently done the opposite of what they should to in order to keep people wanting to use the SLM rather than use in-world commerce.

>Clearly that is why, with direct delivery, there is no longer any need to have land at all to sell on SLM. /sarcasm That's why it's such a brilliant idea. "If" it works as implied "then" it reduces demand for land. But if it works "incorrectly", it increases demand for land. And they just spent the better part of a year assuring not only that it doesn't work correctly, but that it also produces an abundance of other deterrents to continued SLM use, even to users who have not migrated. Seriously, what am I supposed to think about this? >There had always been random issues with the magic boxes. one massively scripted dragon enters your sim where your magic boxes are and boom might have some failed deliveries. Yes. I get that. And it's a potential problem which I think was extra-well exploited starting right around the beginning of August (hmm). That the SLM tended to exacerbate the negative consequences more than Xstreet by doing things like producing double orders on SLM-active items not available in boxes also seems like an interesting coincidence, especially since LL chose not even to change the deactivate/delete instructions after I made such a strong point about how easy a fix that would be for most people trying to avoid the problem. The bottom line is that DD's delivery record isn't any better than the magic box delivery record was in its first 2 weeks of deployment. And the box issue only got worse the more that LL got involved with it, despite the fact that conditions can and could be easily created under which boxes function incredibly well. How well should I expect DD to work after we've given LL another year or two to continue breaking it? >Which is why we place copies in multiple sims. Not exactly. We place copies in multiple sims because LL didn't create specific location to rez them that would be free from borking; something they very easily could have done, considering the huge number of empty sims. The only thing to conclude is that they wanted the borking to happen, even if they (maybe) were not actually producing it in order to create the perception that there was some fundmental problem with box design. Note that they also provided a shopping cart that is only nominally compatible with the magic boxes, pulling yet another "box problem" essentially out of thin air. >The real advantage of DD is faster processing (when fixed) and less load on the grid in general. I will thank you not to abuse the present tense this way henceforth. There is no "is" in terms of DD representing a real advantage at this point. It is not only broken, but it has broken other things as well. Moreover, it was built broken and built to break other things.

>The only thing that seems to happen, or happened, as it is not happened in a couple of days, is a customer bought an item, got it delivered and then got the exact same item delivered again and was charged again in the same minute. >I had that happening quite a bit in the past year with magic box sales - i think it happens at the shopping cart level when things somehow get added twice and people don't notice. It's decreased since they put in that number warning. That was essentially the result I got with my "nonexistent object" test. If an item is unavailable in the magic box but remains active on the SLM (while, for example, the synch button is taking its sweet-ass time), the order gets doubled. Thus my emphatic suggestion to LL that instructions be changed to deactivate items BEFORE deleting them from the box; instructions which have been just as emphatically ignored by LL for over 6 months, because (wait for it...), the usual excuse: " DD is just around the corner anyway". Please let me point out that this is not a box problem, as it did not happen with Xstreet. It's just one a some unknown larger number of subtle downgrades and deterrents to which yo have been subjected since the shift to SLM; in this case, a downgrade and deterrent which has been wilfully continued by LL, as evident from the continued availability to any and all readers of the pertinent thread.

I suspect the whole survey is constructed with the intention of getting to you accept one question as already answered in order to answer another question. Like when the beef industry polled teenage girls on "what kind of beef do you like to eat with your friends".

Please let them fix the vital stuff first. That way, when they don't fix the vital stuff anyway, they can't say you distracted them.

I think your theory is also a valid theory, but it is not necessarily simpler, mainly because it fails to address the question of why so many "accidental" things inevitably add up to produce the same set of effects; errors which erroneously benefit LL's revenues but never erroneously benefit merchant revenues, and new forms of reduced functionality, just as some previous functionality problem seems to have cleared up or begun clearing up due to user behavior. >Boxes that worked in a smaller market did not scale well when applied to the much larger LL-sponsored market. The substantial scripting and server interactions that boxes required were not suitable for large-scale commerce and were subject to failure. The boxes continued to function just fine when rezzed in specific places. The problem was never a box problem, but a sim problem. LL could have chosen to work on the sim problem, but they instead chose to focus on the side of the equation that wasn't broken. Why? >Boxes did in fact have issues from the beginning, including the inability to deliver when the recipient was offline or messages were capped, or they were in busy mode. When scaled up to the millions of transactions that occur in today's marketplace, the servers increasingly dropped scripted transactions and deliveries had a decreasing success rate as the servers struggled to keep up with the load. Reports of non-deliveries increased and "stuck" transactions became common. These problems were neither continuous not universal. The problem was not on the the Xstreet end, not inside the boxes. Again, it was a problem inside SL which LL preferred not to address, even as they replaced (for apparently no other reason) Xstreet with a website of their own creation which only exacerbated the problem. And, again, why? >Unrelated to magic box delivery was the side issue that new users were confused about unpacking boxes, and one of the most frequent questions was why they were wearing a box on their hand. That has been an issue with sales to new users even from the very beginning of SL. And yet LL chose to completely ignore it until it suddenly provided an additional rationale to eliminate magic boxes. And, yet again, why? >All these issues could be solved in one move by implementing a direct server transfer, simplifying the process, eliminating scripted transactions entirely and thus decreasing server load, and as a side benefit, making purchases easier to understand for new users. Not "could"; "should". "Could" is the subjunctive of "can", and there's no "can" here. And now that it's past subjunctive, anyway, it's in the coulda-shoulda-woulda zone. Since there's essentially no end to how high or how low upward or downward counterfactuals can go, it seems to be a moot point that some other possibility arguably existed to what actually happened. They never proved that it worked in Alpha and they never proved that it worked in Beta. So why did they deploy it? That is, yet, yet again, why? >Tests went well enough in the alpha test group that LL called for a larger beta test group and opened up the process to anyone. That the test design was flawed is not only evident in the final outcome, but further supports my belief that the whole process was set up for failure. Before one applies a test, an important thing to do is to test the test, itself. If it can't provide a fail result, it's not much of a test. They can't possibly have done that. And, yet,yet,yet again, why? >A few serious people went to the beta grid to try it and give feedback. I take issue with your implicit charcaterization of the test group as especially serious in some way as compared to those who did not find the risks acceptable. >Unfortunately a common response here on the forums was "hell no, I'm not going to do their testing for them for free." It was fortunate for those who declined, because we didn't find ourselves trapped in a nondisclosure agreement, later worried we might get into trouble if we pointed out or described this week's problems in a way that might be construed as breaking the agreement. I am free to point out whatever there is to point out and describe it however it asks to be described. Our reasons for not participating were not necessarily all the same, and therefor, not entirely as you have characterized them. Some kind of pay might have tipped the balance for me, but probably not. My previous experiences had taught me not to trust LL, and I couldn't participate in good conscience under the suspended belief that they were just going through the motions of testing their new toy, trying to get me to act as an unwitting accessory in its approval for release. >Consequently the test group was not large or varied enough to catch all potential issues or edge cases. OR, the test process was not continued long enough to adjust for the small group fo Kool-Aid drinkers, who, understandably can't have been diverse enough after everyone already being **bleep**-slapped by LL on 13 September. The time to release new code is after it has been tested, not after some kind of an excuse not to test it more has been constructed. They had the opportunity to test it more if they wanted to. But they didn't. And yet,yet,yet,yet again, why? >In fact, I suspect very few people took the time to try it out. Time was far from the only issue. And in my own case, not even a major one. >Beta grid testing was a hassle and inconvenient, it took quite a lot of time, and required a special setup. Many folks haven't ever been on the beta grid and understandably didn't want to start. But, instead of blaming LL for doing nothing about that, you blame the users who didn't want to enter a nondisclosure agreement that would effectively silence them as critics, open their account to any possible amount of fudging, and devote some unstated and possibly open-ended amount of unpaid time to a company that had already repeatedly abused their trust. LL had more access to a more diverse group of merchants before 13 September, but rather than using them to better test DD, they scared people off at pretty much the worst time to scare people off as they made an unannouced code deployment after specifically telling us that they would do no such thing. And, yet,yet,yet,yet,yet again, why? >When the beta group was done and reported issues were resolved, the feature was considered stable and released to the public. Errors occurred. In fact, disastrous errors occurred. The market is now in turmoil. If such a broad and serious range of errors was set to occur, why was the feature nonetheless deemed to be stable? And why use amateur beta testers rather than hire people with pertinent experience if the cost of getting the wrong result is as high as they must have known it could be? Do they not have some kind of accountant or statistician working there at all? And yet,yet,yet,yet,yet,yet again why? >LL has to scramble now to fix it. No they don't. They need to appear to scramble to fix it. Surely, it is easier to fix code in 6 months than in 2 months. And yet they appear to have chosen to fix it in 2 months after not fixing it for 6 months. And yet,yet,yet,yet,yet,yet,yet again why? >I don't expect they're any happier about the problem than the merchants are. Some of them won't be, of course. But those will be the same people who were not part of the basic decision process in the first place. >The failure here is that the tests and QA were not sufficient, OR the success here is that tests were not sufficient. If the tests were not sufficient, the thing to do is not release the feature. Who could possibly have missed out on this basic principle in their education? ALL the Lindens? Is that even possible? Something like 100 people should have been able to point this out to someone in-house. Am I supposed to believe that no one mentioned it, or that, when mentioned, it was deliberately ingored? Which makes better mathematical sense? >and the test group was not varied enough to find all the potential issues. Thus making the test design, itself, a problem. But why hold back a new feature due to bad test design when you can just go ahead and release the feature, then blame users who didn't trust you enough for the fact that it was not ready when released? The only possible logic I can see behind this is that what appears to be a problem left intentionally unsolved is really the solution to some other problem. >There could also be some incompetence involved If it were just one or two things, I might agree. But consistently doing the opposite of what will provide a more functional off-world commerce tool is not a sign of incompetence; it is the sign of competence directed in some direction other than that stated or implied. >but I have no idea who's working on this or what happened; from what I've read in the JIRA the developers seem to know what they're doing. They have said that much all along. Were they right on 13 September? Were they right on 14 February? Were they right on 21 March? When have they been right yet? >In a system as complex as SL, I'd fully expect some glitches and last-minute scrambling. True. But sending us a green bear and telling us to ditch our magic boxes pronto is behavior totally inconsistent with a team that understands the normal process even as well as you or I understand it. Again, it's the opposite of what logic or reason should have led to them to do in order to maintain user confidence in the SLM. Given their massive collective education and experience, how can consistently doing the opposite of the right thing possibly be some kind of unanimous, collective Linden accident? >Unfortunately this one was big and not easily repaired. People got hurt and are justifiably angry. Yes. Mission accomplished. My point, exactly. >This makes a lot more sense to me than a group of cackling businessmen in a back room plotting to make your SL life miserable, and breaking things on purpose on the theory that you'll decide to invest more if they hose the system. That's not what I envision at all. I envision some people looking at the too-large grid not being rented enough to generate the projected land revenues that are keeping them in a job at this unfortunate economic time for them to be unemployed. I envision someone realizing that land rental is in decline due to Xstreet pulling merchants off-world. I envision someone realizing that if they could find a legal way to shut down Xstreet, that might send merchant back in-world. I envision someone realizing that borking sims that are advertised as having boxes on them is step 1... I see means, motive and opportunity. You see a bunch of people repeatedly, accidentally doing the opposite of what they obviously should do. Basic statistical principles do not favor your vision. They favor mine.

Noting the time stamp, I'll see your 10L. 10/10, yes? OK, then... how do we know whether she showed up?

>the only action is NO ACTION Obviously, yes. And yet, every time we open the website, the first thing we see if a green bear urging us to migrate. How hard would it be for them to take that off the screen? The message from LL is to keep migrating. Why?

Even I would think that to be just too sick to be funny.

>How much worse can somebody screw up and still get paid? Careful there. That sounds like a challenge.

>you only lose ANS, if you migrate to direct delivery. Yet another reason why it's urgent to get you away from your boxes. ANS makes it too easy to notice that you didn't get paid.

Also glad to see you're hanging in there and coping well.

Nope. I'm doing just great. But, then again, I've been acting totally paranoid for 8 months, right?

Yes, thanks. That does put a finer point on the matter.

>does not show up in People Search. And I suppose I'm also expected to consider this to be one more "mere coincidence"?

We're all rooting for you.

>Maybe with the next new feature rollout? I expect the next truly new feature rollout will be something that is intended to facilitate in-world commerce. And I expect it will work flawlessly from day one.

>BUT SO far I've not found any way to DELTE inactive objects. Once they are in your marketplace inventory they are stuck there forever, or I'm missing SOMETHING??? You are not missing something. I have come to suspect that this design aspect of the SLM is intended to assure that SLM accounts gradually get congested with nonfunctional data as a subtle deterrent to continued use.

I don't have all the variables. I CAN tell you that the boxes still work and that other people have remigrated successfully. If you're not certain that you want to remigrate, you might at least consider cutting off the sales until something either changes or is made clear enough to you to make a better decision. If you shut the sales off, you lose the sales while they're shut off. If you continue the full perms sales, you risk losing sales of those sales to people who might be getting them cheaper, or for free, instead. As I've already explained in another thread, the permissions leaks might ultimately be good or bad for your total store revenues. But there is a principle involved here, as well as the simple question of whether the numerical result is good or bad. The principle is that if you didn't agree to let LL leak your permissions, you're nonetheless being complicit in the decision they made which allowed the leakage as long as you don't take action, yourself, to plug the leak.

>Okay, this is it -- people need to start losing their so-called jobs. Wow. Did I just hear an echo that's been bouncing around the recesses of cyberspace for about 8 months?

>Overly sensationalist Josh. Is it over sensationalistic to point out that, where Xstreet used to directly provide us with metrics, LL has actually set the data up so that getting it onto an excel spreadsheet is harder than would seem to be consistent with mere design negligence?

Understood. My point was that even if were 99 steps, you would be doing everyone a huge favor by posting them all.