User Experience Interest Group/Transcripts/2009-07-02

Topic & Summary

User Experience Interest Group Discussion for July 02, 2009.

Topic: Gathering UI Metrics.

No summary is yet available for this meeting. Please edit this page to add one.

Transcript

???

[15:13]	Jacek Antonelli:	Topic today is viewer metrics, i.e. collecting data about how people use the viewer. What buttons they use most, how much they use shortcuts vs menu items, other stuff like that
[15:14]	Jacek Antonelli:	Generally collected on a wide scale, and not linked to any individual person. So you can get an impression of how the whole userbase is using the viewer
[15:14]	Charlette Proto:	ah makes me think if the UI interactions can use any logs
[15:14]	Azwaldo Villota:	does LL collect that sort of data?
[15:14]	Jacek Antonelli:	I don't think they do, Azwaldo. But, let's ask Malbers :D
[15:15]	Charlette Proto:	that is what I was thinking of, rather than hanging out in welcome areas like I do
[15:15]	Malbers Linden:	well ,we have all sorts of data
[15:15]	Malbers Linden:	anything that touches the server/region we can track
[15:15]	Charlette Proto:	but do you mine any of it Malbers
[15:15]	Malbers Linden:	the important word there being *can*
[15:15]	Malbers Linden:	we don't keep it all
[15:15]	Mm Alder:	But the UI doesn't touch the server
[15:15]	Jacek Antonelli nods
[15:16]	Malbers Linden:	Mm has it correct
[15:16]	Malbers Linden:	there's a lot of other metrics we are not set up to track
[15:16]	Charlette Proto:	so you wouldn't know if the shortcut keys or menu were used?
[15:16]	Jacek Antonelli:	Right. The viewer doesn't log UI events, and there's nothing (that I'm aware of) in the viewer to send any such data anywhere
[15:16]	Malbers Linden:	the question is which things in the UI would be the most obvious, interesting things to track
[15:16]	Jacek Antonelli:	Quite so, Malbers
[15:16]	Armin Weatherwax:	produces a lot of data, difficult to track ?
[15:17]	Malbers Linden:	difficult to track - No
[15:17]	Malbers Linden:	difficult to mine effectively and store -yeah
[15:17]	Jacek Antonelli:	Let's focus on that for a while -- what sort of data would be useful enough to be worth gathering?
[15:17]	Armin Weatherwax:	ok, mine was what i ment by "track"
[15:17]	Charlette Proto:	what is there was a graph of accessed elements that was logged rather that real time data
[15:17]	Mm Alder:	I think time on task would be one thing to look at, but "task" is not always clear.
[15:18]	Armin Weatherwax:	hi morgaine
[15:18]	Charlette Proto:	hi Morgaine
[15:18]	Malbers Linden:	Hey Morgaine
[15:18]	Azwaldo Villota:	hello Morgaine
[15:18]	Morgaine Dinova:	Hi peeps. OS meeting still going strong, hehe
[15:18]	Jacek Antonelli:	Microsoft did some extensive UI metrics gathering with Office. They recorded how many times every single button, menu, shortcut, or even keyboard key was pressed.
[15:18]	Malbers Linden:	Yes, that would be ideal
[15:18]	Morgaine Dinova:	Ends later every week :-(
[15:18]	Charlette Proto:	precisely what I'm thinking of jacek
[15:18]	Malbers Linden:	we could throw away data that wasn't useful
[15:19]	Malbers Linden:	but there's cost in collecting it
[15:19]	Jacek Antonelli:	So they were able to definitively answer the question: How many people use paste from the shortcut, the menu, or the button. :D
[15:19]	Jacek Antonelli:	(And of course, other questions too...)
[15:19]	Malbers Linden:	so, it is interesting to think about what data would be most valuable to collect (to justify the up-front cost)
[15:19]	Charlette Proto:	what about storing it locally as a graph and uploading on quit
[15:19]	Armin Weatherwax:	my most important point about metrics is that we need to think about whats worth tracking or we end up having gazilliobytes of worthless data
[15:19]	Jacek Antonelli:	Exactly, Armin and Malbers
[15:20]	McCabe Maxsted:	shortcuts definitely need to be counted, if they're ever going to be trimmed down
[15:20]	Jacek Antonelli:	Not everything is useful. A lot of it would just be wasted storage
[15:20]	Malbers Linden:	Jacek made a point early on about the metrics not being tied to an individual
[15:20]	Malbers Linden:	and that is true -- to a point
[15:20]	Mm Alder:	The problem with shortcuts is that they're not configurable.
[15:20]	Jacek Antonelli:	I'd agree that shortcuts would be worthwhile tracking. Pretty much every single UI element -- menu, button, and shortcut
[15:20]	Malbers Linden:	it is VERY interesting to look at differences between 1 hour old residents and 2 year old residents
[15:21]	Jacek Antonelli:	True, Malbers. That's a good point
[15:21]	Charlette Proto:	well anonymous per user data would be the minimun to getting usefull sets
[15:21]	Jacek Antonelli:	Let's hold off on that for a little while, though. We'll talk about usefulness of data vs. user privacy next
[15:21]	Charlette Proto:	the xeperience of the user would be quickly reviealed when the data collections grow
[15:22]	Jacek Antonelli:	Since there are definitely privacy implications of recording every UI interaction the user makes
[15:22]	Jacek Antonelli:	What other sorts of data would be useful to gather, besides how many times shortcuts, menus, etc. are used?
[15:22]	Charlette Proto:	alse one could relate it to age or successful logons etc
[15:22]	Armin Weatherwax:	yeah, malbers ... good point to have variables which are not only ui specific
[15:22]	Armin Weatherwax:	lice psychometrics about personality e.g.
[15:22]	Charlette Proto:	hehe
[15:22]	Malbers Linden:	there are other interesting axes like if people own/rent land, how much L$ they spend
[15:22]	Azwaldo Villota:	such data could inform the design of new user experiences, with tutorials/demos weighted toward most frequently used UI elements
[15:23]	Charlette Proto:	mine would represent a ton
[15:23]	Morgaine Dinova:	While it's interesting, and even important so that beginners can be supported well, some people then make the leap that support for experts should be reduced. While the former is true, the latter is not.
[15:23]	Charlette Proto:	Azwaldo a good point
[15:24]	Jacek Antonelli:	Azwaldo: Exactly. The data would be very useful to have. It could really spur some serious improvement, and give some scientific certainty to any UI changes that are made
[15:24]	Mm Alder:	Azwaldo, the most frequently used elements are probably the most frequently used because they're the best understood.
[15:24]	Charlette Proto:	thso making the greatest leaps are surely alts
[15:24]	Malbers Linden:	Jumping waaaaaayy to far ahead, collecting metrics on the UI could be used to morph the UI in near-realtime to support that particular user
[15:24]	Morgaine Dinova:	Haha
[15:25]	Morgaine Dinova:	Malbers: you mean anyone detected as a newbie should automatically have a box rez on their heads? ;-))))))
[15:25]	Armin Weatherwax:	hm. or make out major groups of user types
[15:25]	Charlette Proto:	also great resource for improving the 'visibility' of functions used by everyone and all the time
[15:25]	Mm Alder:	Malbers, a manual morph would be nice, first.
[15:25]	Jacek Antonelli:	Malbers: It could! It's interesting to think about a UI that adapts to its user's needs. It would be hard to avoid Clippy Syndrome, tho.
[15:25]	Charlette Proto:	hehe Morg
[15:25]	Malbers Linden:	YES, that would be perfect -- the newbie box on head button
[15:26]	Morgaine Dinova:	Mm++
[15:26]	Malbers Linden:	let's not go too far down that rathole though; it's crazy difficult and goes toward Clippy (as mentioned)
[15:26]	Jacek Antonelli makes a note to add "Wear on Head" to the inventory menu -- and switch it with the normal Wear command's position
[15:26]	Jacek Antonelli cackles
[15:26]	Armin Weatherwax:	i attach boxes to the head not earlier than age of 6month
[15:27]	Jacek Antonelli:	Sadly, the default wear position is the hand now. Well, for the past year or more. Accidental box-on-head is a thing of the past, mostly. :( But anyway, we're getting off track
[15:27]	Malbers Linden:	actually, one of hte good uses of metrics is to see which functions get used together most often -- to place them in the UI "closer" together
[15:27]	Charlette Proto:	surely there wouldn't be so much data even if it recorded the counts for all elelments of the interface and uploaded at the session quit
[15:27]	Mm Alder:	Do you think metrics would tell you anything that comments on the forums do not already.
[15:27]	Jacek Antonelli:	Indeed, Malbers. One of the things Microsoft was able to track, was how often command A came after command B. It would give some hints about workflow
[15:28]	Charlette Proto:	precisely Malbers but that wiuld require timestamps on UI events stored
[15:28]	Malbers Linden:	Mm, for different classes of users, Yes. NOt everyone uses the forums to make comments
[15:28]	ATechwolf Foxclaw:	Some elements get clicked once and it stays on teh screen forever, like mini-map, stats, etc.
[15:28]	Jacek Antonelli:	And also hints about which things are too easy to accidently trigger! E.g. if 90% of the time, Command A is followed by "Undo! Undo! Oh God Undooooo!", there's a problem
[15:28]	Armin Weatherwax:	i still believe in experimental design ... have distinct groups and compare
[15:29]	Jacek Antonelli:	Mm: A cool thing about metrics, is that it can reveal things that the users don't even consciously know
[15:29]	Charlette Proto:	and the MiniMap is much more usefull in Snowglobe for sure so it may stay like ATech said
[15:29]	Mm Alder:	I still believe in having the user do his own design. Configurability! :-)
[15:29]	Jacek Antonelli:	Things they aren't even aware they're doing, so they'd never be able to report them on the forums or surveys
[15:29]	Malbers Linden:	Armin, there is a place for that but we also have access to (inherently skewed) data as well
[15:29]	Malbers Linden:	we should look at what it has to say and keep it's skewedness in mind
[15:29]	Mm Alder:	Jacek, like what?
[15:30]	Charlette Proto:	agrees with Jacek
[15:30]	ATechwolf Foxclaw:	And measure how long a blue dialog popup and the button pressed. I bet you find a lot of miniseconds where one is going to click the blue and one pops up over it and it too late to stop the click action and something on the new blue dialog gets clicked and it disapears so fast one can't see what they just click too...
[15:30]	Armin Weatherwax:	experimental design was ment like how science does experiments, have experimental group a and b (for example)
[15:30]	Jacek Antonelli:	ATechwolf: Good point. So maybe something to measure would be how many times each window is open, and also keep a running total of how long it is open?
[15:30]	Charlette Proto:	lots of things we just do and don't pay ant attention too (automatically so to speak)
[15:31]	Morgaine Dinova:	In large part, the "box on head" thing is not a newbie problem at all. It's an SL design problem, in that creators are not given a "Wearable" bit to set in objects, and so all prim objects look the same in inventory, just a box. If there were a "Wearable" bit to set, then boxes could not accidentally be worn accidentally, and the inventory would show a different icon for werarables and for containers. That would be effective ergonomics both for beginners and experts.
[15:31]	Charlette Proto:	sounde like we do need some timestamps
[15:31]	Malbers Linden:	VERY good point Morgain about "wearable bit
[15:31]	ATechwolf Foxclaw:	Once you get a list of what to measure, someone should open a jira task. This would be a good one to try on snowglobe.
[15:32]	Malbers Linden:	Snowglobe could be a good test bed -- but it will give an even more skewed sample than SL
[15:32]	Charlette Proto:	well we all know the Inventory doesn't cut it when it comes to typology of things ATech
[15:32]	Malbers Linden:	i mena, than the LL production viewer
[15:32]	Mm Alder:	In most scientific experiments, you start with a hypothesis, and then do measurements. Do we have any hypotheses that we want to explore?
[15:32]	ATechwolf Foxclaw:	Snowglobe to test the matrax mesureing before pushing it out on the main client to get usefull data.
[15:32]	Jacek Antonelli:	Yeah. Ideally it'd be in the main SL viewer, to get the widest, most representative sampling
[15:33]	Armin Weatherwax:	i think so Mm
[15:33]	Jacek Antonelli:	But yeah, Snowglobe would be good to check that everything's working first
[15:33]	Mm Alder:	Armin, could you give an example?
[15:33]	Charlette Proto:	hypothesis assumes to much and would skew the results, I'd avoid this
[15:34]	Armin Weatherwax:	look at the pie menus, mm
[15:34]	Charlette Proto:	data mining can follow data without any hypothesis
[15:34]	Armin Weatherwax:	we thought they could be better
[15:34]	Mm Alder:	Data mining is a last resort when everything looks confusing.
[15:35]	Morgaine Dinova:	Malbers: can you think of a way to detect prims intended as wearables indirectly? If a heuristic could be found that guesses right most of the time, then maybe an extra Wearable bit on the server might not be needed.
[15:35]	Charlette Proto:	but doesn't it look like LL will dith the pie all together, regardless of how practical it is (even being flawed in design)
[15:35]	ATechwolf Foxclaw:	full prim avatars would foil that algrethem.
[15:35]	Armin Weatherwax:	the question always is when you do not have a hypothesis in the first place is what does it mean
[15:35]	Jacek Antonelli:	There are a lot of mini-hypotheses (e.g. "More people press the Build button than use the B shortcut"). But in general the idea is to collect a lot of information that can be used to learn about how people are using the viewer, not just to test a specific hypothesis.
[15:36]	Mm Alder:	Armin, "better" is too vague. Everything can be better.
[15:36]	Charlette Proto:	agrees with Jacek
[15:36]	Morgaine Dinova:	ATech: very good point. So I guess the Wearable bit isn't optional
[15:36]	Malbers Linden:	I'm not sure Morgaine. Of course, I'm not too concerned with putting an extra bit on the server (personal bias). I'd rather give better tools to our Builders.
[15:36]	Morgaine Dinova:	Malbers++
[15:36]	McCabe Maxsted:	when you design a car, it's best to know how people drive and use the steering wheel, brake, etc. Gathering data on the UI is no different
[15:36]	Morgaine Dinova:	Malbers: your database colleagues might not agree ^_^
[15:37]	Malbers Linden:	I'm sure my DB bretheren would complain, yes
[15:37]	Morgaine Dinova chuckles
[15:37]	Mm Alder:	McCabe, do you think gigabytes of telemetry would telly you something you don't already know?
[15:37]	McCabe Maxsted:	honesty, yes
[15:38]	Charlette Proto:	to me it looks like rich data set (eg timestamps) collected over a shorter time window may be more usefull than having shallow data all the time
[15:38]	Jacek Antonelli:	Mm: You asked McCabe, but my response is: Hell yes. Definitely.
[15:38]	Jacek Antonelli:	Without a doubt.
[15:38]	Armin Weatherwax:	true, mm, a good design would have been have one group with the new design and another without and then have a hypothesis about how the differce is of the use of the gestures menu
[15:38]	Morgaine Dinova:	McCabe: if you gather enough mass statists, you'd probably find that most drivers prefer a driving interface with one hand to their ear. ;-)
[15:39]	Malbers Linden:	One interesting thing that we *knew* was that few people used bubble chat but we didn't know how few. But, digging into some data we have, we figure it's around 5%
[15:39]	McCabe Maxsted:	wow that high?
[15:39]	Malbers Linden:	Those more explicit numbers can help motivate action
[15:39]	Charlette Proto:	very few Malbers
[15:39]	Mm Alder:	How do you find meaningful information in thousands of mouse clicks if you don't know what you're looking for?
[15:39]	Jacek Antonelli:	Quite so, Malbers
[15:39]	Charlette Proto:	but I have met some
[15:39]	Morgaine Dinova:	I've not had bubbles on for years. Think I'll experiment now :-)
[15:40]	Malbers Linden:	even simply for prioritizing bugs, that data is useful
[15:40]	ATechwolf Foxclaw:	I did that the other day. Bubble chat can be usfull in large groups.
[15:40]	Jacek Antonelli:	Mm: We decide what we're going to look for after we have the data. It's expensive (in terms of time and effort) to gather new metrics for every question we want to answer
[15:40]	Charlette Proto:	I tried it about 6 months ago, found it impractical and messy to look (bubbles)
[15:40]	Charlette Proto:	agrees with Jacek again
[15:40]	Jacek Antonelli:	So the idea is to gather as much as we can afford, so that we can detect trends and answer questions that come up without having to gather it all over again
[15:41]	ATechwolf Foxclaw:	But only if nametags was turned off also. Otherwise it got messy fast.
[15:41]	Charlette Proto:	hypothesis up front wuld skew the whole exercise
[15:41]	Morgaine Dinova:	Wow. Immediate observation is that the order of chat lines can't be seen in bubbles. So, chat bubbles are ideal for people who can't handle logic.
[15:41]	ATechwolf Foxclaw:	Mogaine, like proky? lOL
[15:41]	Jacek Antonelli:	Morgaine: Heh, nice loaded statement ;)
[15:41]	Charlette Proto:	not so much Morg, the chat lag does this anyway
[15:41]	Malbers Linden:	and there can be second or third-order effects == people who are 2+ year-old resident but don't own land do X and Y and Z more often
[15:42]	Malbers Linden:	sometimes you can't guess at that kind of data being important
[15:42]	Jacek Antonelli:	Ack, brb, gotta restart the router. Someone send me a notecard of what is said while I'm gone? ???
[15:42]	Morgaine Dinova:	Jacek: not a loaded statement at all. One can't see the order of statements in the bubbles, and so can't see their logical relationship to each other.
[15:42]	Mm Alder:	Malbers, do the survey questions at logon give you any information about how to improve the UI?
[15:43]	Morgaine Dinova:	Haha Charlette, re chat lag :P
[15:43]	Malbers Linden:	not really. the answers are usually rather general. but they are interesting to look at.
[15:43]	ATechwolf Foxclaw:	Never based an idea to remove or change a feature based on persentage of resedents that use it. That one percent repersencts over 10K and can be a very noisy group, expelly if a good percentage of them are sim owners.
[15:43]	Malbers Linden:	and we do track them
[15:43]	Charlette Proto:	you are kidding Mm, when you quit Second Life™ would you bother saying why you did
[15:44]	Malbers Linden:	right ATech, that's why the second-oder data is also so important
[15:44]	Charlette Proto:	noisy is one thing ATech, usage stats another thing
[15:44]	Mm Alder:	Charlette, I would, but I know I'm not normal. :-)
[15:45]	Malbers Linden:	sometimes you gotta piss off 100K residents but other times you caN'T afford to piss of a different 1k of users
[15:45]	Techwolf Lupindo:	begone dobleganger...*clawswipe*
[15:45]	Mm Alder:	So you're saying, Malbers, that you can't rely on numbers. ;-)
[15:46]	Jacek Antonelli:	Back! Notecard please. *makes gimme gimme hands*
[15:46]	Malbers Linden:	data is data -- you have to interpret it carefully
[15:46]	Techwolf Lupindo grins
[15:46]	Malbers Linden:	lies, damned lies, and stats
[15:46]	McCabe Maxsted:	hehe
[15:46]	Techwolf Lupindo:	Gathering good data is importment. This is so one can datamine the same set for different things instaed of trying collect more data.
[15:46]	McCabe Maxsted:	wb jacek
[15:47]	Jacek Antonelli:	Exactly, Tech
[15:47]	Armin Weatherwax:	wb
[15:47]	Morgaine Dinova:	Rate of progress and rate of breaking old things are directly proportional to each other. That's a typical example of needing to piss off the majority of people .... because if you don't then your company and product will become obsolete.
[15:47]	Jacek Antonelli asks again for a notecard of what I missed.
[15:48]	Malbers Linden:	so, Mm, I believe numbers but we also have to apply other business factors to them for them to be information that we can act on
[15:48]	Charlette Proto:	you will get 10 notecards now I think
[15:48]	Jacek Antonelli:	Better than 0. I can datamine them ;)
[15:48]	Charlette Proto:	hehe
[15:48]	Armin Weatherwax gave you New Note.
[15:49]	Malbers Linden thinks Jacek is running her own Metrics experiment.
[15:49]	Jacek Antonelli:	Thanks Armin
[15:49]	McCabe Maxsted:	another case of something numbers will show: just how long it'd take me to actually open the inv, create a notecard, name it propery, and transfer it; right now all I can say is "too long"
[15:49]	Mm Alder:	Will it show that I don't have a clue how to send a notecard? :-)
[15:49]	Techwolf Lupindo:	Why can't I rename a notecard when I am editing it?
[15:49]	McCabe Maxsted:	I think "notecard" is just a linden synonym for "wonky"
[15:50]	Charlette Proto gave you UXIG 090602.
[15:50]	Malbers Linden:	maybe....
[15:50]	Charlette Proto:	woot Mm?
[15:50]	Mm Alder:	?
[15:50]	Morgaine Dinova:	McCabe: if we can include metrics gathering for performance and scalability analysis, that would rock. However, I think this is for HI "gadgets" only, no? Not for the whole user experience, which of course does include performance.
[15:51]	Malbers Linden:	I'd love to ahve all that. But, realistically, we 'll have to start smaller. But where?
[15:51]	Charlette Proto:	so anonymous rich data eg timestamps and avie info like age (alts will confuse it anyway) would be the best start
[15:52]	Malbers Linden:	agree. the 2D UI? the 3D world?
[15:52]	McCabe Maxsted:	the menus and toolbar would be where I would start, but I'm highly biased to toying with them :)
[15:52]	Charlette Proto:	maybe include teleport locations too
[15:52]	Malbers Linden:	I'd *guess* the 2D UI would be easier to instrument
[15:53]	Armin Weatherwax:	also start at not too large samples ... at large sample sizes you get anything statistically significant even if it has no actual meaning
[15:53]	Jacek Antonelli:	What sort of 3D world stuff might be gathered?
[15:53]	Malbers Linden:	I'd guess we already have TP locations as it touches servers
[15:53]	Malbers Linden:	Object interactions, context menus
[15:53]	Morgaine Dinova:	A lot of things already measured/displayed in the Statistic Bar directly affect user experience. Cherry pick from those too
[15:53]	Jacek Antonelli:	Malbers: Ah, right
[15:54]	Malbers Linden:	yes, we know where you are at all times.
[15:54]	Malbers Linden:	but that should be obvious
[15:54]	Jacek Antonelli:	LL already collects some data about performance and stability
[15:54]	Charlette Proto:	if you are worried about the volume of data just look at the freebies in inventories or JIRA that is stored for good, this could be a relatively short periods when the data is gathered
[15:54]	Techwolf Lupindo:	heh...who here has 100K inventory now?
[15:54]	Jacek Antonelli:	We've touched a bit upon some privacy issues, so let's focus on those for a while
[15:55]	Malbers Linden:	it would be nice to have data to prove that NOONE uses the Library
[15:55]	Charlette Proto:	wow no way Tech
[15:55]	Jacek Antonelli:	For example, keeping a record of everywhere the person teleports would be a huge privacy issue
[15:55]	Charlette Proto:	anything over 20k is asking for trouble Tech
[15:56]	Charlette Proto:	not if the identity is withheld Jacek
[15:56]	Jacek Antonelli:	(I'm thinking more here of data that Imprudence et al could be allowed to see. LL already knows where you TP :P)
[15:56]	Malbers Linden:	ah
[15:56]	Morgaine Dinova:	Design is wrong, since inventory performance scales O(N) or higher.
[15:56]	Armin Weatherwax:	vivox does, too, jacek
[15:56]	McCabe Maxsted:	would it be unreasonable to keep avatar names?
[15:57]	Jacek Antonelli:	Charlette: What if someone teleports to "Charlette Proto's Home Island" every day? Or they always happen to teleport there right after they press "Teleport Home" (Hmm, coincidence!). You don't think someone with access to that data couldn't figure out who it was? ;)
[15:57]	Jacek Antonelli:	McCabe: In my opinion, yes.
[15:57]	Jacek Antonelli:	Storing (or at least revealing) avatar names is out of the question, I think.
[15:57]	Charlette Proto:	i don't see much problem with anonymous (with attributes like age) data on TPs etc, we realise that LL knows where we are and wouldaccept location/task relative infoas more valueable
[15:58]	Jacek Antonelli:	As is storing avatar UUID, because those can be easily tracked back to the avatar.
[15:58]	McCabe Maxsted nods. Hm. Communication is something important to track, though. Coud scramble them somehow, all that we need to look at is the relationship, not the actual name
[15:58]	Charlette Proto:	agrees with Jacek on UUIDs and Names
[15:58]	Charlette Proto:	yup
[15:58]	Jacek Antonelli:	If we are going to store information about specific people, my thought is that we would store the SHA1 hash or other encrypted hash of their name or UUID (or both)
[15:58]	Morgaine Dinova:	Totally out of the question, as are TP destinations. We're not doing traffic analysis
[15:59]	Techwolf Lupindo:	Hmmm...do what other large data set goups have done for public access, they can make queries on that data, but do to that data set so huge, could not get any personal data from it. Just find tune the queries to one that make sence or have all queries filter though a Linden.
[15:59]	Malbers Linden:	the "interconnections" betwene people are very valuable -- ho is friends with who, gorups, who gives Inv to who -- but it would be hard to track that without names
[15:59]	Techwolf Lupindo:	find=fine
[15:59]	Mm Alder:	If you already had a hypothesis, then you could do the data analysis in the viewer and just send "yes" or "no".
[15:59]	Charlette Proto:	but location can tell you a lot about task at hand Morgaine
[15:59]	Charlette Proto:	Moragaine has some secrets to hide hehe
[15:59]	Jacek Antonelli:	The hash for a user name or UUID would be constant and unique, but practically impossible to reverse to find out the name
[15:59]	Charlette Proto:	precisely jacek
[15:59]	Jacek Antonelli:	So, IMO, an ideal identifier for people
[16:00]	Jacek Antonelli may or may not have given this quite a bit of thought. >_>
[16:00]	McCabe Maxsted:	hehe
[16:00]	Jacek Antonelli:	This issue in general
[16:01]	Armin Weatherwax:	planing free will participation or forced ?
[16:01]	Techwolf Lupindo:	UUID is the new mark of the beast. lol
[16:01]	Mm Alder:	Any kind of data collection would have to be opt-in. That limits your sample size considerably.
[16:01]	McCabe Maxsted:	I'd assume it'd ask you on install if you want to participate?
[16:01]	McCabe Maxsted:	or first run for non windows people
[16:01]	Jacek Antonelli:	That's another question, Armin. I think the only responsible thing would be to inform them and give them a chance to disable it
[16:01]	Charlette Proto:	naturally if the process and data were compromised one could match the given user to their data if that was available to everyone
[16:02]	Jacek Antonelli:	Explain what data is being collected, and what is not being collected
[16:02]	Mm Alder:	Jacek, it can't be opt-out. It has to be opt-in.
[16:02]	Techwolf Lupindo:	The only way you are going get good date is this:
[16:02]	McCabe Maxsted:	why does it have to be opt-in?
[16:02]	Charlette Proto:	what is the diff Mm????
[16:02]	Charlette Proto:	you have some weird ideas Mm
[16:02]	Jacek Antonelli:	Mm: I disagree. Inform + allow opt-out would be socially responsible, and yield a lot more data
[16:02]	Mm Alder:	Opt-out: you get a revolt. Big brother! Spying! Malware!
[16:03]	Techwolf Lupindo:	"This version of the vewier is collecting usage data in order to improve the user experence. This can be disabled in the preferences panel."
[16:03]	Jacek Antonelli:	Exactly, Tech
[16:03]	Techwolf Lupindo:	popup on start
[16:03]	Morgaine Dinova:	Charlette: I'm sure I could think of a few secrets if pressed. But it's not about secrets of the observed. It's about "none of your fricking business" of the observer. :-)
[16:03]	Charlette Proto:	nobody would enable it if they had to act on it themselves given a choice Mm
[16:03]	Armin Weatherwax:	you can not avoid it anyway, your sample will have more people who are open to try new things and more that are willing to help (2 hypos)
[16:03]	Azwaldo Villota:	must away. thanks again, Jacek, for hosting this. cheers, All.
[16:03]	Jacek Antonelli:	Take care Azwaldo
[16:03]	Morgaine Dinova:	Cyu Az
[16:03]	Mm Alder:	I was sent this notecard today: WARNING: do NOT use the RC viewer, or at least be VERY careful. It is possible that somebody has installed a back door:( Even the main viewer might not be safe:(
[16:04]	Charlette Proto:	hehe morg, not so fussy about my secrets atm
[16:04]	McCabe Maxsted:	well, that is sort of the point. We're effectively spying, but in a harmless way; we want the biggest sample
[16:04]	Techwolf Lupindo:	Do NOT give then option to disable it on the startup popup. Folks will hit NO be instint. Yes, I am thinking like LL here.
[16:04]	Techwolf Lupindo:	be=on
[16:04]	McCabe Maxsted:	hehe
[16:04]	Charlette Proto:	spting for the good of the population hehe
[16:04]	Charlette Proto:	we are big brother I think
[16:04]	Charlette Proto:	1984
[16:04]	Charlette Proto:	hehe
[16:05]	Mm Alder:	1984 is so passe :-)
[16:05]	Jacek Antonelli:	This is another big privacy vs. useful data issue. Frankly, opt-in would severely reduce the number of people who are willing to send us the data. (Yes, just the different of opt-in vs. opt-out makes a *huge* difference.)
[16:05]	Techwolf Lupindo:	There are some nasty ways of getting stuff done and still be legil and/or still ethicial.
[16:05]	Charlette Proto:	not going any further on that Mm (?)
[16:05]	McCabe Maxsted:	it's not malware if we're upfront and honest about what's going on and give people a choice
[16:05]	Mm Alder:	Linden cannot risk alienating their residents any more. :-)
[16:06]	Charlette Proto:	installing a reality mining bot on Mm
[16:06]	Techwolf Lupindo:	Mn, ROFL
[16:06]	Morgaine Dinova:	I think a serious question needs to be asked first. What is on the cards for changing as a result of data capture? Just saying "we won't know until we capture it" is a black cheque for total spyware. Don't go there. Be specific in what you need to observe.
[16:06]	Jacek Antonelli:	E.g. countries with opt-in organ donation policies ("Check this box if you want to be a donor") tend to have 10-20% of the population donate organs. Countries with opt-out ("Check this if you DON'T want to be a donor") have more like 80% of the population as donors.
[16:06]	Charlette Proto:	heard of Reality Mining Mm, it is very NOW
[16:06]	Techwolf Lupindo:	LL allready have tested those waters many time. Adult Content was one of them, opensim price hike was another, they can't do any worse? Can they?
[16:07]	Charlette Proto:	you are right Tech
[16:07]	Charlette Proto:	it would be hard to appear lower than LL atm
[16:07]	Mm Alder:	Charlette, no reality here. :-)
[16:07]	Techwolf Lupindo gets another group chat DING!
[16:07]	McCabe Maxsted:	you know, honestly, my biggest concern would be whether or not if LL decided to collect data, they'd let us look at it
[16:07]	Jacek Antonelli:	Indeed, McCabe
[16:08]	Mm Alder:	Good point McCabe.
[16:08]	Charlette Proto:	Mm, read up on reality Mining and you will remember Orwel no doubt
[16:08]	Jacek Antonelli:	To be honest, this discussion is just as much about third party viewer projects (e.g. Imprudence) collecting data, as it is about LL collecting data.
[16:08]	Charlette Proto:	hi Geneko
[16:08]	Armin Weatherwax:	hm. my thought was when i read about this on sldev first: find a (or some) university (-ies) to let them do research. that takes out a lot of 1984 paranoia in my opinion
[16:08]	Jacek Antonelli:	If LL wants to collect the data *and share it*, that is super peachy. But if not, we're pretty determined to tackle it ourselves.
[16:08]	Geneko Nemeth:	All but forgot about UXIG... overslept...
[16:09]	Jacek Antonelli:	Heya Gen
[16:09]	Morgaine Dinova:	The more I think about this idea of data capture for feedback, the more I think it's a recipe for disaster, and actually helps zero.
[16:09]	Charlette Proto:	hehe that must have benn nice till you woke up Geneko
[16:09]	McCabe Maxsted:	if they don't share it, honestly, it's no different than if they didn't collect it at all; most decisions woud be based on "because we said so" if they were explained
[16:09]	Malbers Linden:	Hi Geneko
[16:09]	McCabe Maxsted:	and it'd be rather ironic that microsoft would be the more open company...
[16:09]	Charlette Proto:	true McCabe
[16:09]	Jacek Antonelli:	Does Microsoft share it's data, McCabe? I didn't think they did
[16:10]	McCabe Maxsted:	they've shared a lot of data that went behind windows 7 decisions
[16:10]	Geneko Nemeth:	But Microsoft doesn't make its (most used) softwares' code freely avaliable.
[16:10]	Jacek Antonelli:	Hrm. Did they share the datasets, or just "this is what we found"?
[16:11]	Techwolf Lupindo:	Windows 7 is Vista SP3. Vista was such a disator that marking knew the only way out of it is to create a new name and dump the tarished name.
[16:11]	Jacek Antonelli:	Morgaine: the benefit is tangible and significant. The risk can be avoided by careful design.
[16:11]	Morgaine Dinova:	Jacek: your evidence for that?
[16:11]	Charlette Proto:	i don't think data could go out of Microsoft
[16:11]	Jacek Antonelli:	For which?
[16:12]	Charlette Proto:	Tech you are right, but I'd like to work for a company that does such a good SP3 and gets to charge the full price for it with such good response
[16:12]	Morgaine Dinova:	Evidence that the benefit is tangible and significant, and evidence that the risk can be avoided by careful design. 3 of them
[16:12]	Armin Weatherwax:	morgaine +3
[16:13]	Charlette Proto:	Win 7 is a marketing marvel besides being a great product 10/10 to Microsoft
[16:13]	McCabe Maxsted checks. Mostly "this is what we've found"
[16:13]	Jacek Antonelli:	The usefulness of UI metrics is well documented in the software design industry. The fact that the privacy risks can be avoided is obvious -- if you don't collect data that can be tracked back to someone, and you give them a chance to opt-out, the privacy issues are avoided.
[16:14]	Charlette Proto:	this is from a position of a total looser with Vista
[16:14]	Charlette Proto:	agrees with Jacek, data is a key to usability work
[16:14]	Morgaine Dinova:	I think it's quite a red herring. Most people already know what problems exist, and they have the benefit of a million eyeballs. In contrast, data being gathered and analysed by a few designers is inherently poor because of the limited eyeball and limited range of experience. So I think a far better approach is the current one: sensible people telling you what their problems are.
[16:15]	Morgaine Dinova:	And it carries no risk.
[16:15]	Geneko Nemeth:	You assume that people are sensible and know their problems, though.
[16:15]	Armin Weatherwax:	thats just another way of data mining, morgaine
[16:15]	Jacek Antonelli:	Exactly, Geneko
[16:15]	Charlette Proto:	heuristics are a very diff beast to raw data Morgaine
[16:15]	Morgaine Dinova:	Not all do, but many do.
[16:15]	Geneko Nemeth:	(And are willing to tell, but if they are sensible and knowledgeable they would probably be willing. XD)
[16:16]	Jacek Antonelli:	Morgaine: Even the most sensible and willing person is not going to take the time to count how many times they use each menu entry or button and tell you that.
[16:16]	Malbers Linden:	I almost never listen to what people SAY in usability studies; i look at what they DO
[16:16]	McCabe Maxsted:	morgaine: that approach is akin to asking the support desk what problems users are experiencing, then figuring out what's wrong with the software that way; that can work to an extent, but it's dirty and not objective and won't reveal important trends
[16:16]	Charlette Proto:	see even a chinese is willing to be spyed on for the good of humanity hehe
[16:17]	Geneko Nemeth:	Mweh Mweh! :O
[16:17]	Morgaine Dinova:	Armin: no, it's very different to have the value judgement distributed among the users, versus centralized value judgements by designers. Not the same thing at all.
[16:17]	Jacek Antonelli:	Data is not a replacement for user feedback, nor is user feedback a replacement for data. They're both important, but so far the data has been nil
[16:17]	Malbers Linden:	right, we ahve lots of opinion and not enough data; a balance is needed
[16:17]	Morgaine Dinova:	I see no evidence for how it can help. How about giving us some examples?
[16:17]	Jacek Antonelli:	If we had data, we could make informed design decisions. Otherwise, it's just guesswork and theory.
[16:17]	Charlette Proto:	agrees with Jacek, opinions work on a different level than raw stats
[16:18]	Charlette Proto:	some people simply don't mention cumbersome aspects of their work because these have become routine
[16:18]	Armin Weatherwax:	true, morgaine, not the same thing, but its both data mining , its just another method to get the data (and in psychological terms a different level of abstraction)
[16:18]	Mm Alder:	If you don't believe what users report, then accept their statements as hypotheses and test them.
[16:18]	Jacek Antonelli:	Morgaine: I never use the B key to open the build tools. I could probably find 15 other people to tell me that they dont use it either. From that, I could figure that nobody must use it, so I'd remove that shortcut. Then get thousands of people screaming "Why did you remove that, I use that all the time!!"
[16:19]	Malbers Linden:	arg. I'm way overtime. Great discussion. Nice to be back for a meeting with all of you.
[16:19]	Morgaine Dinova:	Jacek: exactly. Which shows the extreme danger of UI design based on statistical data collection.
[16:19]	McCabe Maxsted:	take care malbers
[16:19]	Jacek Antonelli:	Thanks for coming, Malbers! Take care
[16:19]	Malbers Linden:	bye
[16:19]	Morgaine Dinova:	Cyu Malbers
[16:19]	McCabe Maxsted:	and remember: sharing is caring
[16:19]	Mm Alder:	Thanks for stopping by Malbers
[16:19]	Geneko Nemeth:	See you Malbers!
[16:19]	McCabe Maxsted:	:)
[16:19]	Charlette Proto:	nice to see U Malbers, thanks for comming
[16:20]	Jacek Antonelli:	Morgaine: No, that shows the danger of UI design based on anecdotal evidence, not statistical data collection. Asking 15 people is not statistical data collection.
[16:20]	Armin Weatherwax:	... thats what malbers says: look at what people are doing and what they are saying is different data, but its all data (and most likely they correlate highly)
[16:21]	Jacek Antonelli:	If we had proper data, those dangers would be *eliminated*, because we'd know for certain how many people used it.
[16:21]	Charlette Proto:	not neccessarily Armin, routines are subconscious
[16:21]	Jacek Antonelli:	And that's just one use for the data. There are many, many others just as compelling.
[16:21]	Morgaine Dinova:	Jacek: you missed the point them. Let's say you use the B key and consider it essential, and data from 100 million people shows that they tried it once and never again. So what, now you delete your essential HI element just because of the wisdom of the masses?
[16:22]	Charlette Proto:	I prefer data because of the easy with which people put up with mandane routines
[16:22]	Mm Alder:	There used to be a saying: You don't need a weatherman to tell you which way the wind is blowing. If you told people you were trying to figure out what is wrong with the UI, they'd ask you in disbelief, "Don't you see it?!"
[16:22]	Jacek Antonelli:	Morgaine: If 100 million people use it once and never again, then it's not an essential HI element. (Unless we're talking a user base the size of the population of the earth)
[16:23]	McCabe Maxsted:	data doesn't lie; people on the other hand...
[16:23]	Charlette Proto:	agrees
[16:23]	Mm Alder:	Data doesn't lie. Interpretations do. :-)
[16:23]	Morgaine Dinova:	Jacek: what constitutes "essential" is a per-person judgement. You're one human with an interface need, not a jello colony (pardon to Malbers).
[16:23]	Charlette Proto:	people (especially in SL) are very opinionated and that is not what we want
[16:23]	Jacek Antonelli:	On the other hand, if the data showed that 10% of the user base uses it frequently, that would allow us to make an informed decision, instead of guessing.
[16:23]	McCabe Maxsted nods. I'd be much happier to argue over an interpretation of data than do guesswork back and forth
[16:24]	Jacek Antonelli:	Especially if we have secondary data to tell us that those 10% are hardcore builders.
[16:24]	Charlette Proto:	just consider the recent change to the pie menu, with usability stats that would have never happened
[16:24]	Morgaine Dinova:	An informed decision based on majority appeal? So now you're going to plaster porn all over the viewer, are you?
[16:24]	Charlette Proto:	hardcore builders hehe
[16:24]	Jacek Antonelli:	Morgaine: Don't be ridiculous.
[16:24]	McCabe Maxsted:	hahaha
[16:24]	Charlette Proto:	hehe Morgaine
[16:25]	McCabe Maxsted:	ahh, *starts developing a porn skin*
[16:25]	McCabe Maxsted:	I'll be Internet Famous!
[16:25]	Jacek Antonelli:	haha McCabe
[16:25]	Armin Weatherwax:	what brings us back to have a proper hypothes so that you know how to interpret data before you have it
[16:25]	Charlette Proto:	have an option for a porn login screen
[16:25]	Morgaine Dinova:	Seriously, I have to wonder about some designers thinking they can make a "one shoe fits all" best UI.
[16:25]	Charlette Proto:	based on user behaviour
[16:25]	Jacek Antonelli:	Armin: I'd rephrase that as "so you know how to interpret data before you start mining it".
[16:25]	Charlette Proto:	opt-in to porn
[16:25]	Mm Alder:	Morgaine++
[16:26]	McCabe Maxsted:	armin: we have tons. Do people use the toolbar more than the menu? Do they keep the advanced menu open? How many clicks are in a standard building session? How often is the build window opened and closed? What constitutes a build session anyway? What is the first ui task people perform on entering a new sim?
[16:26]	Armin Weatherwax:	what does it mean, that turn off typing animations is triggererd only once in a lifetime of a viewer
[16:26]	Charlette Proto:	it has been noted you use Second Life™ for cybering 90% of the time; would you like to make this easier for you
[16:26]	McCabe Maxsted:	the current ui *IS* our hypothesis; it's our guess as to the answers to those questions
[16:26]	McCabe Maxsted:	what data will do is tell us how well our guess has fared, among other things
[16:26]	Morgaine Dinova:	Haha, Charlette
[16:27]	Jacek Antonelli:	Morgaine: We're not talking about one size fits all. But a default UI that serves 90% of the needs of 90% of the people is an undeniable improvement over a viewer than servers 50% of the needs of 50% of the people.
[16:27]	Charlette Proto:	hypothesise when you see the data not before
[16:27]	Armin Weatherwax:	does it mean its a useless checkbox ? everybody turns it off only once ...
[16:27]	Mm Alder:	McCabe, "Do people use the toolbar more than the menu?" is a question. It doesn't say what you intend to do about it.
[16:28]	Jacek Antonelli:	Armin: Obviously, you'd have to interpret the data in context.
[16:28]	Armin Weatherwax:	...
[16:28]	Morgaine Dinova:	Jacek: and both of those are WRONG design goals. Instead, make a viewer that potentially satisfies 100% of the population by being configurable.
[16:28]	Charlette Proto:	well it means the checkbox should be on a 'setup' screen and doesn't need to be very 'visible' Armin
[16:28]	Mm Alder:	Charlette, that's a good way to ensure that the data always fits your hypothesis.
[16:28]	McCabe Maxsted:	yeah, you're right. It's a question. I'd really like to know the answer though
[16:28]	Mm Alder:	Why?
[16:28]	Charlette Proto:	don't bother Mm, I've studied this shit to ma master's level with awards
[16:29]	Charlette Proto:	and we call Design a science Mm
[16:29]	Geneko Nemeth:	Morgaine: I thought we talked about sensible defaults vs. customiziability before.
[16:29]	McCabe Maxsted:	because you can't form a new hypothesis without data; hypotheses exist to explain observations, and as was pointed out earlier, we need to observe first, and are severely lacking in that area
[16:29]	McCabe Maxsted:	another way of saying we can't fix a problem we don't understand
[16:29]	Techwolf Lupindo:	A good data set would include what options have been changed from default. If it turns out over 50% change that option, like IM in chat, then that default should be changed, but nevered removed.
[16:29]	Charlette Proto:	agrees with McCabe
[16:30]	Mm Alder:	McCabe, you can't fix a problem you don't have, either.
[16:30]	Jacek Antonelli:	Techwolf: Quite so
[16:30]	Jacek Antonelli hands Mm the uselessly obvious statement of the day award? :)
[16:30]	McCabe Maxsted:	true. And data should show a lot in that area. For example, what really is a problem, what is a problem but only for a few people, and what *REALLY* is a problem for a lot of people
[16:30]	Charlette Proto:	hehe Jacek 10/10
[16:31]	Mm Alder:	I'd like to thank my ... :-)
[16:31]	Jacek Antonelli:	hehehe
[16:31]	Morgaine Dinova:	Geneko we did talk about that yes, but somehow the idea of mere defaults has been lost entirely. Not once has it been said that the data mining would be used merely yo configure the default setup.
[16:31]	Geneko Nemeth:	And it's better not said, because we need both.
[16:31]	Jacek Antonelli:	Morgaine: Do we have to explain every week that customizability is good?
[16:32]	Techwolf Lupindo:	Big problem for newbies, when they buy something, the vender can't give them a folder to wear, instead they get a box to wear on there head.
[16:32]	Morgaine Dinova:	Jacek: yes, apparently we do, to keep you from pedalling us backwards
[16:32]	Jacek Antonelli rolls her eyes.
[16:33]	Charlette Proto:	well Morg; what changes are waranted based on the data may be the hypothesis we arrive at (hehe) defaults, visibility of functions etc
[16:33]	Geneko Nemeth:	Now, I never said we don't need a viewer that can satisify 90% of the people out of the box, either.
[16:33]	Armin Weatherwax:	yeah, sorry that i was a bit agressive today, its just that things need to e done very carefully for getting *good* data. anyway, late here, nett to say goodbye
[16:33]	Armin Weatherwax:	need*
[16:33]	Jacek Antonelli:	Take care Armin, sleep well
[16:33]	Armin Weatherwax:	:)
[16:33]	Armin Weatherwax:	cu
[16:33]	McCabe Maxsted:	one thing that data should also be able to do: convince the server people that the architecture needs to change. Like morgaine's wearable bit example, it's a huge investment to change something like the inv, and unless you can justify it in a really convincing way there's no reason for developers to do it
[16:33]	Morgaine Dinova:	Charlette: yes, it could be used for that. Shall we make it plain and open? THE DATA MINING WOULD MERELY BE USED FOR CONFIGURING DEFAULTS. Sounds good? I'd be happy.
[16:34]	Charlette Proto:	Tech you are thinking specifics here, we need a generalised approach to improve the aspects of UI
[16:34]	Jacek Antonelli:	Morgaine: I have not ever said, and never will say, that I am opposed to UI customizability. Exactly the opposite -- I think it's a good thing. But customizability is not an excuse to have crappy defaults.
[16:34]	Morgaine Dinova:	McCabe: I'd love metrics to prove that there are issues server-side. But I don't think they fit into the present collection scheme.
[16:34]	Charlette Proto:	Morg, but some of the data may reveal that UI organisation needs to change etc
[16:35]	Charlette Proto:	you can't assume the extent of change needed till you get a look at the stats
[16:35]	Techwolf Lupindo:	1) collect usage data with current viewer, good for setting up default settings. 2)collect data from the new user/newbies mentors about problems newbies encounter all the time. See about fixing it.
[16:36]	McCabe Maxsted nods. Yeah, you're probably right. Hopefully it'll be broadened to include that there, or at least show enough of a deficit in certain areas where people will take these things more seriously
[16:36]	Charlette Proto:	don't see the need for that separation Tech
[16:36]	Charlette Proto:	I've spent last 12 months with newbies and mentors teaching them
[16:37]	Morgaine Dinova hands Charlette two buckets of spirin
[16:37]	McCabe Maxsted:	that's two very different datasets, charlette
[16:37]	Morgaine Dinova:	And aspirin too
[16:37]	Techwolf Lupindo:	some things can't be collected with data. Unless you do a search for "box on head" against the data collected. One might get the idea I lot of folks like putting boxes on head.
[16:37]	Techwolf Lupindo:	I=a
[16:37]	Jacek Antonelli:	hehe
[16:37]	Morgaine Dinova:	Haha
[16:37]	Charlette Proto:	the mentors have very good observations, but they only attempt to teach what they know gets them through and don't neccessarily have any UI design knowledge
[16:37]	Geneko Nemeth:	And then you have furries who usually have a lot of attachments on every points.
[16:38]	Charlette Proto:	Tech is wearing a box on his head, or is this the reason he always wears the same thing
[16:38]	Techwolf Lupindo:	That way ask mentor what they encounter all the time with newbies. Not ask for what they think is a good UI change. As charleete pointed out.
[16:38]	Jacek Antonelli:	On the other hand, if you had data that showed that lots of people wear something on their head, then immediately take it off, rez it on the ground, and open it .... you could come up with an alternate explanation. :)
[16:39]	McCabe Maxsted:	(does right clicking close notices for other people too?)
[16:39]	Morgaine Dinova:	Tech: your very good point equates to "Correlation does not imply causation." And from data gathering, you can easily make correlations with no underlying causation.
[16:39]	Jacek Antonelli:	(Right clicking on what, McCabe?)
[16:39]	McCabe Maxsted:	group notice...
[16:39]	Charlette Proto:	precisely Tech, teaching difficulties are one thing (eg hard to teach UI elements etc) but solutions may be somewhere else
[16:40]	Techwolf Lupindo:	Setting an object for sale sould have an option of giving out folder or object. But that requers a server change, not lickly to happen in my lifetime. Usfull things that can help newbies tend to never get looked at by LL. :-/
[16:40]	Geneko Nemeth:	Actually, that option is there.
[16:40]	Charlette Proto:	Tech we could have product duplication inside vendors eg 'BOXED' vs folder
[16:41]	Geneko Nemeth:	It's just that vendors that give out boxes are easier to make than ones that give out folders.
[16:41]	Mm Alder:	Got to go. Thanks for tolerating me. :-)
[16:41]	Charlette Proto:	byee Mm
[16:41]	Jacek Antonelli:	Tech: My hypothesis is that LL prioritizes changes least useful first ;)
[16:41]	Jacek Antonelli:	Take care Mm
[16:41]	Morgaine Dinova:	Techwolf: well Malbers liked my "Wearable item" bit for creators to set on their prim wearables ... so not all hope is lost :-)
[16:41]	Morgaine Dinova:	Cyu Mm
[16:41]	Techwolf Lupindo:	Or something that more usfull, but harder to implemt. Entire outfits in one box. Just wear the box and the outfit inside is worned.
[16:42]	Charlette Proto:	agrees with Jacek again, just look at Snowglobe
[16:42]	Jacek Antonelli:	I dunno Charlette, HTTP textures is pretty useful. So are some of the other UI changes (although most of those are from Aimee, hehe ❤ )
[16:44]	Charlette Proto:	what LL needs to do pronto is to make a way so one could 'wear' a whole avie at any of the attachment points to make avies align and stay together for as long as they like, eg like a rat on a shoulder
[16:44]	Geneko Nemeth:	POKE!!!!
[16:45]	Charlette Proto:	i agree jacek, but HTTP textures are not a result of the Snowglobe release
[16:45]	Techwolf Lupindo:	More then one wear on one point can be done. Just look up the "evertihng attacted on my ass" bug.
[16:45]	Morgaine Dinova:	1.27 sim is coming out shortly. Is there any viewer angle to it too, something helpful that could be built into viewer?
[16:45]	Charlette Proto:	just seems to me Snowglobe is a launching pad for unancicipated changes from upstairs, just look out for the face gestures now
[16:45]	Jacek Antonelli:	hehe Tech
[16:46]	Jacek Antonelli:	Hrm, I'm not familiar with the sim changes for 1.27
[16:46]	Morgaine Dinova:	Http-in, mainly
[16:46]	Charlette Proto:	hehe I remember when one would TP and the shoes and hair were in the ass
[16:46]	McCabe Maxsted:	http://wiki.secondlife.com/wiki/Release_Notes/Second_Life_Beta_Server/1.27
[16:46]	McCabe Maxsted:	*facepalm*
[16:47]	McCabe Maxsted:	they're removing the "restore from inv" feature
[16:47]	McCabe Maxsted:	how stupid
[16:47]	Jacek Antonelli:	What??
[16:47]	Charlette Proto:	what???
[16:47]	Techwolf Lupindo:	McCabe, explain?
[16:47]	Techwolf Lupindo:	What does that do? I hav't use it I think
[16:47]	Charlette Proto:	really, that is really bad
[16:47]	McCabe Maxsted:	ahh wait nm, I read that backwards haha
[16:48]	Charlette Proto:	hehe good, more offline backup should be allowed
[16:48]	McCabe Maxsted:	"Remove "Save Object Back to My Inventory" features"
[16:48]	McCabe Maxsted:	that one was always really poorly named...
[16:48]	Jacek Antonelli:	heh, ah. Yeah, they disabled the "Save Object Back to My Inventory" a while back. Because it's not working. Y'know, cuz if it's not working, it's better to remove it than fix it.
[16:48]	Techwolf Lupindo:	I remember one featrue that had to be truned off due to inventory lose. I think it was restore from inventory, I seem to remember now. Restoring to a defunk sim/offline sim, the item would be deleted .
[16:49]	Charlette Proto:	some of the permission UI terms should be reviewed
[16:49]	Jacek Antonelli:	They removed Restore To Last Position from the 1.23 client, because it didn't have any confirmation dialog or message about where the object went, so people were losing (i.e. couldn't find) their stuff
[16:50]	Jacek Antonelli:	It took me about half an hour to add a confirmation dialog for it in Imprudence.
[16:50]	Jacek Antonelli:	But I guess that's too much work
[16:50]	McCabe Maxsted:	(do we display an SLurl™ in chat of where it rezed too, btw?)
[16:50]	Charlette Proto:	I need to make coffee before next meet, so BYEE to anyone that pooffs before I return
[16:50]	Jacek Antonelli:	No. I wish we could, but LL hasn't made that possible
[16:51]	Jacek Antonelli:	There's no way to ask the server where it will be sent, or where it was sent
[16:51]	McCabe Maxsted:	ah; shame
[16:51]	Jacek Antonelli:	Minor oversight on LL's part
[16:51]	Morgaine Dinova:	I get the feeling that each character they type has to be signed off by M Linden.
[16:51]	Morgaine Dinova:	In triplicate
[16:51]	Jacek Antonelli:	lol
[16:51]	Jacek Antonelli:	In blood!
[16:51]	Morgaine Dinova:	Ouch
[16:52]	Charlette Proto:	just imagine Morgaine Linden, that would be a flame in a barn full of straw
[16:52]	Morgaine Dinova:	Seriously. it's not fun to be a Linden dev at present, by the many hints in office hours
[16:52]	Jacek Antonelli:	Yeah
[16:52]	Techwolf Lupindo:	jacek, did that confermation include "where" the object was rezed at?
[16:52]	Jacek Antonelli:	Tech: No, unfortunately that's not possible, as I mentioned. No way to find out.
[16:52]	Geneko Nemeth:	Sigh, everything changes...
[16:53]	Jacek Antonelli:	The dialog is pretty much just "Are you sure you wnat to do that? You shouldn't do this if you don't know where the object will go!" and has a link to our wiki page about the feature
[16:54]	Jacek Antonelli:	I really, really *wish* we could get more information about where it will be/as rezzed, but it's just not possible.
[16:54]	Jacek Antonelli:	Without server changes
[16:54]	Techwolf Lupindo:	Jacek, they probley did have that choice, but firgure it better to turn it off untill they have time to put in a real fix istead of half-assing it....wait a minute....
[16:55]	Techwolf Lupindo:	Now why did they truned if off instead of just doing a half-ass job.....
[16:55]	Techwolf Lupindo:	:-)
[16:56]	Jacek Antonelli:	Oh, I'm sure that's why they removed it from the main SL viewer. It's still only half-baked, but at least people won't accidently trigger it, then go crying to support that they lost something.
[16:57]	Techwolf Lupindo:	Later, andrew in 5 minutes
[16:57]	Jacek Antonelli:	(The server part is only half-baked, that is. Thankfully they didn't remove it from the server, because it's actually incredibly useful if you know what you're doing)
[16:57]	Jacek Antonelli:	Yeah, let's adjourn.
[16:57]	Jacek Antonelli:	Thanks for coming everyone! Good discussion. See you next week.