Imnotgoing Sideways

Monti Messmer wrote: Some of the help-groups are busy but not all are very friendly. No telling names but every question i asked left unanswered and they kept on talking rubbish about non viewer things. Monti Signal:noise can be a tricky thing. Help groups and their hangouts are generally balancing volunteers and 'regulars' and in many ways, finding a volunteer is the proverbial needle in a haystack. NCI attempts to resolve this by having a seperate community chat group along with the main Q&A group with shaky success. Though, when landing at our main site in Kuula, the place is very much a community sandbox with a long list of social regulars who aren't really interested in volunteering. =^-^= **I'll admit some guilt in being a rubbish talker. ()y

I do what I can to use every client available at least once. SL has had multiple UIs for quite a long time (remember OnRez?). But, I have seen rifts form in the volunteer communities. (._.) 1) Helpers who have self-determined that "Phoenix is best" and won't offer any help until the person they're helping installs it. 2) Helpers who have self-determined that every TPV is a password stealing virus but also don't like V2/V3 so they give out links to LL's last 1.23.5 build. 3) Helpers who run the most fringe client on a computer that still has AGP graphics and goes on to talk about how "bad" SL got since glow/windlight/voice/shared media/web search came out. 4) Helpers who run the absolute most recent release of the LL client yet never diversify enough to see how other UIs behave or what's coming in the near future. 5)6)7)... I can keep going. (=_=) I don't think it's the UI diversity alone which causes support issues. It's more this cult of personality which follows each project. It's brand loyalty run rampant and many are trying to get technology to stall at their own preferred stage of advancement. (._.) My plea to Mentors, volunteers, support, and helpers is to diversify. Let people pick their own favorite and support them on the use of that. No need to 'convert' people from one interface to the next. If anyone asks me what client I'm running, I half jokingly say "not the one I was using yesterday". (=_=) So, no, more viewers do not result in less support. Shortsighted support associates result in less support. (=_=)y

SL names are serious business. ()

I must say that I'm surprized you were able to get it that far!!! (O_O) Water reflections are generally done at a lower resolution to reduce lag and the waves distort things enough to get away with the cut in quality. I'm not sure how much of your current reflection was accomplished strictly with Windlight settings and/or how many debug settings were involved. My guess is that there would be a quality/blur/resolution/LOD/scale setting in debug in order to dial that in. I'm sure it'll take a bit of hunting. Though, there's a part of me which doubts it's an open variable and may require a HAXd client to adjust the necessary constants. (._.) **Cool name, by the way! =^-^=

Long story, short; pretty much yes. Frequent rolling restarts are masking isues which take a week or more to be observable. (._.) If anything, the lab should (and hopefully does) maintain a small collection of sims which run a certain degree of endurance testing. Getting code reviewed by an external QA is a risk many closed-model software companies won't take. So, it's pretty general in practice to gather performance issues from compiled code only. Not saying it's the best way to work out issues, but, going open sources isn't the beat all. end all either. (>_<) In a way, though, this is a rather old topic. When Havok4 hit the grid, sandboxes began becoming plagued with asset issues. Ghosted prims, flaky inventory performance, and general overall mess. Was it all result of some mad rush of bad code? Likely not. What was occurring was that Havok1 was craching sims so often that most sandboxes would get a daily restart. With the newfound stability of Havok4, old problems became more apparent. (._.) This is a pretty typical growth process for something so complex. Each solution can possibly lead to a number of pre-existting problems being unmasked. So, in a way, what we're experiencing right now isn't as alien as it may seem. This isn't to say that the sim code doesn't need any in depth review. But, given what I got from a brief conversation with Bagman at SLCC, the lab has their work cut out for them. With that in mind, there are no easy answers... (._.)

The choppyness is the result in a change over how alphas are handled. As you can see, it's horribly imperfect. (T_T) Up until now, all alphas other than Linden trees and the avatar mesh have been treated as if they're 8-bit and have 256 potential levels of transparency. Because of this, the client has had the daunting task of deciding which texture is in front while dealing with multiple alpha surfaces. (._.) Now, I believe the feature is called "Automatic Alpha", but someone can correct me. What the client is doing is finding textures with alpha values 0x00 and 0xFF and not much in between. Once it passes some sort of threshold, the client renders it as a 1-bit alpha. Meaning, it's visible or it's not. This is far easier to overlay from one alpha to the next since there's no blending necessary. () But... I don't believe the current threshold has been set at its best. Hair strands in alpha textures do tend to be 'on' or 'off' but still depend on the shades in between to smooth themselves out. Once the client knocks it to 1-bit, that's lost and you get that grainy, pixellated look. (>_<) I haven't explored it much, but, there's a setting either in preferences or a debug setting for Automatic Alpha (someone please correct me). This can disable the feature so things will go back to as they were before.. () Another solution (And, it's hard work) is to find each alpha textured prim and set Transparency to 1%. This will be such a small change in transparency that it'll relatively go unnoticed, but, it'll be enough to "trick" automatic alpha to not trigger and leave your hair strands alone. () ****WARNING!!! - Don't do this to the whole head of hair. The solid textured prims need to stay solid or things just get uglier!!! DX ****

I would have to see an angle shot to see how much detail is in the texture vs the shape. That said, with careful texturing, I could see doing that in 8LI or so. () On the other hand... My low detail stuff generally rides around 0.50LI... I lurves my linksets. =^-^=

There have been attempts to fill this 'much more' statement. Premium-only sandboxes and sporadically updated exclusive freebies come to mind. But, beyond that, yeah. The enthusiasm by the lab to fulfill this 'much more' appears quite lacking. Things are turning around in regards to general quality of experiencing SL, but the promising voice of the LL blog continues to reek of the inability to fllow through. (._.)

"No, thank you." ()y

My experience is completely opposite. I use web-on-prim to share data when collaborating with volunteers and friends. My display name is set to "Immy". I consider the V1 interface to be terrible. The V2/V3 isn't great, but, I find it to be a vast improvement. I make mesh items. They're not crappy. So far, with the tiniest bit of effort, I've been able to make items which look nicer than prim builds, load faster than sculpts, and use far far less resources. () My store remains on the first page search results for my target audience and the two freebies I "sell" on the marketplace move trouble-free at 20 to 50 units per day. =^-^= There ~is~ less server lag. I remember the days when a teleport would take ages, if it didn't time out, and cause chat to stall while the server was working out your arrival/departure... At the same time, any IMs being sent while in teleport would go to Email because the system assumed I was offline. Physics have been smoother and faster than ever before. So much so that I've been able to build a fully physical 2 piston crankshaft assembly which will operate in a sim with 30 avatars present. () Resident from 2007, that is seeing things go in very good ways. =^-^= **can I have your stuff when you leave? ()y

I likes. Been having fun seeing what kinds of savings I get in simple builds. =^-^= ... I still grief you! >8D

I've seen a Belldandy AV, but, the maker wasn't a very practiced clothing maker and it turned out quite bad. (._.)

The avatar mesh UV is an imperfect guide. You're right about the alpha layer falling off before the seam. The thing to do is colour outside the lines. Continue your texture pattern just slightly beyond the UV map to give it a little extra data to pull from to complete things. The UV map you're using is likely to include 'blend' regions as coloured bands just outside the main shape. Those are the areas to fill in. ()y And, once you work with them, you'll notice there are a few spots where two blend regions overlap. Those are big bad trouble spots. (>_<)

Invisiprim has always been an exploited graphics glitch. For shadows and improved shine and alpha behaviour, the glitch finally had to be fixed. That said, I don't think there's any better or other way to block the visibility of water. (._.)

As far as I know, the client doesn't send 'hover' messages to the server. So LSL doesn't have anything to pull an event from. (._.)

It also helps to take a screenshot of your existing AV space and use it as a template for how things are placed. I even got a bit silly and the mini icon which overlaps the background image is a copy of that sgment of the image.

Product:Demand can be hell. ()y

So... You choose to abandon land and then complain that you no longer have access to it. Uhm, good luck with that. ()y

Oh, that's well wicked! I'm so totally gonna expand the hall of doors on my sim!. =^-^=

To see if I'm reading this right... () "Returns 0 and does not move the object if position is more than 10m off region or above 4096m." Does this mean that there is a potential to position an object up to 10m into a neighboring sim, triggering a sim crossing? ()

Oskar Linden wrote: I almost forgot that I get to make all new wiki pages for 2012! Where's my whip?... (=_=)

I like the direction V3 is going. Honestly, I miss the sidebar a bit. I always liked tabbed IMs and grew to prefer tabbed inventory/group/whatever dialogues. I am also extremely anti-pie menu. 2/3rds of my computing day is spent in Visual Studio, MS Office, Chrome, and SolidWorks... Each of which drops a contextual dialogue which I can slide right into. Suddenly faced with a "pie" menu, I'm looking for things in a circle. But, what I'm looking for isn't there. Click "More", still not there. "More" again, still not there... Until I've gone full circle and finally find it on the second "More" page where I overlooked it the first time. (=_=) So, no, you can't PAY me to use any throwback UIs. (>_<) That said, I currently use Niran's (Which happens to have a pie menu option... that I disable). I can surely attest that the frame rate is near the low end of all clients I've used before. But, I found that I'm pretty much incapable or crashing it on my system. In addition, I can turn up every graphic setting and take a high resolution screenshot without immediately going kaput. I can't dislike that. =^-^= When asked which client I'm using, I half jokingly respond "Not the one I was using yesterday". Because I somewhat constantly rotate, try out betas, try out dev builds, and anything else that comes down the pike. I will never say any one client I use is "better" or "best" in any way. But, I can say which is my favorite. At the moment, it's Niran's. Now that I've seen these FPS comparisons, I'm going to try some others out to see if anyone else has achieved the stability I find in Niran's with the added bonus of some FPS on top. () Oh... And it MUST render mesh. I'm having too much fun toying with the stuff. =^-^=

Can't please everyone. () It's still fun to have fun. ()y

When wearing multiple items on the same layer, it's first come - first serve. So, if you have a shirt as undershirt and a bra on the same layer; put the bra on first, then the shirt. This way the shirt will layer on top. If you're selecting a whole folder from outside your ourfit list, it's a gamble. From within your outfit list, the order is remembered. You can go to the list, take off the shirt and wear it again, then save (not save as) the outfit to fix the issue. ()y

Mesh has the advantage of manual control of LOD losses. Depending on how you imported it, a mesh could be 0.50LI or 50LI. Still... Mesh doesn't "replace" anything. Its efficiencies and advantages will make it a preferred building standard in due time, but, all the other techniques will still be there to take advantage of. () One example linked here. Those are two windows. Two separate windows which can be placed anywhere independant of each other. Each window has two faces. One for the frame and one for the glass. Already, I'm talking about something that would require 4 prims if using sculpts and prims. In mesh, when linked, they're 1 prim(LI). I would never make a window any other way, taking that into account. The frame won't alpha clip with things. The windows can change texture via script or whatever. And, they're not locked to any relative position to each other. () Also, I noticed "Without doing any adjustment work in Blender it came in at Land Impact 41 with an upload cost of L$51 - way too expensive from all aspects." Without doing any adjustments!?!?!?!? That's crazy. All those invisible line vertices necessary for the sculpt can be removed for the mesh. Why upload an intentionally wasteful mesh like that? ()