Wulfie Reanimator

Opening tickets is good. If it's related to another ticket (similar issue, more information), it will be linked to that ticket and closed. This is a completely normal way of keeping things organized, since all the info can be viewed from those "central" tickets by anyone who's trying to keep up. It's not useful to have 20 open tickets about the same thing, because then everyone would have to search for them individually, and conversation is split.

Look for marionette or maybe kimono / shrine maiden related AOs. Oracul (in a sim called Kuso) has a couple that could work.

Many merchants simply don't do it. It's a listing violation. I can find linked demos with a simple search for "demo" and price set to zero.

The "landing zone congestion" is a pretty widespread problem, and I'm a part of that problem. 😋 It's generally just much faster and easier to "explore" an area by detaching my camera from my avatar, rather than walking/running/flying around.

Learning how to make LODs is always good. The size or complexity of the model doesn't matter, since the process is the same. 🙂 I would not look at the render weight (or "Complexity") as any sort of indication of how "good" a model is. That algorithm uses some really bad shortcuts on top of being really arbitrary and easy to manipulate. (But just for reference, render weight is measured in the thousands (plural). Being under 1000 is a very very small value.)

It's super subjective. Check if there are other full sets like you're trying to sell. If not, consider prices on the individual items and multiply that by some amount. How many items are in the set? Consider how much money you spent in trying to get the set, but you can sell those other duplicates, so you don't need to recover any "losses" in your pricing. You can always lower the price if it's not selling fast enough.

If you go to imgur.com, you can upload images without an account. (They won't be visible to anyone without a link.)

They've done a little bit more than that recently. 😋

If you search for that product with quotes, you only get that product as a result. (Even if you just search for "hattention brat" with quotes) If you make a general search and sort by "Price: Low to High", of course you'll get only demos. Refining the search with "do not show demos" will correctly hide products that are marked as demos, though there are a lot of demos that aren't listed correctly (which is against the rules).

You don't need a URL. If you press the Print Screen button, you can right-click the text area while typing a new post and select 'paste'. The image will be automatically uploaded and inserted into your post.

No, these question-answer-type threads use a voting system (on the left, the little ^ button above the "score") and the intention is to have users vote up the best answers. Very few people here actually vote on answers. The ability to vote down answers has been disabled by LL.

We're both making a lot of assumptions about how things are done. You could attend one of those meetings and ask them directly. The meetings are held on a fixed schedule and they're open to everyone, the subjects they talk about will be whatever is relevant at the time or what residents bring up.

Nobody is talking. It's just you. (And I'm just here to prevent any future notifications since I got 9 already.)

There seem to be a couple overlapping things going on. First they made changes, which broke things. Then they fixed things. Then they made a blog post about the fixes. Then they made more changes which broke things again. So now it looks like their fix didn't actually do much, which is false. That meeting was held during a time when search was fixed. Nobody mentioned any issues on that day, or the day after that. Speaking generally about web development... and development in general... they have at least one "copy" of the Marketplace where new changes are tested before they're sent to the real Marketplace. Why they don't catch these problems, I don't know, maybe the people testing things don't have enough specifications for what kind of results should be expected. Fuzzy searching means that even if you search for "dess" or "derss", you'll still find dresses. It may also include related words like "gowns" as well (even if they aren't tagged as dresses).

They sure can.

To quote itself: Wonky LSL scripts are one great example of the 'statistical probabilities' at play. While it might not have enough information specifically about LSL scripts, there's a whole lot of similarities with other languages that it can adapt to LSL-style code it's been prompted to write. As long as it finds enough "data" to generate a response, it will, even if it makes no sense to us. (For this reason, a general "just use ChatGPT" solution won't work for SL. It would need to be trained on a specialized dataset about SL, but who knows if there's even enough information out there to get more useful responses. Not to mention how often it'd have to be retrained with more up-to-date information.) The same thing applies when you ask about people. If its dataset contains, say, accusations about a person (even if false), it might give those as a response. Like someone said in another thread, ChatGPT doesn't actually understand what is correct or true. Its job isn't to give truth or verify information, it just regurgitates whatever it has (which is still incredibly useful). Interestingly enough it does actually have some information about some individual people from Second Life. Probably those who have been mentioned in multiple blog posts or news articles. If you start a new chat and only ask about a forum member by name (without giving any context about Second Life), it'll either give a somewhat accurate description or "I'm sorry, but I don't have any information about that person." And since we know ChatGPT can learn from its conversations with the people using the OpenAI service, it's probably possible to feed it false information about things it doesn't know much about.

The main problem we have is that there's no official way for scripts to tell which outfit should be worn. RLV can kinda do it, but RLV isn't part of the official viewer and can't touch anything outside of the RLV folder.

Just to be clear, you can't really "arc" the ray, unless you were to cast and process dozens of individual rays. (Which isn't a very good idea.) You're gonna need the ray in addition to the sensor as Rolig suggested. Depending on the specific use-case, it might be better to ignore avatars during the raycast and just make sure there's no other objects in the way. This is generally a more consistent way to detect line of sight if the target tends to be moving. Depending on their velocity and script lag, the avatar might move out of the way before the ray is cast.

Did you try searching for unicorns?

To be fair, the wiki has a long list of supported features. It's a really good reference for what you can expect to work. https://wiki.secondlife.com/wiki/LlLinksetDataFindKeys (The wiki could use some example function calls, though... note to self.)

I can't tell if lazy quantifier is working, but negative lookahead doesn't work. Mode modifiers can't work since we can't specify the full regex expression (wrapped in slashes).

I wasn't gonna mention dark mode, but yeah. 😄

I think that's what we'd call a tantrum.

The problem is caused by two semi-transparent surfaces (your hair, and your applier tattoo) being drawn over each other in the wrong order, so in this case your hair was drawn first and the tattoo thinks it's being covered by the hair, even though it shouldn't be. Switching to BOM would absolutely help, since then the tattoo would actually become part of your skin texture (which isn't transparent). If your skin and tattoo don't have BOM versions, you could also change the attachment points of your body and hair (to change the priority in which they're drawn on the screen). Read more about it here:

Who says it has to appeal to anybody else? 🙂