Babbage finally blogs about script limits

*laughs* It IS a very good thing, you cute nerd you.

But then, it was always POSSIBLE to write resize scripts that didn't kill sims. People just didn't know how to, and/or didn't take the time to do it.

I'm not talking about what the right technical solution is. I'm talking about a mechanism to get a decent technical solution widely distributed, quickly, with enough publicity that legacy products are taken off the market more quickly.
.

Right -- and with script memory limits in place, your customers won't be able to put on their shoes. Your product won't do well in the marketplace. Yay!

/me smiles and blushes.

At this point, it's probably not worth someone's effort to do, since things are changing, and because it takes quite a while between when a solution is available and when creators get it into most of their products.

Ask a creator of a large line of products how much fun it is to do a script udate for the whole line. Unless that update affects their bottom line, most won't bother. They might start using a new one on new products.

One problem, though, is that the best solutions using today's technology have drawbacks compared to the laggy ones. (For example, taking 20 seconds to resize rather than changing almost instantaneously.)

Now, if this had happened a year or two ago, it would have been great. But I think it's too late at this point. That's JMHO, of course. Feel free to have a contest! I'll even donate a prize (e.g., a fun rug or HUD).

Well, it depends.

Do we know what order things are attached in? What if I make the hat, say, 85% of the limit?

I'd guess that that would give me a good shot at adding load to the sim, just because I (or, rather, the hypothetical sucky dweeb I'm pretending to be in this thread) can and preventing the user from wearing OTHER things that are actually well-behaved.

When they introduce limits, are they gonna give people decent feedback or will it be just "over the limit! denied!!"? Based on the way stuff in SL tends to work, I'd guess we will get the later.

That's a good question. It's clear that property owners will be able to see usage, similar to prim counts today.

I believe it's also important for avatars to be able to see their allocations. I don't know whether it's been specifically addressed. We can't currently see our av prim usage as a total; we have to select each attachment and look.

It sure would be nice to be able to see memory usage along with prim count when editing an object, but I worry about how feasible that would be. Script memory usage changes much more easily than prim count.

You're right, a contest isn't really worth it given the solution that's coming along relatively quickly. It will be fairly simple to create a decent script.

I still think that a kit pushed out to creators is a good idea, as soon as the new commands become available. It would be good if the script came from someone credible, though. Someone with a little bit of a name for making useful open source scripts.

*Looks up, looks down, scratches 'Lear' in the dust with her toe*

I'll find the sample hair and boots if you make the scripts. Deal?

It really is not that important. The average SLer is clueless about the technical side of SL and barely has a grasp on adjusting attachments let alone worry about prim count and script times of their avatar.

They are just going to have to make some hard scripting limits for avatar attachments so that if the attachment is over the script time limit, it simply gives an error and does not work when attached.

There is no way you are going to take SLers who don't even know how to put pictures on a rezzed cube and teach them about script times and avatar prims. Not going to happen. It is a shock these days to run into anyone from 2009 that even knows how to build.

Hard limits are the only way. However, i think the tools to do as you say would be a nice function anyways for those of us who do care and are capable of utilizing SL in a more technically adept manner than the average user. Even a pull down menu that said "Avatar Script Time" would be nice and then it could display the individual script time of your items - which would be extremely informative and helpful. I just don't think it will be helpful to the greater majority.

Are alts created in 2009 allowed in that population sample?

I would *LOVE* to know the stats on a script I was using or creating. For example, I always wear a tail which uses a wag script, but I have no idea how efficient the script is (especially since the same script also talks to a colour-change HUD).

Sure, if you'll promise to handle all the support calls. I've found that no good deed goes unpunished. No doubt Miffy F (author of original MLP) snickered like hell when he saw what I was up to!

memory limits will help LL plan how to scale and how many sims can run on one host but it won't do a thing for crappy scripts gobbling up frame time. And that is what the "bad" resizer scripts do. I use a single script solution when pressed to include a version of shoes with resizer scripts. Barely any time used. And they are mod so you can delete the scripts. But wait! I also include the versions without scripts to save you the trouble of having to delete them.

Consider a vote for http://jira.secondlife.com/browse/SVC-5252 to help keep estates free of script time resource over (ab)users.

Lindens also insisted that the change-over from the pre-Windlight to Windlight-capable renderer would at worst result in the same FPS. It took many months before at least one Linden would admit that the Windlight renderer performs worse under the same conditions as the old one.

Looking at Babbage's office hours does not inspire confidence either:
A whole lot is going to break if the limit is 100 or less per avie.

Their decision to simply stop supporting legacy LSL scripts at some future point will break a whole lot more still, least of all older content that can never be replaced or redone (ie Starax wand).

LL *is* at fault when it comes to crippling the ability to "descript" no modify linksets containing "copy" scripts.

Nice. Another global policy being made by a tiny set of office hour attendees..

I think once he fixes the sim stalls, the calls for script limits will drop and they'll drop a lot. They'll likely plow ahead and break lots of stuff anyway.

Stop looking at this from a technical standpoint. LL have to present a very complex issue in a very very simply way to the end users, so what wwe can expect will be

512 plot ! Perfect first home. 117 prims, 32 scripts !!

(and that 32 INCLUDES your scripted ass and boots and boobs and collar and cuffs and peen)

Ok, Now you see why this is going to suck, and by suck I mean [out of memory, your hair has been removed]

tiny empires hud has ~30 scripts, at least 20 of which are absolutely necessary to displaying hovertext.... and that's just one hud.... a few xcite items and you'd probably double the limit pretty quick.

even email moderated vending servers need at least 2 extra scripts just to pass info and process data in a reasonable time frame, and still be well under the imposed 500/hour.

and 3am PST? on a wednesday? ahahahaha that's not even lunch in GMT and before most alarms go off in the morning for N/S america...

If they implement all the new LSL calls we asked for at his office hours, there won't be any more need for separate hovertext scripts, or color/texture-changing scripts, or prim-move scripts, or whatever. I think script counts will plummet down to around 10-25% of its current amount.

Of course, there's the issue of legacy support, but meh. At some point, they are going to have to cut legacy; it's just not feasible to continue supporting things all the way back to the beginning of time, especially when doing so holds us back from getting the things we really need (llTeleport*?).



Yeah, I have to stay up all night to catch it. I guess I won't be sleeping much tonight. :-/

It's human nature that people don't consider resources until faced with a shortage. People will quickly adapt to the limits; the newly scripted products will make better use of resources. It doesn't seem like at all a bad thing to me.

the only really useful ones I've heard confirmed are get link params, and undelayed set params... I could think of ~50 functions that should be there, but how long till they actually implement is the problem... sadly those are just functions that would fill out the basic get/set/link

I thought the Starax wand was already broken.

It's kind of bad if stuff you own now breaks. Hopefully people will provide updates for the content that's out there now, but I'm not holding my breath on that.

Buh-bye scripted no-mod combat boots.

A bit of content will break, but the world will function so much better that it seems an easy trade.

now there's an exercise in understatement!

We also brought up llSetLinkText (as well as llGetLinkText) and I mentioned llSetLinkTextureAnim, llLinkParticleSystem, and llLinkTargetOmega. I think those (along with llSetLinkPrimParamsNoDelay and llGetLinkPrimParams) pretty much cover the vast majority of prim-related settings that people use extra scripts to accomplish.

Though, I could also see the sound functions added, but it is rare anyone bothers using multiple scripts to accomplish multiple sounds.

I just wish we could get some firm information about dates. I've got several projects on hold, including a major and badly-needed update for a customer, because I want to use llGetLinkPrimitiveParams() and llSetLinkPrimitiveParamsNoDelay() and I really don't want to do the work -- especially the update -- knowing that I'll have to redo it all again a month later.

As I understand it, the implementation is contingent on Viewer 2 or whatever it's called being released. I know it's supposed to be out sometime this quarter, just like it was supposed to be out sometime last year (and Blondin Linden was still saying in November he'd been assured it would be out in beta before the end of December). But I'll feel a lot happier when we actually see a release candidate version at some point.

Before that stupid mono-rez bug appeared, I thought things in SL were the best they had ever been.

Babbage never responded to the blog comments that asked about places where big chunks of regions were Protected land or not being used and I would very much like to see script limits get initially set at, say, 50% over what they think their worst-case numbers should be.

Don't hold your breath though .

This is very similar to what they recently posted:
Except it was posted January 2009.

Before that script limits and monitoring tools were supposed to be available to coincide with the release of the "homestead" product which was January 2009.