Mesh UserGroup meeting notes August 1st

[Charlar Linden]

Agenda and Meeting Notes: https://wiki.secondlife.com/wiki/Mesh/Archive/2011-08-01

Thanks everyone for participating. As with other meetings, we got a lot of good feedback and are investigating how to merge that with our overall priorities and strategy. 

 

We're seeing a lot of new faces, so we may need to get a little stricter until everyone is used to it. For instance, we stick to the agenda. It's important because otherwise people just try to yell their questions and it degenerates into madness. Also, I try to keep things moving so we can get to all the agenda items. Sometimes I might appear to be shutting things down, but it's unfair to take up a majority of time on one issue.

Also, if you attack or personally insult one of the Lindens - or any other participant - I will kick you. The team desires engaged, passionate, discussion, but you must act like everyone else present is not insane, stupid, or some other category. If you can say it with an insult, you can also say it withou and insult.

If you really think that about someone, mute them, or leave, and please IM me with your concerns.

I reserve the right to be a jerk.

 

Charlar

[Nalates Urriah]

Hear, hear!. Thanks.

You may want to consider a sign in the meeting location that simply restates the rules.

Of the meetings I'm attending I find the Mesh Upload Group the most intersting and well facilitated. Especially when one considers how empassioned some are about Mesh Costs.

[Tarius Auxarmes]

Looks like the sticking point for that meeting was the increase in cost for adding a script to something. Thats the one thing that I would like to see addressed as well.

The one question I did not see asked: What is the difference between a server loading a regular prim with a script and the server loading a mesh with a script? I think answering that question would go a long way. Is it possibly because you can not predict the complexity of a given mesh?

[HD Pomeray]

---

Tarius Auxarmes wrote:

Is it possibly because you can not predict the complexity of a given mesh?

---

I think that's exactly what they said in the meeting.  That they have no way to measure the the loading do to scripts.  Some of this is just simply that there is no way to know up front how often a given script is going to perform certain operations.  The other part is they have no tools at this point to analyze each script as it is loaded.   Either way... they are being cautious with the first roll out and we may see things moderate down as time goes on.  I know this is not a popular opinion but, personally, I'm good with that.  Why screw it up from the beginning.  

[Rage Riptide]

I read: "It's a BIG job" when they refered to analyzing scripts. That says to me it's on the table but I'm not always right. That sentence wasn't even correct.

[Tarius Auxarmes]

Well whatever the case, I think a scaling system would probably be better. Basically, the more a mesh costs, the more it would cost to add a script to it. Say something like a linear scale where adding a script to 1 PE mesh wouldnt cost anything up to like where something that costs 256 prims would be doubled by adding a script(to 512). This would mean that something that costs 128 would be increased by 50%(so 128 + 64) when a script is added. Something like this is what should be done rather than an across the board penalty. Heck, have a different sliding rule even, mabe exponential instead of linear, but either way, the more prims something cost, the more complex it is likely to be. I do realize they want to be cautious, but limiting things too much will also screw stuff up.