STOP 'fixing' SL - please!

Whilst I do appreciate that Phil can do what he wants pretty much, and does make some good calls... if you read the thread, and consider the title and its punctuation, it's not actually asking for SL to go into stasis.

A fair number of people are seeing their client fps *fall* update by update. That's called adversely affecting your user base, rather than improving the experience of SL. Take local lighting - on a really top end machine it used to be useable, suddenly in one update it slows machines that used to run with it on at 20+ fps at 4 fps in the same place with the same stuff around.

The known bugs list is also growing, growing, growing. Some of that may be more people, so doing different things, so more are being discovered. But a lot of them are bugs that are known and accepted and have been around for a long, long time.

The thread isn't asking them to stop introducing new things into SL, it's mostly about changing the emphasis to some extent so that the known bugs list goes down too, and to stop and consider the things that make things documentably *worse* for a number of users.

In fairness they've stopped and listened to the feedback about 1.9.1 that is saying precisely "you're killing my frame rate" and "this is breaking X product".

The grey goo fence that was introduced on Wednesday is being shown to hit a number of legitimate products - and there is a dialog about how it can be fixed that is having regular input from the person that introduced it so we can get a sensible balance between allowing the legitimate products that have behaviour that appears similar to a grey goo and yet still offering extra protection to the grid, which most of us agree is also desirable.

The various interations with ripple water - here's something cool and optional. Ah, it hits fps, we'll tweak, the tweak looks horrid, we'll move back to something a bit heavier on fps, but much better looking is another fine example of the sorts of things that can be done - the end point was a system that just about everyone uses because it looks great and although it does impact fps a bit for everyone, it's so good it's worth it and if you really need the fps for something you can always turn it off!

That's the sort of thing this thread was asking for, and LL have responded to it internally it would appear, even though they've not directly come out and said so.

I'd gladly give back most of the changes introduced in the newer versions for a downgrade to... lets say 1.5

Before things started breaking and getting really dodgy.

1.6 introduced the now famous grey boxes of doom

1.7 Not sure what they did here, but it was probably bad

1.8.1 FPS on the preview grid was good until about build 25, then it went into the sh*tter!


I'd still give it all back for less lag and more stability.

LL needs to quit with the "Oh, thats cool" features and work more on the showstoppers. Soft prims? Yay, another way to lag things to hell. I was playing with this in the preview grid, jesus h. christ it causes the fps to tank! It seems like, say, when a cube goes soft, it adds prims to simulate the softness. These dont add to your count, but they do add to your lag.

Lighting system? Ok, this is a Good Idea©, running lighting on hardware. Good LL, have a donut.

Occlusion Culling? Also good. FPS raise is always a Good Thing© in sl. Join me with the rest of the residents in saying "About f*cking time!"


Cmon guys. Innovation is a good thing, but please, PLEASE , PLEASE fix your bugs first. We'd all love you for it.

1.9.1

Polygons, not prims; the cube is the prim; polygons make up the prim/cube. If course flex prims add polygons--but it would be nice to be able to lock the LOD to reduce how many polygons a flex prim uses.

I agree.