Skip to content
This repository has been archived by the owner on Mar 28, 2024. It is now read-only.

[BUG-41491] Please honor the request to cap attachment's render weight to 10 million instead of the current 1 million. #13002

Closed
1 task
sl-service-account opened this issue Mar 7, 2017 · 4 comments
Closed
1 task

[BUG-41491] Please honor the request to cap attachment's render weight to 10 million instead of the current 1 million. #13002

sl-service-account opened this issue Mar 7, 2017 · 4 comments
Labels
project:BUG

Comments

@sl-service-account
Copy link

sl-service-account commented Mar 7, 2017 •
edited by sl-issues-importer bot

Information

With viewer 5.0.2.324126 a change was made to address the request filed in BUG-41034 to limit an attachment's render weight to 10 million. Instead of capping at the requested 10 million, it was capped at 1 million instead.

https://bitbucket.org/lindenlab/viewer-bear/commits/32142916538281831ad19008ed3eb4f039557abf

This value is too low.

I sometimes see users with an ACI slightly over 1 million and turn the viewer's Maximum Complexity setting to 0 to investigate what attachments are causing this high ACI.

This is usually caused by wearing multiple complex mesh that make up a large avatar or by several linksets that make up a detailed dress or outfit made of sculpts or flexible prims.

The problem with a 1 million render weight cap per attachment is that a single worn sculpt lagger can now give its user the same ACI as a user that wears legit, but high ACI content.

The most basic sculpt laggers start off with a render weight around 2 million, but can get up to over 100 million.

The request for 10 million achieves two goals:

  • Prevent laggers on all attachment slots from signed integer flipping of the wearer's total ACI.
    This included enough breathing room for the possibility of an increase of attachments from 38 to 200.

  • To still make it possible to distinguish if an agent's total ACI is the result of wearing multiple complex, but legit content or of wearing a single sculpt lagger.

    With the current 1 million render weight cap, if I drop Maximum Complexity to investigate a slightly over 1 million reading I could be in for a crash.
    With the requested 10 million render weight cap, I would see a much higher total ACI value and would know better.

    Please honor the request to cap attachment's render weight to 10 million instead of the current 1 million.

    Thanks.

    Other Information

    Also, the few changes added in 5.0.2.324126 were only given 3 days, a weekend at that, before being promoted to release...

Links

Related

Original Jira Fields
Field Value
Issue BUG-41491
Summary Please honor the request to cap attachment's render weight to 10 million instead of the current 1 million.
Type New Feature Request
Priority Unset
Status Closed
Resolution Expected Behavior
Reporter Lucia Nightfire (lucia.nightfire)
Created at 2017-03-07T07:21:33Z
Updated at 2017-04-06T10:08:41Z 
{
  'Business Unit': ['Platform'],
  'Date of First Response': '2017-03-15T13:06:56.851-0500',
  'How would you like the feature to work?': "With viewer 5.0.2.324126 a change was made to address the request filed in BUG-41034 to limit an attachment's render weight to 10 million. Instead of capping at the requested 10 million, it was capped at 1 million instead.\r\n\r\nhttps://bitbucket.org/lindenlab/viewer-bear/commits/32142916538281831ad19008ed3eb4f039557abf\r\n\r\nThis value is too low.\r\n\r\nI sometimes see users with an ACI slightly over 1 million and turn the viewer's Maximum Complexity setting to 0 to investigate what attachments are causing this high ACI.\r\n\r\nThis is usually caused by wearing multiple complex mesh that make up a large avatar or by several linksets that make up a detailed dress or outfit made of sculpts or flexible prims.\r\n\r\nThe problem with a 1 million render weight cap per attachment is that a single worn sculpt lagger can now give its user the same ACI as a user that wears legit, but high ACI content.\r\n\r\nBasic sculpt laggers start off with a render weight around 2 million.\r\n\r\nThe request for 10 million achieves two goals:\r\n- Prevent laggers on all attachment slots from signed integer flipping of the wearer's total ACI.\r\nThis included enough breathing room for the possibility of an increase of attachments from 38 to 200.\r\n- To still make it possible to distinguishable if an agent's total ACI is the result of wearing multiple complex, but legit content or a single sculpt lagger.\r\n\r\nWith the current 1 million render weight cap, if I drop Maximum Complexity to investigate a slightly over 1 million reading I could be in for a crash.\r\nWith the requested 10 million render weight cap, I would see a much higher total ACI value and would know better.\r\n\r\nPlease honor the request to cap attachment's render weight to 10 million instead of the current 1 million.\r\n\r\nThanks.",
  'ReOpened Count': 0.0,
  'Severity': 'Unset',
  'Target Viewer Version': 'viewer-development',
  'Why is this feature important to you? How would it benefit the community?': '?',
}
@sl-service-account
Copy link
Author

Oz Linden commented at 2017-03-15T18:06:57Z

We tnink that allowing objects with that high a render weight is to dangerous to nearly all users and is not justified.

@sl-service-account
Copy link
Author

Lucia Nightfire commented at 2017-03-15T20:39:38Z, updated at 2017-03-15T20:49:27Z

Oz, I don't think you understand what the recent cap actually does.
I does not interfere with the default Max Avatar Complexity setting performance at all.

What it does is simply cap an attachment's render cost so there is no way it can flip the ACI signed integer max, but even when the ACI integer is flipped, jelly doll performance is not impacted and still successfully defends against legacy/sculpt objects.

The problem with the 1 million cap is that it will negatively affect users who want to run higher Avatar Complexity values slightly over 1 million.

Since a user wearing bunch of attachments whose total ACI is slightly over 1 million and a user wearing a single sculpt lagger will now have the same end ACI value, the latter scenario will now start having a negative impact on those users.

The 10 million ACI cap would still allow jelly doll performance to not be impacted and protect users that want to run higher Max Avatar Complexity values as well as users that want to run default or lower values.

The 1 million ACI cap will only protect users that run default or lower Max Avatar Complexity settings and put users that want to run higher values at risk.

As long as an active Max Avatar Complexity value is used and the value is under the ACI cap, an ACI cap of 1 million, 10 million, 100 million or 1 billion will not make any difference on jelly doll performance.

This will be even more the case when the Love Me Render deterrent is factored into the equation.

@sl-service-account
Copy link
Author

Lucia Nightfire commented at 2017-03-27T02:49:26Z

Additionally, I think this value needs its own debug setting. This way, the value isn't hardcoded and can be adusted as necessary depending on the user's gfx rendering capabilities or the frequency the user encounters high render cost legit content, not to mention flexibility with testing, but I still implore LL to set the default to the originally requested 10 million value as the current 1 million value puts users that run higher than default max ACI at risk.

@sl-service-account
Copy link
Author

Oz Linden commented at 2017-04-04T19:27:27Z

We've already responded to this a couple of times.

@sl-issues-importer sl-issues-importer bot mentioned this issue Feb 14, 2024
Open
2 tasks
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Assignees
No one assigned
Labels
project:BUG
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@sl-service-account