[Bf-committers] Sculpt Mask Data (lazy initialize?)

Nicholas Bishop nicholasbishop at gmail.com
Sat Sep 22 04:00:59 CEST 2012


Yes, it would only move the delay. I figured entering sculpt mode
would be a better place for that delay, but if the risks of making a
change close to release are not deemed too great I don't mind it being
changed.

-Nicholas

On Fri, Sep 21, 2012 at 5:27 PM, Campbell Barton <ideasman42 at gmail.com> wrote:
> @Nicholas, this happens when entering sculpt mode so there would be
> some slowdown there - I assume, this would just move the delay to the
> first time someone uses the mask brush, it would only be once - since
> this is a very light data layer I doubt it would be noticeable
> slowdown. - compared to undo which does far more heavy stuff for eg.
>
> @Sergey, I wouldn't want to wait until after release with this -
> mainly because it adds data to peoples meshes they probably won't use
> its not horrible but not good practice IMHO - unless the change looks
> like its error-prone, Id like to change the behavior before release.
>
> On Fri, Sep 21, 2012 at 11:48 PM, Sergey Sharybin <sergey.vfx at gmail.com> wrote:
>> I think one more thing:
>>
>> 3) GPU_draw_buffers seems to be also assuming sculpt mask always exists.
>>
>> We could switch to lazy initialization, but i wouldn't want this to happen
>> at this point -- think this could wait for after release.
>>
>> On Fri, Sep 21, 2012 at 7:29 PM, Nicholas Bishop
>> <nicholasbishop at gmail.com>wrote:
>>
>>> I don't have any strong objections, but a couple of things to look at:
>>>
>>> 1) Should check how big of a delay this introduces for the first use
>>> of the mask brush on a dense mesh. Don't want there to be a big delay
>>> that makes masking a second-class citizen among brushes :)
>>>
>>> 2) Implementation issue: the mask code for sculpt makes the assumption
>>> that a mask layer always exists, so if we change that we should
>>> carefully check for problems in both regular and multires sculpting.
>>>
>>> -Nicholas
>>>
>>> On Fri, Sep 21, 2012 at 4:01 AM, Campbell Barton <ideasman42 at gmail.com>
>>> wrote:
>>> > This mail is mostly for Nicholas Bishop but others may have input,
>>> >
>>> > The other day I noticed the default cube had a mask vertex layer -
>>> > this isnt so bad - only storing one extra float per vertex - but seems
>>> > unnecessary.
>>> >
>>> > There are some other data layers that the user had no way to remove
>>> > (skin and sticky), now theres a panel "Geometry Data" - where layers
>>> > can be cleared,
>>> > (note - unlike UVs and VColors - these layer types are singletons - a
>>> > mesh can only have one of each - so the UI is different).
>>> >
>>> > ... back to the point of the mail...
>>> >
>>> > At the moment just entering sculpt mode will add sculpt-mask data
>>> > layer to the mesh verts, and be saved, and get stored in every undo
>>> > step... etc.
>>> >
>>> > Id prefer this was lazy initialized - so only when using the mask
>>> > brush it gets added, any reason not to do this?
>>> >
>>> > --
>>> > - Campbell
>>> > _______________________________________________
>>> > Bf-committers mailing list
>>> > Bf-committers at blender.org
>>> > http://lists.blender.org/mailman/listinfo/bf-committers
>>> _______________________________________________
>>> Bf-committers mailing list
>>> Bf-committers at blender.org
>>> http://lists.blender.org/mailman/listinfo/bf-committers
>>>
>>
>>
>>
>> --
>> With best regards, Sergey Sharybin
>> _______________________________________________
>> Bf-committers mailing list
>> Bf-committers at blender.org
>> http://lists.blender.org/mailman/listinfo/bf-committers
>
>
>
> --
> - Campbell
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> http://lists.blender.org/mailman/listinfo/bf-committers


More information about the Bf-committers mailing list