# Duplicate GetGroup/GetBinding in vtkWebGPUComputeBuffer.h
**URL:** https://discourse.vtk.org/t/duplicate-getgroup-getbinding-in-vtkwebgpucomputebuffer-h/16467
**Category:** Development
**Tags:** webgpu
**Created:** [June 19, 2026, 4:55pm UTC](https://discourse.vtk.org/t/duplicate-getgroup-getbinding-in-vtkwebgpucomputebuffer-h/16467 "2026-06-19T16:55:51Z")
**Posts on this page:** 10
**Page:** 1
### Author:  [@Bengt](https://discourse.vtk.org/u/Bengt)
#### Post date: [June 19, 2026, 4:55pm UTC](https://discourse.vtk.org/t/duplicate-getgroup-getbinding-in-vtkwebgpucomputebuffer-h/16467/1 "2026-06-19T16:55:51Z")
Building VTK **9.6.2** with WebGPU enabled on **Windows / MSVC** fails to compile `Rendering/WebGPU/vtkWebGPUComputeBuffer.h`.
`Group` and `Binding` are declared as `vtkIdType` members but each accessor is defined twice, once hand-written returning `uint32_t`, and once via `vtkGetMacro` returning `vtkIdType`:
```cpp
// lines 91-92
uint32_t GetGroup() const { return this->Group; }
vtkGetMacro(Group, vtkIdType); // expands to vtkIdType GetGroup() const
// lines 104-105
uint32_t GetBinding() const { return this->Binding; }
vtkGetMacro(Binding, vtkIdType); // expands to vtkIdType GetBinding() const
```
After macro expansion each is a redefinition of `GetGroup()` / `GetBinding()` differing only by return type (`uint32_t` vs `vtkIdType`), which is ill-formed:
```auto
vtkWebGPUComputeBuffer.h(92): error C2556:
'vtkIdType vtkWebGPUComputeBuffer::GetGroup(void) const': overloaded function differs only by return type from 'uint32_t vtkWebGPUComputeBuffer::GetGroup(void) const'
vtkWebGPUComputeBuffer.h(92): error C2371: 'vtkWebGPUComputeBuffer::GetGroup': redefinition; different basic types
```
Are we doing something wrong? Cause that must have gone through CI/CD at some point.
Thanks!
---
### Author:  [@Bengt](https://discourse.vtk.org/u/Bengt)
#### Post date: [June 19, 2026, 5:28pm UTC](https://discourse.vtk.org/t/duplicate-getgroup-getbinding-in-vtkwebgpucomputebuffer-h/16467/2 "2026-06-19T17:28:13Z")
We discovered that VTK\_USE\_FUTURE\_CONST=ON triggers it. Without it, vtkGetMacro expands to a non-const getter which doesn’t interfere with the const ones. We can disable it for now but this is a potential bug.
---
### Author:  [@ben.boeckel](https://discourse.vtk.org/u/ben.boeckel)
#### Post date: [June 20, 2026, 12:11pm UTC](https://discourse.vtk.org/t/duplicate-getgroup-getbinding-in-vtkwebgpucomputebuffer-h/16467/3 "2026-06-20T12:11:25Z")
@jaswantp It looks like @tclabault wrote this originally, but we should investigate it. A method changing return type based on `const` is confusing anyways.
---
### Author:  [@jaswantp](https://discourse.vtk.org/u/jaswantp)
#### Post date: [June 20, 2026, 2:58pm UTC](https://discourse.vtk.org/t/duplicate-getgroup-getbinding-in-vtkwebgpucomputebuffer-h/16467/4 "2026-06-20T14:58:02Z")
we should remove the overload that returns `uint32_t`.
---
### Author:  [@Bengt](https://discourse.vtk.org/u/Bengt)
#### Post date: [June 20, 2026, 3:49pm UTC](https://discourse.vtk.org/t/duplicate-getgroup-getbinding-in-vtkwebgpucomputebuffer-h/16467/5 "2026-06-20T15:49:11Z")
Please note that this happens in other WebGPU headers too with the same problem. I forgot which ones but there were at least two others.
---
### Author:  [@mwestphal](https://discourse.vtk.org/u/mwestphal)
#### Post date: [June 22, 2026, 10:04am UTC](https://discourse.vtk.org/t/duplicate-getgroup-getbinding-in-vtkwebgpucomputebuffer-h/16467/6 "2026-06-22T10:04:01Z")
> [@Bengt](#):
>
> VTK\_USE\_FUTURE\_CONST=ON
I dont think VTK\_USE\_FUTURE\_CONST=ON is reader for prime time yet, it is only tested independantly by @seanm , but maybe will be turned on in some CI in the coming months.
---
### Author:  [@Bengt](https://discourse.vtk.org/u/Bengt)
#### Post date: [June 22, 2026, 1:22pm UTC](https://discourse.vtk.org/t/duplicate-getgroup-getbinding-in-vtkwebgpucomputebuffer-h/16467/7 "2026-06-22T13:22:16Z")
Thanks @mwestphal , sounds good.
That said, VTK\_USE\_FUTURE\_BOOL could use the same treatment as some vtkSetClampMacro which are now bool are trying to force int clamping values resulting in a compile time warning (or error if treated as such).
---
### Author:  [@jaswantp](https://discourse.vtk.org/u/jaswantp)
#### Post date: [June 22, 2026, 8:20pm UTC](https://discourse.vtk.org/t/duplicate-getgroup-getbinding-in-vtkwebgpucomputebuffer-h/16467/8 "2026-06-22T20:20:40Z")
should the `uint32_t` getters be deprecated? or removed?
note that deprecating them will not address your issue unless you set `VTK_LEGACY_REMOVE` which may have side effects.
---
### Author:  [@seanm](https://discourse.vtk.org/u/seanm)
#### Post date: [June 22, 2026, 8:50pm UTC](https://discourse.vtk.org/t/duplicate-getgroup-getbinding-in-vtkwebgpucomputebuffer-h/16467/9 "2026-06-22T20:50:18Z")
> [@mwestphal](#):
>
> I dont think VTK\_USE\_FUTURE\_CONST=ON is reader for prime time yet, it is only tested independantly by @seanm , but maybe will be turned on in some CI in the coming months.
Well, I only use a subset of VTK of course, but we’ve been shipping our app built with VTK\_USE\_FUTURE\_CONST for years. It’s ready for prime time in that sense.
---
### Author:  [@ben.boeckel](https://discourse.vtk.org/u/ben.boeckel)
#### Post date: [June 23, 2026, 11:13am UTC](https://discourse.vtk.org/t/duplicate-getgroup-getbinding-in-vtkwebgpucomputebuffer-h/16467/10 "2026-06-23T11:13:38Z")
Even if it isn’t ready, this is a bug regardless of the setting. It’s just that the setting actually uncovered an issue in this case.