TF_MAKE_PP_{ARRAY,LIST,SEQ} and TF_PP_TUPLE_{SIZE,TO_LIST,TO_SEQ} are unused in pxr and introduce boost dependencies

[nvmkuruc]

Description of Issue

The following macros in preprocessorUtils.h are unused in pxr and introduce additional dependencies on BOOST_PP_EMPTY, BOOST_PP_TUPLE_TO_SEQ, and BOOST_PP_TUPLE_TO_LIST.

#define TF_PP_TUPLE_SIZE(tuple) \
#define TF_PP_TUPLE_TO_LIST(tuple) \
#define TF_PP_TUPLE_TO_SEQ(tuple) \
#define TF_MAKE_PP_ARRAY(...) \
#define TF_MAKE_PP_LIST(...) \
#define TF_MAKE_PP_SEQ(...) \

Steps to Reproduce

N/A

System Information (OS, Hardware)

N/A

Package Versions

N/A

Build Flags

N/A

[spiffmon]

The TF_MAKE_XXX macros are, in fact, unused in our entire codebase. The TF_PP_TUPLE_TO_XXX macros, however, are used in execution libraries that are likely to become part of OpenExec, so it's not immediately clear that moving them out of pxr now would be a win...

[tallytalwar]

Filed as internal issue #USD-7953

[nvmkuruc]

The TF_MAKE_XXX macros are, in fact, unused in our entire codebase. The TF_PP_TUPLE_TO_XXX macros, however, are used in execution libraries that are likely to become part of OpenExec, so it's not immediately clear that moving them out of pxr now would be a win...

It's definitely something that can be waited on until the other preprocessor usage is addressed, but I might argue that there would still be a benefit to having a boost-free USD even if OpenExec wasn't.

[spiffmon]

The TF_MAKE_XXX macros are, in fact, unused in our entire codebase. The TF_PP_TUPLE_TO_XXX macros, however, are used in execution libraries that are likely to become part of OpenExec, so it's not immediately clear that moving them out of pxr now would be a win...

It's definitely something that can be waited on until the other preprocessor usage is addressed, but I might argue that there would still be a benefit to having a boost-free USD even if OpenExec wasn't.

An excellent point, @nvmkuruc , given that OpenExec is a ways out.

[nvmkuruc]

TF_NUM_ARGS can be added to the list of unused macros with the integration of #2530

[nvmkuruc]

@jesschimein @spiffmon @tallytalwar I think this can be closed now with 24.05.