[libsersi] Add new library

[crhowell3]

• Changes comply with the maintainer guide.
• The name of the port matches an existing name for this component on https://repology.org/ if possible, and/or is strongly associated with that component on search engines.
• Optional dependencies are resolved in exactly one way. For example, if the component is built with CMake, all find_package calls are REQUIRED, are satisfied by vcpkg.json's declared dependencies, or disabled with CMAKE_DISABLE_FIND_PACKAGE_Xxx.
• The versioning scheme in vcpkg.json matches what upstream says.
• The license declaration in vcpkg.json matches what upstream says.
• The installed as the "copyright" file matches what upstream says.
• The source code of the component installed comes from an authoritative source.
• The generated "usage text" is accurate. See adding-usage for context.
• The version database is fixed by rerunning ./vcpkg x-add-version --all and committing the result.
• Only one version is in the new port's versions file.
• Only one version is added to each modified port's versions file.

This PR introduces the libsersi package, a C++ implementation of the IEEE 1278.1a-1998 Distributed Interactive Simulation standard.

[crhowell3]

@microsoft-github-policy-service agree

[LilyWangLL]

The CI pipeline error could be fixed by run command .\vcpkg x-add-version opendis6 --overwrite-version.

[crhowell3]

The CI pipeline error could be fixed by run command .\vcpkg x-add-version opendis6 --overwrite-version.

I ran this command and updated the version database in my latest commit, but the microsoft.vcpkg.pr (x86_windows) CI job is still failing. Is there something else I need to do?

[LilyWangLL]

Usage test passed on x64-windows.

[BillyONeal]

I'm a bit nervous about the name, as searching for opendis points to https://github.com/open-dis/open-dis-cpp rather than this library. The name 'opendis6' points to this repo on repology, but only from Conan and the author of the Conan recipe is also @crhowell3 . I'm torn between saying 'it's in repology, it's already out there' and asking for crhowell3-opendis6 instead.

Tagging vcpkg-team-review for a 2nd opinion

[crhowell3]

@BillyONeal Maybe this is too drastic of a suggestion, but what if I were to just change the name of my package to avoid confusion with the other opendis library? I could name it something like dis-cpp or disprotocol, maybe even just dis6. Thoughts? Alternative name suggestions?

[BillyONeal]

@BillyONeal Maybe this is too drastic of a suggestion, but what if I were to just change the name of my package to avoid confusion with the other opendis library? I could name it something like dis-cpp or disprotocol, maybe even just dis6. Thoughts? Alternative name suggestions?

The problem is that there's nonzero potential confusion with https://github.com/open-dis . (That is, we want someone who does vcpkg install opendis6 to not be surprised when they get something not owned by that org)

I think odds are pretty high we take the name as written here, we just want more than one maintainer to OK any time anything is worth questioning because once we 'give away' a name like that, we can't take it back. Whereas we can always rename crhowell3-opendis6 to opendis6 should the library in question become the unambiguous defacto meaning of that name in the future.

[crhowell3]

Looks like the pipeline is getting an error stating that the package directory is nonexistent. I pushed a fix to my upstream main branch, but I was only able to confirm it was resolved locally using vcpkg install libdis6 --head. Just using vcpkg install libdis6 does not work. Is there something I need to do to correct that?

Fixed in the latest commits

[JavierMatosD]

@data-queue, @vicroms, @AugP, and below discussed this today.

We don't count trailing numbers as a distinguisher between ports. For example, we expect libfoo6 to be related to libfoo5 and libfoo. Additionally, a lib prefix isn't considered a distinguisher between ports. For example, we don't want libpng to be different that png. Therefore, this name is taking the stem dis.

@BillyONeal - Still concerned about the name being confused with the open-dis project. A user cannot type the name you picked and think they got open-dis and dis or even libdis6 isn't sufficiently different enough to disambiguate between projects. I'm weakly against this. I still propose <org>-name

@ras0219-msft - For example, we wouldn't accept the name libecmascript6. dis is less known than ecmascript but using the name of the standard as the name of the library is problematic unless blessed by owners of the standard.

Given the discussion above, we would still like a more unique name.

[crhowell3]

I understand. I'll try to think of a package name that does not incorporate the dis stem so as to avoid confusion with the open-dis library. If I can't think of anything, then I'll just go with <org>-name as you all have proposed. Thank you for taking the time to review and comment.

[crhowell3]

@JavierMatosD, @data-queue, @vicroms, @AugP and whomever else may be involved:

Here is my final name proposition: libsersi (from serialize and simulation). If this is sufficient, I'll change it upstream. If not, then I am fine moving forward with the <org>-name suggestion. In either case, thank you for being patient with me and for being as helpful as you have been.

[JavierMatosD]

@JavierMatosD, @data-queue, @vicroms, @AugP and whomever else may be involved:

Here is my final name proposition: libsersi (from _ser_ialize and _si_mulation). If this is sufficient, I'll change it upstream. If not, then I am fine moving forward with the <org>-name suggestion. In either case, thank you for being patient with me and for being as helpful as you have been.

@crhowell3, I think libsersi is good. Thank you for your patience and understanding. Make sure to mark "ready for review" when you've applied the changes.