Remove server+client flags from independent standalone execs

[j-rivero]

🦟 Bug fix

Summary

The PR removes from the standalone executables the parameters (--render-engine and --render-engine-api-backend) that affects the server and client. This parameter makes sense to be used when the gz sim combined utility is being run to affect both part of the application. The standalone executables ship the customized versions of these --render-engine-server and --render-engine-gui.

There is still the open issue of how should behave the standalone executables with respect of the input of sdf files. The server can process them as the input, this is the more natural way. In the combined run (server + client) the server is waiting for the client to send a SDF file from the initial selection screen if no file is passed. For the standalone client I'm inclined to remove the input of the sdf file and just leave it to connect to the server but I can probably be convinced of other options.

Checklist

• Signed all commits for DCO
• Added tests
• Updated documentation (as needed)
• Updated migration guide (as needed)
• Consider updating Python bindings (if the library has them)
• codecheck passed (See contributing)
• All tests passed (See test coverage)
• While waiting for a review on your PR, please help review another open pull request to support the maintainers
• Was GenAI used to generate this PR? If so, make sure to add "Generated-by" to your commits. (See this policy for more info.)

[j-rivero]

There is a failure in a test:

2025-07-22T19:50:59.5759620Z (2025-07-22 19:50:50.997) [info] [TriggeredPublisher.cc:742] Service call [/srv-test-1] succeeded
2025-07-22T19:50:59.5759813Z /github/workspace/test/integration/triggered_publisher.cc:799: Failure
2025-07-22T19:50:59.5759904Z Expected equality of these values:
2025-07-22T19:50:59.5759999Z   static_cast<std::size_t>(pubCount)
2025-07-22T19:50:59.5760174Z     Which is: 10
2025-07-22T19:50:59.5760252Z   recvMsgs0.size()
2025-07-22T19:50:59.5760322Z     Which is: 9
2025-07-22T19:50:59.5760505Z /github/workspace/test/integration/triggered_publisher.cc:800: Failure
2025-07-22T19:50:59.5760590Z Expected equality of these values:
2025-07-22T19:50:59.5760681Z   static_cast<std::size_t>(pubCount)
2025-07-22T19:50:59.5760745Z     Which is: 10
2025-07-22T19:50:59.5760818Z   recvMsgs1.size()
2025-07-22T19:50:59.5760882Z     Which is: 9
2025-07-22T19:50:59.5761068Z /github/workspace/test/integration/triggered_publisher.cc:803: Failure
2025-07-22T19:50:59.5761221Z Expected equality of these values:
2025-07-22T19:50:59.5761290Z   pubCount
2025-07-22T19:50:59.5761360Z     Which is: 10
2025-07-22T19:50:59.5761485Z   std::count(recvMsgs0.begin(), recvMsgs0.end(), true)
2025-07-22T19:50:59.5761550Z     Which is: 9
2025-07-22T19:50:59.5761725Z /github/workspace/test/integration/triggered_publisher.cc:804: Failure
2025-07-22T19:50:59.5761811Z Expected equality of these values:
2025-07-22T19:50:59.5761878Z   pubCount
2025-07-22T19:50:59.5761943Z     Which is: 10
2025-07-22T19:50:59.5762059Z   std::count(recvMsgs1.begin(), recvMsgs1.end(), true)
2025-07-22T19:50:59.5762124Z     Which is: 9
2025-07-22T19:50:59.5762337Z (2025-07-22 19:50:56.205) [debug] [SimulationRunner.cc:577] Exiting postupdate worker thread (0)
2025-07-22T19:50:59.5762533Z [  FAILED  ] TriggeredPublisherTest.MultipleServiceForOneInput (6083 ms)

[azeey]

Makes sense to keep the flags that affect both the GUI and server only in the combined executable.

[azeey]

Homebrew failures don't seem to be related to this PR. These tests have started failing in other PRs as well (https://build.osrfoundation.org/job/gz_sim-ci-pr_any-homebrew-amd64/1962) from #3005.