Closes#858.
If a project is named `default.project.json`, it acts as an `init` file
and gains the name of the folder it's inside of. If it is named
something other than `default.project.json`, it gains the name of the
file with `.project.json` trimmed off. So e.g. `foo.project.json`
becomes `foo`.
When enabled, the `baseurl` of the session is written to
`workspace:SetAttribute("__Rojo_ConnectionUrl")` so that the test server
can connect to that session automatically.
This works for Play Solo and Local Test Server. It is marked
experimental for now (and disabled by default) since connecting during a
playtest session is... not polished. Rojo may overwrite things and cause
headaches. Further work can be done later.
If the sync lock is claimed in Team Create, the user cannot sync.
Therefore, a sync reminder notification is unhelpful as it is calling to
an invalid action.
Windows and macOS runners consume GitHub Actions minutes at [2x and 10x
the rate of Linux runners,
respectively](https://docs.github.com/en/billing/managing-billing-for-github-actions/about-billing-for-github-actions#minute-multipliers).
This is a bit concerning now that there are two Windows jobs and two
macOS jobs, introduced in #825.
This PR aims to reduce the cost by:
* Adding [rust-cache](https://github.com/Swatinem/rust-cache/) to reduce
the amount of time spent. I'm aware there were some concerns raised
about CI caches in general in #496 - are they still a blocker?
* Removing the unnecessary Windows and macOS MSRV build jobs. If an MSRV
build fails on one platform due to usage of new language features, then
it will fail on all of them.
@Kampfkarren may have to change this repository's required status checks
before this PR can be merged
This is a fairly important test verifying whether the action of moving a
folder into a watched folder is correctly detected and processed. It was
disabled in
b43b45be8f.
The fact that it failed indicates a possible bug in change processing,
so in this PR, I'll re-enable the test, investigate why it fails, and
fix it.
This PR adds macOS and Windows jobs to the CI workflow. This allows us
to see when changes break functionality on any supported platform, which
is particularly important for changes that involve the file system or
file watcher.
Right now, serve tests will fail when Rojo is built with the FSEvent
backend. The cause is essentially due to the fact that `/var` (where
temporary directories for serve tests are located) on macOS is actually
a symlink to `/private/var`. Paths coming from FSEvent always have
symlinks expanded, but Rojo never expands symlinks. So, Rojo's paths
during these tests look like `/var/*` while the FSEvent paths look like
`/private/var/*`. When Rojo's change processor receives these events, it
considers them outside the project and does not apply any changes,
causing serve tests to time out.
To work around this, we can call `Path::canonicalize` before passing the
project path to `rojo serve` during serve tests. Rojo does need to
better support symlinks (which would also solve the problem), but I
think that can be left for another day because it's larger in scope and
I mostly just want working tests before addressing #609.
Brings over some changes to rbx_dom_lua to validate attribute names
before calling `Instance:SetAttribute`. This should prevent Rojo from
falling over when it attempts to sync an attribute with an invalid name.
These warnings always appear for properties like `Capabilities`,
`SourceAssetId`, etc. and tend to scare users who are syncing models.
This information is now surfaced in the patch visualizer, so I think
these warnings can be demoted to debug logs.
Due to the rewrite of the plugin's core sync loop and the change for the
Notify backend on MacOS, along with all the other changes in 7.4.0, it
makes sense for us to use a release candidate before actually cutting a
proper `7.4.0` release.
This modifies Rojo's build script to throw a fit if we're building a
plugin with a semver incompatible version. In the process, it moves the
version of the plugin to a file named `Version.txt` that's parsed at
runtime. This should be minimally invasive but it's technically worse
for performance than the hardcoded table and string we had before.
This feels better than a CI check or just manually verifying because it
makes it physically impossible for us to forget since Rojo won't build
with it being wrong.
