You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In (Start to) handle cli hyperlinks in the terminal used for pkg dev tasks #5231, I've made positron the publisher of positron-r. That's not written in stone, but I'm quite sure it should not be vscode. For "normal" (meaning: not built-in) extensions, this really is a publisher in the marketplace sense and, when such a URI is activated, it leads to activation and/or installation of said extension.
If we offer a way for folks to launch Positron without any non-built-in extensions, similar to --disable-extensions, you could imagine the publisher field possibly being relevant.
Unknown unknowns: what happens if we switch all of "our" (for some definition of "our", see below) extension to have positron or posit as the publisher? Literally, what happens with our tests?
More questions:
What should the publisher be? positron? posit?
The Shiny-related extensions are published by Posit in both the VS Code Marketplace and in the OpenVSX Registry, so I suspect it would be confusing to use posit for extensions, like positron-r, that are built-in.
Which extensions should we change publisher for? The positron-* extensions seem like a clear yes, although positron-python is a special duck. Anything else?
Docs on publishing, which include a wee bit on the significance of the publisher field (though not much relevant to our case):
I'm talking about this, e.g.:
positron/extensions/positron-r/package.json
Line 6 in e842d83
A possibly incomplete list of technical implications:
An extension can register a URI handler, which is something I'm implementing for positron-r, starting with (Start to) handle cli hyperlinks in the terminal used for pkg dev tasks #5231. An extension can only address URIs addressed to it, which means they must start like so:
In (Start to) handle cli hyperlinks in the terminal used for pkg dev tasks #5231, I've made
positron
thepublisher
of positron-r. That's not written in stone, but I'm quite sure it should not bevscode
. For "normal" (meaning: not built-in) extensions, this really is a publisher in the marketplace sense and, when such a URI is activated, it leads to activation and/or installation of said extension.If we offer a way for folks to launch Positron without any non-built-in extensions, similar to
--disable-extensions
, you could imagine thepublisher
field possibly being relevant.Unknown unknowns: what happens if we switch all of "our" (for some definition of "our", see below) extension to have
positron
orposit
as the publisher? Literally, what happens with our tests?More questions:
positron
?posit
?Posit
in both the VS Code Marketplace and in the OpenVSX Registry, so I suspect it would be confusing to useposit
for extensions, like positron-r, that are built-in.publisher
for? The positron-* extensions seem like a clear yes, although positron-python is a special duck. Anything else?Docs on publishing, which include a wee bit on the significance of the
publisher
field (though not much relevant to our case):https://code.visualstudio.com/api/working-with-extensions/publishing-extension
The text was updated successfully, but these errors were encountered: