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
On the version of rye I published there is currently no support for local Python installations. This is pretty much intentional because I really want the experience to not involve manually downloading Pythons. However there are also situations where this is necessary. In the original implementation I had, the way non automatically managed Python installations worked was with registering them.
Link a Python:
ln -s /path/to/python ~/.rye/py/my-dst@3.11.0
And to pick a default distribution for what gets picked up when you write 3.11.0 another symlink does:
ln -s ~/.rye/py/my-dist@3.11.0 ~/.rye/py/3.11.0
I however never really fleshed out the details and it was pretty brittle. In particular the only way I found to reliably pin a project to pypy for instance was to write pypy@madeupversion into .python-version which obviously is not very portable.
Remaining tasks:
enable rye to work with a local toolchain entirely (eg: make the self venv work with a supplied one)
find a good way to pin to specific installations without throwing stuff into the .python-version file that other toolchains cannot work with
The text was updated successfully, but these errors were encountered:
On the version of rye I published there is currently no support for local Python installations. This is pretty much intentional because I really want the experience to not involve manually downloading Pythons. However there are also situations where this is necessary. In the original implementation I had, the way non automatically managed Python installations worked was with registering them.
Link a Python:
And to pick a default distribution for what gets picked up when you write
3.11.0
another symlink does:I however never really fleshed out the details and it was pretty brittle. In particular the only way I found to reliably pin a project to pypy for instance was to write
pypy@madeupversion
into.python-version
which obviously is not very portable.Remaining tasks:
self
venv work with a supplied one).python-version
file that other toolchains cannot work withThe text was updated successfully, but these errors were encountered: