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
As a "reproducible document archive" is it in scope for DAR files to include an environment specification (packages, runtime versions, etc.), or is that expected to be left to other tools, such as pip requirements.txt and/or conda environment.yml as neighboring files, not in the archive itself? If so, how is that specified? For Jupyter, we consider this to belong as a sibling of the document, but I think the scope of Jupyter notebooks is meant to be a bit narrower than DAR.
The text was updated successfully, but these errors were encountered:
One idea was to only provide a runtime identifier and a version. And then there would need to be some kind of registry to resolve and download/install that runtime. This is probably very related to Binder and we're open to any solution that's does the job in a generic way. Like not locking into some concrete technology too much.
In an abstract way there could be a runtime provider, which i can ask: "please provide me runtime x at version y and tell me when ur ready to connect". We'd want a solution that can be used locally on a machine, or as an online service to connect to.
Relevant to jupyterhub/repo2docker#309
As a "reproducible document archive" is it in scope for DAR files to include an environment specification (packages, runtime versions, etc.), or is that expected to be left to other tools, such as pip requirements.txt and/or conda environment.yml as neighboring files, not in the archive itself? If so, how is that specified? For Jupyter, we consider this to belong as a sibling of the document, but I think the scope of Jupyter notebooks is meant to be a bit narrower than DAR.
The text was updated successfully, but these errors were encountered: