Andreas Klöckner
Andreas Klöckner
Sorry for the long silence here. We should be able to get this in. If worst comes to worst, I'll run the UIUC instance on this branch for a bit.
> Looks like checks are failing since the testing container doesn't know where/what the inducer/runoc image is. I don't think that's the case. The tests don't rely on Docker being...
> Got it. Don't worry about the timeline, then. You sure? I'm pretty sure we can still get it by then.
Unifying them seems like the right idea. `CODE_QUESTION_CONTAINER_PORT` maybe?
Honestly, neither. All shared logic should definitely reside in the superclass. (I don't deal well with duplicated code.) Variant data (such as image or command names) should be queryable by...
I.e. use inheritance-based dispatch whenever you're tempted to say insert an `if`.
Trying to merge master into this gives me pretty puzzling results, as if this branch were trying to repeat changes from #667. Not even merging e77748c65007179309961f884af3e71148e624af (the merge commit for...
@davis68, I saw your comment on #631. So the status of this PR is that it has been gutted to produce #674 and #667, and the remaining Octave changes are...
Here's a pretty important question that occurred to me: Can you find/add a link in source to nbconvert's policy on whether it's designed to be safe with untrusted input?
Currently, in order for anyone to view a flow, they need to have a `FlowSession`. If they get a randomized version of the flow, then this records their randomization and...