oliverbock

Hi @mboudreau, yes, something like that: our devs make PRs that are merged onto master, and release branches are taken weekly from master for stabilisation until QA passes it for...

Hi @ccrolf, so far as I can tell, using Jira deployments would require an API call to specify which deployment any particular Jira is in. We could do that automatically...

@sOoN92, @AndrewRedican might not be accepting this PR because it includes code to delete most of README.md and updates package-lock.json, etc. Perhaps you should cut it back to just the...

@sOoN92, it makes sense to me.

After inspecting the evaluate source code I noticed that the error callback is chained off withCallingHandlers()'s error callback, which means that you can call sys.calls() and get a stack that...

frame is the `count` of the stack depth. You record it before the call to `evaluate()` ``` frame

@flying-sheep, the constant (22 in my case) will vary depending on how you are calling evaluate(). It may also change if the implementation of any intermediate methods change. In your...

I think I have found a workaround: ``` evaluate::replay(eval(quote(evaluate::evaluate(input=input)), envir=e)) ```

No, I fooled myself, the above workaround does not work. The best I can come up with is to modify the code to assign print.dog into the global environment: ```...

Hi @yihui, Thanks, I understand your point. I think in my case that I can work around my problem by using envir=globalenv(). In some respects this is a righteous solution...