'out' variable can occasionaly be treated as int and resulting in parameter error for str.replace(something,out).

[avengerx]

it gives an error if you have an argument like:
start: 1961010100
and it treats 1961010100 as an integer.
Maybe this is not the better solution for it, but the bug is pointed out.

This solution fits WORKSFORME status.

[avengerx]

The second commit (8f5b70a) raises (shows) the following inconsistency:
On documentation it is said you can just

bosun restart:name=expbase,restart=19611101,finish=1966010100

And the error displayed is:

ERROR: requested warm start model in '1961011001' but it is already partially run up to '1961110100'. Did you mean to change namelist.yaml value to this?

Thus, ignoring by any means what was specified on the commandline, using only what values are set on namelist.yaml.

By the way, the 'restart' section on the documentation is very hard to read, it could be improved by code-blocking commands...

[luizirber]

Thanks for the PR! But please keep the two commits separated, despite being related they are doing different things.

[avengerx]

Which of the three commits in the pull request are you referring to?

[luizirber]

519bd78 should be one PR, and 8f5b70a and 0cf68eb another one.

[avengerx]

Not my fault. I've made a single pull request and github is adding the others by itself. I really created this pull request with only the first commit.

Blame github for this. :)

You know, I did not git clone it to my host, I am editing files directly from github. I even tried to create another pull req when I built the second commit but, it disallowed me, when I looked here, there was the other commit already in this pull request.

I think github might have changed the behavior of pull reqs so they are on one instead of several prs from same remote branch, or something like it.