[JOSS] Paper-specific comments

[matbesancon]

Stochastic processes have random and uncertain outcomes and decisions must be made at different stages of the process.

Decisions must be made because it's a stochastic optimization problem, stochastic processes on their own do not imply decisions.

be approximated into simpler ones,

be approximated by simpler ones?

But still there is no an open-source

still there is no open-source (not no an)

Process distance extends and generalizes

The process distance extends...

The formatting of the lines 1. Generation of scenario trees and scenario lattices from stochastic pr ocesses using the stochastic approximation procedure: and Generation of scenarios based on limited data using conditional density estimation: are weird, the authors should not use inline code for plain English text.

[matbesancon]

Linking with the JOSS review issue: https://github.com/openjournals/joss-reviews/issues/1912

[kirui93]

Hello @matbesancon Thank you for your insightful comments.

I do not understand what exactly you mean by the following lines:

The formatting of the lines 1. Generation of scenario trees and scenario lattices from stochastic pr ocesses using the stochastic approximation procedure: and Generation of scenarios based on limited data using conditional density estimation: are weird, the authors should not use inline code for plain English text.

[kirui93]

@matbesancon Please have a look at the paper again. I have made the corrections you indicated.

[kirui93]

Hello @matbesancon

Did you have time to look again on the corrections you had made?

[matbesancon]

Hi, sorry last week was too dense to manage to take a significant look, I'll do that today

[kirui93]

Okay, I am sorry for the tight schedule. Thank you for having time to have a look at it today.

[matbesancon]

For the footnotes on page 2, prefer a link to the current latest version, because latest is going to change. For this, it seems you don't have documentation for the specific versions online (instead of just dev). I think you will need to modify your .travis file to adapt the documentation phase, but not sure exactly how to do it

[matbesancon]

If you prefer I can post a new issue to track this point

[kirui93]

What you have just noted is what I have been trying for long. I don't know specifically how possible to get the documentation for the specific versions online but I will still try it.

[kirui93]

Is it possible to let it the way it is now and then as time goes by I will get a solution to it?

[matbesancon]

Ok, the advantage is that the link on the paper to the tutorial will remain a valid link afterwards, that's why I think we should do it before the publication of the paper

[kirui93]

Would it not be better to have the documentation as latest as the user wants the latest documentation rather than the specific documentation?

[matbesancon]

users will change to the latest documentation, but it will be easier for the readers of the paper not to arrive on a dead link, with the link explicitly mentioning the version, they can then change to latest if they want

[kirui93]

Okay. Let me see what I can do. My .travis.yml file has the same structure as most of the Julia packages with the specific versions of the documentation. I don't really understand why i am not getting this specific versions.

[matbesancon]

Maybe you can check out how Documenter.jl itself does it: https://github.com/JuliaDocs/Documenter.jl/tree/master/docs

[kirui93]

Documenter.jl doesn't state explicitly how to get these specific versions https://juliadocs.github.io/Documenter.jl/stable/man/hosting/

[matbesancon]

Does this help: https://juliadocs.github.io/Documenter.jl/stable/man/hosting/#Documentation-Versions-1

[matbesancon]

Otherwise check Distributions.jl, we specify that we deploy all versions: https://github.com/JuliaStats/Distributions.jl/blob/master/docs/make.jl

[kirui93]

Otherwise check Distributions.jl, we specify that we deploy all versions: https://github.com/JuliaStats/Distributions.jl/blob/master/docs/make.jl

Exactly, this is what I have been contemplating to do. Let me try it out now.

[kirui93]

I am afraid this hasn't worked. I have followed the instructions well but still it is only dev appearing for the documentation.

[matbesancon]

does it work when you build locally? You did lots of changes in the last commit I'm not sure which could have an effect

[kirui93]

Yes it does work locally but online it doesn't show it.

[matbesancon]

so it means the problem is likely in your travis

[kirui93]

I don't think it is in my travis since if you compare with the travis of Distributions.jl, you will see that there are similarities. Maybe it is more than travis. I will still keep looking into it.

[matbesancon]

Curious, maybe ask help on the documentation channel on slack

[matbesancon]

This line is weird for example https://travis-ci.org/kirui93/ScenTrees.jl/jobs/643322303#L466

[kirui93]

@matbesancon what I am getting as replies from slack is that when you make release of the package, these versions comes with it.

[kirui93]

This line is weird for example https://travis-ci.org/kirui93/ScenTrees.jl/jobs/643322303#L466

This ain't no more for the last commit I made.

[kirui93]

What am getting from slack is that the stable documentation will be available after I make a patch release. Also with this issue deploying skipped tags, they propose also the same solution. I don't want to make a release now since I have made many releases. I will make the last release immediately when the paper is accepted so that the version will be adapted to the paper.

[kirui93]

Hello @matbesancon ,

[matbesancon]

the issue does not seem resolved, you should make a new release to see if the documentation appears for it

[kirui93]

Okay let me do it. Hopefully it won't disappoint.

[kirui93]

Hello @matbesancon I think I have found it. So you said I should link to the stable version of the package? Is it this one? https://kirui93.github.io/ScenTrees.jl/stable/

[matbesancon]

Yes that's great! You should also link this one on your README. Also I think the version should be v0.2, since you did some breaking changes with the function name changes from Camel Case to snake case

On Sun, Feb 2, 2020 at 9:52 PM Kipngeno Kirui notifications@github.com wrote:

Hello @matbesancon https://github.com/matbesancon I think I have found it. So you said I should link to the stable version of the package? Is it this one? https://kirui93.github.io/ScenTrees.jl/stable/

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/kirui93/ScenTrees.jl/issues/14?email_source=notifications&email_token=AB2FDMV2U5MLEMA2IGUXUQDRA4XCHA5CNFSM4KKH5FH2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKSAV6I#issuecomment-581176057, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB2FDMQ5UENU7HVL6L55XQTRA4XCHANCNFSM4KKH5FHQ .

-- Mathieu Besançon

[kirui93]

Okay let me fix the README and also bump it to version v0.2 now.

[matbesancon]

In the mean time I think this issue can be closed