pynac development changes

[Ralf Stephan]

Hello,

let me summarize here some changes to how Pynac development is thought to happen.

In short, there is now a github branch for accompanying Sage changes (the ones that

finally go into the upgrade ticket), and there is a wiki page suggesting a development

protocol that contains a freeze phase before a new release if necessary.

The github branch https://github.com/pynac/sage was created to collect Sage patches

(both code and doctests) to make the Sage test suite pass with Pynac master.

Developers can add to it via pull requests and the release manager will use it for the

upgrade ticket. It should be synced with Sage develop. Short Sage patches can be submitted

as diff snippets inside Pynac PR or ticket comments, we are not fixed in how we collect 

these.

The procedure wiki page for developers is at 

https://github.com/pynac/pynac/wiki/Development-protocol

Regards,

[William Stein]

Hi Ralf,

This is really interesting! Are you guys basically developing a part
of a branch of Sage on github, then periodically merging it back?

Can you please report on how this goes since I would love if Sage
could move entirely to github someday.

William

> --
> You received this message because you are subscribed to the Google Groups
> "pynac-devel" group.
> Pynac web site: http://pynac.org
> ---
> You received this message because you are subscribed to the Google Groups
> "pynac-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to pynac-devel...@googlegroups.com.
> To post to this group, send email to pynac...@googlegroups.com.
> Visit this group at https://groups.google.com/group/pynac-devel.
> For more options, visit https://groups.google.com/d/optout.



--
William (http://wstein.org)

[Ralf Stephan]

It's down to me manually taking the diff and putting it into the Pynac upgrade trac ticket but yes, everything up to that point can happen on github.

[Benjamin Hackl]

Couldn't you also just accept the PR to the branch containing doctest changes on the sage-fork?

At least I don't really understand why you would make copying a part of your workflow, but maybe I'm overseeing something.

[Benjamin Hackl]

... nevermind; I thought you meant you always have to take the diff and merge it into the sage fork. (I just assume that you *would* also accept PRs there.)

But still, you can use the very same branch and push it to trac, no need for manually copying anything. :-)

Benjamin

Am Freitag, 4. März 2016 19:51:24 UTC+2 schrieb Ralf Stephan:

[Ralf Stephan]

You are right but I have a habit of doing thematic commits which makes reviewing easier. Also cuts down on merge commits. OTOH authorship is lost on the Sage changes. That would be an argument. I'd like to hear the opinion of the long-time reviewers like Jeroen or Karl-Dieter on this. Are commit message and structure important for review, or do you just skim the diff?