Re: What is the Multigrain status?

[Daniel Price]

Hi Mark,

Good. At least we force the issue and pick up the pieces. I got sick last week so didn’t manage to clear the time to do it.

Given that we seem to get an “ok” approximately once per month now it may be time to change the way we pull in changes. For example the way the pipelines runs on branches is good, because you’re sure they work before you merge them. Pretty sure we could do the same with pull requests (i.e. only merge pull requests that pass pipelines).

Cheers,

Daniel

> On 13 Apr 2018, at 12:36 am, Mark Hutchison <marka...@gmail.com> wrote:
>
> Hi everyone,
>
> I heard no complaints about the "new-multigrain" branch so I have gone ahead and merged it into master. Fingers crossed that I didn't break too much along the way...you have permission to spam me with hate mail if I have.
>
> Cheers,
>
> Mark
>
> On Wed, Mar 21, 2018 at 3:29 AM, Mark Hutchison <marka...@gmail.com> wrote:
> Hi everyone,
>
> I searched for hours trying to find a good way of reverting the online repo such that we could (i) retain them in history, (ii) prevent the reverted commits from undoing not only my own commits, but the original commits in Master once we finally merge Multigrain back into the main code, and (iii) preventing people who have already pulled the online branch from having to do anything special (e.g. due to rewritten history). The best solution that I could find that meets all three of these criteria was to make a new branch. Once everyone is happy with the new branch, then the old branch can be deleted.
>
> The new branch is unimaginatively named "new_multigrain", and as of 5 minutes ago, should be up to date with Master. Of course as we have found out, the merging is complicated and it is possible that I have made some mistakes. Feel free to test things out and make sure that I haven't broken any of your routines.
>
> Cheers,
>
> Mark
>
> On Tue, Mar 20, 2018 at 11:38 AM, Mark Hutchison <marka...@gmail.com> wrote:
> Hi Daniel,
>
> There is no need to apologise. You have done some incredible work and fixed a bunch of my mistakes along the way. I can't count how many times I have had to abort a merge with Master because it got too complicated. Like Christophe said, I probably shouldn't have left it for so long. This merge was a bit weird anyway because the dust fraction parameterisation changed. In any case, I can't promise that my merge is error free either. That is the nice thing about working in a team...we can help catch each other's mistakes. My only purpose in writing is to make sure I was up to date with what you guys were doing and that I have accounted for everyone's contributions. If I were better at Git, perhaps I could have found a cleaner way without all of the e-mails.
>
> I will go ahead and revert the branch and push my local changes. Hopefully, when the dust has settled, we'll still have a Multigrain branch on which to work. I will hold off merging Multigrain into Master until we either work out the sqrt(dust-to-gas-ratio) parameterisation for multiple dust phases or decide to use the arcsin parameterisation.
>
> Cheers,
>
> Mark
>
> On Tue, Mar 20, 2018 at 4:34 AM, Daniel Mentiplay <daniel.m...@monash.edu> wrote:
> Hi Mark,
>
> I naively attempted to merge master into multigrain, and hit a few roadblocks. Many lines of code were merged automatically by git but there were quite a few lines that I had to merge by hand. I may have made several mistakes. In the end, I don't think I'm the person to do this. I don't have enough knowledge about the 1-fluid method and its implementation in phantom. I just thought it would be good to give it a go, especially around the time of the phantom workshop, and so we could move on with implementing multi-fluid multigrain.
>
> I'm happy for you to take over the merge, if you're happy to do so. Sorry if I've left the branch in a bit of a messy state.
>
> Cheers,
> Daniel