BLD: update commit hashes to point at tip of scipy mckib2:gcc48to52

[mckib2]

• see CI: bump gcc from 4.8 to 5.x scipy/scipy#13347
• testing to see if wheels using gcc5 need extra compiler/linker arguments as in C++14 pypa/auditwheel#125
• the scipy branch currently does not use any special compiler/linker arguments -- this is a fact-finding mission

[tylerjereddy]

A few thoughts here:

1. Why not open the PR against master? Unless there's a special reason to do otherwise, we'd normally do that and backport from there if we really need a backport.
2. If you really want to build your specific feature branch, you can try setting BUILD_COMMIT to the name of your feature branch (if you open a PR against master you'll notice that BUILD_COMMIT points at master.
3. I don't see anything special about your feature branch at the moment in terms of things the wheels repo would care about--I assume your plan is to make meaningful changes in that feature branch in response to any issues that crop up here in the wheels repo.

[mckib2]

@tylerjereddy Thanks, I think with your feedback I understand what's going on a little bit better.

1. Yes, I can open a new PR against master
2. If I open a PR against master, does DAILY_COMMIT need to change to the branch name? or do I just need to make BUILD_COMMIT=mckib2:boost?
3. I was pointing to the wrong branch -- I will need to point to the actual feature branch that we need testing for (ENH: Boost stats distributions scipy/scipy#13328) and modify the scripts here to use gcc5 and then mirror those changes in mckib2:gcc48to52 if any are needed

I will open a new PR tomorrow, thanks!

[tylerjereddy]

Just the build commit to temporarily point to your branch

…

On Sun, Jan 17, 2021 at 21:17 Nicholas McKibben ***@***.***> wrote: Closed #110 <#110>. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#110 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB4JOZW62Q3KFTHLQTU7L7TS2OY6FANCNFSM4WGTBKTA> .

[rgommers]

@mckib2 in your new PR, you may want to delete .travisci.yml to avoid triggering TravisCI until other CI systems are happy - we're very short on TravisCI credits.

[mckib2]

@rgommers Yes, I can do that. I ran out of time today, hopefully tomorrow I can another PR open and try again.