Should we send Travis messages to ghc-builds?
shumovichy at gmail.com
Thu Jul 21 12:20:12 UTC 2016
On Thu, 2016-07-21 at 00:15 +0200, Ben Gamari wrote:
> Ben Gamari <ben at smart-cactus.org> writes:
> > [ Unknown signature status ]
> > Hello everyone,
> > While picking up the pieces from a failed merge today I realized
> > that we
> > currently spend a fair bit of carbon footprint and CPU cycles
> > making
> > Travis test GHC yet the results of these tests aren't pushed
> > anywhere.
> > Would anyone object to having Travis push notifications of changes
> > in
> > red/green state to ghc-builds at haskell.org? Perhaps this will allow
> > some
> > of us to react more quickly to regressions.
> Actually Thomas points out that we indeed used to do this and yet
> stopped because it meant that users would fork the repository, enable
> Travis build on their fork, and then inadvertantly spam the list. So,
> perhaps we shouldn't do this.
I think it could be controlled by an environment variable set in travis
UI ( https://travis-ci.org/ghc/ghc/settings ). Repository fork will not
clone the variables.
> - Ben
> ghc-devs mailing list
> ghc-devs at haskell.org
More information about the ghc-devs