Looks like there is a difference between git's "core.autocrlf" configuration
between machines. On windows1 it's set to "true" while on the new machines it's
Any specific reason to prefer one over the other? Why not leave it at the
The Brooklyn project expects that the files are in CRLF format and that's
configured in .gitattributes, but for some reason "autocrlf=input" takes
precedence. I'll need to investigate that further.
> On 18.10.2016 г., at 6:44, Gavin McDonald <ga...@16degrees.com.au> wrote:
> HI All,
> As some of you know, we have been working on replacements for
> Windows1(hudson-win.apache.org <http://hudson-win.apache.org/>) and Windows2
> (je-win2012.apache.org <http://je-win2012.apache.org/>) Jenkins slaves.
> We are pleased to announce that windows-2012-1 and windows-2012-2 are online
> and operational.
> Please do start using these now.
> The new nodes are direct replacements for windows1 and windows2 nodes which
> as of this notice are now
> deprecated and will be turned off at the end of this month - 31st October
> Please move your jobs to the Generic ‘Windows’ label if you have not already
> done so, or tie your jobs to the
> new slaves if you wish to start testing on those before the cut off date.
> Any jobs still tied to the deprecated slaves by the cut-off date will be
> moved to the Windows generic label by
> Gav… (ASF Infra Team)