geode-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Patrick Rhomberg <>
Subject Re: Spotless Check failing on Windows?
Date Mon, 12 Nov 2018 18:11:03 GMT
Spotless has only been checking our gradle for maybe a week or two.  It
enforces dependency configurations using parentheses and single-quotes so
they would be easier to sort while some of us made dependencies explicit.
(Also, a future spotless improvement, once I can shake the bugs out).

The issue you're hitting is exactly what the lineEndings='unix' is supposed
to alleviate.  We specify lineEndings='unix', although looking at the
documentation now, DiffPlug "highly recommend[s] that you do not change
this value" from its default of GIT_ATTRIBUTES.  (This has been the case
since shortly after its introduction in October of 2016.)

You could try changing the lineEndings value to GIT_ATTRIBUTES and update
our .gitattributes to define a less-narrowly-defined eol.  (It currently
appears to only target *.java files.)  I don't have a Windows environment
up and running to test this on, but if you wanted to poke at these corners,
it's likely where the issue lives.  Otherwise, we could coerce a platform
check into spotlessCheck.upToDateWhen so that it is skipped on Windows, but
I'd be morally opposed to that.

Imagination is Change.


On Sun, Nov 11, 2018 at 8:47 PM, Owen Nichols <> wrote:

> The pipeline does actually build on Windows, but with -x spotlessCheck
> Someone (maybe whoever’s touched th spotless rules most recently) might
> know what to do. I tend to agree that the same source code ought to be able
> to pass spotless on both Linux and windows; and our pipeline should
> probably be tweaked to ensure this gets tested.
> On Sun, Nov 11, 2018 at 8:35 PM Michael Oleske <> wrote:
> > Hi Geode Dev Friends!
> >
> > I was building the latest off of develop (hash 75d7ed8a49d6) in
> PowerShell
> > on my Windows box and noticed during build that the spotless check
> failed.
> > It strangely fails on line endings (\r\n fun times) on what seems to be
> all
> > (or at least 34 of them) build.gradle files.  This seems odd to me since
> I
> > didn't realize that spotless check actually checked more than just java
> > files.  I assume no pipeline picked this up since the pipelines would
> build
> > once and pass the jar around rather than build on Windows and Linux
> (since
> > JVM and all that).  I haven't tracked down when this started happening to
> > me (though I can if that is of interest to people).
> >
> > If anyone has any thoughts I'd like to hear (even if they are just why
> > Windows why!)
> >
> > -michael
> >

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message