Hi,

Did you already do the work to list what it implies?
For instance I know interactive mode can be needed on these platforms (for
debug purposes mainly), that travis requires more work for logs (size +
timeout "ping" log to not have the job killed) etc...
So wonder what is the overlap (if it is just -B it is not worth to be
extreme) and if we have that status.
A small alternative which is interesting is to not enable/disable anything
but log it is likely desired: "you are running on X, you should probably
set -X1 -X2, or -no-X1 -noX2 if you dont want this message to show again".

Romain Manni-Bucau
@rmannibucau <https://twitter.com/rmannibucau> |  Blog
<https://rmannibucau.metawerx.net/> | Old Blog
<http://rmannibucau.wordpress.com> | Github <https://github.com/rmannibucau> |
LinkedIn <https://www.linkedin.com/in/rmannibucau> | Book
<https://www.packtpub.com/application-development/java-ee-8-high-performance>


Le dim. 12 juil. 2020 à 19:41, Michael Osipov <[email protected]> a
écrit :

> Am 2020-07-12 um 17:42 schrieb Martin Kanters:
> > Hi all,
> >
> > When setting up GitHub Actions for Maven (core) two weeks ago, I
> initially
> > forgot to enable the -B/--batch-mode flag.
> > While this gave no issues and was easy to fix, it could be avoided by
> > detecting if Maven is executed on popular CI environments and in that
> case
> > automatically enabling batch mode.
> > Popular CI tools such as GitHub Actions [1] and GitLab CI [2] provide
> > predefined environment variables, which could be used for this purpose.
> > When this would automatically activate, we log this on INFO level. We
> would
> > also need an option to turn batch-mode off to override this automatic
> > behavior.
> >
> > The reason why I haven't created a JIRA issue yet is because I realize I
> do
> > not know if this fits the usual workings for Maven.
> > Based on best guesses (de-)activating certain functionality can in my
> > vision be a quality-of-life improvement for our users.
> > On the other hand I have the feeling currently Maven does favor users to
> > explicitly set their CLI flags.
> > Just wanted to know your opinion. If I get positive feedback I will
> create
> > a JIRA issue.
> > Otherwise we can consider alternatives (only log a hint: "CI run
> detected,
> > did you want to run in --batch-mode?") or leave it like it is.
>
> So, I consider explicitly adding zillions of CI support is the wrong way
> to go. We need to be agnostic just like we are IDE-agnostic.
> What you are basically looking for is just like in Subversion:
> >   --non-interactive        : do no interactive prompting (default is to
> prompt
> >                              only if standard input is a terminal device)
> >   --force-interactive      : do interactive prompting even if standard
> input
> >                              is not a terminal device
>
> by default Subversion will try to determine both. You need native code
> here to determine whether stdout/stdin is attached to a terminal. There
> you will have your auto detection. Java's Console is supposed to do
> this, but is is too limited.
> Luckily, we do have Jansi in tree which provides use with:
> POSIX:
>
> https://github.com/fusesource/jansi-native/blob/master/src/main/java/org/fusesource/jansi/internal/CLibrary.java#L74-L76
> Windows:
>
> https://github.com/fusesource/jansi-native/blob/master/src/main/java/org/fusesource/jansi/internal/WindowsSupport.java#L49-L57
>
> This is what you are looking for.
>
> Michael
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
>

Reply via email to