Excellent response... thanks.

Poking around, I appreciate your POC comment the most. I will only use maven-polyglot in my recreational programming as a learning experience until there becomes some compelling advantage otherwise.

1. good point... I concur
2. good point... yes, understandable synergy problems
3. good point... yes, many people claim things are easier, better than
   they really are
4. good point... yes, the sophistication of our tools increasingly
   makes diagnostics and troubleshooting more difficult

Working in a Gradle shop, many of these points also apply to my frustration with Gradle, and SBT from previous shops...

Cheers, Eric

On 2023-10-09 2:35 p.m., Greg Chabala wrote:
I have looked into maven-polyglot before, and come to the conclusion that
it is a proof of concept, and something that few if any people actually
use, because:

    1. Professional programmers are not actually offended by using XML in
    the POM, only novices would complain about such a thing.
    2. All tooling will expect a pom.xml, e.g. IDEs, CI tools, linters, etc.
    If any of those things are able to handle a polyglot POM in non-XML, that's
    a tiny miracle.
    3. maven-polyglot gets mentioned when people complain about Maven using
    XML, as 'look, you don't have to, you can use whatever you want' but no one
    actually does.
    4. The first suspect when something doesn't work right in your build
    will always be maven-polyglot, because no one uses it, so it's
    potentially not compatible with every plugin. Then you get to convert back
    to pom.xml and try again. So better to skip the effort and stay with XML in
    the first place.

Reply via email to