Eliotte, I hope all your concerns are addressed. Please approve PR if so, and with that, assume you "lifted" your -1 :)
Thanks T On Wed, Apr 5, 2023 at 1:35 PM Guillaume Nodet <[email protected]> wrote: > I've raised a new PR to re-deprecate StringUtils and inline it's usages: > > https://github.com/apache/maven-resolver/pull/277 > > Le mer. 5 avr. 2023 à 13:30, Elliotte Rusty Harold <[email protected]> a > écrit : > > > That's still a halfway measure. And we shouldn't move the public > > StringUtils package without a deprecation cycle anyway. This is public > > API that can be used in projects that are not part of this repo or the > > Apache Maven project. So I propose: > > > > 1. Keep StringUtils in deprecated form for now, to be removed later, > > possibly in a major version update. > > 2. Replace usages of StringUtils with JDK methods. > > > > On Wed, Apr 5, 2023 at 7:23 AM Tamás Cservenák <[email protected]> > > wrote: > > > > > > I agree with you regarding duplication, this really adds "yet another" > > > StringUtils to classpath. > > > > > > OTOH, I do see a value to drop non trivial (by size) dependency for > > sparse > > > use. > > > > > > How about this proposal: > > > given un-deprecated StringUtils is used ONLY in (and below) > > > package org.eclipse.aether.internal.impl MOVE the StringUtils to this > > > package. So to say, make it resolver "internal thing", and not "offer" > it > > > via org.eclipse.aether.util package to anyone. I know this would not > > > prevent use of it outside of the resolver, but if anyone uses it, he > > would > > > clearly violate [1] and is left "on his own". > > > > > > [1] > > > > > > https://maven.apache.org/resolver/api-compatibility.html#package-level-contracts > > > > > > T > > > > > > On Wed, Apr 5, 2023 at 1:13 PM Elliotte Rusty Harold < > [email protected] > > > > > > wrote: > > > > > > > I'm -1 on this release. I specifically disagree with the > undeprecation > > > > of > > > > > > > maven-resolver-util/src/main/java/org/eclipse/aether/util/StringUtils.java > > > > in https://github.com/apache/maven-resolver/pull/275 > > > > > > > > We need fewer duplicate StringUtils classes in random packages, not > > > > more of them. If we don't want to depend on commons-lang3 then the > use > > > > can be replaced with JDK methods rather than by introducing yet > > > > another StringUtils class. > > > > > > > > > > > > On Wed, Apr 5, 2023 at 7:05 AM Tamás Cservenák <[email protected]> > > > > wrote: > > > > > > > > > > Howdy, > > > > > > > > > > The Resolver 1.9.8 is done: > > > > > > > > > > > > https://issues.apache.org/jira/issues/?jql=project%20%3D%20MRESOLVER%20AND%20fixVersion%20%3D%201.9.8 > > > > > > > > > > If anyone has/knows/feels there is more to be done in 1.9.8, yell > > here > > > > > please. > > > > > > > > > > Also, if anyone has free cycles, and this mail has no "more to be > > done" > > > > > response in 24h, feel free to release 1.9.8. Otherwise I will do it > > in > > > > the > > > > > 2nd half of next week. > > > > > > > > > > The 1.9.8 release should solve last "known issues" for Maven 3.9.x > > (azure > > > > > and locking timeouts), so it will be followed by Maven 3.9.2 > release. > > > > > > > > > > Thanks > > > > > T > > > > > > > > > > > > > > > > -- > > > > Elliotte Rusty Harold > > > > [email protected] > > > > > > > > --------------------------------------------------------------------- > > > > To unsubscribe, e-mail: [email protected] > > > > For additional commands, e-mail: [email protected] > > > > > > > > > > > > > > > > -- > > Elliotte Rusty Harold > > [email protected] > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: [email protected] > > For additional commands, e-mail: [email protected] > > > > > > -- > ------------------------ > Guillaume Nodet >
