https://github.com/apache/grails-core/pull/16011 has been updated

On 2026/07/22 10:54:49 James Fredley wrote:
> I have had very little time to work on this, but clearly think it is 
> important.  
> 
> Circling back to flags, since I was very confused.
> 
> cliAutoProvision came from https://github.com/apache/grails-core/pull/15948 
> and defaults to true which means auto-adding grails-core-cli, grails-console, 
> and discovered *-cli companions
> 
> legacyCommandSupport was also requested at some point and gpt-5.6 didn't like 
> it.  But it will be added nested under cliAutoProvision so, cliAutoProvision 
> wins and legacyCommandSupport is only checked if cliAutoProvision=true.  I 
> think legacyCommandSupport should default to true.  I am sure there will be 
> disagreement, but if we always agreed, we would be redundant.  
> 
> The flag for cliAutoProvision was already there, I was not aware of it, but 
> my PR had no intention of touching the split CLI side of this.  My initial 
> intention was they would both work by default, no flags, since I did not know 
> about the flag, but having flags is fine.  The Legacy PR, until early AM 
> today, ended up tied to cliAutoProvision, but now will be based on 
> cliAutoProvision and legacyCommandSupport.  The Split CLI Plugin commands 
> will work exactly as they did in 
> https://github.com/apache/grails-core/pull/15948.  legacyCommandSupport is 
> just a new flag for Grails 7 Plugin Commands working in Grails 8.
> 
> https://github.com/apache/grails-core/issues/16035 - there appears to be some 
> additional work, if we want plugin CLI-only script-commands excluded from 
> plugin's normal runtime JARs, like was done for Grails commands 
> (ApplicationCommand implementations).
> 
> James
> 
> On 2026/07/21 22:33:36 James Fredley wrote:
> > I don't think we are that far apart and I am willing to remove my -1 vote 
> > on M4.  
> > 
> > Once https://github.com/apache/grails-core/pull/16011 is refreshed in a few 
> > hours, we will likely be down to opt-in vs opt-out.  
> > 
> > https://github.com/apache/grails-core/pull/16011 feels closest to 
> > https://github.com/apache/grails-core/pull/15409#issuecomment-4051635625, 
> > although it appears we did remove those for 8.0.x
> > 
> > GRAILS 8 - BREAKING CHANGES THAT HAVE A FLAG
> > ============================================
> > 
> > Two kinds:
> > 
> > OPT-OUT OF THE BREAK
> >   Grails 8 does the NEW thing by default.
> >   You set a flag if you still want the OLD (Grails 7) behavior.
> > 
> > OPT-IN TO THE BREAK / NEW THING
> >   Grails 8 keeps the conservative or off path by default.
> >   You set a flag if you want the NEW thing (or a temporary old stack).
> > 
> > 1) OPT-OUT OF THE BREAK
> >    (new default ON - set flag to restore Grails 7)
> > --------------------------------------------------
> > 
> > GORM unconstrained domain properties are nullable
> >   Flag:  grails.gorm.default.nullable
> >   8.0:   true  (nullable - NEW)
> >   Old:   grails.gorm.default.nullable = false
> >   Also:  grails.gorm.default.constraints = { '*'(nullable: false) }
> > 
> > Accept header honored for all clients (including browsers)
> >   Flag:  grails.mime.disable.accept.header.userAgents
> >   8.0:   unset (honor Accept - NEW)
> >   Old:   set back to Gecko, WebKit, Presto, Trident
> > 
> > MIME types supplied by the framework
> >   Flag:  grails.mime.types  (full block still REPLACES defaults)
> >   Flag:  grails.mime.mergeDefaults
> >   8.0:   framework defaults when you declare nothing
> >   Note:  mergeDefaults=true is optional if you want to ADD types
> >          without re-listing everything
> > 
> > Jackson 3 is the default mapper stack
> >   Flag:  spring.jackson.use-jackson2-defaults
> >   8.0:   Jackson 3 / JsonMapper (NEW)
> >   Old:   spring.jackson.use-jackson2-defaults = true
> >          (temporary migration aid for Jackson 2-style defaults)
> > 
> > Hibernate native query date/time types (java.time)
> >   Flag:  hibernate.query.native.prefer_jdbc_datetime_types
> >   8.0:   java.time types (NEW)
> >   Old:   prefer_jdbc_datetime_types = true
> > 
> > Hibernate 7 Oracle float mapping
> >   Flag:  hibernate.dialect.oracle.use_binary_floats
> >   Old:   use_binary_floats = false
> > 
> > Hibernate 7 array JDBC type
> >   Flag:  hibernate.type.preferred_array_jdbc_type
> >   Old:   preferred_array_jdbc_type = VARBINARY
> > 
> > DevTools live reload off by default (Boot 4)
> >   Flag:  spring.devtools.livereload.enabled
> >   8.0:   false (NEW Boot default)
> >   Old:   true under development if you still want it
> > 
> > Groovy/Java parameter names preserved for compilation
> >   Flag:  grails { preserveParameterNames = ... }
> >   8.0:   true (NEW)
> >   Old:   preserveParameterNames = false
> > 
> > Boot defaultViewResolver removed for Grails web apps
> >   Flag:  grails.web.removeDefaultViewResolverBean
> >   8.0:   removed (effective true)
> >   Old:   removeDefaultViewResolverBean = false
> >   Note:  spring.gsp.removeDefaultViewResolverBean still honored,
> >          deprecated
> > 
> > Boot welcome-page mapping removed
> >   Flag:  grails.web.removeWelcomePageMapping
> >   8.0:   removed
> >   Old:   removeWelcomePageMapping = false
> > 
> > Boot catch-all static resource mappings off
> >   Flag:  spring.web.resources.add-mappings
> >   8.0:   false (NEW for Grails)
> >   Old:   add-mappings = true
> > 
> > Form content filter on PUT/PATCH/DELETE
> >   Flag:  spring.mvc.formcontent.filter.enabled
> >   8.0:   true (NEW; DELETE included)
> >   Old:   enabled = false
> > 
> > Auto-applied Grails BOM (replaces Spring DM plugin)
> >   Flag:  grails.bom  /  deprecated grails.springDependencyManagement
> >   8.0:   grails-bom applied automatically
> >   Old:   bom = null
> >          or springDependencyManagement = false
> > 
> > TagLib closure deprecation warning
> >   Flag:  grails.taglib.warnDeprecatedClosures
> >   8.0:   warn (default)
> >   Old:   warnDeprecatedClosures = false
> > 
> > Undertow max HTTP post size hardened
> >   Flag:  server.undertow.max-http-post-size
> >   8.0:   2MB
> >   Old:   -1 for unlimited
> > 
> > 
> > 2) OPT-IN TO THE NEW THING
> >    (default OFF or conservative - set flag to get new behavior)
> > --------------------------------------------------------------
> > 
> > Groovy invokedynamic
> >   Flag:  grails { indy = true }
> >   Default: false
> > 
> > Compile controllers/services/taglibs with @GrailsCompileStatic
> >   Flag:  grails { compileStatic { controllers/services/tagLibs = true } }
> >          or compileStatic { all = true }
> >   Default: all false
> >   Note:  per-class @CompileDynamic still wins
> > 
> > Hibernate 7 (instead of default Hibernate 5)
> >   Flag:  dependency / BOM choice (grails-hibernate7, hibernate7 BOMs)
> >   Default: Hibernate 5
> > 
> > Micronaut integration
> >   Flag:  add grails-micronaut + Micronaut-compatible BOM as
> >          enforcedPlatform; JDK 25 required if used
> >   Default: off
> > 
> > MIME merge onto framework defaults (instead of replace)
> >   Flag:  grails.mime.mergeDefaults = true
> >   Default: false
> > 
> > James
> > 
> > On 2026/07/21 20:19:39 James Daugherty wrote:
> > > I've made my case in the discussion thread I opened, and the shift being 
> > > asked here is the opposite of every other major change we've done.  Its 
> > > also arbitrary compared to past breaking changes.  The inconsistency is 
> > > frustrating & tantamount to: we effectively can't ever deprecate an API 
> > > in the future because it breaks plugins. This needs resolved before we 
> > > proceed as it goes against our major release policy on breaking changes.
> > > 
> > > There's not sufficient communication happening to resolve this so I think 
> > > we should cancel the vote until there's time to discuss or these issues 
> > > can be worked through.  
> > > 
> > > On 2026/07/21 19:58:23 James Fredley wrote:
> > > > M4 target was 7/22, it being early without 
> > > > https://lists.apache.org/thread/nq8f7yqtfp13wjbckkjxt1y82lrwkjq2 
> > > > address, really caught me off guard.
> > > > 
> > > > 2. Release 8.0.0-M4 by Wednesday (July 22) next week
> > > > - would include the undertow support + reproducible fixes + cli split
> > > > - last release of new features or breaking changes
> > > > 
> > > > If we are committed to getting some form of Grails 7 plugin in Grails 8 
> > > > backwards compatibility in RC1 or M5, I will be willing to change my 
> > > > vote to 0.
> > > > 
> > > > https://github.com/apache/grails-core/pull/16011 - is being updated 
> > > > right now to hopefully somewhere closer to a comprise.  
> > > > 
> > > > I think the number of hours and days of week we have worked on this as 
> > > > volunteers was historically unsustainable and is not something I have 
> > > > been able to maintain this year and will not be doing going forward.  
> > > > 72 hour turn around is the ASF expectation and that's what I will try 
> > > > to do.   Many time sooner, but I can't look at Apache Grails every day 
> > > > of the month.   
> > > > 
> > > > James
> > > > 
> > > > On 2026/07/21 13:29:06 James Fredley wrote:
> > > > > I am a -1 until we figure out a path for end users that does not 
> > > > > require them waiting on plugins to be re-released to use plugin 
> > > > > commands in Grails 8.  Grails 7 is EOL and users have to move to 
> > > > > Grails 8 basically upon release, let's not break their ability to use 
> > > > > Plugin Commands from the existing already published Grails 7 plugins, 
> > > > > if we do not have to.  
> > > > > 
> > > > > https://github.com/apache/grails-core/pull/15948 was merged before a 
> > > > > path was identified and we only addressed 1 of 4 paths that provide 
> > > > > commands.
> > > > > 
> > > > > https://github.com/apache/grails-core/pull/16011 - provides a path, 
> > > > > we can debate that path, but I am firm that it should be the default 
> > > > > for the reason in the first paragraph.  Let's not hurt the framework. 
> > > > >  
> > > > >  
> > > > > James
> > > > > 
> > > > > On 2026/07/21 09:35:43 Mattias Reichel wrote:
> > > > > > Hi Everyone,
> > > > > > I am happy to start the VOTE thread to release Apache Grails 
> > > > > > 8.0.0-M4!
> > > > > > 
> > > > > > Release notes for the release are here:
> > > > > > https://github.com/apache/grails-core/releases/tag/v8.0.0-M4
> > > > > > 
> > > > > > The tag for this release is:
> > > > > > https://github.com/apache/grails-core/releases/tag/v8.0.0-M4
> > > > > > 
> > > > > > Tag commit id: cb35496301c5423a557034394ce814dc1e1a36fa
> > > > > > 
> > > > > > The artifacts to be voted on are located as follows (r86111):
> > > > > > Source release:
> > > > > > https://dist.apache.org/repos/dist/dev/grails/core/8.0.0-M4/sources
> > > > > > Binary distributions:
> > > > > > https://dist.apache.org/repos/dist/dev/grails/core/8.0.0-M4/distribution
> > > > > > 
> > > > > > Release artifacts are signed with a key from the following file:
> > > > > > https://dist.apache.org/repos/dist/release/grails/KEYS
> > > > > > 
> > > > > > Please vote on releasing this package as: Apache Grails 8.0.0-M4.
> > > > > > 
> > > > > > Reminder on ASF release approval requirements for PMC members:
> > > > > > https://www.apache.org/legal/release-policy.html#release-approval
> > > > > > 
> > > > > > Hints on validating checksums/signatures (but replace md5sum with 
> > > > > > sha512sum):
> > > > > > https://www.apache.org/info/verification.html
> > > > > > 
> > > > > > Details of our release process are documented at:
> > > > > > https://github.com/apache/grails-core/blob/HEAD/RELEASE.md
> > > > > > 
> > > > > > As a reminder, most of our tools support an env `GRAILS_REPO_URL` so
> > > > > > that the tools can work with staged artefacts. Set it to the staging
> > > > > > repo like this: `export
> > > > > > GRAILS_REPO_URL=https://repository.apache.org/content/groups/staging`
> > > > > > 
> > > > > > The vote is open for a minimum of 72 hours and passes if a majority 
> > > > > > of at least
> > > > > > three +1 PMC votes are cast.
> > > > > > 
> > > > > > [ ] +1 Release Apache Grails 8.0.0-M4
> > > > > > [ ] 0 I don't have a strong opinion about this, but I assume it's ok
> > > > > > [ ] -1 Do not release Apache Grails 8.0.0-M4 because...
> > > > > > 
> > > > > > Here is my vote:
> > > > > > +1 (binding)
> > > > > > 
> > > > > 
> > > > 
> > > 
> > 
> 

Reply via email to