Hello. Le mer. 13 nov. 2019 à 14:49, sebb <seb...@gmail.com> a écrit : > > +1 for CP > > CP versions are entirely optional;
Not "entirely". [We already had this discussion.] It happened that a fix for <something> was available in a newer CP but that <something else> in CP broke some components (e.g. when CP functionality is "enhanced" based on assumptions not valid for all components). Then, that *suddenly* unsupported component has to cherry-pick and duplicate (in its local POM) what exists in that newer CP which it cannot use anymore. Gilles > a new version is only used if a > component chooses to use it. > > Is this also true of commons-skin? > If so, then +1 > > If not, then extra care needs to be taken to ensure backwards compatibilty. > There should probably be a formal vote on the actual changes in that case. > > > On Wed, 13 Nov 2019 at 13:32, Alex Herbert <alex.d.herb...@gmail.com> wrote: > > > The recent release of RNG has highlighted some issues with the commons > > parent configuration for multi-module builds and a desirable change to > > commons skin. > > > > 1. [parent] JaCoCo aggregate reports are included. > > > > Aggregate reports show coverage of dependencies. Since most commons > > components are stand-alone this should be disabled and the report set > > updated to have non-aggregate reports. > > > > An example of the output is shown for the recently release BCEL: > > > > https://commons.apache.org/proper/commons-bcel/project-reports.html > > > > The aggregate report is empty. > > > > > > 2. [parent] japicmp does not allow clean opt-in > > > > The japicmp maven plugin is immature. If set to skip the report it still > > creates a menu entry in the site reports page. See this release of > > Collections: > > > > https://commons.apache.org/proper/commons-collections/project-reports.html > > > > The link for the japicmp report is a blank page. > > > > > > For collections this could be fixed by running the report. > > > > For a multi-module build using japicmp any module with no code (i.e. a > > pom) has this empty entry if japicmp is enabled as the skip > > functionality does not work. > > > > > > The solution is to move the reporting section from the main > > configuration into the japicmp profile. This allows opt-in on a > > per-module basis to japicmp. > > > > > > 3. [skin] Customisation of the site <head> section > > > > RNG uses formulas in the site documentation. Ideally this would be > > rendered latex using MathJax [1]. > > > > This cannot be included in the site descriptor for the project as the > > <head> entry for each page is generated by commons-skin. This adds > > javascript to allow pretty print of code inside <pre "prettyprint"> > > tags. I would like to add a javascript tag to allow inclusion of MathJax. > > > > > > I have rendered the RNG site using these changes and staged it here: > > > > mvn package site site:stage -Dcommons.release.version=1.2 > > -DcomparisonVersion=1.2 > > > > https://home.apache.org/~aherbert/commons-rng-1.4-site/index.html > > > > > > Top level reports page does not have japicmp: > > > > https://home.apache.org/~aherbert/commons-rng-1.4-site/project-reports.html > > > > Components do: > > > > > > https://home.apache.org/~aherbert/commons-rng-1.4-site/commons-rng-simple/project-reports.html > > > > (There are also no jacoco aggregate reports.) > > > > > > You can view how MathJax is rendered on this page: > > > > https://home.apache.org/~aherbert/commons-rng-1.4-site/developers.html > > > > (search for 'will render an in-line formula') > > > > > > AFAIK an update to commons-skin to include MathJax will not effect sites > > that do not contain the \( or \[ characters in plain text on site pages. > > > > > > I propose to: > > > > - Update commons-skin to add a MathJax script to the <head> section > > > > - Release commons-skin (last release was May 2015) > > > > - Update commons-parent: > > > > - Use the new commons-skin > > - Remove JaCoCo aggregate reports > > - Reconfigure japicmp for clean opt-in > > > > > > Alex > > > > > > [1] https://www.mathjax.org/ > > > > > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > > For additional commands, e-mail: dev-h...@commons.apache.org > > > > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
