Hi all, As the PR got quite complex with its many changes and moves of content, I merged the doc update yesterday, you can check the result at https://sling.apache.org/documentation/bundles/models.html.
if you find any inconsistencies or plain errors, please let me know, I plan to do another PR on top with minor improvements. Jörg Am Mi., 21. Dez. 2022 um 15:41 Uhr schrieb Jörg Hoh <jhoh...@googlemail.com >: > > Many thanks for the feedback Stefan! I updated the PR. > > Can someone else help with the conversion of the @Via examples to the > injector-specific annotations? > > Jörg > > Am Mi., 21. Dez. 2022 um 10:05 Uhr schrieb Stefan Seifert > <stefan.seif...@diva-e.com.invalid>: > >> thanks for the work! >> >> to make it easier to comment i've created a pull request with your >> changes and added some cosmetic comments: >> https://github.com/apache/sling-site/pull/106 >> >> overall, looks very good to me. i think i've never used the @via >> annotation, cannot help much here. >> >> the new "performance" section is also a good read, which was not >> summarized in that why before afaik. >> >> stefan >> >> >> >> > Am Do., 8. Dez. 2022 um 10:45 Uhr schrieb Stefan Seifert >> > <stefan.seif...@diva-e.com.invalid>: >> > >> > > +1 >> > > >> > > using @Inject is not only a performance problem, it also may lead to >> > > unexpected results if an injector steps in you are not aware of. >> > > >> > > stefan >> > > >> > > > -----Original Message----- >> > > > From: Jörg Hoh <jhoh...@googlemail.com.INVALID> >> > > > Sent: Thursday, December 8, 2022 10:25 AM >> > > > To: Sling Developers List <dev@sling.apache.org> >> > > > Subject: [Sling Models documentation] discourage the use of @Inject >> > > > >> > > > Hi, >> > > > >> > > > currently the the Sling Models documentation is a bit confusing; the >> > > > examples show only the use of the "@Inject" annotation, while more >> > > > down (at >> > > > [1]) in an not obvious location the use of @Inject is discouraged. >> > > > Also, when I benchmarked the use of Sling Models with different >> > > > annotations [2] I found that in some situations @Inject can have a >> > > > huge performance penalty compared to the injector-specfic >> > > > annotations, so I would also argue from this perspective, that we >> > > > should heavily discourage the use of @Inject. >> > > > >> > > > But at the top we state as one of the design goals of Sling Models >> > > > to use standard annotations (which are mostly discouraged at [1] in >> > > > favor of custom annotations). >> > > > >> > > > In the context of SLING-11711 I plan to rework the documentaiton >> > > > and >> > > all >> > > > examples to use injector-specific annotations, remove the design >> > > > goal of standard annotations, and mention the discouraged >> > > > annotations only in a separate section as discouraged (not >> > deprecated!). >> > > > >> > > > WDYT? >> > > > >> > > > Jörg >> > > > >> > > > >> > > > [0] https://sling.apache.org/documentation/bundles/models.html >> > > > [1] >> > > > https://sling.apache.org/documentation/bundles/models.html#hints-1 >> > > > [2] https://cqdump.joerghoh.de/2022/11/28/sling-model-performance/ >> > > > [3] https://issues.apache.org/jira/browse/SLING-11711 >> > > > -- >> > > > Cheers, >> > > > Jörg Hoh, >> > > > >> > > > https://cqdump.joerghoh.de >> > > > Twitter: @joerghoh >> > > >> > >> > >> > -- >> > Cheers, >> > Jörg Hoh, >> > >> > https://cqdump.joerghoh.de >> > Twitter: @joerghoh >> > > > -- > Cheers, > Jörg Hoh, > > https://cqdump.joerghoh.de > Twitter: @joerghoh > -- Cheers, Jörg Hoh, https://cqdump.joerghoh.de Twitter: @joerghoh