I do. The previous API Review had some logic to it. I don't believe we can do without it entirely.
I see a spectrum of options here: * A bugfix with no external impact may be added by a single committer directly. * low impact API / SPI changes must have some API review to them, ie. at least 2 committers involved. Ideally one of the them with some experience in that area. * high impacting API / SPI changes must have proper voting. * also, visible UI changes should have some voting in. Good UX is also about saying no. Otherwise there will always be this little menu item / toolbar icon to squeeze. --emi On Tue, Jun 20, 2017 at 10:58 AM, Sven Reimers <[email protected]> wrote: > Do you envision a voting for new API/SPI's? > > Seems if we see a need we should clarify the process snd the gatekeepers... > > -Sven > > Am 20.06.2017 07:37 schrieb "Emilian Bold" <[email protected]>: > >> For small contributions an ICLA is not mandatory but since you want to >> work on a bigger feature it would certainly help. There's only one >> ICLA but signing one won't make you a committer without a vote. >> >> While your patch may be slim it seems to introduce a SPI so it would >> require an API Review (http://wiki.netbeans.org/APIReviews ). >> >> API Reviews are a old pre-Apache process but I believe we will have >> something similar under Apache too. We cannot just allow hooks all >> over the place if they impact the overall architecture. >> >> >> --emi >> >> >> On Tue, Jun 20, 2017 at 4:19 AM, Ross Lamont <[email protected]> >> wrote: >> > Hi folks, >> > >> > Given that there are a lot of balls in the air with regard to apache >> transition and Netbeans 9 release schedule, what is the correct process for >> submitting a patch, and what chance of getting it into Netbeans 9? >> > >> > Specifically: >> > 1. Do I create a bug/feature in Jira or in Bugzilla? >> > 2. I have not yet signed a contributor agreement. Do I sign the old >> Oracle one, or do I only deal with the Apache ICLA? Is there a different >> ICLA for contributor vs committer? >> > 3. Is there any sort of code freeze (soft or otherwise) in place at the >> moment for Netbeans 9.0, 8.3 or 8.2.1? The patch I’m proposing will be slim >> - it is just a refactoring to replace some hardwired construction of >> Cookies with a factory interface and Service Provider semantics (affecting >> XMLDataObject). This will allow me (and others) to develop new XML >> validation algorithms as a plugin in isolation from the main build. >> > >> > Cheers >> > Ross >>
