You're protesting the approach, not the result. As I understand it. I'm 100% on board with results-based protests. 0% for approach-based. :)
If Patrick can use his AI-based kung-fu to get things to where they're working, don't break backwards compatibility / old branches, have a sane sustainable workflow, and have updated docs that look good and are maintainable on trunk - God Speed. We should all be rooting for him. If you have specific "this approach isn't going to work because of non-obvious, non-documented constraints A-Z" based concerns, please by all means air them. Those are things we should write down and codify in the repo so people that come along with interest and energy have the context they need to know what constraints they're working with. If it's instead "this is making too many changes at once to a thing", I'd contend that the thing he's looking to make changes to is something that's not being heavily actively maintained and that has a lot of gaps. And finally, if it *is* being heavily worked on actively and I'm just not aware, please don't hesitate to correct and inform me. I'm just thinking there's a big iceberg that's been laying around people have been avoiding for a long time and now Patrick's willing to take a flamethrower to it. On Fri, Apr 3, 2026, at 12:51 PM, Mick wrote: > > > If someone's active, wants to do it, and there's enough people who agree > > with them (and nobody actively protesting), we should let things progress. > > > > Yes, I'm actively protesting :-) but not out of principle, out of pragmatism. > It has to work, currently I am unsure how it can. My objection is avoiding > us progressing down rabbit holes we later regret. > > Trying to be helpful… some questions to get going… > What version of antora are you using (we are currently stuck on 2.3) ? How > are the auto-generated configuration and nodetool pages included ? Is the > in-docs search still working ? are you using the cassandra-website ui-bundle > ? are you including the current djencks/asciidoctor-openblock hack ? > > > Probably the simplest test of feasibility is to get the following (from > cassandra-website) working: > > ./run.sh website build -g -u > cassandra:https://github.com/pmcfadin/cassandra.git -b > cassandra:cassandra6-docs-workzone > > > This can be done in parallel, but I don't believe should wait (because there > could be functionality this depends on that simply won't work), and I will > help out.
