Thomas' questions are legitimate and I think we need to answer those in order to take a well-informed decision.
Today, we have no sense of the status of spi+multi-user/multi-tenant on OSIO. With a plan like the one proposed, we don't know if it's going to take few weeks or few months before we are able to deploy 6.x on OSIO.
I agree that the introduction of the new SPI on master, should not break anything that is already running.
If we are not able to run the new version of Che with the spi on OSIO with the same capabilities than the version we are deploying today - then it means that the SPI is not ready and should not be merged.
So I'd like to understand where we are, before moving forward and merging the SPI.
For our teams, our main priority must be to get OSIO running the latest version of Che asap. That's critical to us. All the rest, around single/multi assemblies, is less important for now.
I don't see why we would even do a release of che 6, if we know that we are not able to run it on OSIO.
@Sergii: since mario's team is rushing to get 5.19 multi-tenant on OSIO, can you lead that analyze?