Hi
If this would be addressed by jessie-backports, there would potentially be a significant wait until it's available, and as you say, it is a package that is somewhat important strategically - tiered storage is becoming more important with SSDs $/GB projected to be far higher than HDD for many years and conversely moving further and further ahead on performance (we are using Intel NVMe PCI SSDs for the cache in RAID5 - one of the crucial benefits of bcache is that all writes to the cache are sequential, and therefore very RAID5 friendly as this mitigates the write penalty). We've been using bcache with Jessie (by building the userspace tools from source) for some time now, but realistically very few are going to test this in earnest unless it is in 'testing' - we've got the opportunity to do that now before release if the package is unblocked - the package itself is small and has few dependencies unlike all the kernel code which will be in Jessie. Basically I'm suggesting that it would only be a minor burden on the release team as it's such a lightweight package - the reasons for not getting it into testing a week sooner seem to have been minor and not code-related, and in the end it only just failed to make the cut by 3 days! Is it possible to release on probation? In other words if it becomes a burden or apparent that it's the source of RC bugs then it just gets ejected again? If you do unblock the sooner you do this the more testing it will get before release. Kindest regards Jack
