GCC, the whole toolchain, myriad dependencies, the ways that Python has patterend itself
after Java. Add in the way that the major Linux distributions are moving targets and
building / running on just one of them is a huge task, not to mention multiple versions of
each. And the way that systems running the same nominal release rarely are completely
identical, due to midstream package updates. Heck I’ve seen a mostly-well-run operation
that nonetheless had 80 different kernels running due to this.
Re the community, remember that “they” *are* the community.
Ceph is a big complex hunka burning love that many of us get for FREE. It by and large
works beautifully and helps us feed our addictions to food and shelter. Stability and
performance continually improve, and we’re largely free of the baggage of proprietary
solutions. Ceph is simultaneously sliced bread and a cosmic love pulse matrix.
So long as packages can be built, those who want to manage the traditional way still can.
The things we quibble about are usually minor, which is a testament to just how much we
take for granted.
So let’s discuss things that can be made *even better*, and try to respect the those who
do the heavy lifting.
— aad
But I think what Sage meant was e.g. different
versions of GCC on the
distributions and not being able to use all the latest features needed
for compiling Ceph.