CMake¶

Summary¶

I'm not sure if this is worth a separate blueprint, but, maybe as part
of some sort of general tools or processes session, I would like to
discuss the possibility of switching our build system to CMake. There
is a preliminary branch from The Linux Box on GitHub [1] and a thread
on ceph-devel on this topic [2].
The numbers presented in [2] are a bit misleading, since AFAICT not
everything was built, and, as mentioned in that thread, we could
probably get some of the speedup by not building two versions of
everything with the existing autotools system, but, at least to me,
speed is not the only issue. libtool is a huge PITA as it breaks the
ability to export source trees to VMs which leads to bloated VM images,
not to mention the need to compile everything separately for each VM.
CMake is also better supported by IDEs, which would benefit people who
use them, IMHO has a better language and somewhat easier to work with.
What I'm looking to get out of this is to gather opinions, see if
people would be open to this, and perhaps discuss some of the details,
e.g. gitbuilders, the oldest platform that needs to be supported, etc.
[1] https://github.com/linuxbox2/linuxbox-ceph/commits/cmake
[2] http://thread.gmane.org/gmane.comp.file-systems.ceph.devel/17836/

Owners¶

• Name (Affiliation)
• Name (Affiliation)
• Name

Interested Parties¶

• Ilya Dryomov (Inktank)
• Noah Watkins
• Dan Mick
• Loic Dachary <loic@dachary.org>

Current Status¶

Detailed Description¶

Work items¶

Coding tasks¶

1. Task 1
2. Task 2
3. Task 3

Build / release tasks¶

1. Task 1
2. Task 2
3. Task 3

Documentation tasks

1. Task 1
2. Task 2
3. Task 3

Deprecation tasks¶

1. Task 1
2. Task 2
3. Task 3