On 28/06/11 12:26, Dan McGee wrote:
On Mon, Jun 27, 2011 at 9:26 PM, Allan McRae<firstname.lastname@example.org> wrote:
Rename alpm_mountpoint_t to pmmountpoint_t to keep naming consistent
with every other struct we have.
What if I said "f*** it, it's 4.0, we should fix them all to be alpm_
like they should have always been"? I'm tempted to actually go the
other way- this one doesnt matter so much since it is internal to the
library, but we are very inconsistent with exposing PM_, pmxxx_t, etc.
when all our functions are rightly named alpm_*().
My thoughts are that I am glad I stopped writing my GTK3 libalpm
frontend when I realised what was happening to the API!
I am actually quite comfortable with the distinction that alpm_* is a
function and pm*_t is a type. But saying that, converting the pmxxx_t
to alpm_xxx_t does seem an overall more consistent approach. Our list
type is already prefixed alpm_ and now we have exposed a lot more of
those structs it is probably worth changing them.
And now is the best time to do this. So lets say I am more in favor of
doing this than not (yay for committed answers!). I would be quite
happy to prepare the patchset to do it.