[MacPorts] #32482: rtmpdump: update to 2.4
MacPorts
noreply at macports.org
Sat Jan 7 23:47:45 PST 2012
#32482: rtmpdump: update to 2.4
---------------------------+------------------------------------------------
Reporter: dargo@… | Owner: captsolo@…
Type: update | Status: new
Priority: Normal | Milestone:
Component: ports | Version:
Keywords: haspatch | Port: rtmpdump
---------------------------+------------------------------------------------
Changes (by ryandesign@…):
* keywords: => haspatch
* cc: ryandesign@… (added)
Comment:
Attached is a Portfile patch and updated patchfiles to update to 2.4.
Please evaluate and indicate whether it can be committed, or attach
alternate patches.
Upstream now defines the library version number as 0; in our old
patchfile, since upstream had not defined a library version number, we had
set it the same as the program version number. Since with this update the
library version number will change from 2.3 down (!) to 0, all ports
linking with librtmpdump will need their revisions increased to force a
rebuild.
Upstream changes in 2.4 include support for building OS X dylibs, and
building the shared library by default (in addition to the static one) so
some parts of the patchfiles were no longer necessary. Upstream neglected
to use the `-install_name` argument when creating the dylib so I retained
that from the old patch. I did not retain the `-current_version` and
`-compatibility_version` arguments we were using in our old patch, because
0 is the default anyway, and I recall badness on Tiger when these
arguments are set to zero. Upstream would be wise to select a library
version number greater than zero.
Other dylib creation arguments in our old patchfile I did not retain are
`-Wl,-undefined`, `-Wl,dynamic_lookup`, and `-Wl,-single_module`, because
I do not know their purpose. Meanwhile, dylib creation arguments upstream
is using that we were not using are `-flat_namespace`, `-undefined
suppress` and `-fno-common`; I do not know their purpose either.
I retained the part of our patchfiles that changed the VERSION variable
from e.g. "v2.4" to just "2.4". I do not know why we are doing this. I
imagine it was at least partly to ease computation of the library version
number; since we no longer compute the library version number, perhaps we
can stop munging the VERSION.
Since this version's distfile does not show up in the directory listing,
the livecheck still thinks 2.3 is the latest version. Upstream would be
wise to fix their directory listing.
--
Ticket URL: <https://trac.macports.org/ticket/32482#comment:7>
MacPorts <http://www.macports.org/>
Ports system for Mac OS
More information about the macports-tickets
mailing list