We've been using phyml (http://atgc.lirmm.fr/phyml/) for a long time, for our phylogeny research. I've installed it on several architectures, so I now propose a single ebuild with all the KEYWORD for archs I tested. I know they should be ~arch, but since everything is working fine... The version available for download at phyml home page is 2.4.4 and only binaries are distributed. If one needs the source has to write to the author. He will give version 2.4.5 that has an important bug fix. Phyml is released with GPL-2.
Created attachment 92926 [details] ebuild for sci-biology/phyml-2.4.5 I've tested on my systems, it works fine, installs correctly.
Created attachment 92927 [details] emerge --info for architectures on which the ebuild has been tested
(In reply to comment #1) > Created an attachment (id=92926) [edit] > ebuild for sci-biology/phyml-2.4.5 > > I've tested on my systems, it works fine, installs correctly. > the proposed ebuild misses: RESTRICT="fetch" I'm veery sorry.
Whats up with this bug report? On ~x86 I get: # emerge -uN phyml Calculating dependencies... done! >>> Verifying ebuild Manifests... >>> Emerging (1 of 1) sci-biology/phyml-2.4.5 to / >>> Downloading 'http://distfiles.gentoo.org/distfiles/phyml_v2.4.5.tar.gz' --2008-03-01 12:48:29-- http://distfiles.gentoo.org/distfiles/phyml_v2.4.5.tar.gz Resolving distfiles.gentoo.org... 64.50.238.52, 216.165.129.135, 156.56.247.195, ... Connecting to distfiles.gentoo.org|64.50.238.52|:80... connected. HTTP request sent, awaiting response... 200 OK Length: 92143 (90K) [application/x-gzip] Saving to: `/usr/portage/distfiles/phyml_v2.4.5.tar.gz' 100%[=============================================================================================================================================================================================>] 92,143 134K/s in 0.7s 2008-03-01 12:48:30 (134 KB/s) - `/usr/portage/distfiles/phyml_v2.4.5.tar.gz' saved [92143/92143] * checking ebuild checksums ;-) ... [ ok ] * checking auxfile checksums ;-) ... [ ok ] * checking miscfile checksums ;-) ... [ ok ] * checking phyml_v2.4.5.tar.gz ;-) ... [ ok ] >>> Unpacking source... >>> Unpacking phyml_v2.4.5.tar.gz to /var/tmp/portage/sci-biology/phyml-2.4.5/work >>> Source unpacked. >>> Compiling source in /var/tmp/portage/sci-biology/phyml-2.4.5/work/phyml_v2.4.5 ... cc -O2 -march=pentium4 -mmmx -msse -msse2 -fomit-frame-pointer -pipe -DUNIX -DPHYML -c main.c cc -O2 -march=pentium4 -mmmx -msse -msse2 -fomit-frame-pointer -pipe -DUNIX -DPHYML -c utilities.c cc -O2 -march=pentium4 -mmmx -msse -msse2 -fomit-frame-pointer -pipe -DUNIX -DPHYML -c optimiz.c cc -O2 -march=pentium4 -mmmx -msse -msse2 -fomit-frame-pointer -pipe -DUNIX -DPHYML -c lk.c cc -O2 -march=pentium4 -mmmx -msse -msse2 -fomit-frame-pointer -pipe -DUNIX -DPHYML -c bionj.c cc -O2 -march=pentium4 -mmmx -msse -msse2 -fomit-frame-pointer -pipe -DUNIX -DPHYML -c models.c cc -O2 -march=pentium4 -mmmx -msse -msse2 -fomit-frame-pointer -pipe -DUNIX -DPHYML -c free.c cc -O2 -march=pentium4 -mmmx -msse -msse2 -fomit-frame-pointer -pipe -DUNIX -DPHYML -c options.c cc -O2 -march=pentium4 -mmmx -msse -msse2 -fomit-frame-pointer -pipe -DUNIX -DPHYML -c simu.c cc -O2 -march=pentium4 -mmmx -msse -msse2 -fomit-frame-pointer -pipe -DUNIX -DPHYML -c eigen.c cc -o phyml main.o utilities.o optimiz.o lk.o bionj.o models.o free.o options.o simu.o eigen.o -lm -O2 -march=pentium4 -mmmx -msse -msse2 -fomit-frame-pointer -pipe >>> Source compiled. >>> Test phase [not enabled]: sci-biology/phyml-2.4.5 >>> Install phyml-2.4.5 into /var/tmp/portage/sci-biology/phyml-2.4.5/image/ category sci-biology >>> Completed installing phyml-2.4.5 into /var/tmp/portage/sci-biology/phyml-2.4.5/image/ strip: i686-pc-linux-gnu-strip --strip-unneeded -R .comment usr/bin/phyml * checking 1 files for package collisions >>> Merging sci-biology/phyml-2.4.5 to / --- /usr/ --- /usr/bin/ >>> /usr/bin/phyml >>> sci-biology/phyml-2.4.5 merged. #
The current ebuild in tree is a QA nightmare. sci-biology please advise. [Note: the ebuild attached here doesn't sound perfect either to be honest, the -static filtering sounds strange (and the package should not use -static by default anyway), and the fetch restriction seems unwarranted if the author allows redistribution (GPL-2 would).] At any rate something should be done.
Revision bumped for QA fixes. Package license wording is ambiguous, states just "GPL" without specifying the version. Package does not contain documentation. dawe, if you need any other arches, please advise. Flameeyes, do you see any remaining QA problems?
Ebuild looks better to me.
