Activity for Remko Scharroo
-
Remko Scharroo posted a comment on ticket #5087
Thanks for the explanation of the procedure. And thank you for your confidence in my commits. I'll still submit PRs for the more complicated updates, as I trust your eyeballs better than mine :)
-
dcw-gmt-1.1.3-1
-
Thanks for the description. I've added the pull request on github.com/fink/fink-distributions
-
Yes, I do have a GitHub account (I'm maintaining some software distributed through GitHub). I'll have to see how to set up a pull request. I'm sure SourceTree will help me with that.
-
gmt5-5.4.2-2 (depends on netcdf-c13-shlibs)
-
Closed. This has been superceded by ticket #5084.
-
gmt5-5.4.3-1 (depends on netcdf-c13-shlibs)
-
gmt-4.5.16-2 (depends on netcdf-c13-shlibs)
-
Closed. This ticket has been superceded by ticket #5083.
-
gmt-4.5.17-1 (depends on netcdf-c13-shlibs)
-
fftw3-3.3.7-1 (depend on netcdf-c13-shlibs)
-
gdl-0.9.7-2 (depends on netcdf-c13-shlibs)
-
fftw3-3.3.6-2 (depend on gcc7-shlibs)
-
fftw3-3.3.6-2 (depend on gcc7-shlibs)
-
cdo-1.9.2-1 (depends on netcdf-c13-shlibs)
-
cod-1.9.2-1 (depends on netcdf-c13-shlibs)
-
eccodes-2.6.0-1 (depends on netcdf-c13-shlibs)
-
ncview-2.1.7-2 (depends on netcdf-c13-shlibs)
-
nco-4.7.1-1 (depends on netcdf-c13-shlibs)
-
nco-4.7.1-1 (depends on netcdf-c13-shlibs)
-
gmt-4.5.16-2 (depends on netcdf-c13-shlibs)
-
gmt5-5.4.2-2 (depends on netcdf-c13-shlibs)
-
netcdf-cxx4-4.2.1-4 (depends on netcdf-c13-shlibs)
-
netcdf-cxx4-4.4.4-4 (depends on netcdf-c13-shlibs)
-
netcdf-fortran7-4.4.4-4 (depends on netcdf-c13-shlibs)
-
netcdf-c13-4.5.0
-
For the lack of any better solution I've removed the C/R on nco*-shlibs. But I did change and better hightlight a little PostInstallScript that will display something like this: Checking for older library versions... * You can remove the following older versions of the developer tools and shared library if you haven't built anything against them yourself: nco469-dev nco469-shlibs
-
nco-4.7.0-1
-
Finally got to work on this again. I followed up what Hanspeter suggested, minus the building of a static library. The reason for this is that nco contains a number of executables, all linked against this library and that would result in a much larger package. So I kept the creation of /sw/lib/libnco-4.7.0-dylib, but removed libnco.dylib, libnco_c++....dylib and the whole include directory. There is thus no way a user can accidentally link against this library. I also kept the Conflict/Replaces to...
-
Hi Hanspeter, I am also not happy with the proliferation of the ncoXXX-shlibs and ncoXXX-dev packages. I just continued what Alexander used to do before me. I think he was concerned about removing a dylib that people would have linked to outside of the nco package. Nco is not particularly clear about their backward compatibility of the dylib to older versions. That's probably why Alexander was cautious. In Ubuntu, I assume that the unversioned dylib is just a symbolic link (as we do). So if I upgrade...
-
nco-4.7.0-1
-
lapack-3.7.1-1, lapack-3.6.1-2, lapack-3.6.0-3, lapack-3.5.0-7
-
wip-2p3-2008 (depends on gcc7-shlibs)
-
eccodes-2.5.0-2 (depends on gcc7-shlibs)
-
grib-api-fortran-1.23.1-2 (depends on gcc7-shlibs)
-
netcdf-fortran7-4.4.4-3 (depends on gcc7-shlibs)
-
pgplot-5.2.2-14 (depends on gcc7-shlibs)
-
fftw3-3.3.6-2 (depend on gcc7-shlibs)
-
openmpi-1.10.7-2 (depend on gcc7-shlibs)
-
nco-4.6.9-1
-
eccodes-2.5.0-1
-
See new packages that were linked to grib-api, now to eccodes cdo-1.9.0-1, #5033 gdl-0.9.7-1, #5034
-
gdl-0.9.7-1 (linked against eccodes)
-
grib-api (1.23.1) and grib-api-fortran (1.23.1)
-
I have added the "Replaces:" lines in the attached files. Also please note that I have posted cdo-1.9.0-1 that is linked against eccodes-shlibs. Next step is to do the same with gdl.
-
cdo-1.9.0-1 (linked to eccodes)
-
eccodes-2.4.1
-
eccodes-2.4.1
-
Let's try again. I did take up all your suggestions (thanks!): Depends and BuildDepends: Spread over multiple lines, and alphabetized. Versioned libraries (updated from the Debian patch in order not to make static libraries). Put links to versioned libraries in eccodes and eccodes-fortran. Used openjp2.7 instead of openjpeg1. Updated ctest command, hopefully to make it work for you as well. Removed Provides: lines. Find encloded the new PatchFile and updated Info file. I hope this works better. It...
-
From your description then it looks to me that Conflicts and Replaces are synonyms, or that they should always come in pairs. I still do not see how I can claim that grib-api Replaces eccodes, as it clearly does not. Yes, I realise eccodes can currently not exist when building cdo. There is a new version of cdo out, I believe, that can be configured with eccodes instead of grib-api. So that would need to be done. I don't know gdl. Would need to dig into that one.
-
About the tests. I've seen those failures when during ctest the new binaries cannot be found. That's why I put in the PATH and DYLD_LIBRARY_PATH. Maybe I should make the PATH more specific to avoid looking somewhere else and finding the wrong binaries. It will be difficult to make sure it will work for you too, because it does work for me, even when the old grib-api was still installed. I had not managed myself to make the library versioned. I'll follow your suggestion. I had looked to get things...
-
I considered adding Replaces: eccodes, but that should not be done in my opinion. While eccodes replaces grib-api, grib-api does not replace eccodes. What I want to say is that grib-api is provides by eccodes as a subset (compatible headers), but eccodes has more (eccode headers for example) that grib-api does not provide. So a user should not be able to replace eccodes by grib-api, which Replaces: grib-api implies. (Same with grib-api-bin and eccodes-bin, as well as grib-api-fortran and eccodes-fortran)....
-
eccodes-2.4.1
-
grib-api (1.23.1) and grib-api-fortran (1.23.1)
-
gmt-4.5.16-1
-
gmt5-5.4.2-1
-
gmt5-5.4.2-1
-
gshhg-2.3.7-1
-
wip-2p3-2007 (depending on gcc6)
-
cdo-1.8.2-1
-
Sorry, Alexander, I had absolutely no intend to snark. I only mentioned the jump in version to justify making a new version in the first place. I'd be happy to take over the package, since I actually do use it often. I've entered myself as maintainer in the version attached here.
-
nco-4.6.6-1
-
As usual, you're very astude, Hanspeter. The /usr/X11R6 reference is just a left over from the past, and I did not think about updating it to /opt/X11. Done in the attached version.
-
You are right, Hanspeter. wip should depend on pgplot with gcc6, i.e. I updated to Depends line to pgplot (>= 5.2.2-13). In addition, I tried with readline7, which works fine. The updated wip.info file is attached.
-
openmpi-1.10.6-1
-
This works for me too on 10.12 with "fink -m". Since nobody is claiming this package as maintainer, I am volunteering and added my name. In addition, I adapted the package to now use gcc6 instead of gcc5, and updated DeskPackaging to include information about the "serious warning" produced by Fink. Attached my updated info file openmpi-1.10.6-2
-
arpack-ng-3.4.0-1 (dependency on gcc6)
-
fftw3-3.3.6-1 (dependency on gcc6)
-
fftw3-3.3.6-1 (dependency on gcc6)
-
gmt5-5.4.1-1
-
gmt5-5.4.1-1
-
grib-api-1.21.0-1 and grib-api-fortran-1.21.0-1 (dependency on gcc6)
-
grib-api-1.21.0-1 and grib-api-fortran-1.21.0-1 (dependency on gcc6)
-
wip-2p3-2007 (depending on gcc6)
-
pgplot-5.2.2-13 (depending on gcc6)
-
netcdf-fortran7-4.4.4-2 (dependency on gcc6)
-
pgplot-5.2.2-13 (depending on gcc6)
-
wip-2p3-2007 (depending on gcc6)
-
netcdf-c11-4.4.1.1-1, netcdf-c11-shlibs-4.4.1.1-1, netcdf-bin-4.4.1.1
-
Thanks, Hanspeter, you are right. opj_malloc.h is indeed not used at all except of...
-
gmt5-5.3.3-1
-
cmake-3.7.2-1
-
I don't know whether it was in response to this, but cmake-3.7.2-1 is already in...
-
ghostscript-9.21-1
-
gmt5-5.3.2-1
-
Closed because it suggests 1.0.8 but it should be 1.0.7 which I uploaded separat...
-
matplotlib-basemap-py 1.0.8-1
-
matplotlib-basemap-py 1.0.7-1
-
lapack361-3.6.1-1, lapack360-3.6.0-2 and lapack350-3.5.0-5
-
I also tested with gcc6. Here the updated lapack361.info file.
-
There was really no strong reason to stick with gcc5, I did that because I do not...
-
There was really no strong reason to stick with gcc5, I did that because I do not...
-
lapack361-3.6.1-1, lapack360-3.6.0-2 and lapack350-3.5.0-5
-
grib-api-1.18.0 and grib-api-fortran-1.18.0
-
grib-api-1.18.0 and grib-api-fortran-1.18.0
-
grib-api-1.18.0 and grib-api-fortran-1.18.0
-
gv-3.7.4-2 (libxt dependency added)
-
ddd-3.3.12-8 (to fix missing libncurses dependency)
-
netcdf-4.1.1-1
-
Any progress on releasing this version? It works for me on MacOS 10.12
1 >