DevHeads.net

Postings by Elliott Sales de Andrade

Removing macros.R

Hi,

(This message is cross-posted to the Fedora devel and R SIG Fedora
mailing lists.)

Currently, R-core installs macros.R which defines one macro:
%_R_make_search_index /usr/lib/rpm/R-make-search-index.sh

This script does nothing, and there is no mention of this macro in our
guidelines. The macro is used by one package, R-systemfit. The script
is not used.

According to the R %changelog, it was disabled in 2009.

I would like to clean this up, and will do so in approximately a week,
and/or whenever the mass rebuild finishes.

Automating R package dependencies

Hi R-interested packagers and others,

Recently, I've been looking at how RPM can automatically determine
Provides and Requires [1]. I have since implemented this for R using
an R script [2] and some file attributes [3]. Following other
languages' Provides, I have namespaced them as R(packageName). It then
adds corresponding Requires, Suggests, and Enhances.

Additionally, R package versions commonly contain dashes. In order to
work in RPM, these are replaced with dots. For the automated
Provides/Requires, I have used the *real* versions instead.

So now the question is how to apply this.

xtl tests fail on ppc64le/armv7hl with gcc9

Hi,

I've had some issues with tests for xtl not working on
ppc64le/armv7hl. These tests work fine on Fedora 29, but fail on 30+
[1]. Unfortunately, koschei does not check these platforms, but I'm
fairly certain these started failing when gcc was updated to 9.0.0.
Upstream is also suspicious that this might be a gcc9 issue [2].

Based on the errors, it's as if the code that's supposed to run to
modify strings just isn't run. It's as if the compiler has determined
that it should optimize away the relevant code.

lmdb fails to map large DBs on i686/s390x

Hi,

I've been having some issues with LMDB on i686/s390x when building
python-zarr and its dependants. I filed a bug report [1], but the
maintainer suggested asking here to get more reach. I have replicated
my original message from the report below:

The documentation for LMDB [2] states that the map size should be
"chosen as large as possible". Consequently, the zarr developers have
chosen a map size of 2**40 on 64-bit systems and 2**28 on 32-bit
systems [3]. When the package is built on a 64-bit system [4],
everything works fine.

License change of R-gsl: GPLv2+ to GPLv3

Hi,

Upstream has made a new release that changes the license from GPLv2+
to GPLv3. I intend to build this later today. Since its main use and
linkage was to GSL which is GPLv3 only, this is effectively the
license the built work is already under anyway.

Unannounced soname bumps: proj and geos

Hi,

The recent bump of proj in Rawhide from 4.9.3 to 5.2.0 also bumped the
soname from libproj.so.12 to libproj.so.13.
The bump of geos in Rawhide from 3.6.1 to 3.7.1 also bumped the soname
from libgeos-3.6.1.so to libgeos-3.7.1.so.

These bumps should be announced *before* the build has been made.

Fortunately, as I was already investigating the possibility of
updating proj, I have the list of proj-dependent packages already. I
was not looking at geos, but hopefully the list below includes
everyone.

%find_lang for non-standard locations

Hi,

The general guidelines currently state that translations should be
found using %find_lang; this macro provides a listing of all *.mo
files but not their containing directories.

Specifying just the files makes sense when they are in the global
location (/usr/share/locale), but not so much when they are in, e.g.,
the Python site-packages directories.

python?-sip-api is no longer available, breaking python-wxpython4

Hi all,

I am attempting to build the latest matplotlib but it is failing to
install python3-wxpython4 [1]:

- nothing provides python3-sip-api(12)(x86-64) = 12.5 needed by
python3-wxpython4-4.0.1-9.fc29.x86_64

I attempted to rebuild python-wxpython4 [2], but that did not help. It
was then that I found out that python?-sip-api was removed on purpose
[3], meaning python?-wxpython4 can no longer be installed.

Is this an oversight? Should python?-sip-api be Provided by a
different subpackage? Should python?-wxpython4 Require on something
else?

R-tweenr license change

R-tweenr 1.0.0 has changed license from GPLv2+ and WTFPL to MIT and WTFPL.

Haskell failures: relocation refers to local symbol "" [1], which is defined in a discarded section

Hello all,

All Haskell packages seem to be broken now, e.g., ghc-optional-args [1] or anything on koschei [2].

License change in git-annex

git-annex-6.20180316 has changed license from GPLv3+ to AGPLv3+.

Technically, this was true when I enabled the webapp option in Rawhide earlier, but I did not realize it. However, AGPL code has been merged into the main application so the entirety of the package is now AGPL regardless of that build option.