Package Details: lammps 20201029-1

Git Clone URL: https://aur.archlinux.org/lammps.git (read-only, click to copy)
Package Base: lammps
Description: Large-scale Atomic/Molecular Massively Parallel Simulator
Upstream URL: https://lammps.sandia.gov/
Keywords: physics simulation
Licenses: GPL
Submitter: gucong
Maintainer: physkets
Last Packager: physkets
Votes: 9
Popularity: 0.176726
First Submitted: 2013-04-05 23:32
Last Updated: 2020-10-29 06:33

Required by (0)

Sources (1)

Pinned Comments

physkets commented on 2020-04-15 06:03

The comments here are only for issues with the AUR package and the PKGBUILD.
For problems with LAMMPS (including installation/compilation), you could try asking in the ##lammps IRC channel at freenode.net
https://lammps.sandia.gov/irc.html

Latest Comments

1 2 3 Next › Last »

sgsaenger commented on 2021-02-12 16:39

Fair enough, thanks anyways!

physkets commented on 2021-02-12 15:37

Hi @sgsaenger , I prefer not to make any changes to the default configuration, and decided that the bare setup is best. Since this is an AUR build-script, people are free to tack-on whatever options they like. You could add this package as an 'optional' dependency, and instruct users of your package to add those options. People installing stuff from the AUR should at-least be able to do that anyway.

sgsaenger commented on 2021-02-12 13:47

Hi, thanks for maintaining this package.

I have a package depending on lammps (vipster) and would welcome if this could depend on a global lammps installation. Would you mind adding -DBUILD_SHARED_LIBS=ON -DLAMMPS_EXCEPTIONS=ON to the build?

Also, are you aware of the cmake presets provided by lammps? I guess you like to keep the unbiased nature of lammps, but it may be useful to at least include minimal.cmake to cover most basic use cases (currently includes MOLECULE, KSPACE, RIGID and MANYBODY).

Thanks for considering!

physkets commented on 2020-04-15 06:03

The comments here are only for issues with the AUR package and the PKGBUILD.
For problems with LAMMPS (including installation/compilation), you could try asking in the ##lammps IRC channel at freenode.net
https://lammps.sandia.gov/irc.html

physkets commented on 2020-04-15 06:00

@Scatola Put all your options above the comment. I should make that a bit more obvious in that statement.

Scatola commented on 2020-04-14 18:25

@physkets First of all thanks for maintaining this package! I'm trying to build and install LAMMPS with additional packages I need (in particular MOLECULE, KSPACE, and RIGID), hence I'm using your PKGBUILD to do so.

I've modified the build function as follow (following what you have done):

build() {
    cd "${pkgname}-stable_${_pkgver}/build"
    cmake ../cmake \
        -DCMAKE_INSTALL_PREFIX="/usr" \
        -DCMAKE_INSTALL_LIBDIR="lib" \
        -DCMAKE_INSTALL_LIBEXECDIR="/usr/lib" \
        "${_feature_args[@]}" \
        # Add options for additional packages
        -DPKG_MOLECULE=yes \
        -DPKG_KSPACE=yes \
        -DPKG_RIGID=yes \
    make

But once I go with "makepkg -si" I get:

line 78: -DPKG_MOLECULE=yes: command not found

And the building is aborted. What is the problem?

AG_Caesar commented on 2019-03-20 09:59

@physkets tnak you for your work! It is working perfectly now!

physkets commented on 2019-03-18 12:46

The issue with the date should now be fixed. I hard-coded it as 12Dec2018, and I will change it along with a version bump.

physkets commented on 2019-03-18 11:40

I just changed the versioning scheme, switched to building with cmake, and added a mechanism to build the documentation natively. There has not been a version bump, so you do not have to build and install this version, if you don't want to. Although if you do, it will just be simpler when the next version bump happens.

Also note that setting _ENABLE_KIM=1 will not work right now, because the dependency package kim-api packages v2 of the API, while the current stable lammps uses v1. This will begin working fine from the next stable version. If you need it right now, use the lammps-beta AUR package.

physkets commented on 2019-03-18 11:21

@AG_Caesar That seems odd. It should be Dec, and not Dez. Might be a locale thing. What is yours? I might have to specify that.

Can you try this and tell me what the output is: ?

LC_TIME=C date -d 20181212 +%-d%b%Y