The following files in src/fortran/ have preprocessor directives, and should be .F90 and not .f90:

angles.f90
bonds.f90
chains.f90
dmtx.f90
gr.f90
initchains.f90
initrings.f90
molecules.f90
msd.f90
parameters.f90
rings-guttman.f90
rings-king.f90
rings-primitive.f90
sk.f90
spherical.f90

With all the recent pull requests merged, I can now run make GCCVER=12 GTKV=4 atomes natively on macOS, with Gtk+4 and GCC installed through Homebrew, and it compiles directly.

However, there is an OpenGL issue. I can view a structure, rotate it, but there is a color or lighting issue:

I can pick atoms and they get a cyan color, so I think the issue is not color, but lighting or shading.

Also, colors are OK in wireframe mode:

Hello, Dr. Slookeur
I cannot wait to install the latest Atomes (version 1.1.9). Thanks for your effort some bugs have been fixed, however, I met with a new issue and I don't know whether it's a bug or there is just something wrong with my device. I can open an XYZ file with only one frame but Atomes (version 1.1.9) would crash when I open an XYZ file with more than one frame (e.g. a file with lots of MD steps).
I upload a ZIP file with three XYZ files, test.xyz is an MD simulation with one step, test1.xyz and test2.xyz both are MD simulations with more than one step. Can you test it and help me find the issue? Thank you very much.
In addition, my PC's CPU (with GPU) is AMD 5600G, RAM is 16GB, ROM is 1TB.
Sincerely,
Sailang
test.zip

After I compile, I see that make has created some weirdly named files:

Untracked files:
  (use "git add <file>..." to include in what will be committed)
	"-DPACKAGE_DATA_DIR=\"\""
	-O0
	-O3

I think they are created because of these rules in Makefile:

$(OBJ)gtk-misc.o:
        $(CC) -c -MT $(OBJ)gtk-misc.o -MD -MP -MF $(CPPFLAGS) $(CFLAGS) $(DEFS) -o $(OBJ)gtk-misc.o $(GUI)gtk-misc.c $(INCLUDES)

and

$(OBJ)global.o:
        $(CC) -c -MT $(OBJ)global.o -MD -MP -MF $(CFLAGS) $(DEFS) -o $(OBJ)global.o $(SRC)global.c $(INCLUDES)

The -M options output dependency rules, but I am not sure if they are actually used. But, the -MF option is supposed to specify the file where the rules are written, and because it's used as -MF $(CPPFLAGS) or -MF $(CFLAGS), then it creates a file that is named as the first flag of CPPFLAGS (in my case, -DPACKAGE_DATA_DIR=\"\") or CFLAGS (in my case, -O3 or -O0).

This is made worse by the fact that this first flag, treated as a file named, is not actually used by the compiler as a flag.

Hi Dr. Sébastien Le Roux,

Thank you for your efforts to develop this excellent software.
I am writing to ask if and how I can run atomes to analyze the rings data and output the results by a pure command-line method? Because I want to use it to do the ring's analysis but which is only part of the post-process of the simulation, and if I can run it just by command-line interaction, it will be convenient to run by a shell scipt on HPC clusters.

Best,
Fan

Hello Dr. Sébastien Le Roux,

I met with a problem when I opened an xyz file and set the box size and boundary condition. I don't know why my laptop directly crushed and was restarted automatically. I don't know if it's my laptop's problem or Atomes' problem, actually I never met with this problem. I tried 3 times, but failed every time.
Here is some information in case you need to test it:
Atomes version: 1.1.14
Windows System Version: Windows 11 23H2 22631.3447
My laptop is Lenovo XiaoXinPro 14 APH8, the cpu is AMD Ryzen 7 7840HS with Radeon 780M Graphics.
The attachment is the XYZ file for your test.

dump_IPG40_40000_sample.zip

By the way, I wonder if you have the plan to add the feature of supporting lammps dump format file input? The request originated from two reasons:

1. LAMMPS is quite popular recently.
2. Yes I can dump xyz format file use dump_modify in LAMMPS, but I found that there are some restrictions for XYZ file. For instance, in my system, I have type1: Fe3+, type2: Fe2+, type3: P, type4: O, this can be specifically defined in LAMMPS and output in LAMMPS dump file. However, for XYZ file, the valence of element can not be defined, which means Fe3+ and Fe2+ are exactly the same. There is no doubt we will lose some information. That's why it came across in my mind if you will add features in Atomes to support read LAMMPS dump file.

Thanks a lot.

Best regards,
Fan

Dear Dr. Sébastien Le Roux
There is a confusing problem when I used Atomes. I have a *.XYZ file which has been attached (You need to change hh.txt to hh.XYZ, because I couldn't upload an XYZ file here). I could open it by Atomes in my Ubuntu 20.04, however, I couldn't import the file by Windows Atomes.
hh.txt

I've tried to open a new project with a file.xyz but "Error code: 2" showed up. Please kindly show me how to fix this.

calc/cpmd/cpmd_print.c:247:36: warning: duplicated 'if' condition [-Wduplicated-cond]
calc/cpmd/cpmd_print.c:251:36: warning: duplicated 'if' condition [-Wduplicated-cond]

Source code is

if (tmp_cpmd -> calc_type == 0)
{
i = TSTPO;
}
else if (tmp_cpmd -> calc_type == 0)
{
i = TSTPG;
}
else if (tmp_cpmd -> calc_type == 0)
{
i = TSTPC;
}

Suggest code rework.

Hello,

I saved my project (output.apf) and when trying to reopen it, atomes closes and there's written on the shell 'Segmentation fault (core dumped)'.
What could be the source of this error?

Thanks a lot

The Makefile for Linux has:

  LIBS = -Wl,--export-dynamic $(LGTK) -lm -lgfortran

This is not an option accepted by the macOS linker. I can change the Makefile to avoid its use on macOS, but I am not sure why it is needed on Linux in the first place. The GNU ld manual says:

When creating a dynamically linked executable, add all symbols to the dynamic symbol table. The dynamic symbol table is the set of symbols which are visible from dynamic objects at run time. If you do not use this option, the dynamic symbol table will normally contain only those symbols which are referenced by some dynamic object mentioned in the link. If you use dlopen to load a dynamic object which needs to refer back to the symbols defined by the program, rather than some other dynamic object, then you will probably need to use this option when linking the program itself.

I have not seen anywhere in the code the loading of dynamic objects (with dlopen). Is it really needed?