Panda tools in a Pip installation

General Discussion
1

Under Ubuntu Linux, I see that Panda’s various tools–multify, PStats, etc.–are at time of writing included in pip-based installations, in a folder called “panda3d_tools”. However, they don’t seem to be installed to the system as runnable commands, and attempting to run “multify” at least (I haven’t tried the others, I don’t think) results in an error to the effect that it can’t find the Panda library files. (The files do exist; they’re just in the sibling-directory named “panda3d”.)

I’m guessing that I can at least get multify running by copying it into a directory along with various “.so” files, either just dumping all of them in there with it, or attempting to determine which specific ones it wants.

However, is there a more elegant and general way to get these various tools working?

For reference, I’m using Ubuntu 18.04.3, and Panda 1.10.5.dev103.

2

Perhaps they can be called from the Python interpreter. Or is it still in the plans.

3

Hmm… They appear to be executables, not Python files, so I doubt that they’re called from the interpreter.

It may well be that having these tools be installed to the system under Linux is indeed still in the works. :/

4

Hmm, I realized that they are in the form of packages, because pip is a python package manager.
You can register the full path to the executable file, not a short name.

5

I’ll admit that I’m not sufficiently familiar with pip and Python packaging to be clear on what that means in this context. ^^;

As far as I can see, they’re simply executables, just stored in the “site-packages” directory.

6

Aha! If I install then as root (i.e. via sudo, and with the “-H” switch to sudo), then they seem to be installed to the system!

7

In windows it works.

8

If I’m understanding you correctly, I tried that, but while the program ran, it complained about lacking certain libraries (that were in a sibling directory).

However! I posted again just above indicating that I seem to have solved the problem–in short, they work if you install as root, via sudo. :slight_smile:

1 Like
9

pip will automatically create links to the executables.

If you install as a local user, it will install them to something like ~/.local/bin. You must ensure that this directory is on your PATH.

10

Aaah, I see. I hadn’t thought to check my PATH. Thank you for that! :slight_smile:

For now, I’m happy enough with Panda installed under root permissions, I think–but I’m glad to know what it was that I missed when I was installing under my user permissions!

11

I’m probably a weird user, but I always get rid of PATH. My teams look like this, and that suits me perfectly.

C:\Panda3D-1.10.4.1-x64\bin\egg2bam.exe scene.egg scene.bam
pause

C:\Panda3D-1.10.4.1-x64\python\python.exe main.py
pause

call "C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Auxiliary\Build\vcvars64.bat" x86_amd64

cl /Ot /I C:\PandaSourse\built_x64\include /EHsc main.cxx /link /out:C:\Demo_Shaders\bin\hello_demo.exe ^

C:\PandaSourse\built_x64\lib\libp3framework.lib ^
C:\PandaSourse\built_x64\lib\libpanda.lib ^
C:\PandaSourse\built_x64\lib\libpandaexpress.lib ^
C:\PandaSourse\built_x64\lib\libp3dtool.lib ^
C:\PandaSourse\built_x64\lib\libp3dtoolconfig.lib

pause

:slight_smile:

12

That’s fair; each to their own, after all. :slight_smile:

I prefer to have the convenience of being able to simply call a command by name alone, myself.

I will admit that I’m not sufficiently familiar with these elements of Linux to be confident in messing with them too much, at least without a solid reason to do so!