The download link for the arm version doesn't seem to work

It seems curl | sudo sh no longer works and the link is broken. It may be temporary but any ideas??

1 Like no longer supports Arm version see below, taken from download page:-

Earlier releases have been removed because we can only support the current versions of the software. To update old code, read the [ changes ]( page. Changes for each release can be found in [revisions.txt]( If you have problems with the current release, please [file a bug]( so that we can fix it. Older releases can also be [built from the source](

See also P4 on Pi4 (is not a joke)

1 Like

Thanks for a quick reply!

I see that ‘Download - Processing for Pi’ still has this method as one of installation options. Are you saying the information in this page is no longer valid (it probably is…)? If it is the case, would Processing be available on Raspberry Pi only for the expert users (as page suggests) unless we result to downloading the whole RP image…? All good if it is the case but I just want to confirm before I tell others. :smiley:

I didn’t know that link existed, so yes currently best option is to download tarball and extract that. The image is quite an old version of RaspberryPI targetting RaspberryPI 3. You don’t need to be much of an expert to install the tarball. In principle it should be possible to modify source of processing4 to get it to run on the rasbpberrypi, since PiCrate (a ruby processing implementation) and py5 (a python implementation) both run fine on the raspberryPI. NB: The install of py5 takes quite a bit of time on the pi.

1 Like

Thanks for the info. I’ll have a go and see what happens. I was totally unaware of this so it’s really good to know. Thanks again…!

Why does py5 take a long time to install on the pi?

I am going to guess that it takes a long time because py5 depends on numpy, and if you install through pip and use the source, you have to sit through the numpy compilation process, which is slow, especially on the pi. There are also two other libraries (line_profiler and noise) that can require compilation.

Installing using miniconda might be faster, as it can pull compiled packages from conda-forge. I think. I’m not that experienced using Python on raspberry pi’s, but do remember being frustrated by long installation times.

Once you get it working, py5 should work well on the pi because it uses linux.

1 Like

I have found it does indeed work well once installed, I just wanted to warn people to be patient, I know that unless your an old timer like me people expect immediate gratification. Like I used to compile my own linux kernel on a pentium II back in the day.


I bet there were no instructions for that linux kernel compilation.