Ah ha, wait a second, my apologies, I completely misunderstood this response. Never mind - when I quit and restart Pd the behaviour is exactly the same on Mac and Linux, and I now understand that this is how it should be. Previously I was just closing the *patch* on Mac rather than quitting Pd entirely. On Linux of course closing the patch was closing Pd entirely no matter what.

Many thanks,

Yann

On 11 Mar 2022, at 07:01, Yann Seznec <yann@yannseznec.com> wrote:

OK thanks, it’s good to know the expected behaviour. It certainly works that way on Mac, but I’m reasonably certain that it is not working that way on Linux (at least on the most recent version available on raspberry pi). 

Images attached that show the behaviour on Linux - I launch the simple random patch twice on Linux and it produces the exact same stream of numbers.

So perhaps this is a bug that I should file? Perhaps someone else can confirm?

Cheers,

Yann

<Screenshot 2022-03-10 at 17.20.15.png>
<275635929_10103224057679409_2406650352563642422_n.jpg>
On 11 Mar 2022, at 05:26, Miller Puckette <msp@ucsd.edu> wrote:

Each time a [random] is created it gets a new (pseudo-random) seed - so if
you want total repeatability you should exit and restart Pd.  Then you should
see exactly the same behavior on linux and on Mac.

cheers
Miller

On Thu, Mar 10, 2022 at 12:04:15PM +0100, Yann Seznec wrote:
Hi! Here's thing I’m confused about…if I use [random] to generate some numbers on startup, it appears to have different behaviour on macOS and on linux (running on a raspberry pi).

With a very simple patch generate a stream of random numbers using [random] (with no seed) and printing to console, on macOS it will create a different random set every time I launch the patch. On linux, it will generate the same set of numbers each time the patch is launched. It feels to me like the [random] object on linux is using the same seed each time, whereas on macOS it is either using a new seed each time or just not using a seed at all (if that’s possible, I’m clearly no random number specialist).

Any thoughts appreciated. Thanks!
Yann


_______________________________________________
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.puredata.info_listinfo_pd-2Dlist&d=DwIGaQ&c=-35OiAkTchMrZOngvJPOeA&r=XprZV3Fxus2L1LCw80hE4Q&m=vNtFkc2FjZDtwqIAgsEDP9Guvogt_cL8daFu1mGXVJv2iRxIu_-NXQ_c8fR_v8tv&s=wdaa7geCOAzHg2vPDAzMrvXU1pfbV3jp-FtB5Wxt6iA&e=