hi IOhannes, thanks again, this makes alot of sense, ill investigate (including doing a new build on a fresh sd card and not overclock it this time)
On Thu 21. Nov 2019 at 14:18, IOhannes m zmoelnig zmoelnig@iem.at wrote:
On 21.11.19 12:05, iftah gabbai wrote:
if by flyspeck you mean " ` " its formatting errors that are there because i had no easy way of exporting the print out of pd to a "real" machine
that should be easy enough: $ pd -stderr .... 2>pd.log copy the "pd.log" to your "real" machine. tada.
or, if running with a GUI, you could also just go the the Pd-console window and hit <Ctrl>+<Shift>+<S> (aka "Save as...") to save the contents of the Pd-console as a txt-file.
On 21.11.19 12:05, iftah gabbai wrote:
is me doing: pd -nogui (various alsa tags to get sound going) -path /usr/lib/pd/extra/cyclone /etc/mydir/mypatch.pd &
for what it is worth, you should switch to using [declare -path cyclone] in your patch anyhow.
also worth mentioning, not sure what happend but i dont get the ELF error anymore, instead im getting : "svf~ error: ... couldn't create error: load_object: symbol "svf_tilde_setup" not found"
my bet is a filesystem corruption. e.g. the "svf~.pd_linux" is not resolved correctly, and points to this or that data chunk. if the datachunk happens to be an ELF-object (but not the real binary for [svf~]), then you would get the "svf_tilde_setup" not found error. if the datachunk is not an ELF object (e.g. a wav-file), then you would get something like the former error.
interestingly, pd is also not able to locate pd~ with the following
error:
"pd~: can´t stat /usr/lib/pd/pd",
this looks a bit like a broken Pd-installation. iirc, the "pd" binary should be both in /usr/bin/pd (so you can start it with "pd") and /usr/lib/pd/bin/pd (so pd~ can find it).
but of course, if your filesystem is corrupt, all kind of weird things can happen.
fgadmst IOhannes
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list