good morning,
in PD version 0.52.2
I'm trying to study this nice generative patch by Giovanni Dettori, "Perotinus". The patch includes some audio stuff saved as different array contents, therefore it is a bit heavy (45 mb or so). So I put here the link from which you can download it. [http://pdpatchrepo.info/patches/patch/6](link url)
If, i.e., I cancel some tabplay~, to put them into a subpatch or abstraction PD gets stuck, and I can't save the patch with a new name.
I get this message (in Italian) trying to close pd (English translation below):

ERRORE: impossibile terminare il processo con il PID 14216.
Motivo: Questo processo può essere terminato solo forzatamente ( con l'opzione /F ).
ERRORE: impossibile terminare il processo con il PID 14216.
Motivo: Questo processo può essere terminato solo forzatamente ( con l'opzione /F ).
while executing
"exec taskkill /pid $::connect_pid"
(procedure "::pd_connect::nuke_pd" line 4)
invoked from within
"::pd_connect::nuke_pd"
("after" script)

=======
ERROR: Unable to terminate the process with PID 14216.
Reason: This process can only be terminated forcibly ( with the /F option ).
ERROR: impossible to terminate the process with PID 14216.
Reason: This process can only be terminated forcibly ( with the /F option ).
while executing
"exec taskkill /pid $::connect_pid"
(procedure "::pd_connect::nuke_pd" line 4)
invoked from within
"::pd_connect::nuke_pd"
("after" script)
= = = =

Any help or suggestion to work around this issue? many thanks for your help - stefano

@stefano_zorzanello Beautiful sounds......
It was built in 2009 and so opens perfectly in Pd Extended.

But your problem........?
The error message originates in pd_connect.tcl so somehow you have broken the socket connection between Pd and the GUI.... or the TCL application thinks that it is broken.
The ::pd_connect::nuke_pd command is new in Pd 0.52
Maybe.... just maybe...... if you replace pd_connect.tcl in the Pd/tcl folder with the previous version pd_connect.tcl the problem will be solved....
I have no great expectations, but it is possible, as it is a new function that depends on a 2 second timeout......
Make a backup of the original first.

The "after" script originates in the tcl86 binary.....

Someone else might have an idea.....?
David.

Hi David,
thank you very much for your reply. I replaced the pd_conect.tcl in the tcl folder with the one you sent, but pd does't start giving the following report (see attached image):

Any idea?
thanks
stefano

@stefano_zorzanello So other files have probably also changed, but that confirms that wish and pd are not connecting through localhost ports.
Does Pd start when you go back to the original pd_connect.tcl...?
Do any other patches open correctly?
Are you running Windows?

If none will open....
..... and whatever your OS is.... open the task manager and check in the "processes" tab that "wish" or "pd" (in windows wish86.exe and pd.exe.. or....pd.com) are not running.... one or the other might still be in memory, but hung.
When we used to shut down completely at the end of the day such problems would be resolved.
Maybe try a hard restart of the OS (total shutdown and start it again).
You could check your host file...... here is how to find it........ https://www.howtogeek.com/27350/beginner-geek-how-to-edit-your-hosts-file/
There should normally be no entries in Windows....... every line should be commented with a # at the start of the line.....
"# something"
but as this happened while patching I doubt there is a problem with it.

Also check that Pd is still allowed by the firewall to communicate on your private network.
And that it has not been blocked by your antivirus for suspicious activity.

David.

HI David, thanks again.
btw I'm on Windows 10. I'd rather not get into the system host and so on, because I easily would get lost, because I'm not so expert in this kind of aspects...

I seems anyway to be a problem of too many graph - arrays with large lists of elements saved as content...
I managed have PD not crashing by building some abstractions instead of clusters of [tabplay~] objects, each one containing directly the [graph] object and the related [tabplay~] inside the abstraction itself...

Certainly I did not solve the PD problem, form a general point of view, but at least I can work on the patch.

Thank you very much
and warm regards
stefano