Unable to edit recording - missing file .ap

[el bandido]

It will take some time to make a proper comparison. I will do it....Wait for it

 

[el bandido]

Hello @el bandido , can you please give me a favor?

Can you please try reproduce the issue with openATV?

So let's do the same thing.
(1) Build OpenATV using OE 6.0 and ATV master enigma2 branch.
Install image and record FTA channel on CNN feeds, 97w ku.
Record the channel and get 5 record files.

(2) Edit ATV enigma2 -master branch. Remove one line from filepush.cpp:
m_protocol = m_stream_id = m_session_id = m_packet_no = 0;
See: https://github.com/EB-TNAP/enigma2-ATV/commit/d9834d99626ed54a180e615bc264dd28bb89447b
Build and Install image using OE 6.0 and edited ATV enigma2-master branch.
Record the exact same FTA channel on CNN feeds, 97w ku.
Get 3 record files.

Warning: Permanently added '192.168.1.160' (ED25519) to the list of known hosts.
root@sf8008:~# cd /hdd/movie
root@sf8008:/hdd/movie# ls -l
-rw-r--r--    1 root     root     295630000 May  1 05:15 20260501 0509 - NS Ch4 - instant record.ts
-rw-r--r--    1 root     root          4544 May  1 05:15 20260501 0509 - NS Ch4 - instant record.ts.ap
-rw-r--r--    1 root     root             0 May  1 05:15 20260501 0509 - NS Ch4 - instant record.ts.cuts
-rw-r--r--    1 root     root           121 May  1 05:09 20260501 0509 - NS Ch4 - instant record.ts.meta
-rw-r--r--    1 root     root        290816 May  1 05:15 20260501 0509 - NS Ch4 - instant record.ts.sc
-rw-r--r--    1 root     root     293545268 May  1 06:10 20260501 0605 - NS Ch4 - instant record.ts
-rw-r--r--    1 root     root             0 May  1 06:10 20260501 0605 - NS Ch4 - instant record.ts.cuts
-rw-r--r--    1 root     root           121 May  1 06:05 20260501 0605 - NS Ch4 - instant record.ts.meta
root@sf8008:/hdd/movie#

The 4 variables have to be set to 0 else record files may be lost on certain channels. OBH has these variables set to 0 TWICE. One of them should be eliminated.

I have no clue as to WHY OBH suddenly required setting these variables to 0.
ATV has been setting those same 4 variables to 0 for years. So I imagine something changed somewhere else in OBH... It is hard for me to trace because everyone has a different enigma2 source code.

 

[el bandido]

This filepush script becomes more interesting when you look at it. While OpenATV-master and OpenPLi-scarthgap filepush.cpp are not identical, they at least resemble each other. Consider a diff of atv and pli filepush.cpp today:

+++ b/filepush.cpp    2026-05-02 12:53:41.851056000 -0400
@@ -1,6 +1,5 @@
 #include "filepush.h"
 #include <lib/base/eerror.h>
-#include <lib/base/esimpleconfig.h>
 #include <errno.h>
 #include <fcntl.h>
 #include <sys/ioctl.h>
@@ -20,7 +19,7 @@ eFilePushThread::eFilePushThread(int io_
       m_blocksize(blocksize),
       m_buffersize(buffersize),
       m_buffer((unsigned char *)malloc(buffersize)),
-      m_messagepump(eApp, 0, "eFilePushThread"),
+      m_messagepump(eApp, 0),
       m_run_state(0)
 {
     if (m_buffer == NULL)
@@ -359,7 +358,7 @@ eFilePushThreadRecorder::eFilePushThread
                                                                                              m_overflow_count(0),
                                                                                              m_buffer_fill(0),
                                                                                              m_stop(1),
-                                                                                             m_messagepump(eApp, 0, "eFilePushThreadRecorder")
+                                                                                             m_messagepump(eApp, 0)
 {
     m_protocol = m_stream_id = m_session_id = m_packet_no = 0;
     CONNECT(m_messagepump.recv_msg, eFilePushThreadRecorder::recvEvent);

Then you have three notable commits, March 2, March 24, and April 5:

History for lib/dvb/filepush.cpp - OpenPLi/enigma2

Framebuffer based frontend for DVB functions on Linux settop boxes - History for lib/dvb/filepush.cpp - OpenPLi/enigma2

github.com

So it appears that OpenPLi looked at OpenVix filepush.cpp and decided some edits were in order. But then a short time later reverted those edits. And at the same time OBH has what OpenVix has. Sort of along for the ride I guess, which there is nothing wrong with that...

But look at the difference between ATV and OBH filepush.cpp. The difference is so great that it will not fit on this canvass without getting an error(3.4KB). OpenVix took a different path with this script for some reason. OpenPL looked at it, tried it, and discarded it according to the commits. And I would only ask WHY? Did VIX make bad edits and that is the reason PLi reverted? Maybe this is something to look at by someone that knows...