I don't know about "should" or not, but it's not the only pre-installed plugin that's not in the easy-4-ui source. EPGSearch and AutoTimers are commonly-used pre-installed plugins that aren't in the easy-ui-4 source.adoxa wrote:... Since it's included with the release, shouldn't its source be part of easy-ui-4, too?
mcut.cc source code and bugs
-
- Wizard God
- Posts: 32709
- Joined: Tue Sep 04, 2007 13:49
- Location: Canberra; Black Mountain Tower transmitters
mcut.cc source code and bugs
Peter
T4 HDMI
U4, T4, T3, T2, V2 test/development machines
Sony BDV-9200W HT system
LG OLED55C9PTA 55" OLED TV
T4 HDMI
U4, T4, T3, T2, V2 test/development machines
Sony BDV-9200W HT system
LG OLED55C9PTA 55" OLED TV
Re: SDK 16.1.20170311 available
Code: Select all
root@devt4:~# opkg search *mcut*
enigma2-plugin-extensions-moviecut - 2.0+git6295+146ea49-r10.3
root@devt4:~# opkg info enigma2-plugin-extensions-moviecut
Deprecated version constraint '>' was used with the same meaning as '>='. Use '>>' for LATER constraint.
Package: enigma2-plugin-extensions-moviecut
Version: 2.0+git6295+146ea49-r10.3
Depends: libgcc1 (>= 5.3.0), libstdc++6 (>= 5.3.0), libc6 (>= 2.23)
Status: install ok installed
Section: base
Architecture: inihdp
Maintainer: OE-Alliance team
MD5Sum: f5f061fb2dd8e18dd0b89dc6d849a211
Size: 29162
Filename: enigma2-plugin-extensions-moviecut_2.0+git6295+146ea49-r10.3_inihdp.ipk
Source: git://github.com/beyonwiz/enigma2-plugins.git;protocol=git;branch=2.3 file://pluginnotwanted.patch
Description: extensions-moviecut Perform the cuts specified with the Cutlist editor
Installed-Size: 82605
Installed-Time: 1489233857
Code: Select all
opkg install enigma2-plugin-extensions-moviecut-src
Re: SDK 16.1.20170311 available
That's what I did, but the binary I generated from source is apparently different to the binary that comes with the firmware, since the original will stop cutting prematurely (only rarely, but still) whereas the generated one will keep going.
Re: SDK 16.1.20170311 available
Have you got a test case?
I had a look at mcut.cc and it's a pretty sad piece of code. Lots of signed 32-bit arithmetic in use, so I suspect that the failures will be on any quantities that overflow about 2GB value.
I had to fix some 32-bit issues in createapscfiles.cc, which was written by the same author.
I had a look at mcut.cc and it's a pretty sad piece of code. Lots of signed 32-bit arithmetic in use, so I suspect that the failures will be on any quantities that overflow about 2GB value.
I had to fix some 32-bit issues in createapscfiles.cc, which was written by the same author.
Re: mcut.cc source code and bugs
BTW: The mcut.cc code has not seen any changes for about 6 years.
I'm pretty sure that editing recordings on enigma2 boxes is a very rare thing, so most of the time bugs and shortcomings related to such activities would go unnoticed. That's not to say that it shouldn't be fixed. Having a test case will help. A pull-request with a fix would be even better.
I'm pretty sure that editing recordings on enigma2 boxes is a very rare thing, so most of the time bugs and shortcomings related to such activities would go unnoticed. That's not to say that it shouldn't be fixed. Having a test case will help. A pull-request with a fix would be even better.
Re: mcut.cc source code and bugs
Sorry, I'm not about to go through all the hassle of setting up a PR for a single-line change - it seems putting brackets around the shift to remove the warning is what fixed the problem. However, I don't seem to have any bad recordings atm (having deleted the test case when it worked) to actually confirm it.
Re: mcut.cc source code and bugs
OK, if creating a pull request is too much overhead and you don't have a test case, can you at least be a bit more specific about what you changed to fix the problem? Perhaps a diff of that single line change?
Re: mcut.cc source code and bugs
Code: Select all
--- mcut.org 2017-03-11 11:43:26 +1000
+++ mcut.cc 2017-03-25 17:25:00 +1000
@@ -200,7 +200,7 @@
if (endp)
break;
if (buf0[1] - buf1[1] > 45000 || buf1[1] - buf0[1] > 900000) {
- if (absless(buf1[1] + ((long long int)1)<<33 - buf0[1], 900000))
+ if (absless(buf1[1] + (((long long int)1)<<33) - buf0[1], 900000))
time_offset -= ((long long int)1)<<33;
else
time_offset += buf1[1] - buf0[1];
@@ -622,7 +622,7 @@
int main(int argc, char* argv[])
{
- int f_ts, f_out, f_cuts, f_cutsout, f_ap, f_apout, f_sc, f_scout, f_meta, f_metaout, f_eit, f_eitout;
+ int f_ts, f_out, f_cuts, f_cutsout, f_ap, f_apout, f_sc, f_scout = -1, f_meta, f_metaout, f_eit, f_eitout;
char* tmpname;
const char* suff = 0;
char* inname = 0;
Re: mcut.cc source code and bugs
Thanks. I applied a few more tweaks and created a pull request upstream. It was merged and I pulled it down into beyonwiz fork.
It should appear in the next beta build.
It should appear in the next beta build.