newer ffmpeg broken?

[Selur]

Using ffmpeg build at the end of last year:
ffmpeg version N-112974-gd324415563-g5256b2fbe6+2 Copyright (c) 2000-2023 the FFmpeg developers
built with gcc 13.2.0 (Rev3, Built by MSYS2 project)
configuration: --pkg-config=pkgconf --cc='ccache gcc' --cxx='ccache g++' --ld='ccache g++' --extra-cxxflags=-fpermissive --extra-cflags=-Wno-int-conversion --disable-autodetect --enable-amf --enable-bzlib --enable-cuda --enable-cuvid --enable-d3d11va --enable-dxva2 --enable-iconv --enable-lzma --enable-nvenc --enable-zlib --enable-sdl2 --enable-ffnvcodec --enable-nvdec --enable-cuda-llvm --disable-doc --enable-libmp3lame --enable-libopus --enable-libvorbis --enable-libvpx --enable-libx264 --enable-libx265 --enable-libdav1d --enable-libaom --disable-debug --enable-fontconfig --enable-libass --enable-libbluray --enable-libfreetype --enable-libmfx --enable-libmysofa --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvo-amrwbenc --enable-libwebp --enable-libxml2 --enable-libzimg --enable-libshine --enable-gpl --enable-avisynth --enable-libxvid --enable-libopenmpt --enable-version3 --enable-librav1e --enable-libsrt --enable-libgsm --enable-libvmaf --enable-libsvtav1 --enable-mbedtls --extra-cflags=-DLIBTWOLAME_STATIC --extra-libs=-lstdc++ --extra-cflags=-DLIBXML_STATIC --extra-libs=-liconv --disable-w32threads
libavutil 58. 33.100 / 58. 33.100
libavcodec 60. 35.100 / 60. 35.100
libavformat 60. 18.100 / 60. 18.100
libavdevice 60. 4.100 / 60. 4.100
libavfilter 9. 14.100 / 9. 14.100
libswscale 7. 6.100 / 7. 6.100
libswresample 4. 13.100 / 4. 13.100
libpostproc 57. 4.100 / 57. 4.100
Hyper fast Audio and Video encoder
usage: ffmpeg [options] [[infile options] -i infile]... {[outfile options] outfile}...
and:
ffmpeg_working -y -i "G:\TestClips&Co\test.avi" -an -sn -vf zscale=rangein=tv:range=tv -pix_fmt yuv420p -f yuv4mpegpipe j:\tmp\working.y4m
the output is fine.

using an ffmpeg verision I build yesterday:
ffmpeg version N-113316-g7f20bca34d-g564a15b2ee+2 Copyright (c) 2000-2024 the FFmpeg developers
built with gcc 13.2.0 (Rev3, Built by MSYS2 project)
configuration: --pkg-config=pkgconf --cc='ccache gcc' --cxx='ccache g++' --ld='ccache g++' --extra-cxxflags=-fpermissive --extra-cflags=-Wno-int-conversion --disable-autodetect --enable-amf --enable-bzlib --enable-cuda --enable-cuvid --enable-d3d11va --enable-dxva2 --enable-iconv --enable-lzma --enable-nvenc --enable-zlib --enable-sdl2 --enable-ffnvcodec --enable-nvdec --enable-cuda-llvm --disable-doc --enable-gmp --enable-libmp3lame --enable-libopus --enable-libvorbis --enable-libvpx --enable-libx264 --enable-libx265 --enable-libdav1d --enable-libaom --disable-debug --enable-fontconfig --enable-libass --enable-libbluray --enable-libfreetype --enable-libmfx --enable-libmysofa --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvo-amrwbenc --enable-libwebp --enable-libxml2 --enable-libzimg --enable-libshine --enable-gpl --enable-avisynth --enable-libxvid --enable-libopenmpt --enable-version3 --enable-librav1e --enable-libsrt --enable-libgsm --enable-libvmaf --enable-libsvtav1 --enable-gnutls --extra-cflags=-DLIBTWOLAME_STATIC --extra-libs=-lstdc++ --extra-cflags=-DLIBXML_STATIC --extra-libs=-liconv --disable-w32threads
libavutil 58. 36.101 / 58. 36.101
libavcodec 60. 37.100 / 60. 37.100
libavformat 60. 20.100 / 60. 20.100
libavdevice 60. 4.100 / 60. 4.100
libavfilter 9. 17.100 / 9. 17.100
libswscale 7. 6.100 / 7. 6.100
libswresample 4. 13.100 / 4. 13.100
libpostproc 57. 4.100 / 57. 4.100
Universal media converter
usage: ffmpeg [options] [[infile options] -i infile]... {[outfile options] outfile}...
and
ffmpeg -y -i "G:\TestClips&Co\test.avi" -an -sn -vf zscale=rangein=tv:range=tv -pix_fmt yuv420p -f yuv4mpegpipe j:\tmp\broken.y4m
the output is broken.
working:
YUV4MPEG2 W640 H352 F25:1 Ip A1:1 C420mpeg2 XYSCSS=420MPEG2 XCOLORRANGE=LIMITED
FRAME
QQQQQQQQTTTTTTTTTTTTTTTTVVVVVVVVVWWXXYYZXXXXXXXX\\\\\\\\[[[[[[[[\\\\\\\\]]]]]]]]^^^^^^^^ZZZZZZZZ]]]]]]]]ZZZZZZZZ^^]]\[[ZZZZZZZZZ[ZZZ[ZWTLMMNNOOPMNOPRSTUSSSSSSSSYXXWVUTTXXXYZZ[[[[[[[[[[]^^__``adeghhgedghiklmnnppopqtwyvwy|��������������������������������S$%!!#!*S~����������������~|{zzwrn}{eORdpvg[OQ^konrqoljgedcdefhikkjjjjjjjjijjjifdbccccccccdddddddd````````aaaaaaaaaaaaaaaabbbbbbbbccccccccffffffffggggffffhhhhhhhhjjkkllmmllllllllpqrtttssqqrstuuvyzz{|}~}}}}}}}}~��������������������������}|{zyxwvuutvvutsrqqqqponmlljjihgfeeddccbbaa`^]^_`_]]]]]]]]]]\\[[ZZZVUUTTSSRPPPPPPPPPPPPPPPPMMMMMMMMJJJJJJJJLKJIGFEDEEFGGHIIFFFFFFFFCCCCCCCCCBBAA@@?QQQQQQQQTTTTTTTTTTTTTTTTVVVVVVVVVWWXXYYZXXXXXXXX[[[[[[[[[[[[[[[[\\\\\\\\]]]]]]]]^^^^^^^^]]]]]]]]]]]]]]]][[[[[[[[ZZZZZZZZZZZZZZZZ[ZZZ[ZVTLMMNNOOPLLMOPQRSSSSSSSSSTTTTTTTTVVWWXYYY[[[[[[[[]^^__``a^`bdefedlllmmnnnssrqqsuvz{|}�������������������������������nD(
broken:
YUV4MPEG2 W640 H352 F25:1 Ip A1:1 C420mpeg2 XYSCSS=420MPEG2
FRAME
~~~~~~~~~~~~���������������~~~~~~~~~~��������������������������������������������������������������������������������������������������jjjjijidccccdddd````aaaaaaaabbbbccccffffggffhhhhjklmllllprtsqrtuyz|~}}}}~�������������|zxvuvusqqpnljigedcba`]__]]]]]\[ZVUTSPPPPPPPPMMMMJJJJLJGEEFGIFFFFCCCCCBA@
(output seems to be broken, even when using '-loglevel quiet' and '-f rawvideo')

the broken output looks like this:
https://i.ibb.co/sQs0Td5/grafik.png (https://ibb.co/wCp9nFJ)

both versions were build with mbas (https://github.com/m-ab-s/media-autobuild_suite) does anyone else have this issue with a new build ffmpeg? (trying to figure out whether this is a. a bug in ffmpeg b. a bug in mbas or c. something going awry with my build system)

Cu Selur

[richardpl]

I confirm its really broken. FFmpeg becomes corporate low quality code dumpster. Stop using it, switch to gstreamer.

[Selur]

Hmm,... seems like it's related to '-vf zscale=rangein=tv:range=tv' without it, the output is fine, independent of the version.

[ProWo]

There are problems with JSON too.
6.1.1. is ok.

[Selur]

Just noticed there was a fix for vf_zscale just now (https://git.ffmpeg.org/gitweb/ffmpeg.git/commit/bfa1b7577dd646e84acafd0c82a8c2c6fe9c2a0a)
=> will build a new ffmpeg version and report later whether that commit fixed the problem

[Selur]

Yes, that commit fixes the problem.
Nice!

Cu Selur

[VoodooFX]

What about ffmpeg v6 is producing different wavs than v5?


ffmpeg -i input.mp3 -ar 16000 -ac 1 -c:a pcm_s16le output.wav

[microchip8]

What about ffmpeg v6 is producing different wavs than v5?


ffmpeg -i input.mp3 -ar 16000 -ac 1 -c:a pcm_s16le output.wav


add -flags +bitexact

[VoodooFX]

add -flags +bitexact

Has no effect with both v5 & v6.


ffmpeg -i input.mp3 -ar 16000 -ac 1 -c:a pcm_s16le -flags +bitexact output.wav

[richardpl]

It is just "-bitexact"

[VoodooFX]

It is just "-bitexact"

This has effect, but there are still same differences in wavs. [minus a diff in the headers]


Download the test files: https://we.tl/t-XyKI8cnSy4


ffmpeg -i original.aac -ar 16000 -ac 1 -c:a pcm_s16le output.wav


EDIT:
There are 30 one byte [±1] differences. [Windows]
Btw, strangely, a Linux user reported 41 differences with same(?) Linux versions.

[Selur]

got a question about bitexact:
-bitexact (input/output)

Enable bitexact mode for (de)muxer and (de/en)coder source: https://ffmpeg.org/ffmpeg-all.html#Advanced-options
What does this actually mean? For encoders, I get that the output might not be always deterministic, but especially for demuxers and decoders shouldn't the output always the same?

Cu Selur

[VoodooFX]

So, no one knows why wavs are different?
Those one byte differences propagate into the huge differences in AI. Found it when an user reported various hallucinations and I couldn't reproduce it.

[lvqcl]

Different compilers, different floating point optimizations, etc.

MP3 decoding is not required to be bit-exact and different decoders can produce slightly different results.

[Selur]

Okay, but shouldn't demuxers be always output the same?

[VoodooFX]

Different compilers, different floating point optimizations, etc.

MP3 decoding is not required to be bit-exact and different decoders can produce slightly different results.

Actually it's about AAC, do the same stuff applies for it too?

[lvqcl]

Actually it's about AAC, do the same stuff applies for it too?
Yes. MP3, AAC, Vorbis, Opus, etc.

[FranceBB]

That's not the only problem in the latest master.
This leads to a memory leak:

ffmpeg.exe -hide_banner -i "raw_video.h264" -i "Doc Ljubicic.avs" -map 0:0 -c:v copy -map 1:1 -c:a pcm_s24le -ar 48000 -f mxf "semi-final_output.mxf"

yet it worked just fine in a version from 2020 I had lying around (lavc58.82.100)...

[Selur]

Using:
ffmpeg -y -loglevel debug -noautorotate -nostdin -hwaccel auto -threads 1 -i "C:\Users\Selur\Desktop\Test\Doc Martin (S09E01) To The Lighthouse-001.mkv" -map 0:0 -an -sn -vf scale,zscale=rangein=tv:range=tv,scale -pix_fmt yuv420p -vsync 0 -f rawvideo NUL
ffmpeg version N-113317-g02fb3a775e-gbfa1b7577d+2 Copyright (c) 2000-2024 the FFmpeg developers
built with gcc 13.2.0 (Rev3, Built by MSYS2 project)
configuration: --pkg-config=pkgconf --cc='ccache gcc' --cxx='ccache g++' --ld='ccache g++' --extra-cxxflags=-fpermissive --extra-cflags=-Wno-int-conversion --disable-autodetect --enable-amf --enable-bzlib --enable-cuda --enable-cuvid --enable-d3d11va --enable-dxva2 --enable-iconv --enable-lzma --enable-nvenc --enable-zlib --enable-sdl2 --enable-ffnvcodec --enable-nvdec --enable-cuda-llvm --disable-doc --enable-gmp --enable-libmp3lame --enable-libopus --enable-libvorbis --enable-libvpx --enable-libx264 --enable-libx265 --enable-libdav1d --enable-libaom --disable-debug --enable-fontconfig --enable-libass --enable-libbluray --enable-libfreetype --enable-libmfx --enable-libmysofa --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvo-amrwbenc --enable-libwebp --enable-libxml2 --enable-libzimg --enable-libshine --enable-gpl --enable-avisynth --enable-libxvid --enable-libopenmpt --enable-version3 --enable-librav1e --enable-libsrt --enable-libgsm --enable-libvmaf --enable-libsvtav1 --enable-gnutls --extra-cflags=-DLIBTWOLAME_STATIC --extra-libs=-lstdc++ --extra-cflags=-DLIBXML_STATIC --extra-libs=-liconv --disable-w32threads
libavutil 58. 36.101 / 58. 36.101
libavcodec 60. 37.100 / 60. 37.100
libavformat 60. 20.100 / 60. 20.100
libavdevice 60. 4.100 / 60. 4.100
libavfilter 9. 17.100 / 9. 17.100
libswscale 7. 6.100 / 7. 6.100
libswresample 4. 13.100 / 4. 13.100
libpostproc 57. 4.100 / 57. 4.100
...
[AVFilterGraph @ 000002266992f780] Setting 'rangein' to value 'tv'
[h264 @ 0000022669b5e980] ct_type:1 pic_struct:0
[AVFilterGraph @ 000002266992f780] Setting 'range' to value 'tv'
detected 32 logical cores
[Parsed_scale_0 @ 0000022675c60540] w:iw h:ih flags:'' interl:0
[Parsed_scale_2 @ 000002266a05ea80] w:iw h:ih flags:'' interl:0
[graph 0 input from stream 0:0 @ 0000022669b5d1c0] Setting 'video_size' to value '1920x1080'
[graph 0 input from stream 0:0 @ 0000022669b5d1c0] Setting 'pix_fmt' to value '23'
[graph 0 input from stream 0:0 @ 0000022669b5d1c0] Setting 'time_base' to value '1/1000'
[graph 0 input from stream 0:0 @ 0000022669b5d1c0] Setting 'pixel_aspect' to value '1/1'
[graph 0 input from stream 0:0 @ 0000022669b5d1c0] Setting 'colorspace' to value 'bt709'
[graph 0 input from stream 0:0 @ 0000022669b5d1c0] Setting 'range' to value 'tv'
[graph 0 input from stream 0:0 @ 0000022669b5d1c0] Setting 'frame_rate' to value '24000/1001'
[h264 @ 0000022669b5e980] nal_unit_type: 8(PPS), nal_ref_idc: 3
[h264 @ 0000022669b5e980] nal_unit_type: 6(SEI), nal_ref_idc: 0
[h264 @ 0000022669b5e980] nal_unit_type: 1(Coded slice of a non-IDR picture), nal_ref_idc: 0
Last message repeated 3 times
[graph 0 input from stream 0:0 @ 0000022669b5d1c0] w:1920 h:1080 pixfmt:nv12 tb:1/1000 fr:24000/1001 sar:1/1 csp:bt709 range:tv
[h264 @ 0000022669b5e980] ct_type:1 pic_struct:0
[format @ 00000226698b4040] Setting 'pix_fmts' to value 'yuv420p'
[AVFilterGraph @ 000002266992f780] query_formats: 6 queried, 15 merged, 0 already done, 0 delayed
[Parsed_scale_0 @ 0000022675c60540] picking yuv420p out of 53 ref:nv12 alpha:0
[Parsed_zscale_1 @ 0000022675c5ff80] picking yuv420p out of 53 ref:yuv420p alpha:0
[Parsed_scale_0 @ 0000022675c60540] w:1920 h:1080 fmt:nv12 csp:bt709 range:tv sar:1/1 -> w:1920 h:1080 fmt:yuv420p csp:bt709 range:tv sar:1/1 flags:0x00000004
[Parsed_scale_2 @ 000002266a05ea80] w:1920 h:1080 fmt:yuv420p csp:unknown range:tv sar:1/1 -> w:1920 h:1080 fmt:yuv420p csp:unknown range:tv sar:1/1 flags:(null)
[Parsed_zscale_1 @ 0000022675c5ff80] code 3074: no path between colorspaces
Last message repeated 31 times
[vf#0:0 @ 0000022669b54480] Error while filtering: Generic error in an external library
[vf#0:0 @ 0000022669b54480] Task finished with error code: -542398533 (Generic error in an external library)
[vf#0:0 @ 0000022669b54480] Terminating thread with return code -542398533 (Generic error in an external library)
[vost#0:0/rawvideo @ 0000022669d609c0] Encoder thread received EOF
full output: https://pastebin.com/4WUdrte4
works if I remove "-hwaccel auto"

am I missing something? Is it just me, or can others confirm this too?

[Ritsuka]

I don't think zscale supports biplanar formats, vf_zscale.c lists only planar ones.
Anyway, what's "zscale=rangein=tv:range=tv" supposed to do?

[Selur]

Anyway, what's "zscale=rangein=tv:range=tv" supposed to do?
make sure that content is interpreted as tv scale.
I don't think zscale supports biplanar formats, vf_zscale.c lists only planar ones.
Strange, I have been using this code for years,...

I switched to an older ffmpeg build:
ffmpeg version N-112974-gd324415563-g5256b2fbe6+2 Copyright (c) 2000-2023 the FFmpeg developers
built with gcc 13.2.0 (Rev3, Built by MSYS2 project)
configuration: --pkg-config=pkgconf --cc='ccache gcc' --cxx='ccache g++' --ld='ccache g++' --extra-cxxflags=-fpermissive --extra-cflags=-Wno-int-conversion --disable-autodetect --enable-amf --enable-bzlib --enable-cuda --enable-cuvid --enable-d3d11va --enable-dxva2 --enable-iconv --enable-lzma --enable-nvenc --enable-zlib --enable-sdl2 --enable-ffnvcodec --enable-nvdec --enable-cuda-llvm --disable-doc --enable-libmp3lame --enable-libopus --enable-libvorbis --enable-libvpx --enable-libx264 --enable-libx265 --enable-libdav1d --enable-libaom --disable-debug --enable-fontconfig --enable-libass --enable-libbluray --enable-libfreetype --enable-libmfx --enable-libmysofa --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvo-amrwbenc --enable-libwebp --enable-libxml2 --enable-libzimg --enable-libshine --enable-gpl --enable-avisynth --enable-libxvid --enable-libopenmpt --enable-version3 --enable-librav1e --enable-libsrt --enable-libgsm --enable-libvmaf --enable-libsvtav1 --enable-mbedtls --extra-cflags=-DLIBTWOLAME_STATIC --extra-libs=-lstdc++ --extra-cflags=-DLIBXML_STATIC --extra-libs=-liconv --disable-w32threads
libavutil 58. 33.100 / 58. 33.100
libavcodec 60. 35.100 / 60. 35.100
libavformat 60. 18.100 / 60. 18.100
libavdevice 60. 4.100 / 60. 4.100
libavfilter 9. 14.100 / 9. 14.100
libswscale 7. 6.100 / 7. 6.100
libswresample 4. 13.100 / 4. 13.100
libpostproc 57. 4.100 / 57. 4.100
Hyper fast Audio and Video encoder
and that one works fine.

[Ritsuka]

Well, the whole things was a bit (maybe more) broken before, so it actually worked for the wrong reasons. Try to convert first to yuv420. Anyway, using zscale just to set the color range looks so wrong, isn't there a better way in FFmpeg cli to override the input range?

[richardpl]

zscale is for conversion not metadata modification. Even people with IQ < 50 understand this.

[Selur]

For those interessted, it works again in:
ffmpeg version N-113431-g377391b414-g763e31a8d3+2 Copyright (c) 2000-2024 the FFmpeg developers
built with gcc 13.2.0 (Rev3, Built by MSYS2 project)
configuration: --pkg-config=pkgconf --cc=gcc --cxx=g++ --ld=g++ --extra-cxxflags=-fpermissive --extra-cflags=-Wno-int-conversion --disable-autodetect --enable-amf --enable-bzlib --enable-cuda --enable-cuvid --enable-d3d11va --enable-dxva2 --enable-iconv --enable-lzma --enable-nvenc --enable-zlib --enable-sdl2 --enable-ffnvcodec --enable-nvdec --enable-cuda-llvm --disable-doc --enable-gmp --enable-libmp3lame --enable-libopus --enable-libvorbis --enable-libvpx --enable-libx264 --enable-libx265 --enable-libdav1d --enable-libaom --disable-debug --enable-fontconfig --enable-libass --enable-libbluray --enable-libfreetype --enable-libmfx --enable-libmysofa --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvo-amrwbenc --enable-libwebp --enable-libxml2 --enable-libzimg --enable-libshine --enable-gpl --enable-avisynth --enable-libxvid --enable-libopenmpt --enable-version3 --enable-librav1e --enable-libsrt --enable-libgsm --enable-libvmaf --enable-libsvtav1 --enable-gnutls --extra-cflags=-DLIBTWOLAME_STATIC --extra-libs=-lstdc++ --extra-cflags=-DLIBXML_STATIC --extra-libs=-liconv --disable-w32threads
libavutil 58. 36.101 / 58. 36.101
libavcodec 60. 38.100 / 60. 38.100
libavformat 60. 20.100 / 60. 20.100
libavdevice 60. 4.100 / 60. 4.100
libavfilter 9. 17.100 / 9. 17.100
libswscale 7. 6.100 / 7. 6.100
libswresample 4. 13.100 / 4. 13.100
libpostproc 57. 4.100 / 57. 4.100
Universal media converter
usage: ffmpeg [options] [[infile options] -i infile]... {[outfile options] outfile}...

[StainlessS]

Even people with IQ < 50 understand this.

Must be nice to have a brain the size of a planet (or maybe its just the ego).

[manolito]

Must be nice to have a brain the size of a planet (or maybe its just the ego).

:p:goodpost:

[FranceBB]

Yeah well, that's the kind of replies people usually get in the FFMpeg bug tracker.
Luckily Doom9 is a safe haven for both users and developers, but every now and then there are FFMpeg-related questions and FFMpeg-guys style related answers...
Don't get me wrong, those guys do quite remarkable work and single handedly support lots of features by constantly committing code, but they can be harsh most of the times.
I know the feeling 'cause unfortunately, every now and then, I find some non Avisynth related bugs that affect the stuff I work on and I have to head to those dreadful places only to be battered... :(
(I mean, they generally fix the bug I report, but I also get battered in the meantime...).

Still, I'm glad to see that the regressions are being fixed.
I might as well try to see if mine got fixed as well (I mean the raw_video.h264 muxing issue in mxf). https://trac.ffmpeg.org/ticket/10966

[richardpl]

Doom9 is already known as very low quality community.

[tebasuna51]

Doom9 is already known as very low quality community.

Thats is a clear rule 4 violation. You gain a strike.

[richardpl]

You are free to censor because you have power in your hands.

[Grimsdyke]

Aaah, insulting people/the community and when called out for it complain about censorship. That's classy ...

[Emulgator]

Those small pinches of irony ?
Well, here at my place a BigThankYou from me towards the FFmpeg developers for 23 years of coding, builds and stuff seems in order.

[rwill]

Two things from my side ..

1) richardpl wrote "Doom9 is already known as very low quality community.". He did not write that he thinks Doom9 is one but is perceived as such by some people, maybe in certain bubbles. I can confirm that Doom9 has this reputation in certain bubbles by interacting with people outside of Doom9.

2) tebasuna51 wrote that richardpl made a rule 4 violation and thus gained a strike. On the sanctions page it says: "A violation of rules 4 or 6 will result in two strikes (per offense) rather than one, because we feel that violating them does serious harm to this community.".

[huhn]

rule 17.
making this rule 3 breach.

the moderation of this forum is very fair calling this out without retroactive censoring show this yet again.