Re: [FFmpeg-devel] [PATCH 5/8] lavu: add a JSON writer API (WIP)

2023-04-29 Thread Nicolas George
Nicolas George (12023-04-28): > Signed-off-by: Nicolas George > --- > libavutil/Makefile | 1 + > libavutil/json.c | 368 +++ > libavutil/json.h | 470 + > 3 files changed, 839 insertions(+) >

Re: [FFmpeg-devel] [PATCH 08/21] fftools/ffmpeg_filter: add filtergraph private data

2023-04-29 Thread Nicolas George
Nicolas George (12023-04-28): > And during the work of turning all into threads, opportunities to split > the structure in more logical ways with less code noise will almost > certainly present themselves. I had intended to not reply further on this, but I just had a severe case of

Re: [FFmpeg-devel] [PATCH 08/21] fftools/ffmpeg_filter: add filtergraph private data

2023-04-28 Thread Nicolas George
elegant. But if I failed to convince you or others with this mail, I will not fight this anymore. Regards, -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmp

Re: [FFmpeg-devel] [PATCH 2/8] lavu: new AVWriter API

2023-04-28 Thread Nicolas George
d one > previously in the same sentence). Disagree on this one: the last part should still be in the sentence that starts with “Note that”. There is no problem in having multiple semicolons to separate parts on the same level. Regards, -- Nicolas George signature.asc Description: PGP

Re: [FFmpeg-devel] [PATCH] lavu: header and documentation for AVWriter

2023-04-28 Thread Nicolas George
o the header. Of course. If you notice remarks that I neglected to take into consideration, please remind me of them. > And please, do not ignore or dismiss any of them. No one is out to block > your work for the sake of it. By the way, thanks for intervening on my behalf on Monday. Regards, --

Re: [FFmpeg-devel] [PATCH] lavu: header and documentation for AVWriter

2023-04-28 Thread Nicolas George
it is the main reason I write “significant more support than opposition” and not “no opposition at all”. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listin

[FFmpeg-devel] [PATCH 8/8] lavf/dump: use av_disposition_write()

2023-04-28 Thread Nicolas George
It changes the output to have underscores instead of spaces. Signed-off-by: Nicolas George --- libavformat/dump.c | 37 + 1 file changed, 1 insertion(+), 36 deletions(-) Note: I consider the change to backspaces good: now we can copy-paste from dump

[FFmpeg-devel] [PATCH 7/8] lavf/options: add av_disposition_write()

2023-04-28 Thread Nicolas George
TODO APIchanges entry Signed-off-by: Nicolas George --- libavformat/avformat.h | 14 ++ libavformat/options.c | 34 ++ 2 files changed, 48 insertions(+) This and the next patch show the kind of fruits become low-hanging thanks to the platform

[FFmpeg-devel] [PATCH 6/8] lavu: add JSON writer test (WIP)

2023-04-28 Thread Nicolas George
Signed-off-by: Nicolas George --- libavutil/Makefile | 1 + libavutil/tests/json.c | 139 +++ tests/fate/libavutil.mak | 4 ++ 3 files changed, 144 insertions(+) create mode 100644 libavutil/tests/json.c diff --git a/libavutil/Makefile b

[FFmpeg-devel] [PATCH 5/8] lavu: add a JSON writer API (WIP)

2023-04-28 Thread Nicolas George
Signed-off-by: Nicolas George --- libavutil/Makefile | 1 + libavutil/json.c | 368 +++ libavutil/json.h | 470 + 3 files changed, 839 insertions(+) create mode 100644 libavutil/json.c create mode 100644

[FFmpeg-devel] [PATCH 4/8] lavf/dump: use a writer

2023-04-28 Thread Nicolas George
Signed-off-by: Nicolas George --- libavformat/avformat.h | 21 +++ libavformat/dump.c | 318 + 2 files changed, 185 insertions(+), 154 deletions(-) This is meant to be an example, I chose a piece of code that does a lot of strings but does

[FFmpeg-devel] [PATCH 3/8] lavu/writer: add test

2023-04-28 Thread Nicolas George
Signed-off-by: Nicolas George --- libavutil/Makefile | 1 + libavutil/tests/.gitignore | 1 + libavutil/tests/writer.c | 192 + tests/fate/libavutil.mak | 4 + tests/ref/fate/writer | 36 +++ 5 files changed, 234 insertions

[FFmpeg-devel] [PATCH 2/8] lavu: new AVWriter API

2023-04-28 Thread Nicolas George
Signed-off-by: Nicolas George --- doc/avwriter_intro.md | 186 libavutil/Makefile| 2 +- libavutil/writer.c| 458 +++ libavutil/writer.h| 488 ++ 4 files changed, 1133 insertions(+), 1

[FFmpeg-devel] [PATCH 1/8] lavu: add macros to help making future-proof structures

2023-04-28 Thread Nicolas George
Signed-off-by: Nicolas George --- libavutil/extendable.h | 59 ++ 1 file changed, 59 insertions(+) create mode 100644 libavutil/extendable.h FFReservedPadding is used by the WIP JSON writer. diff --git a/libavutil/extendable.h b/libavutil/extendable.h

Re: [FFmpeg-devel] [PATCH 12/21] fftools/ffmpeg_filter: keep track of filtergraph input timebase

2023-04-28 Thread Nicolas George
gt; } InputFilterPriv; Just put it in InputFilter. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or

Re: [FFmpeg-devel] [PATCH 11/21] fftools/ffmpeg_filter: add InputFilter private data

2023-04-28 Thread Nicolas George
ame as FilterGraph: compleely useless change. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or

Re: [FFmpeg-devel] [PATCH 09/21] fftools/ffmpeg_filter: make graph_desc private

2023-04-28 Thread Nicolas George
Anton Khirnov (12023-04-27): > It is not used outside of ffmpeg_filter. > --- > fftools/ffmpeg.h| 1 - > fftools/ffmpeg_filter.c | 22 ++ > 2 files changed, 14 insertions(+), 9 deletions(-) Completely useless change. -- Nicolas George signature.

Re: [FFmpeg-devel] [PATCH 08/21] fftools/ffmpeg_filter: add filtergraph private data

2023-04-28 Thread Nicolas George
o deal with ABI stability or types declared in private headers. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visi

Re: [FFmpeg-devel] [PATCH 2/2] lavu/avassert: include config.h

2023-04-26 Thread Nicolas George
Hendrik Leppkes (12023-04-26): > This is an installed header, it cannot depend on config.h Thanks for spotting it, that was counter-intuitive since config.h is almost indispensable to its workings. Updated version attached, similar to what exists in common.h. Regards, -- Nicolas George F

[FFmpeg-devel] [PATCH 2/2] lavu/avassert: include config.h

2023-04-26 Thread Nicolas George
Fix setting the assert level. Signed-off-by: Nicolas George --- libavutil/avassert.h | 1 + 1 file changed, 1 insertion(+) diff --git a/libavutil/avassert.h b/libavutil/avassert.h index 51e462bbae..8f3f72c80c 100644 --- a/libavutil/avassert.h +++ b/libavutil/avassert.h @@ -28,6 +28,7

[FFmpeg-devel] [PATCH 1/2] lavc/intrax8: fix an assert

2023-04-26 Thread Nicolas George
Signed-off-by: Nicolas George --- libavcodec/intrax8.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavcodec/intrax8.c b/libavcodec/intrax8.c index e4c8b96c9c..c5c6727282 100644 --- a/libavcodec/intrax8.c +++ b/libavcodec/intrax8.c @@ -109,7 +109,7 @@ static inline void

Re: [FFmpeg-devel] [PATCH] lavu: header and documentation for AVWriter

2023-04-25 Thread Nicolas George
antagonizing people, and because I knew a new job would mean little time to work on it soon, I refrained from pushing the header, but consider it done anyway: the fact that the implementation of AVWriter is not present in FFmpeg is now a bug that needs to be fixed.) Regards, -- Nicolas George

Re: [FFmpeg-devel] [PATCH v2] fftools/ffmpeg_mux: fix reporting muxer EOF as error

2023-04-25 Thread Nicolas George
is no special case for EOF, unlike demuxers, and therefore this patch is a waste of code. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To

Re: [FFmpeg-devel] [PATCH 23/25] fftools/ffmpeg_filter: add filtergraph private data

2023-04-24 Thread Nicolas George
nformation hiding”, but an actual immediate benefit. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link

Re: [FFmpeg-devel] [PATCH v2] fftools/ffmpeg_mux: fix reporting muxer EOF as error

2023-04-24 Thread Nicolas George
change, because AVERROR_EOF was not allowed as a return code from a muxer: muxers return error codes, and AVERROR_EOF, despite being negative, is not an error. But in the end, you are right that the reasons given here make this patch wrong. Regards, -- Nicolas George signature.asc Descriptio

Re: [FFmpeg-devel] [PATCH 23/25] fftools/ffmpeg_filter: add filtergraph private data

2023-04-24 Thread Nicolas George
> several years. Nobody except you opposes them and the likelihood of > reversing them is extremely low. This whole thread is a pointless waste > of time that could be spend doing something actually useful. It is never too late to stop making things worse. -- Nicolas George signatu

Re: [FFmpeg-devel] [PATCH 23/25] fftools/ffmpeg_filter: add filtergraph private data

2023-04-24 Thread Nicolas George
ot need to be exposed? I have explained time and again: I oppose to any change that requires us to remember or check which structure a given field belongs to when it is not already obvious by its semantic. And again, there is nothing exposed to hide here. -- Nicolas George signature.

Re: [FFmpeg-devel] [PATCH 23/25] fftools/ffmpeg_filter: add filtergraph private data

2023-04-24 Thread Nicolas George
rude. > Also note that exactly the same pattern is used in ffmpeg_demux, > ffmpeg_mux, and ffmpeg_enc (and soon in ffmpeg_dec). If I had noticed it at the time, I would have objected the same way. -- Nicolas George signature.asc Description: PGP signature

Re: [FFmpeg-devel] [PATCH 23/25] fftools/ffmpeg_filter: add filtergraph private data

2023-04-24 Thread Nicolas George
at all, it is at least severely premature. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-deve

Re: [FFmpeg-devel] [PATCH 23/25] fftools/ffmpeg_filter: add filtergraph private data

2023-04-24 Thread Nicolas George
ot needed is a fundamental notion of mediocre programming. This patch only adds noise. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To un

Re: [FFmpeg-devel] [PATCH 04/25] fftools/ffmpeg_enc: always use video frame durations when available

2023-04-24 Thread Nicolas George
, is the timestamp of the next frame, or the timestamp of EOF for the last frame. Filtered frame durations are only a hint. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.or

Re: [FFmpeg-devel] [PATCH 23/25] fftools/ffmpeg_filter: add filtergraph private data

2023-04-24 Thread Nicolas George
ng you need in FilterGraph. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-deve

Re: [FFmpeg-devel] [PATCH v2] fftools/ffmpeg_mux: fix reporting muxer EOF as error

2023-04-24 Thread Nicolas George
F is intended, it written in the documentation. As it is, we should av_assert0(ret != AVERROR_EOF) in av_interleaved_write_frame() and fix the possible failures. Regards, -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmp

Re: [FFmpeg-devel] [PATCH v2] fftools/ffmpeg_mux: fix reporting muxer EOF as error

2023-04-23 Thread Nicolas George
Marton Balint (12023-04-23): > https://ffmpeg.org/pipermail/ffmpeg-devel/2023-February/306247.html Looks like something that sould return AVERORR(EPIPE). Regards, -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mail

Re: [FFmpeg-devel] [PATCH 3/3] avfilter/vf_uspp: 1000% faster with threads

2023-03-16 Thread Nicolas George
not 10?) Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg: reset pointer, which could be reused by Android and iOS

2023-03-03 Thread Nicolas George
ss what an extended libavfilter with movie source and sink would be. Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] [PATCH] libavformat/tcp: add local_addr/local_port for network option

2023-03-02 Thread Nicolas George
ing. Regards, -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.

Re: [FFmpeg-devel] [PATCH] libavformat/tcp: add local_addr/local_port for network option

2023-02-28 Thread Nicolas George
s->listen = 1; > } > +if (av_find_info_tag(buf, sizeof(buf), "local_port", p)) { > +s->local_port = strtol(buf, NULL, 10); > +} > +if (av_find_info_tag(buf, sizeof(buf), "local_addr", p)) { > +av_freep(>local_addr); > +s->local_addr = av_strndup(buf, strlen(buf)); > +} > if (av_find_info_tag(buf, sizeof(buf), "timeout", p)) { > s->rw_timeout = strtol(buf, NULL, 10); > } Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] [PATCH 2/3] lavc/assenc: set AV_CODEC_CAP_SINGLE_SUB_RECT

2023-02-22 Thread Nicolas George
ling the subtitles encoding API. Of course, other developers might disagree, give it a few days. Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit li

Re: [FFmpeg-devel] [PATCH 2/3] lavc/assenc: set AV_CODEC_CAP_SINGLE_SUB_RECT

2023-02-20 Thread Nicolas George
andles multiple rectangles than inventing yet another annoying flag. A single frame can be encoded into multiple packets, so that should not be hard. Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/m

Re: [FFmpeg-devel] [PATCH 3/3] ffmpeg: respect AV_CODEC_CAP_SINGLE_SUB_RECT

2023-02-20 Thread Nicolas George
t->frames_encoded++; > > -subtitle_out_size = avcodec_encode_subtitle(enc, pkt->data, > pkt->size, sub); > -if (i == 1) > -sub->num_rects = save_num_rects; > +subtitle_out_size = avcodec_encode_subtitle(enc, pkt->data, > pkt->size, _su

Re: [FFmpeg-devel] [PATCH 1/3] lavc/codec.h: add AV_CODEC_CAP_SINGLE_SUB_RECT

2023-02-20 Thread Nicolas George
flag is present but several rectangles are given? Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.o

Re: [FFmpeg-devel] [PATCH] tests: actually test yadif's 10 and 16-bit functions

2023-02-20 Thread Nicolas George
v420p10le,yadif=0 > +fate-filter-yadif16: CMD = framecrc -flags bitexact -idct simple -i > $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -flags bitexact -pix_fmt > yuv420p16le -frames:v 30 -vf scale,format=yuv420p16le,yadif=0 Moving scale before

Re: [FFmpeg-devel] [PATCH 1/2] lavfi/Makefile: Dont compile unused files.

2023-02-11 Thread Nicolas George
fdef ARCH_X86_64 > X86ASM-OBJS-$(CONFIG_NLMEANS_FILTER) += x86/vf_nlmeans.o > +endif > X86ASM-OBJS-$(CONFIG_OVERLAY_FILTER) += x86/vf_overlay.o > X86ASM-OBJS-$(CONFIG_PP7_FILTER) += x86/vf_pp7.o > X86ASM-OBJS-$(CONFIG_PSNR_FILTER)

Re: [FFmpeg-devel] [PATCH 1/2] doc/dev_community: add the addresses of the committees

2023-02-09 Thread Nicolas George
Thilo Borgman (12023-02-08): > From: Nicolas George > > Omitting the .org from the address should be protection enough > against spam spiders. > > Signed-off-by: Nicolas George > --- > doc/dev_community/community.md | 4 > 1 file changed, 4 insertions(+)

Re: [FFmpeg-devel] [PATCH] avdevice: add dxgigrab

2023-02-09 Thread Nicolas George
Timo Rothenpieler (12023-02-09): > How do you get access to the d3d hwdevice, given this lives in lavd, which > has no provisions for that? Would it work better if it was in the same library? Regards, -- Nicolas George ___ ffmpeg-devel mailin

Re: [FFmpeg-devel] [PATCH] avdevice/xcbgrab: enable window resizing

2023-02-09 Thread Nicolas George
the patch. Are you removing support for using shared memory entirely to implement this feature? If so, can we see some benchmarks comparisons? Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/lis

Re: [FFmpeg-devel] [PATCH] avfilter: use ff_inlink_make_frame_writable()

2023-02-08 Thread Nicolas George
; > if (s->outfilter != FILTER_NONE) { > out = av_frame_clone(in); > -av_frame_make_writable(out); > +ret = ff_inlink_make_frame_writable(link, ); > +if (ret < 0) { > +av_frame_free(); > +return ret; > +

Re: [FFmpeg-devel] [PATCH] avfilter: use ff_inlink_make_frame_writable()

2023-02-06 Thread Nicolas George
frame_writable(link, ); > +if (ret < 0) > +return ret; Ditto? > } > > ff_filter_execute(ctx, compute_sat_hue_metrics16, _huesat, > diff --git a/libavfilter/vf_telecine.c b/libavfilter/vf_telecine.c > index e8de63bbcf..227de6f733 100644 > --- a/libavfilter/vf_telec

Re: [FFmpeg-devel] [PATCH] lavfi: get rid of FF_INTERNAL_FIELDS

2023-02-03 Thread Nicolas George
s actually public-or-private-because-of-a-comment / private-because-of-a-ifdef. But even if the distinction really was public / private, I consider it unacceptable if we can do otherwise. And we can. Regards, -- Nicolas George ___ ffmpeg-devel mai

Re: [FFmpeg-devel] [PATCH] Waveform Archiver decoder and demuxer

2023-02-01 Thread Nicolas George
easons for it have less annoying solutions. Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmp

Re: [FFmpeg-devel] [PATCH] lavfi: get rid of FF_INTERNAL_FIELDS

2023-01-31 Thread Nicolas George
No: read the second half of the previous paragraph: two structures with common first fields are compatible types. What we have been using is a deliberately supported construct. Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpe

Re: [FFmpeg-devel] [PATCH] lavfi: get rid of FF_INTERNAL_FIELDS

2023-01-31 Thread Nicolas George
o a solution that does not require that. Your solution (a) has this flaw, plus relies on an extension that is probably not available on all supported platforms (Microsoft I am looking at you). My favor goes to (d). Probably with a stricter test for the (c) part because the inter

Re: [FFmpeg-devel] [PATCH] lavfi: get rid of FF_INTERNAL_FIELDS

2023-01-31 Thread Nicolas George
Nicolas George (12023-01-31): > > * it prevents filterlink internals from being visible in a > > public header, where they have no business being > > * it is a step towards hiding more of lavfi internals from public > > headers > > * the same pattern is alr

Re: [FFmpeg-devel] [PATCH] lavfi: get rid of FF_INTERNAL_FIELDS

2023-01-31 Thread Nicolas George
other people does not seem like your forte either. Therefore, I add this point: * If this change is applied, FFmpeg needs to find somebody else to maintain the core of libavfilter. (And I predict you will not.) -- Nicolas George signature.asc Description: PGP signature __

Re: [FFmpeg-devel] [PATCH] lavfi: get rid of FF_INTERNAL_FIELDS

2023-01-31 Thread Nicolas George
and pts=5 must be 2. This is a bad joke. -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] [PATCH] lavfi: get rid of FF_INTERNAL_FIELDS

2023-01-31 Thread Nicolas George
. Go ahead, have a vote. I no longer care. If it goes against me, that will make one less thing to consider my responsibility. We should never have asked ourselves “how should we change to entice forkers back”, we should have asked “how must the forkers change if they want to be ac

Re: [FFmpeg-devel] [PATCH] lavfi: get rid of FF_INTERNAL_FIELDS

2023-01-31 Thread Nicolas George
the fork, and we all know what happened there. And who is to blame. -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] [PATCH 4/5] lavfi/graphparser: reimplement avfilter_graph_parse* using new API

2023-01-30 Thread Nicolas George
to somebody who might be at the early stages of working on the same thing. Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or e

Re: [FFmpeg-devel] [PATCH 3/5] lavfi: add a new filtergraph parsing API

2023-01-30 Thread Nicolas George
t; +for (unsigned out = 0; out < f->nb_outputs; out++) { > +char *label = (out < p->nb_outputs) ? p->outputs[out]->label : NULL; > + > +// skip already linked outputs > +if (f->outputs[out]) > +continue; > + > +if (label) { > +AVFilterParams *po = NULL; > +unsigned idx = find_linklabel(seg, label, 0, idx_chain, > idx_filter, ); > + > +if (po) { > +ret = avfilter_link(f, out, po->filter, idx); > +if (ret < 0) > +return ret; > + > +continue; > +} > +} > + > +// if this output is unlabeled, try linking it to an unlabeled > +// input in the next non-disabled filter in the chain > +for (size_t i = idx_filter + 1; i < ch->nb_filters && !label; i++) { > +AVFilterParams *p_next = ch->filters[i]; > + > +if (!p_next->filter) > +continue; > + > +for (unsigned in = 0; in < p_next->filter->nb_inputs; in++) { > +if (!p_next->filter->inputs[in] && > +(in >= p_next->nb_inputs || !p_next->inputs[in]->label)) > { > +ret = avfilter_link(f, out, p_next->filter, in); > +if (ret < 0) > +return ret; > + > +goto cont; > +} > +} > +break; > +} > + > +ret = inout_add(outputs, f, out, label); > +if (ret < 0) > +return ret; > + > +cont:; > +} > + > +return 0; > +} > + > +int avfilter_graph_segment_link(AVFilterGraphSegment *seg, int flags, > +AVFilterInOut **inputs, > +AVFilterInOut **outputs) > +{ > +int ret; > + > +*inputs = NULL; > +*outputs = NULL; > + > +if (flags) > +return AVERROR(ENOSYS); > + > +for (size_t idx_chain = 0; idx_chain < seg->nb_chains; idx_chain++) { > +AVFilterChain *ch = seg->chains[idx_chain]; > + > +for (size_t idx_filter = 0; idx_filter < ch->nb_filters; > idx_filter++) { > +AVFilterParams *p = ch->filters[idx_filter]; > + > +if (p->filter_name) { > +ret = fail_creation_pending(seg, p->filter_name, __func__); > +goto fail; > +} > + > +if (!p->filter) > +continue; > + > +ret = link_inputs(seg, idx_chain, idx_filter, inputs); > +if (ret < 0) > +goto fail; > + > +ret = link_outputs(seg, idx_chain, idx_filter, outputs); > +if (ret < 0) > +goto fail; > +} > +} > +return 0; > +fail: > +avfilter_inout_free(inputs); > +avfilter_inout_free(outputs); > +return ret; > +} > + > +int avfilter_graph_segment_apply(AVFilterGraphSegment *seg, int flags, > + AVFilterInOut **inputs, > + AVFilterInOut **outputs) > +{ > +int ret; > + > +if (flags) > +return AVERROR(ENOSYS); > + > +ret = avfilter_graph_segment_create_filters(seg, 0); > +if (ret < 0) { > +av_log(seg->graph, AV_LOG_ERROR, "Error creating filters\n"); > +return ret; > +} > + > +ret = avfilter_graph_segment_apply_opts(seg, 0); > +if (ret < 0) { > +av_log(seg->graph, AV_LOG_ERROR, "Error applying filter options\n"); > +return ret; > +} > + > +ret = avfilter_graph_segment_init(seg, 0); > +if (ret < 0) { > +av_log(seg->graph, AV_LOG_ERROR, "Error initializing filters\n"); > +return ret; > +} > + > +ret = avfilter_graph_segment_link(seg, 0, inputs, outputs); > +if (ret < 0) { > +av_log(seg->graph, AV_LOG_ERROR, "Error linking filters\n"); > +return ret; > +} > + > +return 0; > +} This is all I have found for now, but I still need to look at the tricky parts of the code more carefully. -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] [PATCH] lavfi: get rid of FF_INTERNAL_FIELDS

2023-01-30 Thread Nicolas George
and with field is public to prefix them with link-> or li->, it would not have this issue; avoiding this requirement was a prime goal of the current implementation. At least you did not add an indirection like on some other places. -- Nicolas George ___

Re: [FFmpeg-devel] [PATCH] frame durations for framesync

2023-01-30 Thread Nicolas George
Paul B Mahol (12023-01-30): > libavfilter support overlapping frames and frames with gaps, look into > aresample filter and its async options. aresample is a very special case. And for audio filters, the duration is carried by nb_samples, not duration. -- Nicolas

Re: [FFmpeg-devel] [PATCH] frame durations for framesync

2023-01-30 Thread Nicolas George
ke a point, you will have to explain your position with a lot more than four words. -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ff

Re: [FFmpeg-devel] [PATCH] frame durations for framesync

2023-01-30 Thread Nicolas George
Paul B Mahol (12023-01-30): > It is not redundant, it describes how much frame lasts. > There is no other way to derive it. The duration of a frame is equal to the difference between its timestamp and the timestamp of the next frame. -- Nicolas

Re: [FFmpeg-devel] [PATCH 3/4] lavfi/framesync: add syncing via external timestamp map

2023-01-30 Thread Nicolas George
n a giant string with newlines in 12 3 4 > > it: definitely no. > Why not? > 'No' with no reasoning and no suggested alternative is not much of an > argument. There were four arguments. -- Nicolas George ___ ffmp

Re: [FFmpeg-devel] [PATCH] frame durations for framesync

2023-01-30 Thread Nicolas George
it entirely. Let us just clear it in buffersrc and be done with it. Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-r

Re: [FFmpeg-devel] [PATCH] frame durations for framesync

2023-01-30 Thread Nicolas George
t; --- > libavfilter/framesync.c | 20 > libavfilter/framesync.h | 15 +++ > 2 files changed, 31 insertions(+), 4 deletions(-) Looking into it right now, but I am pretty sure it cannot work. Regards, -- Nicolas George __

Re: [FFmpeg-devel] [PATCH 1/5] lavfi/avfilter: export process_options()

2023-01-27 Thread Nicolas George
aightforward. No, it is not. I have seen likely flaws, but I need time to make sure. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe

Re: [FFmpeg-devel] [PATCH 1/5] lavfi/avfilter: export process_options()

2023-01-27 Thread Nicolas George
Anton Khirnov (12023-01-27): > If nobody has further comments, I would like to push this set next week. I have already said I would have comments. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffm

Re: [FFmpeg-devel] [PATCH 3/4] lavfi/framesync: add syncing via external timestamp map

2023-01-27 Thread Nicolas George
s on output frames you need to force timestamps on input frames. And serializing timestamps in decimal in a giant string with newlines in it: definitely no. -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/ma

Re: [FFmpeg-devel] [PATCH 2/4] lavfi/framesync: reorder functions to avoid a forward declaration

2023-01-27 Thread Nicolas George
illogical. For example you separated ff_framesync_preinit() and ff_framesync_init() with functions that have nothing to do with initialization. Moving framesync_advance() closer to its call point would probably be a much better idea. So: not like this. -- Nicol

Re: [FFmpeg-devel] [PATCH 1/4] lavfi/framesync: use a local variable to shorten code

2023-01-27 Thread Nicolas George
fs->frame_ready = 1; > -if (fs->in[i].state == STATE_EOF && > -fs->in[i].after == EXT_STOP) > +if (in->state == STATE_EOF && in->after == EXT_STOP) > framesync_eof(fs); >

Re: [FFmpeg-devel] [PATCH 4/5] lavfi/graphparser: reimplement avfilter_graph_parse* using new API

2023-01-20 Thread Nicolas George
Nicolas George (12023-01-20): > - Writing a large patch for a part of the code you know well or for a > new feature is fine, of course. In fact, even for code we know well, we should consult, as long as we are not the only one who knows it well: as long as other developers might also have

Re: [FFmpeg-devel] [PATCH 4/5] lavfi/graphparser: reimplement avfilter_graph_parse* using new API

2023-01-20 Thread Nicolas George
m the get go. Also, it would have been the POLITE thing to do. It would have been the thing to do if one considers me as an equal. -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To

Re: [FFmpeg-devel] [PATCH 4/5] lavfi/graphparser: reimplement avfilter_graph_parse* using new API

2023-01-20 Thread Nicolas George
ime is very limited these days. By the way, I am baffled that you would write a 30k+ patch on code you do not maintain and do not know well without consulting the maintainer who knows it. Regards, -- Nicolas George signature.asc Description: PGP signature _

Re: [FFmpeg-devel] [PATCH v4] avformat/dvd: new dvd:// protocol for reading dvd folder/images

2023-01-17 Thread Nicolas George
ce (chapters, languages, late probing, etc.). Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] [PATCH 1/8] lavfi/vf_scale: allow passing options to swscale directly

2023-01-09 Thread Nicolas George
Anton Khirnov (12023-01-08): > Avoid using the dictionary mechanism, which is non-introspectable. As > this is the only user of AVFilter.init_dict(), this callback can now be > removed. Series looks good to me, thanks. Regards, -- Nicol

Re: [FFmpeg-devel] [PATCH v3] avformat/dvd: new dvd:// protocol for reading dvd folder/images

2023-01-09 Thread Nicolas George
but also menu data: will it be mixed along with the movie or skipped? - Are the timestamps monotonous over a whole title? - Does ffmpeg correctly detect all the subtitles streams? - Does ffmpeg correctly detect the language of audio and subtitles streams? - Are chapters detected? Regar

Re: [FFmpeg-devel] [PATCH] avformat/rtpproto: add support for RTP/UDP socket reuse

2023-01-05 Thread Nicolas George
ing with SO_REUSEADDR would not have been more straightforward anyway: the demuxer for each individual stream would have received the packets for all streams and would have to be taught to distinguish. Regards, -- Nicolas George ___ ffmpeg-devel m

Re: [FFmpeg-devel] [PATCH] avformat/rtpproto: add support for RTP/UDP socket reuse

2023-01-03 Thread Nicolas George
ot;rtp" also? Because nobody made the UDP socket context a child object of the RTP context. Regards, -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailma

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_ffplay_ffprobe_cmdutils: add -mask_url to replace the protocol address in the command with the asterisk (*)

2023-01-03 Thread Nicolas George
Wujian(Chin) (12023-01-03): > Please review it again, thanks!! You still treat the option differently without a good reason. You still do not protect credentials in options. Not acceptable. -- Nicolas George signature.asc Description: PGP signat

Re: [FFmpeg-devel] Would a crypto file be acceptable?

2022-12-28 Thread Nicolas George
ike disabling safe. Heuristics to not require safe in cases where it is quite safe would be much more acceptable than a redundant ad-hoc format. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-deve

Re: [FFmpeg-devel] Would a crypto file be acceptable?

2022-12-28 Thread Nicolas George
red it. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subj

Re: [FFmpeg-devel] [PATCH] avformat/rtpproto: add support for RTP/UDP socket reuse

2022-12-26 Thread Nicolas George
it is not a hack. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_ffplay_ffprobe_cmdutils: add -mask_url to replace the protocol address in the command with the asterisk (*)

2022-12-26 Thread Nicolas George
35,10 @@ int main(int argc, char **argv) > WriterContext *wctx; > char *buf; > char *w_name = NULL, *w_args = NULL; > -int ret, input_ret, i; > - > +int ret, input_ret, i, mask_flag; > +char **argv_copy; > init_dynload(); > +mask_flag = ge

Re: [FFmpeg-devel] Would a crypto file be acceptable?

2022-12-26 Thread Nicolas George
c keys would probably not be exploitable, but it is a half measure, too specific to a particular use case. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/li

Re: [FFmpeg-devel] Would a crypto file be acceptable?

2022-12-26 Thread Nicolas George
giving options without giving options, which is a recurring topic. Regards, -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsub

Re: [FFmpeg-devel] 答复: 答复: [PATCH] fftools/ffmpeg_ffplay_ffprobe_cmdutils: add -mask_url to replace the protocol address in the command with the asterisk (*)

2022-12-24 Thread Nicolas George
oper, and even less "what an excellent idea, let us do the work for you". -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel T

Re: [FFmpeg-devel] Would a crypto file be acceptable?

2022-12-23 Thread Nicolas George
Mark Gaiser (12022-12-23): > They actually point to N files. Could also be just 1. Wow, what an interesing and crucial distinction for the discussion. Goodbye. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mail

Re: [FFmpeg-devel] Would a crypto file be acceptable?

2022-12-23 Thread Nicolas George
o several files, yours is about a single file. Second difference: they exist out there, yours you are trying to invent just for that. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org htt

Re: [FFmpeg-devel] 答复: 答复: [PATCH] fftools/ffmpeg_ffplay_ffprobe_cmdutils:add -mask_url to replace the protocol address in the command with the asterisk (*)

2022-12-23 Thread Nicolas George
Wujian(Chin) (12022-12-23): > If you think my idea is wrong, please give your specific advice, I already have. -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.

Re: [FFmpeg-devel] 答复: [PATCH] fftools/ffmpeg_ffplay_ffprobe_cmdutils:add -mask_url to replace the protocol address in the command with the asterisk (*)

2022-12-23 Thread Nicolas George
Wujian(Chin) (12022-12-23): > I've modified most of the issues, and I've explained some of the issues that > don't. > If you don't accept my explanation, do you have any other better suggestions > and methods? I have already made a more detailed comment in the first thread. -

Re: [FFmpeg-devel] [PATCH] avformat/rtpproto: add support for RTP/UDP socket reuse

2022-12-22 Thread Nicolas George
ons are applied recursively automatically. Regards, -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above,

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_ffplay_ffprobe_cmdutils:add -mask_url to replace the protocol address in the command with the asterisk (*)

2022-12-22 Thread Nicolas George
-- > 6 files changed, 126 insertions(+), 14 deletions(-) This new patch still has issues not addressed since the first review. Regards, -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ff

Re: [FFmpeg-devel] 答复: [PATCH] fftools/ffmpeg_ffplay_ffprobe_cmdutils: add -mask_url to replace the protocol address in the command with the asterisk (*)

2022-12-22 Thread Nicolas George
k_url and copy argv in this place. Otherwise, do you have any > other suggestions? Use the normal options parsing system. Regards, -- Nicolas George signature.asc Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https:

Re: [FFmpeg-devel] Would a crypto file be acceptable?

2022-12-22 Thread Nicolas George
within [1]. My proposed format would be: > > --- > CRYPTO-VERSION:1 > CRYPTO-KEY:URI:. > CRYPTO-IV:URI:. > encrypted_file > --- The concat demuxer can already contain options, and despite is name it can be used with a single file. Regards, -- Nicola

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_ffplay_ffprobe_cmdutils: add -mask_url to replace the protocol address in the command with the asterisk (*)

2022-12-19 Thread Nicolas George
; from other users to completely avoid leaking the password. Unfortunately, as far as I know, we do not have any mechanism of the kind. That would be useful. Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_ffplay_ffprobe_cmdutils: add -mask_url to replace the protocol address in the command with the asterisk (*)

2022-12-19 Thread Nicolas George
Gyan Doshi (12022-12-19): > Not the case. They are used for 'end', 'finish', retries ...etc Insufficient vigilance before accepting these changes. But not a reason to accept more. Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffm

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_ffplay_ffprobe_cmdutils: add -mask_url to replace the protocol address in the command with the asterisk (*)

2022-12-19 Thread Nicolas George
r some amount of time, until this is done, it would expose the full > plaintext URL still, no? This is unavoidable. Still, having sensitive information visible for a fraction of a second is better than having sensitive information visible for the length of a playback or transcoding process. Regard

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_ffplay_ffprobe_cmdutils: add -mask_url to replace the protocol address in the command with the asterisk (*)

2022-12-19 Thread Nicolas George
, "-mask_url")) { > +argv[1] = argv[0]; > +maskFlag = 1; > +argc--; > +argv++; > +} > > #if HAVE_THREADS > ret = pthread_mutex_init(_mutex, NULL); > @@ -4056,8 +4063,10 @@ int main(int argc, char **argv) > #endif > > show_banner(argc, argv, options); > -parse_options(NULL, argc, argv, options, opt_input_file); > - > +argv2 = copy_argv(argc, argv); > +parse_options(NULL, argc, argv2, options, opt_input_file); > +if (maskFlag) > +param_masking(argc, argv); > if (do_show_log) > av_log_set_callback(log_callback); > > @@ -4173,6 +4182,7 @@ end: > av_freep(_format); > av_freep(_intervals); > av_hash_freep(); > +free_pp(argc, argv2); > > uninit_opts(); > for (i = 0; i < FF_ARRAY_ELEMS(sections); i++) Regards, -- Nicolas George ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".

<    1   2   3   4   5   6   7   8   9   10   >