00:13:40
##forth
<cleobuline>
next i will change the management on the memory varaiable string and array , instead of a array il will use linked list
09:09:08
##forth
<xentrac>
yeah, I have a collection of recipes for it in ffmpeg-notes.md in pavnotes2
09:09:46
##forth
<KipIngram>
It doesn't help that the options seem to have evolved over time, and there's still lots of older, no longer correct info online.
09:09:57
##forth
<KipIngram>
I will take a look at your notes - thanks.
09:10:53
##forth
<xentrac>
To see what a video is, `ffprobe f.mp4`; To set a bandwidth budget for video use `-b:v`; To set a bandwidth budget for audio use `-b:a`; To reduce bandwidth further, reduce the framerate and resolution and audio budget; etc.
09:10:53
##forth
<xentrac>
Hmm, I think they're usually pretty good about maintaining backwards compatibility
09:11:11
##forth
<KipIngram>
I'm using it on an rtsp stream from a camera.
09:12:04
##forth
<KipIngram>
I've currently got a command that will catch the stream and save it in 15-minute timestamped segments. That works well, but it will run for a few hours and then drop - I assume due to some network error.
14:19:34
##forth
<cleobuline>
mforth: STRING MACRON
14:19:57
##forth
<cleobuline>
mforth: " Macron est un saint homme " MACRON !
14:20:14
##forth
<cleobuline>
mforth: MACRON @ PRINT
14:20:14
##forth
<mforth>
Macron est un saint homme
14:42:01
##forth
<cleobuline>
mforth: WORDS
14:42:01
##forth
<mforth>
.S . + - * / MOD DUP DROP SWAP OVER ROT >R R> R@ = < > AND OR NOT CR EMIT VARIABLE @ ! DO LOOP I WORDS LOAD CREATE ALLOT ." CLOCK BEGIN WHILE REPEAT AGAIN SQRT UNLOOP +LOOP PICK EMIT CR CLEAR-STACK PRINT FORGET STRING " 2DROP DP DOUBLE FACT POW FIBONACCI COUNTDOWN TUCK SUM_SQUARE CUBE SUM_CUBES RECUNACCI CAT :D PGCD PRIME? SEED COUNT NUMS STARS INIT-RANDOM RANDOM RAND INIT-NUMS SHUFFLE-NUMS
15:11:42
##forth
<KipIngram>
:-)
15:51:03
##forth
<cleobuline>
mforth: 1000000 NUM-TO-BIN
15:51:03
##forth
<mforth>
11110100001001000000
15:51:48
##forth
<cleobuline>
mforth: 1000000 10 << NUM-TO-BIN
15:51:48
##forth
<mforth>
111101000010010000000000000000
17:10:53
##forth
<cleobuline>
mforth: WORDS
17:31:36
##forth
<cleobuline>
multi: LOAD "test.fth"
17:31:43
##forth
<cleobuline>
multi: 1 31 << 1 - PRIME? .
17:32:40
##forth
<cleobuline>
mforth: LOAD "test.fth"
17:32:54
##forth
<cleobuline>
mforth: 1 31 << 1 - PRIME?
17:33:05
##forth
<cleobuline>
mforth: .
17:33:05
##forth
<mforth>
1
17:35:30
##forth
<KipIngram>
xentrac: Turns out that ffmpeg needs the reconnect option specified as part of the URL. So generally 'ffmpeg <input options> -i rtsp://ip:port/channel?reconnect=1 <output options>'. And the help/docs don't make that AT ALL apparent. I was trying to put it in <input options> and it just wasn't working. It seems to do the trick now that I have it in the right place, though.
17:36:23
##forth
<KipIngram>
So not only is there a MOUNTAIN of options and flags, but there are several places you can put them in the command, with little guidance as to what goes where.
17:37:38
##forth
<KipIngram>
I tried to ask about it in #ffmpeg, and got called a troll. That was a first. Definite difference between the atmosphere there vs. #kicad - the #kicad guys are just as friendly as they can be.
17:55:46
##forth
<KipIngram>
I don't think the person that did that was prominent in the channel though - just someone that woke up on the wrong side of the bed today, I guess.
18:04:30
##forth
<thrig>
or the previous 640 askers were help vampires
18:08:58
##forth
<KipIngram>
Possibly, I guess, but even so a "help vampire" isn't a "troll" - those are entirely different things.
19:36:51
##forth
<cleobuline>
mforth: 1 127 << 1 - PRIME? .
19:36:51
##forth
<mforth>
1
20:44:41
##forth
<xentrac>
aha, ?reconnect=1, amazing
20:45:16
##forth
<xentrac>
I wonder why video decoding projects are so hostile
20:45:41
##forth
<xentrac>
Gabucino in MPlayer was particularly notorious
20:46:31
##forth
<thrig>
glibc has also had a few spicy maintainers
20:51:03
##forth
<xentrac>
Ulrich really isn't in the same league as Gabucino
20:52:37
##forth
<xentrac>
ugh, Firefox just got upgraded out from under me
21:12:26
##forth
<KipIngram>
You mean the upgrade changed something critically damaging to your use case?
21:14:03
##forth
<KipIngram>
Yeah, that reconnect=1 made all the difference - it went from being "usable but unpleasantly erratic" to "pretty much solid as a rock."
21:14:48
##forth
<KipIngram>
The glitches don't vanish - I still get a few skips, a second or two here or there of gray screen, but the overall pattern is solid.
21:15:34
##forth
<KipIngram>
I can just take what I've got now and background it, so that no console window is consumed, and that'll be long-term good.
21:20:53
##forth
<xentrac>
no, I just couldn't load any new web pages until I restarted it
22:17:07
##forth
<cleobuline>
mforth: : 1 127 << 1 - PRIME? .
22:17:22
##forth
<cleobuline>
mforth: ;
22:17:34
##forth
<cleobuline>
1 127 << 1 - PRIME? .
22:17:39
##forth
<cleobuline>
mforth: 1 127 << 1 - PRIME? .
22:17:39
##forth
<mforth>
Error: Stack underflow
22:18:04
##forth
<cleobuline>
mforth: 1 127 << 1 - PRIME? .
22:18:04
##forth
<mforth>
0
22:19:34
##forth
<cleobuline>
mforth: 1 127 << 1 ) PRIME?
22:19:34
##forth
<mforth>
Unknown word: )
22:19:51
##forth
<cleobuline>
mforth: 1 127 << 1 - PRIME? .
22:19:51
##forth
<mforth>
1
22:20:02
##forth
<cleobuline>
better
22:29:12
##forth
<cleobuline>
i am working on a new version with linked list for string variable and array