> Low preset:vo=opengl:deband
hwdec=auto
> Medium preset:vo=opengl-hq
hwdec=auto
> High preset:vo=opengl-hq:scale=ewa_lanczossharp:cscale=ewa_lanczossoft
hwdec=auto
> Highest preset:vo=opengl-hq:scale=ewa_lanczossharp:cscale=ewa_lanczossoft:prescale-luma=superxbr:prescale-downscaling-threshold=1.5
> Insane preset:vo=opengl-hq:scale=ewa_lanczossharp:cscale=ewa_lanczossoft:prescale-luma=nnedi3:prescale-downscaling-threshold=1.5
>insane preset
step it up senpaivo=opengl-hq:scale=ewa_lanczossharp:dscale=ewa_lanczossoft:cscale=ewa_lanczossoft:tscale=catmull_rom:tscale-clamp:scale-antiring=1:dscale-antiring=1:cscale-antiring=1:tscale-antiring=1:prescale-luma=nnedi3:nnedi3-neurons=128:nnedi3-window=8x6:prescale-passes=5:deband:deband-iterations=16:dither-depth=auto:temporal-dither:interpolation:blend-subtitles=yes:target-prim=bt.709:target-trc=bt.1886:3dlut-size=512x512x512
video-sync=display-desync
framedrop=no
What's your opinion about this madshi's post?
http://forum.doom9.org/showthread.php?p=1764384#post1764384
Personally I can't see any difference between Jinc and SSIM-100 with Anti-Bloating.
>>54033963
>3dlut
>no icc profile
lrn2rtfm
>>54033963
Maybe I will add this one but as "oil-painting preset". ^^'
Does anyone know if they improve their antiring filters?
Is if useful for filters like ewa_lanczos?
>>54033988
add :gamma-auto:icc-profile-auto
>>54034037
yes
>>54033967
Personally I find it ludicrous that they're trying to take their bullshit algorithms and applying them to *downscaling* as well.
I mean, downscaling is literally the easiest thing you can do; since you're just removing existing information (rather than having to come up with new information). The only thing they're doing is distorting the image in order to increase their subjective opinion of what looks better. (And obviously, it's the one that sounds the fanciest, is the newest, and takes the longest to compute)
These scalers are basically the videophile equivalent of audiophiles who use vacuum tubes in their amplifies because they think it gives the sound a more “natural” feeling. (When in reality, it's just adding distortion because vacuum tubes are literally an inferior technology)
After about ten seconds of video playback mpv dims; livestream, local files. How do I stop that from happening?
running on arch, fx6300 & a hd 7870
>>54034423
it's not theirs, they just implemented it
https://graphics.ethz.ch/~cengizo/imageDownscaling.htm
>>54034427
add the line"stop-screensaver=yes"
to your configuration file, typically ~/.config/mpv/mpv.conf
>>54034564
no dice
would cairo-compmgr change my screen settings? that's the last thing I remember installing before this dimming thing started
>>54034538
>it's not theirs, they just implemented it
Of course not, nobody in this circle has invented anything - especially not the ones that try and make it seem like they did. (madVR, MPDN)
>>54034670
AFAIK madshi is the first one that came up with smoothmotion idea
Why the kodi team doesn't integrate the mpv's scalers and debanding filter?
Should I change anything?
Specs: i5 4670k, 8gb, gtx660 2gb, Windows 8.1;
vo=opengl-hq:interpolation:scale=ewa_lanczossharp:cscale=ewa_lanczossoft
video-sync=display-resample
Do you think I can run a higher preset?
Is there a way to get mpv-progressbar to work withMOUSE_BTN0 cycle pause?
It kills seek via the bar. Works fine with the other osc. I'm an idiot btw.
>>54035928
it is better to add a shader
https://github.com/mpv-player/mpv/wiki/User-Scripts#pixel-shaders
than use any of those 2 latest presets
>>54036110
Are they as good as madVR ones?
I really like crispen edges in image enhancements.
I just got this and used the insane preset.
Are there any other recommended things I should be using?
>>54037083
"Crispen Edges" is FineSharp
>>54037083
>crispen edges
it is based on FineSharp
Also One-Pass Adaptive Sharpen is very similar to it
>>54037121
The ones in the OP are shit, use:vo=opengl-hq:scale=ewa_lanczossharp:cscale=ewa_lanczossoft:interpolation
video-sync=display-resample
Also if you didn't build mpv yourself it's probably not even working because nnedi wasn't enabled, which means your whole vo line will be invalidated and you'll actually be using worse than the "low" preset in OP. Nnedi is shit and a meme.
>>54037300
and if I'm too lazy to build it myself should I just be using mpc?
>>54037367
nightly builds have nnedi3
https://sourceforge.net/projects/mpv-player-windows/files/
>>54037367
I meant the "insane" preset won't work unless you build it yourself. The one I posted will work fine.
>>54037398
>windows
Please leave this board.
How can i tell if a shader is working?
I dont see any difference using any of them
>>54037404
thanks senpai. I'll try that out as soon as I get home.
>>54037463
Print screen comparison of the same frame. Then squint a lot.
>>54037457
>mpc
do you use it on linux? that's right.
>>54037463
If you can't tell it's working, remove it and move on with life.
I got a few questions.
How would I implement a right click menu for mpv?
How can I remove the video title in the osc?
How can I make mpv cache entire youtube videos to ram?
>>54038225
Right click menu, only works on linux I think
https://github.com/nezumisama/mpvmenu
Dont know if there is a simpler way to do this, i dont see a command for it. To remove the video title:
Download
https://github.com/mpv-player/mpv/blob/master/player/lua/osc.lua
save it in mpv/scripts/osc.lua
add osc=no to mpv.conf
Then find the layout you use, if you use default ctfl+f "classic box layout". Go down to the part that says title row and delete the bit that says lo = add_layout("title") and the three lines underneath it.
>>54038225
>How can I make mpv cache entire youtube videos to ram?
Increase the cache size
>>54037493
You can also use imagemagick to check if there's any difference between both images, you would be better off using static images instead of videos to do the comparison
Is there anyway to get Hardware Decoding on YouTube without using mpv?
Should I maybe replace bjin's 2-pass adaptive sharpen with the latest version?
Can anybody update the lumasharpen shader?
My netbook can't handle finesharp or adaptive. :(
>>54037300
would this work fine on a macbook air (Intel HD5000) ? or will it judder to shit
Why doesn't wm4 like switching monitor framerate?
https://github.com/mpv-player/mpv/issues/552
Almost all video players do this natively!
Can't use ewa_lanczos because my gpu is too old.
What's the sharpest scaler to use on my old machine?
>>54037145
Can I use finesharp as pre-shaders?
>>54039476
Should just werk if you use the HTML5 player and force h264
>>54039540
Yes
>>54039733
That sounds pretty low end, try tweaking the scaler sharpness instead of relying in another shader
>>54039908
Probably not, it will probably overheat and throttle bretty fast
>>54040685
Except they don't, the few that are able to do it (the only one I know is Kodi) do it in hacky as fuck ways, then there's madVR that's not a video player and the only other that can do it
>>54041552
Lanczos, though you can tweak most scalers to be sharper
vo=opengl-hq:backend=dxinterop:interpolation:tscale=triangle:correct-downscaling:sigmoid-upscaling:deband:scale=ewa_lanczossharp:dscale=ewa_lanczossharp:cscale=ewa_lanczossharp:icc-profile-auto
video-sync=display-resample
framedrop=vo
hwdec=d3d11va-copy
Any reason to use different settings for upscaling to 3440x1440? With my own testing and comparing ewa_lanczossharp is the best scaler in terms of retaining detail, I never drop frames and have found that tscale=triangle works magic on smooth dolly/panning shots.
whats the best settings to use for a Thinkpad x220t with XFCE+Compton? I'm using the medium preset, but videos seem to get all "choppy" when the camera pans.
/mpv/ how do I interpolate video at a rate that matches my monitors refresh rate, and I don't mean a refresh rate divisible by it, I mean something to simulate the same thing SVP does.
There's a lot of options removed or changed options I saw when looking into this, everything I try doesn't give an error but also doesn't actually change the frame rate.
>>54044775
add interpolaton to your vo
what will I gain by addingframedrop=voto my config file?
>>54046368
Nothing. That's the default setting for that option.
shameless plug
https://gist.github.com/haasn/7919afd765e308fa91cbe19a64631d0f
^- set of configuration to make mpv behave more like an image viewer. (I made it replace sxiv as my daily driver)
>>54037457
Fuck off with your muh linucks superiority. You really need that to feel good about yourself? You know how sad that is.
People use what they use and open source software is wonderful in that it usually compiles for most OS.
Most of us actually just want to use our computers, not spend months in configuration files. We do have real lives and don't spend all our waking minutes glued to the monitor.
>>54047059
>Most of us actually just want to use our computers, not spend months in configuration files. We do have real lives and don't spend all our waking minutes glued to the monitor.
Then I don't understand why you'd prefer to use Windows, an OS that requires hours of configuration, tweaking, reverse engineering and struggling with mingw, libs etc. to properly set up; over something like Ubuntu, which just works out of the box.
>>54047101
advertising bloatware
People who don't know enough to do more than point and click around their OS of choice shouldn't be talking about what OS is better than another.
Ignoring the obvious fact that any time spent working on a proprietary OS that will some day cease to exist and will become unsupportable due to being proprietary is a massive waste of time and resources.
M$ doesn't even have a stable next-gen FS out yet and nix has had some for atleast over a decade. Windows is wayyyyyyyy behind the curve in terms of good OS-level features.
>>54047240
>ReFS/WinFS has been in development hell for almost two decades now
I wonder how you fuck up that much
>>54047240
They don't even have a fucking stable desktop environment. I installed the latest Windows update and now explorer will simply not work at random, this is not an isolated issue and nobody knows how to fix it.
The good thing though is that Windows has a rich logging and event system, so I can just see what's causing explorer to crash and hang. Just kidding all the fields are blank.
A lot of people are speculating this has something to do with the Windows search service and that rebuilding the database fixes it, not only do I have the windows search service disable but I also have Cortana turned off too and it's still happening.
What a really really good operating system. I am thiiiiiiis mad.
>>54047363
WinFS is dead and was never even released.
ReFS has been around for a shorter time than even BTRFS which is still not officially labeled as stable. While ZFS the goto "stable" next-gen FS was released in 2005.
You're a dumb fuck.
>>54047544
I want to strangle Sun for their management incompetence and Oracle for their intentional asshole management. If it wasn't for the licensing hell surrounding ZFS everything would be using it right now.
>>54047563
*If it weren't for richard stallman's communist endeavours everything would be using ZFS/dtrace/etc right now
>BUT YOU NEED TO HAVE YOUR RIGHT TO THE MEANS OF PRODUCTION/SOURCE CODE REMOVED SO WE CAN BE FREE FROM CAPITALISM
>>54047582
I am still shocked that such an extreme ideal is so common. I don't think that man has done anything other than hinder progress of everything.
>>54047563
>>54047582
If HammerFS 2 goes well and gets picked up by OpenBSD maybe Apple will join in as well since they also need a new FS.
>>54036006
Progressbar dev doesn't go on 4chan, ask him on Github.
>>54047608
So would you rather live in a world where:
1. Everything is opensource because it needs to be, no history or innovations are lost. This also gives you the ability to do things such as maintain software long after the original maintainer has died or given up on the software.
or
2. Everything is close source and proprietary, any innovations or code history are lost when the business stops making money. If you buy something and the vendor decides to leave you hanging the only logical choice is buy something else.
Is it really that hard to see why people push the GPL etc? What is bad about world 1 exactly?
>>54047687
>false dichotomy
I'd rather live in the real world.
>>54047705
And a real world where things are forced to be opensource is bad how?
>>54047714
Take your bullshit to the nearest license wars thread. You people are the absolute worst, always trying to argue or push your agenda, exemplifying my point of getting in the way of progress.
Dear /g/ I'm an idiot from /a/. When I browse to C:\Users\%username%\AppData\Roaming\mpv there are no signs of any conf files there. Is there supposed to be a conf?
>>54047730
wow, you actually have a point, sorry for being a silly faggot man. i am going to jack off to my little pony and traps now. cya
>>54047730
Lmao if you're too dumb to actually make a decent case for your stupid beliefs don't respond in the first place k?
>>54047736
Make one named mpv.conf, I don't think it's generated by default.
>>54047753
Thanks anon.
>>54047736
Idk about windows but on linux the mpv folder in ~/.config/ isn't there by default either you have to create it and then put the mpv.conf inside. So it's probably the same situation for Windows.
>>54047743
No worries, enjoy your time.
>>54047746
He's the one responding to me, upset that I don't like his prophet and fantasy land.
>>54047762
I guess that's the most likely explanation. Thankyou kindly.
>>54042228
MPDN, mpc-hc, madVR, MediaPortal, they all have framerate auto switcher.
>>54047544
ReFS is pretty much what came out of the corpse from WinFS, they been trying to come with a new filesystem for almost 20 years
>>54047582
Why would you post if you have no idea about what you're talking about?
>>54047792
You said video player, madVR is a renderer, MPDN needs extensions to do frame rate switching, MPC-HC and forks use madVR too
Ie. all of them need an external program, just like mpv needs a script and another program
Except MediaPortal, forgot about that one
>>54047792
This is the nature of directshow environment.
However mpc-hc has a framerate switcher.
There are api or built-in external program for each operating systems to change the display framerate.
Why not using it?
All shitty htpc boxes I owned (4/5 years ago) has this feature too.
Just know the existence of this little media player from kde planet (picture related).
http://davidrosca.blogspot.be/2016/04/audio-volume-improvements-in-plasma-57.html
>>54046915
So how do you use that? With an alias that opens mpv with this set of config files? Or is[image]enough to toggle it automatically?
Also mpv starts noticeably slower on my computer than feh or imv (although it's still very fast), so I'm not yet convinced to do the same.
>>54049406
Seems like it's just a frontend to pulseaudio? pavucontrol is pretty similar if you want a GUI.
>>54047510
>They don't even have a fucking stable desktop environment. I installed the latest Windows update and now explorer will simply not work at random, this is not an isolated issue and nobody knows how to fix it.
I installed Windows 10 and the start menu crashed whenever I clicked on “All Apps”.
How can I have MPV as a player in popcorn time?
>>54040685
What problems are there with using the script? There's little motivation to change the status quo.
>>54049566
>With an alias that opens mpv with this set of config files?
Exactly,alias mvi='mpv --profile=image'
>Also mpv starts noticeably slower on my computer than feh or imv (although it's still very fast), so I'm not yet convinced to do the same.
Afaik none of those support color management, lanczos upscaling, >8 bit rendering or other features of mpv, though.
>>54049971
No problem but damn that's really annoying to download scripts, check them, ...
But if you say it needs shitty hacks to have this feature natively I trust you.
>>54050010
>No problem but damn that's really annoying to download scripts, check them, ...
Meh, if you think this is a hassle, I can't imagine what something like installing a browser is like for you.
>But if you say it needs shitty hacks to have this feature natively I trust you.
kek it sounds like the absence of this feature frustrates you
>>54049870
If you have MPV in your PATH it should be configured by default
>>54050007
>Afaik none of those support color management, lanczos upscaling, >8 bit rendering or other features of mpv, though.
Yeah that's a pretty big draw, the configurability of mpv is also much better imo.
I suppose it's ffmpeg initialization that takes the most time when launching mpv?
>>54050035
> What something like installing a browser is like for you.
For web browsing, I did this some times ago:yaourt -S firefox firefox-extension-https-everywhere firefox-extension-omnibar firefox-extension-ublock-origin firefox-extension-umatrix
I still have the default bookmarks and search engines.
I am a lazy guy, hopefully it's smart laziness.
> It sounds like the absence of this feature frustrates you.
Yeah it's a shame for the best media player. ^^'
Can any one provide the version of MPV that works with SVP on Windows?
Smart laziness...
>>54050196
Firefox extensions are in the AUR? What's the benefit, and how is that gonna work when mozilla requires signing?
>>54050273
Smart laziness: if annoying mozilla popups, install chromium...
>>54050273
I am not a paranoid guy who check any in/out web requests.
I live in a very restricted government who even news websites are forbidden.
So I don't care about signing certificate or keys.
I only get rid of things that slowdown my computer like ads.
>>54050097
>I suppose it's ffmpeg initialization that takes the most time when launching mpv?
I'm not entirely sure. A large part is probably also GLSL shader compilation and generating the 3DLUT for color management.
If you want to help cut down startup time it would probably be rather straightforward to collect some stats and help identify which parts are slowest in the critical path.
It's not a huge priority for me, I would rather have an image viewer that presents a very high quality image and works in my (10-bit) environment over one that starts faster.
>>54050383
Also, it's not like mpv takes insanely long either.$ time mpv ~/cms/video.mkv -vo opengl
Playing: /home/nand/cms/video.mkv
(+) Video --vid=1 (*) (h264)
VO: [opengl] 100x100 yuv420p10
Resolution: 100x100, Framerate: 25.000 Hz
Exiting... (End of file)
mpv ~/cms/video.mkv -vo opengl 0.07s user 0.07s system 96% cpu 0.145 total
With-vo nullthat goes down to about 60ms. So you have about ~100ms of time needed to initialize the GLX context and create a window, and only like 50ms to set up ffmpeg and decode a (small) frame.
Could anyone make screenshot comparison of these 2 adaptive-sharpen shaders:
> two-passes
https://gist.github.com/igv/4792d0abab41d436ac1a51bb171f8c2f
> one-pass
https://gist.github.com/igv/8a77e4eb8276753b54bb94c1c50c317e
>>54033942
Assuming I only watch 1080p REMUXes on a 1080p screen, is there a noticeable difference between MPC-HC and mpv? Not trolling, just curious and considering trying mpv.
>>54050581
Sure. Original
>>54050682
Here's the 1-pass shader
>>54050697
Here's the 2-pass one. For some reason it doesn't seem to change much from the original. I double checked, but maybe somebody else can find the error?
I ran mpv asmvi ~/cms/upscalers/baby/original.png --fullscreen=no --geometry=512x512 --video-unscaled=no -vo opengl-hq:no-icc-profile-auto:dither-depth=8:no-sigmoid-upscaling:target-prim=bt.709:target-trc=srgb:scale=ewa_lanczossharpmvi ~/cms/upscalers/baby/original.png --fullscreen=no --geometry=512x512 --video-unscaled=no -vo opengl-hq:no-icc-profile-auto:dither-depth=8:no-sigmoid-upscaling:target-prim=bt.709:target-trc=srgb:scale=ewa_lanczossharp:post-shaders=/mem/adaptive-sharpen.glslmvi ~/cms/upscalers/baby/original.png --fullscreen=no --geometry=512x512 --video-unscaled=no -vo opengl-hq:no-icc-profile-auto:dither-depth=8:post-shaders='[/mem/adaptive-sharpen-pass1.glsl,/mem/adaptive-sharpen-pass2.glsl]':no-sigmoid-upscaling:target-prim=bt.709:target-trc=srgb:scale=ewa_lanczossharp
and took the screenshot asimport -depth 8 Xpass.png.
>>54050682
>>54050697
>>54050717
Literally the same
>>54050717
You have to add format=rgba16f in your vo line for the 2-passes adaptive-sharpen.
>>54050428
Yeah I suppose you're right. I'll give a try to mpv-as-image-viewer, it sounds like a solid idea.
>>54050804
Good catch; I didn't realize the script needs values outside of [0,1].
Updated 2pass.png
>>54050682
>>54050873
The difference between those two is so tiny, I bet any amount of money no one here would notice any difference on a movie in a double blind test.
>>54050908
1pass shader has everything the same as 2pass except "Better overshoot contol" and "Improved low threshold for sharpening"
what's the best config for anime that isn't nnedi3
>>54051143
>best that isn't NNEDI3
redundant
>>54047059
>being this ignorant
Has anyone seen this thread where the author of madvr is accusing the mpv coder for stealing what he did and not contributing back?!
http://forum.kodi.tv/printthread.php?tid=223175&page=151
Continues some pages later, search for madshi.
>>54053345
>Also the quality of some of the key algorithms that mpv does have doesn't match madVR's quality (yet?). I could give you a detailed list of all quality differences, but if I gave you such a list, I already know that you'd forward it to the mpv devs to use as a checklist for their future development, so I'm not going to do that.
He sounds like a real dickhead.
>And when they found a bug in one of my algorithms, they didn't bother to report it to me, instead they used it to advertize their own solution as being "superior".
what is he referring to?
>>54053483
>You don't know the background/history. Recently mpv has started copying a lot of ideas from madVR, including some algorithms I invented myself (things nobody else ever did before), without giving any credit. And when they found a bug in one of my algorithms, they didn't bother to report it to me, instead they used it to advertize their own solution as being "superior". I only heard about it months later from a user. Furthermore, mpv fans are going through many forums now, trying to convince madVR users to switch to mpv because they say it now would have all important madVR features and at better quality (which is both not true). Considering this background, maybe you can understand that I don't feel like helping mpv copy even more ideas/algorithms.
Would you like it if somebody came along, copied all the good ideas from Kodi (but no source code, to make it legal), named it something else, without giving any credit to you, and then tried to steal users away from Kodi by claiming that their software had all the important Kodi features and at better quality (when both is not true)?
>>54053501
madvr jinc implementation was shit
he then went on to said they copied his jinc
but his stupid shit is closed source (and he doesn't own a mathematical function)
>>54053483
Did MPV really steal his ideas?
>>54053507
He really wants a monopoly then I guess.
>>54053581
reminder this outburst came after it was revealed wm4 got hired by plex and is making $100k+/year
yet madshi is still trying to figure out how to monetize his "product"
kek open sores won this round
But yea, mpv didn't steal anything. It wouldn't be possible anyway since madvr is proprietary closed source
>>54053345
>I'm not hurt, I'm not at war, I'm not bitter.
>I do believe that if I posted a detailed description of where madVR's algos are better than mpv's in a public forum, it would rather quickly come to the attention of the mpv devs, and would be used to close the gap. Again, this is not war, or bitterness, just a fact of life. I do see mpv as competition, so I don't want to help them
serious business
>>54053581
mpv developer here and this entire thread is pretty much gold standard comedy
>>54053824
>I do see mpv as competition, so I don't want to help them
Does that include the part where I asked him via e-mail how his anti-ringing algorithm worked and he just went ahead and explained it to me in detail?
(Note that mpv doesn't include the algorithm because it didn't seem to improve image quality in my own testing)
>>54054017
it seems it doesn't improve image quality in madvr either
this is madvr + spline + ar + adaptive sharpen + ab + ar
>>54054017
and this is mpv + spline + ar + adaptive sharpen
How do you all get to understand all the configs and stuff? I try to read all the posts on this thread but hardly get something off the top of my head.
I think is the cryptic names, or the video/audio concepts that I don't know.
>>54054113
>>54054134
Why is the crop/zoom different? It almost seems like you introduced some off-by-one in your comparison
>>54054179
Normally I try and document the most confusing concepts on the mpv wiki. Is there something in particular you think should be added? I can't really tell what people need help understanding and what not.
>>54054190
it's a madvr's thing, it's always been 1 pixel wider, it has correct height with dxva scaler only.
>>54054209
I read the manuals but hardly keep anything in memory. Typing something is not guaranteed to understand it. Like "vo", what is that acronym supposed to mean? I wouldn't complain if there is a dictionary at hand.
Maybe the lack of videos with someone explaining what this and that does helps too, a tutorial sticks easier.
>>54054274
>Like "vo", what is that acronym supposed to mean? I wouldn't complain if there is a dictionary at hand.
Well that kind of stuff is most easily looked up in the manual.
>Video
>--vo=<driver1[:suboption1[=value]:...],driver2,...[,]>
>Specify a priority list of video output drivers to be used. For interactive use, one would normally specify a single one to use, but in configuration files, specifying a list of fallbacks may make sense. See VIDEO OUTPUT DRIVERS for details and descriptions of available drivers.
So I recall superxbr not being recommended before. Did they "fix" it? What really changed? Is it just all around better/more efficient?
>>54054335
The tutorial is still a good idea. Gotbletu made many people into otherwise obscure software.
>tfw I installed alpine linux (uses musl)
>tfw I can't use le closed source nvidia driver because of that
>tfw my placebo mpv config lags because of that
pls nvidia devs.
On a serious sidenode, is there anything I could do about this, like somehow statically link mpv against a glibc?
>>54054969
erm I mean statically link nvidia agains glibc or something
>>54033942
>need to read a 173 pages manual to use it
This is why linux will never be the main desktop OS.
>>54055096
And alhamdulillah for that, no developers dumbing down software to cater to normies.
>>54055096
It's better to have a manual than having nothing...
Except poor changelog, there are nothing for madvr, it's absolutely undocumented.
However I think somebody should create a "noobs guide" in the official mpv wiki.
For people who don't understand madshi or haasn:
Don't forget that madshi and haasn live in Germany.
They earn huge amount of money, have 2 days of weekends and so on.
So they have time to spend/waste in personal projects.
For free or open source software developers:
In some countries, free or open source software are real menace for jobs.
Some of my friends CAN'T understand this concept at all.
They are sure there are hidden crappy things behind.
>>54055216
>It's better to have a manual than having nothing.
This is true, but if their software doesn't have a proper UI at least they should make a quick start guide or something like that.
>>54055171
Grow up.
>>54055435
>Don't forget that madshi and haasn live in Germany.
>They earn huge amount of money, have 2 days of weekends and so on.
lol
>>54055451
>This is true, but if their software doesn't have a proper UI at least they should make a quick start guide or something like that.
You mean like this?
>>54055588
I was talking more about settings, but that helps I guess.
>>54055632
But normal people don't need to change the settings :^)
>>54055632
>>54055588
This is a pretty good starting point for configuring mpv. It comes with a GUI and everything.
https://github.com/tdaede/mpv-rice
>SVP mpv builds are behind a paywall
great
>>54055766
>SVP
Hope that shit gets sued soon.
I discussed it with RMS and he confirmed that what they are doing is blatantly violating the GPL, and forwarded the author of MVTools (the GPL'd project which SVP stole their code from) towards a lawyer specializing in GPL violations.
There was recently some noise about it over at their forums, hopefully this is leading somewhere.
http://www.svp-team.com/forum/viewtopic.php?pid=58240
>>54055814
The MVTools stuff seems like a violation but selling a build of MPV is legal under the GPL, I don't think build scripts have to be forced open either, but I could be wrong on that.
>>54055900
>The MVTools stuff seems like a violation but selling a build of MPV is legal under the GPL
Right, it just proves even more and more just how much of an asshole the SVP team is.
Also fun fact, the “SVP mpv” builds were ridiculously exploitable, since the code they patched into it enables remote code execution just by visiting a web site in your browser as long as mpv is running. (No javascript required)
The mpv developers pestered them enough about it to make them fix it, though, so hopefully that should be the past.
Just shows to go what lengths they'll go to. They'll not only steal other's code and sell it, but also modify it with shitty backdoor's code.
>>54055944
Any evidence?
>>54056079
Of the RCE backdoor?
https://github.com/mpv-player/mpv/pull/2917
Proof of concept inside
>>54056144
I don't think it's fair to call that a backdoor, it is a security hole but it's not intentional, that is the critical difference.
I don't mean to defend this person but I know firsthand how fucking terrible it is working with Windows named pipes, it is infinitely easier to work with TCP sockets so I can understand how someone could implement this while being ignorant to security. Named pipes allow for much more granular control such as direction, authorization, domain restrictions, etc. but they're a pain in the ass and every library I've used with them doesn't even work unless you use bi-directional, byte only, unrestricted pipes with varying limits on the instances which at that point throws all the security out anyway so you're left having to write your own code to handle it all.
That's what should be done but it sure is a pain and much more appealing to just drop in something that does work and expect the user to control their network.
If it could be proven that this was done intentionally that'd be another story but even the dev doesn't seem to want it, he just wants IPC to work and it doesn't seem like it's his fault that the supported commands can be used in a way that's not desired. Security and convenience are hard to deal with sometimes especially in cases like this where you can easily be ignorant to potential vectors of attack. People make mistakes.
>>54056383
>I don't think it's fair to call that a backdoor, it is a security hole but it's not intentional, that is the critical difference.
They intentionally implemented it even though they were warned about the security implementations.
>>54056448
Fair enough, that's bad. I figured the implementation was done, shipped and then the merge was requested.
>>54056475
They merged it because they were impatient about the safer named-pipes branch being completed.
muchos gracias
>>54057872
do you create these yourself>?
