I will get a makerel and watch a good dolphin sex flick.
The leader of the FFmpeg open source project has resigned amid ongoing turmoil among the project's developers. FFmpeg, a set of cross-platform, open source libraries for playback of video formatted according to standards created by the MPEG organization, was founded in 2000 by French developer Fabrice Bellard, working under …
No impossible. Both words have to begin with the same letter.
These are all valid names.
Bad blood & project management disagreements seems a poor reason for allowing a fork to happen. Technical differences of opinion, fair enough, licence issues like that which resulted in Openoffice & LibreOffice I can understand as well.
Duplication because team members just can't get on seems ultimately pointless in the long run. One of the teams are wasting their time.
It should come down to merit, but ffmpeg has the advantage of the recognised name.
Some may disagree, but it seems (to me) that this is exactly the type of thing/attitude that has kept Linux from being a desktop alternative to MS operating systems.
The incredible incapacity of the Open Source movement to harness and line up the awesome amount of resources at it's disposal and put it to work in the same direction: to develop a Linux desktop OS that would allow us to leave the MS ecosystem once and for all.
Much more incredible when you see the path being taken by M$ and W10 and realise the consequences of what is brewing.
Just my $0.02, YMMV.
Indeed. Your point is quite well covered in last year's "Why Linux Sucks" presentation by Bryan Lunduke https://www.youtube.com/watch?v=5pOxlazS3zs
... as a dev of all platforms, master of none, and [primarily] Linux user at home myself, i always like his 'Linux sucks presentations, as they always cut to the heart of such things that slow down its paths.
That being said, I think the case in point of ffmpeg/libav is more of an issue for distro managers than end-users. In my last job, [a Debian platform] I was asked by my manager [Who'd heard the rumours] to initiate a migration from ffmpeg -> avconv/libav for the back end transcoding... after a very short time of investigation I felt it was best to maintain the status quo... nothing was borked, nothing was going away, and in a matter of weeks Debian turned back to ffmpeg. "Phew, that was close", I thought, "I nearly had to update a few scripts"
Does that, perhaps coupled with other issues in Bryan Lunduke's presentation, mean that Linux is a less attractive desktop O/S than Windows... I think that's down to end-user familiarity and choosing an O/S which runs what your friend/colleague's does... and the end-user of any O/S, unless they're a techie, doesn't care if their video conversion is using this library or that.
Sure, it slows down the forward development. But, I'd rather have a slower forward development that forks and splinters and by natural selection lets the crap ideas die, than the closed corporate directed Fordism of Windows / OSX. We shouldn't be in so much of a hurry to get a new O/S that we worry if such debates take time.
That's not to bash Windows and OSX for what they are good at, which is in the first instance a really stable and integrated business system and the second instance [for me in my own world] the system which happens to run the best audio production software platform. I have several Linux machines, a couple of OSX machines and a couple of Windows machines... I use each to what each is best geared for.
Bottom line: these sorts of arguments/battles/forks happen in closed source too, we just don't hear about them so much. If the transparency of Open Source means that its arguments turn people off, I personally shrug and say "so what" ... I far prefer argument and the slowing down of progress than dictatorship or trend following any day. Whether that means Linux is/n't ready for the [average user] desktop or not, doesn't really bother me, and I'm not really sure why it is ever a talking point other than for punditry sake.
LibAV People's Front? ... splitters.
Beer = something to wash down these Otter's Noses.
>Some may disagree, but it seems (to me) that this is exactly the type of thing/attitude that has kept Linux from being a desktop alternative to MS operating systems.
No forks generally are good as they increase choice (don't have to rely on the one GUI Microsoft decides for the desktop for example) and keep projects honest. Microsoft being in the right place at the right time and getting the network effect early on (my software only runs on Windows so why switch) plus being friendlier for most of its existence for Grandma plus some OEM shenanigans is why Linux never did much on the desktop.
Is either of them really clean of MPEG and similar patent issues? I think there was some part of the library that wasn't purely legitimate in those terms, and Linux and other FOSS fans just pretended not to notice, because Linux would be a lot less fun without video playing.
I think VLC player also has either or both of these libraries in its guts, as well as mplayer.
Niedermayer doesn't want to play anymore, fine. At least the ball ain't his, and he can't take it with him.
But ouuhhh, careful, he might retuuuuurn! Unless humanity changes for the better... he mightn't.
Hey, I have an idea! Now that systemd is running perfectly (since introduction, mind'sh), can't that team take over ffmpeg?! It'll be for the greater good.
Who, claiming a rational mind, would ever suggest the systemd team to touch anything else? I thought my sardonic post was (as Not That Andrew) points out, obvious in nature – maybe not trolling, but obviously not serious.
But to elucidate, I do not care for anyone who threatens to leave a project for what I perceive, kindergarten antics. Regarding the condition of his return, who the hell is he? It is stressful being a project admin, I can see that. But go out in a fiery blaze, or keep your wangsty trap shut, and the doom and gloom to yourself – you are a leader, after all.
Systemd, in my opinion, is a big heaping pile of bullcrap. Like the Euro, users of the end product had no say in its establishment. Greybeards know why it sucks, and hipster nerds have no clue.
I can point out at least 3 equivalents that ruin computing the same way systemd does. My downvotes will exponentially rise, and disagreeable douchbags will focus on those than the real issue.
Niedermayer may be a good and reasonable person who was worn down by his community. Not that the article showed that, and if he wrote those excerpts, I wouldn't want to find out more about him.
Being a developer is usually a thankless occupation, be it paid or not. Most people don't care who helped with Firefox, the linux kernel, or ffmpeg. Notorious people usually leave an everlasting expression (Mr Torvald), but I'll be damned if I remember the name of any systemd developer. And I looked them up a mere 24 hours ago.
So, popping up and posting diva drivel as a team lead does not make me scratch my head in amazement. There is no Charlton Heston moment at the beach coming along, because our ways led to apes running the show, now. They kinda are, but that's not the point.
You deserve ridicule if you are doing a dramatic heel-turn.
I am puzzled by the Ars Technica style downvotes, but I wear them as a badge if you disagree with me. I shan't provide any excremental excretions.
The troll icon exists, yes, but the intent was satanic in nature.
I know nothing of the technical or political issues at play here. But I do know that in practice libav seems to be buggy whereas ffmpeg always just works. No wonder distros are switching to ffmpeg.
Last time I checked, libav could not encode DNxHD without introducing 'snow' artifacts, and its handling of H.264 non all-I input was, shall we say, suboptimal. Switched to ffmpeg, suddenly everything worked as expected.
Guess which one I chose for production video editing systems.
Not really a field favored by French univ as far as I remember my studies.
It can be challenging in term of signal processing, and French engineering school are rather good at maths (e.g. Fabrice Bellard, the guy mentioned as the founder of ffmpeg, went to ecole polytechnique, see his homepage for the rest of his top notch works: http://bellard.org/).
Also look at INRIA if you wonder why some French people are rather good in image processing/rendering (INRIA is a good contributor to siggraph)
Thanks. I should have mentioned INRIA. Many moons ago (97 I think), I do recall looking for video-conference software to use with linux and IVS (http://www-sop.inria.fr/rodeo/ivs/) were one of the few choices (along with VIC http://ee.lbl.gov/vic/ ) available.
So, yes, not video per se, that's just collateral :).
One of the positives of open source and free licensing is that (like in OpenOffice's case) the project can be forked. This is great because another project with a different aim can move away from the main project without having to start over.
Unfortunately (as in this case) one of the negatives of open source and free licensing is that people fork projects for the sake of having their own project (i.e. with exactly the same aim!) - this is counter-productive and there is no logical reason for it.
In this case the reason is clear - they've had a fight (like kids in a playground) and they are not playing with each other any more. Now they've realised that it's a real pain in the neck to not play together they've decided they'd rather not play at all than resolve the differences and move on.
I say let him go, maybe it will prompt the groups to get back together and act like grown-ups!
> Whether that means Linux is/n't ready for the [average user] desktop or not, doesn't really bother
> me, and I'm not really sure why it is ever a talking point other than for punditry sake.
I have been at this since I was in my early 30's and have just turned 60, so I don't think I am an 'average user', but then that's just my own idea of myself.
And it does bother me a bit. I have been trying for the longest time to do with Linux what I have been doing with W2000/XP since early 2000 (15 years ago ... ) but never been able to.
Get my two (CRT/LCD) monitors to work as a shared desktop or be able to set up a new box, drop in a wireless card/dongle and and have it just 'work' without having to spend hours looking around the web to get a fix, only to find that the bug was reported, looked at, assigned and abandoned two years before.
So with wireless it's been rather a tug of war up to some time ago and with the 2 monitors it's been a severe waste of my time. So I install XP and these two basic (to me) items just work.
I can change every single W32 app I use for a Linux equivalent but these two basic issues are dealbreakers. Eventually I won't bother anymore, I'm sure I'm not alone in this.
> ... one of the negatives of open source and free licensing is that people fork projects for the sake > of having their own project (i.e. with exactly the same aim!) - this is counter-productive and there
> is no logical reason for it.
Maybe it's that they're all 'prima donnas' ...
Again, just my $0.02
I'll take your 2 cents and raise you 2 pennies in return :)
re: Dual head monitors and WiFi...
Oddly, with Dual head monitors, my experience with Linux has been the opposite: I have had far fewer problems doing that sort of thing on Linux than on Win' - I certainly had multi monitor setups [3 of] on Mandrake [when it was still Mandrake], before it was easy on Win'
WiFi, I have to agree... but then it's far less of a problem than it was.
... Though, to hoist the argument by its own petard, I personally think the bigger issue with both WiFi and monitor drivers under Linux has been the nature of Closed Source drivers made available for Win, but not out-of-the-distro for *nix... so, from your argument [and I agree with it, especially from your own experiences] we might conclude that the reason for Linux not being ready for the mainstream, is less due to the Open Source dev community and forking etc, and more due the nature of a FLOSS Licence and how that restricts closed source driver distribution as being part of the general release. That's more of an attitude/idealism issue really.
I wish there was a RMS icon :)
My first foray into linux in late '99 resulted in having to buy a rather expensive pcmcia modem card because my cheap laptop had a wintel modem card.
True, with linux you have to research a little more thoroughly before you buy kit to ensure you are not purchasing something that will end up either driving you back to win or spends its potential useful life gathering dust in a cupboard. Any IT person would research a new purchase thoroughly anyway.
When it comes to multi-monitor, in my opinion, it's only truly useful in tiling setups. One desktop encompassing all screens is just such a pain when trying to juggle more than one persistent visible application with others.
There was a mixed bag when it came to software modems. I actually found that I could get mwave modems working relatively well in Thinkpads around the same time as you had trouble. Admittedly at the time, Redhat 7 and 8 (not RHEL) did not have the mwave driver in their repositories, but it was available, and built relatively easily on mainstream Linux distros, and worked quite well.
I did have an HP Riptide sound/modem card combination card that was more troublesome, and I completely gave up on that, both for sound and modem.
I've not tried to get integrated modems in laptops working recently. It's all a bit pointless since 3G dongles and public/guest WiFi are so common.
>> Get my two (CRT/LCD) monitors to work as a shared desktop or be able to set up a new box, drop in a wireless card/dongle and and have it just 'work' without having to spend hours looking around the web to get a fix,
This is a bit odd, as an example I did set up a scraps PC with two old nvidia cards (NV40) (whatever I found in the storage room) and four monitors not long ago, the critter is even running OpenGL applications using the 4 screens simultaneously.
I did use Ubuntu 14.04 and although I had to tamper with the xorg config it was because of my own choosing (old school) rather than lack of GUI tools.
That PC is also connected to the network via wireless for convenience, and it uses a £10 edimax usb dongle
So I find strange that since 2000 you haven't managed to have a desktop spawning two monitors.
I'm not saying you could not have issues or a weird piece of hardware that is problematic or unsupported.
These days hardware issues with Linux tend more in line with: "Game X with Graphics card Y does not perform as well as the windows version" or things like closed wireless card with poor driver support loses connection.
I normally avoid those issues following the recipe: "Doesn't work with Linux, no buy"
"Doesn't work with Linux, no buy"
Yep. Me too.
If the manufacturers of this world are ever going to catch on, it's when they see a segment of the market resolutely refusing to buy their products, or worse, buying their products and posting scathingly negative reviews when they realise the company has a whiff of O/S fascism.
If someone, somewhere, was posting a list of recommended hardware for a Linux machine, then, as a manufacturer of cutting edge technology, I would be ashamed to not get a mention.
Shouldn't be a problem in Linux these days, but as a guide that's just as applicable to BSD as Linux. Your CRT may involve extra effort.
1) Research the wireless card before buying it. Don't just expect it to work.
2) For a dual monitor configuration, use the Nvidia binary drivers and double check their compatibility. Buy a card on that compatibility list, and run both monitors off the same card.
If you can't run Nvidia binary drivers (OpenBSD, FreeBSD) research the cards first, and on the BSDs multiple monitors *must* be run off the same card.
SteamOS does not like CRTs or 4:3 monitors in some cases (Nvidia card, can't remember which). That problem doesn't exist with normal Linux or the BSDs.
I can tell you that Nvidia cards happily run dualhead under BSD and Linux, in addition to Windows.
If you're connecting to your monitor through a KVM switch, and sometimes when you're not, the graphics card may not be able to identify the resolutions supported by the CRT. You'll have to dig into the joys of modelines and horizontal/vertical frequencies in xorg.conf to fix that.
I love my CRTs, but they are more hassle to use these days. If you want an easy life, use a TFT with DVI.
I found that Linux could do dual monitors before Windows could. Of course it required fiddling with the xfree86.conf file. When Windows introduced the capability, it was soon easier to get working in Windows, though not always exactly the way you wanted it to.
More recently it's gotten a lot easier to do dual monitors in Linux as well. I am typing this from a dual monitor setup using an AMD card with the open source drivers. I have also used this card with the closed source drivers (when the card was newer and needed them to get video playback without tearing) on dual monitors before.
I don't often have wifi issues with Linux anymore. Often any issues are about missing firmware in certain distributions, which is solvable. The more user friendly distributions usually just work, though occasionally I will run across a card that is a problem. I ran across a very old laptop (about 8 years) that had a card that didn't just work several months ago, and I just swapped it for another card from the same time period that I happened to have rather than try to make that one work. Other than that, it's been about 5 or 6 years since I had an issue, and that issue was that the card was so new, support hadn't been integrated into the distribution I was using yet.
I'll raise you my projection machine.. VGA monitor and HDMI projector. Windows insists that the HDMI should be monitor #1 and I can change the settings to do roughly what I want desktop wise.. until someone disables the projector and the screen I've setup for the projector is now locally on my monitor and the start bar is somewhere else. This means the projector must be powered up before the machine is.
The fix? There isn't one. Just trust Windows to do what's best and deal with the result. At least in Linux I can edit xorg.conf and permanently force things to work the way that I want although I find it's getting less often that I need to touch that file...
> At least in Linux I can edit xorg.conf and permanently force things to work the way that I want
And that in a nutshell is why I use Linux. I have to go through incredible pain sometimes to get something to work, but once it works, it works. It's fixed. It's doing what I want/need. And it stays that way.
With Windows, I sometimes end up at "nope, can't do it" but sometimes I manage to get it working. For a week. Or perhaps a month. Until it shits the bed for whatever reason and doesn't work any more, and I'm screwed because I'm depending on using that feature/setup to get work done, and SUDDENLY I CAN'T GET WORK DONE ANY MORE. So there's a panic and I'm stuck dicking around with Windows instead of getting work done.
@oiseau: I am 10 years windows free.
Wifi dongles has nothing to do with linux and has much to do with the dysfunctional nature of hardware/firmware and Microsoft skewing the market with their generic FUD. Also some vendors blacklist wlan cards (so if you have crap one, you can't even change it for one with support chipset). So this is not a linux thing, but it is solvable in most instance.
The monitors are a solved problem, you must be unlucky!
I am using Debian and KDE and it just works. My desktop is
6400 x 2160, maximum 16384 x 16384
If you have 2 fixed monitors , save the edid.bin for both of them and it will always work. CRTs did not have proper EDID (I think? Any experts) so you need to hand tune a little bit, but that has become far nicer since I started using Linux.
Linux (or more accurately FOSS) is a way of life. If you keep your personal data separate from the system install, you can maintain continuity across distributions and machines - I'm on number 5!
Wifi dongle/card issues has to do with them being difficult to program, and the documentation lacking. Windows is the best platform for wireless, followed by Linux (mostly, coders from the wireless manufactures contribute code, so it's in good shape), with the BSDs unfortunately trailing last - because practically all BSD code isn't written by the wireless vendor, must be BSD licensed, and the hardware documentation is poor to non existent.
The monitors are only a 'solved problem' insofarasmuch as there are solutions to issues with CRTs, not that it's easy. You're using multiple TFT monitors, so of course it works.. (at least until the EDID becomes corrupt, oh is *that* fun to fix (the easiest way to fix it is to re-connect a Windows 8 box where the monitor correctly worked. Windows 8 will re-write the EDID and save a *lot* of hassle))
Please don't say something 'just works' when you don't actually have personal experience or definite knowledge of it.
Most recent CRTs do have an EDID, delivered via DDC. There are also a (very small) number of monitors/TVs with either DVI-A or HDMI connections. Of course if it's ancient enough to be using only RGBHV BNC, there's no way the card can know what the monitor is capable of.
Unfortunately, some drivers and cards are better at extracting this information than others, and the manufacturers don't really care about CRTs any more. For reliable operation you must go old school and edit xorg.conf to specify horizontal and vertical refresh, plus the modelines - as even if you're lucky enough to have the EDID working with a decent resolution, it's not uncommon for a non ideal refresh to be used.
@BinkyTheMagicPaperclip: Not to be picky, but the "it just works" is correct, assuming the hardware works properly. I believe that is correct and I have had quite a lot of experience plug-n-play monitors, but the "not working as specced" is the issue. The CRT one I will give you is more difficult to be sure as I haven't seen one in almost a decade, but I still think the situation has improved.
I will reassert, wifi is a pain but has improved. Multi-modern-monitors (especially using KDE/Nvidia) just works. CRT may need tuning...
Who said anything about Linux? I was talking about open source and freely licensed software where the devs are usually in control rather than big corporates - Open Office for example wasn't Linux only - I've used it comfortably on Windows for years.
Also - I happen to be a Linux user (All of my servers are on Linux of one variety or another) - I think you are missing the point.
>Unfortunately (as in this case) one of the negatives of open source and free licensing is that people fork projects for the sake of having their own project (i.e. with exactly the same aim!) - this is counter-productive and there is no logical reason for it.
It may seem that way at first (and probably is in most cases including this one) but much with the OpenBSD and NetBSD (also a butthurt split) the code bases can diverge and end up eventually serving different markets.
Biting the hand that feeds IT © 1998–2021