Login
You're viewing the f.duriansoftware.com public feed.
  • Oct 3, 2026, 12:49 AM

    @jpetazzo dude, semantics. Does it matter? The point here is that without #ffmpeg (either standalone or as a library, either server-farm side or client side), you wouldn't have watched a steamed video for the past two decades, not through a browser, not through vlc, not otherwise. @lbloch

    💬 1🔄 0⭐ 2

Replies

  • Oct 3, 2026, 6:55 AM

    @otte_homan @lbloch of course it matters :)

    When we write or say something to educate folks, it's important to pick the right words and make sure they understand precisely what we're talking about and what we mean, and that we don't give them misconceptions.

    In that case, instead of "he wrote the code that streams every [...] video" personally I wouldn't have used the word "stream" since it's not clear if we're talking server-side or client-side; and furthermore, in video players, ffmpeg and libavcodec are generally not involved in the streaming side (as in "moving bytes from server to client"). They are, however, involved in decoding. And they're not only involved for YT, Netflix, TikTok.

    Furthermore, while Bellard's achievement and contributions are absolutely fantastic, it's also disingenuous to say something like "he wrote the code that streams every video".

    The rest of the text does a much better job (mentioning he started the project and what the project does and how it's involved in streaming), but the short summary in the beginning is just sensationalist viral crap (and probably slop).

    Oh, also, I'm old enough to remember RealPlayer and I've deployed video streaming servers before ffmpeg existed, so your last point is also plain wrong. Video streaming existed before ffmpeg (and it sucked for all kinds of reasons).

    What would probably be more accurate would be to say that ffmpeg and libavcodec used a permissive open source license, and were so powerful and well-designed that virtually no equivalent library managed to take their place 25 years later. Personally, that's the part that I find the most impressive: dude wrote code that could handle encoding, decoding, transcoding, work with both network and disk streams (which have very different constraints) simultaneously, mix and mux codecs that weren't even meant to work together in the first place, and would keep accommodating new formats decade after decade.

    Now, in the end, the "author" of the original post merely copy-pasted it from Twitter, where it was copied/adapted from somewhere else (I've seen better variants of that post making rounds months ago).

    💬 0🔄 1⭐ 4
  • 💬 0🔄 0⭐ 1
  • 💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0