Live data from Hacker News

AAC and Debian

tookmund.com

1–10 of 25 posts

Re: AAC and Debian

#2
The problem is with spectral band replication in aac-he and parametric stereo in aac-he-v2 both of which are impressive but not transparent and should only be used if you are trying save bandwidth

Re: AAC and Debian

#3
Interesting. Could FFmpeg’s AAC en/decoder not be used? There’s an explicit flag necessary to include libfdk-aac when compiling FFmpeg and it makes the resulting binary not able to be shared.

However, without that flag, FFmpeg still has (their own? VideoLAN’s?) an AAC en/decoder that works fine.

Re: AAC and Debian

#4
Any reason they didn't use the debian apt source repositories instead of git? I tend to try and use deb-src whenever possible, to stay as close to Debian truth as possible.

Re: AAC and Debian

#6
This kind of thing is why I’ve only ever used Debian as a server distro until the Raspberry Pi started “fixing” various aspects of it… My Linux laptops ran Ubuntu for a solid decade until Fedora 36 turned me around (and I’ve been happy since).

But the key thing is that Debian tends to lag on user amenites (hardware, drivers, codecs, etc.) to a point where one wonders if they actually use the distro on modern hardware...

Re: AAC and Debian

#7

Interesting. Could FFmpeg’s AAC en/decoder not be used? There’s an explicit flag necessary to include libfdk-aac when compiling FFmpeg and it makes the resulting binary not able to be shared. However, without that flag, FFmpeg still has (their own? VideoLAN’s?) an AAC en/decoder that works fine.

The issue, I believe, is more around the AAC patents and less around the actual library.

Re: AAC and Debian

#8
From the article:

    "This is a slightly unusual situation, as the fdk-aac library is licensed under what even the GNU project acknowledges is a free software license. However, this license explicitly does not grant any patent rights."
Here, in the US at the very least, this would seem to be less of a problem given De Forest Radio, 273 U.S. 236 (1927):

    "Any language used by the owner of the patent, or any conduct on his part exhibited to another from which that other may properly infer that the owner consents to his use of the patent in making or using it, or selling it, upon which the other acts, constitutes a license and a defense to an action for a tort."
And logically this case makes sense. If I've granted you all these copyrights (MIT and BSD), just to surprise you/someone with a patent suit, that seems not equitable. Large companies are more comfortable with Apache 2 given it's explicit patent grant, but I wouldn't be surprised if users of certain Rust software (which is often MIT + Apache 2 licensed) had more expansive patent rights under the MIT license, given how narrow the rights granted by the Apache 2 are.

However, here, the facts are not as stated in the article. The FDK license does not just "explicitly not grant", as the article notes, it puts you on notice that you must obtain a patent license. That is not like the MIT license. The MIT license is silent re: a patent license.

Re: AAC and Debian

#10

From the article: "This is a slightly unusual situation, as the fdk-aac library is licensed under what even the GNU project acknowledges is a free software license. However, this license explicitly does not grant any patent rights." Here, in the US at the very least, this would seem to be less of a problem given De Forest Radio, 273 U.S. 236 (1927): "Any language used by the owner of the patent, or any conduct on his…

Franhoufer is not the owner of the patents in question, most of them are likely owned by the usual gang . They cannot license it to you even if they wanted to.
Post reply on HN