Talk:Binary blob

This is still OpenBSD specific, even after I added a sentence about the FSF. Fans of other OSes are invited to expand this article! -- Joachim

I really don't think that is going to happen, noone outside the OpenBSD community really cares, just look at how many major operating systems even mention binaries as a problem? Hell, Scott Long of FreeBSD regularly mocks OpenBSD's efforts and claims their reverse engineered work is illegal, while Mark Shuttleworth of Ubuntu has said he will seek out more in order to give his users the "best" experience. Much like in the OpenBSD release 3.9 song, blob is taking over the world and the people are loving it. Janizary 23:22, 23 June 2006 (UTC)[reply]

GNewSense cares pretty hard. I think Ututo have a similar policy and I've heard that Fedora are considering the same idea. --Gronky 22:34, 17 October 2007 (UTC)[reply]

I know why there not considered a binary blob, but there is a project called openbios which replaces the bios (which is firmware?) with a open bios. I suppose in reality to exclude firmware completly would mean free open operating systems would take a lot longer to develop and many consider firmware less of a issue than drivers, as there not directly communicating with the operating system but with a driver. My point was a little food for thought..

Allix Fri Oct 13 19:56:00 BST 2006

The difference is that the BIOS runs on the main CPU and is general-purpose, the firmware on specific cards do not and are not. That firmware is really part of the hardware. You want open hardware, it is important; open software, it's API is important but it itself is not. NicM 19:15, 13 October 2006 (UTC).[reply]

The current article says that blobs are accepted into the Linux kernel "as a fast route to the missing or enhanced functionality these blobs provide". And i wonder - how can it be legal? Who allowed this "fast route"? This must have been discussed somewhere, maybe at the LKML.

According to the GPL, code licensed under GPL can only be merged and/or linked with code that is licensed under GPL or a compatible license. So how can a binary blob be legally distributed, if it is not "Accompan[ied] with the complete corresponding machine-readable source code", where "source code" is defined as "the preferred form of the work for making modificatto it"?

I am not a kernel driver developer, but clearly, an undocumented binary blob is not "the preferred form of the work for making modifications to it".

I've been wondering about this ever since i first read the first announcement of gNewSense. --Amir E. Aharoni 07:28, 15 July 2007 (UTC)[reply]

Closed-source binary drivers are allowed to be loaded into the kernel, although it does "taint" the kernel. There has of course been much discussion of the legality of this, Google should turn it up (http://kerneltrap.org/node/1735 is one example). Note that this Binary blob article does not (although maybe it should) cover the official policy/licensing of the Linux kernel, such as it is, it just notes that some Linux distributions are happy to distribute binary modules. Linux developers have generally not gone out of their way to encourage binary blobs, complaints about the tainting scheme and frequent ABI changes have basically met with "tough shit" (see, eg, http://www.linuxdevcenter.com/pub/a/linux/2004/09/02/driver_ease.html which has some quotes from Linus on LKML). NicM 12:36, 16 July 2007 (UTC).[reply]

The article should state clearly what is decided at the kernel level and what is decided at the distro level. --Amir E. Aharoni 13:07, 16 July 2007 (UTC)[reply]

Sure, feel free to fix it, there seem to be quite a few suitable cites. NicM 18:32, 16 July 2007 (UTC).[reply]

I am not sure that i should do it - i am a mere Slashdot reader and user of GNU/Linux who just happens to be curious about the goings-on of the community. This is not enough to write precise information. I really don't want to get into the trouble of writing something wrong about the management of the kernel and misrepresenting it. --Amir E. Aharoni 19:37, 16 July 2007 (UTC)[reply]

Not willing to spout your opinion as if definative, sheese, what kind of Slashdot reader are you? 74.13.46.118 18:45, 17 July 2007 (UTC)[reply]

I share my opinions on my blog and i try to avoid it on Wikipedia.

On /. i mostly RTFA, i hardly ever comment. --Amir E. Aharoni 20:14, 17 July 2007 (UTC)[reply]

Is this the same blob as in Binary large object? (in which case "binary blob" would be a pleonasm). Apokrif 17:03, 9 September 2007 (UTC)[reply]

• Depends on who is talking, usually I use it as a blob, as in a pile of gunk, a load of garbage, a slimeball, a disgusting little mess, since if I were using an acronym, I'd be using BLOB, not blob. Janizary 19:36, 9 September 2007 (UTC)[reply]

Actally, that would be BLOb, but whatever. :-) 76.126.134.152 (talk) 09:01, 8 June 2008 (UTC)[reply]

Does this mandate an update to this article? Specs on X.org Slashdot article -- Kl4m ^{Talk Contrib} 02:17, 13 September 2007 (UTC)[reply]

I updated the article to remove ATi's name as they have in fact made a significant amount of information on their cards' hardware public.Pgk1 (talk) 03:44, 18 February 2008 (UTC)[reply]

Is it accurate to say that this is a free software community issue? Is the term used outside of the free software community, and is this problem an issue (is it considered a problem and are people working on it) outside of the free software community? --Gronky 22:35, 17 October 2007 (UTC)[reply]

How could it possibly be a problem outside of the free software community? If you expect all vendors to provide source code in addition to compiled binaries, you are, ipso facto, a member of the free software community. It's kind of like asking if lactose intolerance is considered a problem outside the mammalian community. 76.126.134.152 (talk) 08:56, 8 June 2008 (UTC)[reply]