| |||||||||||||
FSFLA: Linux kernel is "open core"FSFLA: Linux kernel is "open core"Posted Nov 9, 2010 13:42 UTC (Tue) by hmh (subscriber, #3838)In reply to: FSFLA: Linux kernel is "open core" by rilder Parent article: FSFLA: Linux kernel is "open core"
No. This is not a technical wording bikeshed paint color party.
If a driver requires firmware to get the device work, and that firmware is not compliant to the FSF's definition of free software[1], it is considered "bait". Whether the firmware is distributed in a separate place makes absolutely no difference whatsoever.
FSFLA is doing their job (whether they could have worded that text better or not is up to discussion, I suppose). They have their reasons, and anyone who knows the embedded device market, knows damn well that we need as much pressure as we possible can gather to keep the damage contained.
It is not in their job description to gather for the immediate needs of users, or anything else like that. Which is quite correct, the usefulness of FSF-like bodies is always in the long-term picture. We have other bodies taking care of short-term compromises (sometimes to the detriment of long-term benefits, and sometimes as a useful counterpoint to keep the balance).
[1] I prefer Debian's, so I tend to separate the various definitions of what is free software.
(Log in to post comments)
FSFLA: Linux kernel is "open core" Posted Nov 9, 2010 17:06 UTC (Tue) by foom (subscriber, #14868) [Link]
> Whether the firmware is distributed in a separate place makes absolutely no difference whatsoever.
Unless of course the separate place is non-volatile memory that comes with the hardware. Then it's perfectly okay.
FSFLA: Linux kernel is "open core" Posted Nov 9, 2010 20:11 UTC (Tue) by hmh (subscriber, #3838) [Link]
I don't know what the FSFLA and FSF thinks about this, but IMO what should make a difference is whether it is field-upgradeable or not.
Being an EE, I also have other criteria for field-upgradeable units, as I am perfectly aware that in some cases, open firmware would be usable only if almost all of the inner working of the hardware was documented to the public in the first place, at which point we're talking about open hardware.
Personally, I have no issues with self-contained non-free firmware as long as the API to interface with it is fully documented, and support channels are in place to get fixes for that firmware for at least 5 years after the product is available to the public. I don't expect the FSF to agree with this one...
|
|||||||||||||