It sparked my interest.
Ethically, there's no freer software than public domain.
For myself, it makes it simpler to look into lifting out parts of the library to complement where avr-libc is lacking, which is in a lot of places. I don't have to worry over the fine line, a single line of attribution in the documentation of the final hardware device is more than is asked for.
As to if I used something BSD licensed instead?
If the license is somehow included with the software, well, the end user would never be able to see it, as it would be stored in EEPROM, and take about half the memory.
Strings are a pain in the butt for memory management, and having a pop-up of some sort with the BSD license, would use about a third of available memory.
Should it then be included in the documentation? Possibly, but included a no-warranty clause amongst documents for a warranty-covered device is likely to cause some strife amongst the purchasers of said device.
Including it only in online documentation doesn't seem to fit either the letter or the spirit of the license.
Practically, and legally, where to include the LICENSE is hard to tell.
But I don't have to hire a lawyer with this.
I can just use it.
Lots of Mac apps I used have an "about" window with a trail of near-identical BSD and MIT licenses. It always seems absurd to me.
If someone puts out a code library with no usage restrictions, why require this turgid legal notice to be attached? Public domain dedication seems a much more elegant solution to me.
Which means that there's a non-zero chance that distributing the code without the CC0 license attached might cause copyright problems in those places. In the FAQ they recommend that you treat the CC0 declaration as a license, with boilerplate at the top of each file and a COPYING file in the distribution:
https://wiki.creativecommons.org/wiki/CC0_FAQ
Should this be done for binaries as well as source? Beats me.
Also note that the OSI haven't certified it, citing concerns with the patent clause:
THIS SOFTWARE IS PROVIDED BY <COPYRIGHT HOLDER> ''AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL <COPYRIGHT HOLDER> BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
And as the CC0 doesn't stringently demand to be included with the software, I can be sure that including it within the standard literature for clients, or even amongst the spec sheets, is acceptable.
Copyright has made public domain painful, but CC0 is as close as we can get right now, (that I know of).
Copyright notice requirement means you can't just copy-paste a single function.
Whereas the OSI model relies on historical artifacts that are difficult to understand and whose meaning is viewed differently even by experts.
I'd like to see more software move to the CC model of licensing and perhaps vainly hope that eventually the OSI will create a model that uses the same kind of layered approach.
Most other c libraries don't have it, the few ones that do (musl for example) are linux-only usually...
Unless I'm missing something...
[0] https://bitbucket.org/pdclib/pdclib/src/95be959efe2e/opt/pth...