libera/#commonlisp - IRC Chatlog
Search
11:17:04
green_
IANAL, but I don't think Kent's understanding of copyright back then is correct in this new Oracle v. Google world, where APIs are copyrightable.
11:57:01
paulapatience
green_: I thought the conclusion of that lawsuit was that Google's usage of the APIs was fair use
12:54:43
green_
There was a previous decision that decided that they were copyrightable. The fair use question only makes sense in the context of copyrighted APIs.
12:58:11
paulapatience
What I read (on Wikipedia, admittedly), was that since it was fair use anyway, they didn't need to decide whether the APIs are copyrightable.
13:00:28
nij-
Another question, sorry to interrupt, is there any way to properly loop over a sequence? Per SPEC (https://novaspec.org/cl/t_sequence), a sequence is not just a vector or a list. There may be other subtypes. So (loop .. in ..) and (loop .. across ..) are not exhaustive.
13:02:35
scymtym
there is no standard way to CL:LOOP over a sequence. (map nil (lambda (element) ...) sequence) is one alternative
13:03:47
nij-
"A sequence is an ordered collection of elements, implemented as either a vector or a list." https://novaspec.org/cl/17_1_Sequence_Concepts
13:08:56
scymtym
could be clarified in wscl, peferrably along with item 8. in https://www.cliki.net/ANSI%20Clarifications%20and%20Errata
13:23:58
bike
modern consensus is pretty much that you can have other kinds of sequences, though support for actually doing it is not universal
13:24:57
green_
paulapatience, you are right that SCOTUS didn't have to evaluate the API copyright issue, because Google had a fair-use out, but I think that means the Federal Circuit court's ruling that APIs are copyrightable still holds.
13:26:22
nij-
Right. So it's better to use #'map not #'loop when one needs to iterate through a sequence.
13:27:19
green_
I am no expert, but Google did reach out to me in preparing for the fair use argument, as I was heavily involved in GNU Classpath back in the day.
13:49:40
skin
yesterday, I learned that a collection in rdf, a semantic web technology, is built using first rest and nil a lot like car cdr and nil. https://www.w3.org/TR/rdf-schema/#ch_collectionvocab
18:31:47
pfdietz
I have my ow copy, but it's mostly for playing with and extending the random tester.
19:54:09
nij-
I'm confused with this description "load sequentially executes each form it encounters in the file named by filespec. " https://novaspec.org/cl/f_load
19:55:03
nij-
Or does each form gets loaded into the image (and say (load-time-value ..) will expand), and then each form is executed?
20:14:29
bike
::notify nij- please get a bouncer or something, or at least don't drop off less than five minutes asking a question if you actually want an answer
20:15:16
bike
::notify nij- to summarize, the loader will recreate all serialized/dumped objects, evaluate load time value forms, and then behave as if it was executing each form in sequence