Paul Ganssle @pganssle@qoto.org

@valleyforge @freemo They did, and he did mean to say "a", but I don't think that was ever the question.

I think the question was always whether he accidentally said "for man", flubbing his line.

@freemo What misinterpretation? I feel like the quote has always been interpreted as meaning "one small step for A man", since if you leave out the "a" the whole phrase doesn't make sense (since "man" and "mankind" are synonyms).

I thought that the main controversy was whether he misspoke (i.e. forgot to say "a"), the "a" was lost in static or possibly that in his accent "for a man" runs together and the "a" gets elided.

Someone (Peter Ford Shannon) wrote a paper about it, arguing that he indeed say "a"; the PDF is here: https://web.archive.org/web/20061123073155/http://www.controlbionics.com/images_EEPRa/Electronic%20Evidence%20and%20Physiological%20Reasoning%20Identifying%20Elusive%20Vowel-01%20Final.pdf

I'm not terribly convinced by the waveforms he shows, but it definitely feels plausible that the "a" gets elided when you say that phrase; without explicitly trying to enunciate the syllables, "for man" and "for a man" come out as sounding very similar when I try it.

"Now, I see your birth plan involves Mike standing guard at the door with a spear because of... prowling sorcerers?"
"Yes, they can be very tricky."

Do you hate the feeling of being a useless spectator while your mate is laboring to bring your child into the world? Consider hiring "prowling sorcerers" who you can fend off with a spear!

(Note: This will likely not be covered by insurance.)

@schlink It's weird that they talk about the face masks everyone wears now to prevent disease proliferation but don't mention that the mask leaves your mouth uncovered, making it useless for that purpose.

@freemo Isn't Kumquat a citrus fruit?

@pschwede Unclear doesn't mean the information doesn't exist or isn't accessible, it just means that it requires extra effort to figure out.

In any case, thanks for pointing me at that. I'm kinda shocked to find that it's just hard-coded, not even locale-dependent. Seems to me like if you live in China or Ireland or Israel `date` will just be silently wrong in some cases.

@pschwede Yeah, what part of that determines how to map three-letter abbreviations to offsets?

@toast Seems like they are getting better about this sort of thing, apparently they have also approved the "Zero-clause BSD" license, which is basically BSD but with no restrictions.

I am curious to know how these things are thought of in legal terms, though. I had a vague sense that licenses are contracts and adherence to the restrictions placed in even permissive licenses serves as the consideration received by the owner of the copyright. I suspect that analysis is wrong at least in the case of 0BSD, unless there's some other source of consideration I'm missing.

@toast To be honest, I think the OSI's mission is a lot more reasonable when you see it through the lens of "only corporations big enough to sue or be sued care about license compliance." The rest of us already live in the post-copyright anarchistic paradise where we'll use any source code or program we can get our hands on (except for me, obviously, I always abide by all copyright laws and sumptuary codes 😇).

The only time it really matters is if you may want it to end up in the supply chain of a corporation like that (which could in fact be your *own* corporation, two employers from now). In that case, having just a few standard licenses makes it a lot easier to analyze things and a lot cheaper for small companies to comply.

@toast And I'm not certain I'll have time to read a 10k word essay in a timely manner, but if you've got a distribution list I'm happy to be on it.

@toast I usually do CC-0 for anything I can and Apache 2.0 for everything else — partially because Apache 2.0 satisfies the corporations who, more often than not, own the code I produce and partially because for all practical purposes Apache 2.0 might as well be public domain and in a lot of corporate environments Apache 2.0 is *easier* to adopt than something with a public domain dedication. In the end, licenses tend to only matter to corporations and other OSS projects, both of whom tend to prefer something simple like "Apache 2.0 is OK". Individuals don't care about license compliance because it's vanishingly unlikely that being out of compliance would hurt them.

To the extent that I own any code, I'm almost always willing to grant individual waivers of the requirements on request.

Apparently the OSI approved the Unlicense as a "Special Purpose" license a few months ago, but I do not see what the special purpose is anywhere on their main site. There's some more background in this mailing list post:

https://lists.opensource.org/pipermail/license-review_lists.opensource.org/2020-June/004890.html

I'm still not clear what the "special purpose" is that they're suggesting, though. Seems like they don't want to be seen recommending it because it's somewhat weirdly drafted and is partially a public domain dedication.

@toast CC-0 was not approved but also not rejected, and most of the OSS lawyers I've talked to have said it's good but not for software.

I was mildly surprised that OSI goes so far as to say that you should not use "open source" to refer to anything except OSI-approved licenses ( https://opensource.org/faq#avoid-unapproved-licenses ). I would think that they'd be OK with you using the term "open source" to refer to things that meet the OSD (https://opensource.org/osd-annotated) but aren't approved for license proliferation reasons.

@toast Looks like Unlicense is OSI approved as a "special purpose license": https://opensource.org/licenses/unlicense

Maybe it came out of a conversation with a frustrated just-arrived time traveler:

"What's today?"
"Thursday"
"No, the date!"
"August 27th."
"Be more specific!"
"Uh... 12:34:56? AM? Central standard time?"
"THE YEAR, MAN!"
"Oh, 2020."

Also, who picked that as the default date format? Was it someone trying to make American MM-DD-YY formats look sane?

Day of week, then month, then day of month, then time, then time zone indicator *then* year.

For me "CST" maps to "Central Standard Time" (UTC-5) rather than Cuba (UTC-6) or China (UTC+8):

$ TZ=UTC date --date="Thu Aug 27 12:34:56 AM CST 2020" --iso=min
2020-08-27T06:34+00:00

Interestingly, that time doesn't exist in America/Chicago, but it still applies UTC-5!