Well, just taking rs3746544 as an example, the SNPedia page has this chart at the bottom of the right-hand column:

Notice there are only three colors there. Just eyeballing it, it seems most people have either the AA or AC genotype, with CC making up a pretty small minority.

But the fact that TT doesn’t register at all on that chart and otherwise just isn’t mentioned anywhere on the page at all makes me think I must be missing something. (Which wouldn’t surprise me. It is genetics, after all. There are definitely tons of rabbit holes to fall down in that area of study.)

Someone else mentioned in a comment that it must be a mutation. (I assume they mean a mutation that happened recently enough – few enough steps up my ancestral lineage – that sources like SNPedia have just never seen it before.) And if I only had one example, and if it was only one allele and not on both strands, I might be inclined to agree.But I have… a lot of examples of that. Just doing some napkin math, I got a rate of about 16.5% “impossible” or undocumented genotypes. There’s no way that for 16.5% of well-known/well-studied SNPs I happen to have a completely undocumented double-strand mutation, right?

(Just to explain my methodology for coming up with that 16.5% figure, I took all the SNP’s listed on these two pages: one and two, and found all SNPs for which a) genotypes are documented on the page and b) I’ve got a genotype for that SNP in at least one of the data files I got from either 23andme or AncestryDNA (or both). That got me 139 examples. Of those, 23 of my SNP’s didn’t match any of the genotypes listed on the Eupedia pages. For a rate of 23/139=0.1654676…~=16.5% .)

(Side note: oddly of the 23 “mismatch” examples I mentioned, my genotype doesn’t have a single allele in common with the documented possible alleles for the SNP. For example, I don’t have any AT’s where the documented alleles are AA, AC, and CC. My genes either match the documented alleles or have no alleles in common with the documented genotypes. Which seems even stranger.)

(Another side note: if 23andme and AncestryDNA didn’t agree on the genotype, I’d be inclined to think it was an error on one of their parts, but I haven’t found any specific SNPs where they disagree with each other yet.)

So, to get back to your main question:

Does “possible” mean “normal values for x% of the population?”

By “impossible”, I mean I haven’t been able to find any documentation of anyone else having the same genotype as I have for that particular SNP. And that makes me feel like I’m almost definitely not understanding something.

I kindof doubt that the genotypes I have for these mismatches are actually exceedingly rare or completely undocumented or anything. I think probably someone knows exactly why I’d be seeing the results I’m seeing (and it probably isn’t tons of obscure mutations or anything.) So, honestly, “impossible” is probably bad wording.

It seems like the opportunity to use this image comes up nearly daily lately:

I can definitely see a lot of good applications for this way of doing things.

It does seem like I often run across “error handling” code that literally just catches a bunch of different exception types and throws a new exception with the same content from the caught error just reworded, adding literally zero helpful information in the process.

It’s definitely the case that sometimes the exact sort of crash you’d get if you didn’t handle errors is exactly the best sort of exception output the program could do given its particular use case and target audience. Or at least it might be best to let the error be handled much further away in the call stack.

Well, 50% is “at least 15%”.

I see your granddaddy long legs and raise you a great granddaddy long legs.

(I’ve never heard them called “great granddaddy long legs”, just “granddaddy long legs” and “daddy long legs”. But I think I’ll start calling them “great granddaddy long legs” now.)

That’s why I smuggle in my own chicken wings.

Open Source that shit. (If you want to get it to a more “done” state first, that’s fine.) The world can always use more FOSS.

I can’t imagine you’re the only one in this situation. If I were in your shoes, I’d search for similar stories online and see if I could get a sense of how friendly the company is to swapping OSs. For some companies, changing the OS is a complete deal breaker. Other companies are pretty willing to assume the issue was indeed strictly hardware and had nothing to do with changing the OS, and thus will go ahead and do the repair.

If you find that company is more like the former, install Windows. If not, just start the warranty repair process.

That “occasional phase” is called “keto diet.”

Nothing about copyleft causes the “owner” to not hold the copyright on a work.

Copyright gives the holder (either the author or the party to which the copyright is assigned) a few specific (but broad) exclusive rights to the work: reproduction, preparation of derivative works, distribution, public performance (which probably doesn’t so much apply to software), and public display (also not applicable to software, so much). (And then there’s circumvention, but that’s yucky and irrelevant to this case, so we’ll ignore it.)

“Exclusive” means nobody is allowed to do any of those things except the copyright holder (unless the copyright holder licenses those rights to others, but we’ll get to that.)

The copyright holder can give/sell/transfer the copyright to someone else (in which case the previous holder is now excluded from doing with the work all the things in the first paragraph above because someone else now holds all those exclusive rights), but that’s not what the AGPL does.

The copyright holder can also license any or all of the exclusive rights in the first paragraph to some person or party (or in the case of an “open license” like the AGPL, to everyone).

The AGPL licenses rights like distribution and preparation of derivative works to others (under certain conditions(/covenants) like “you can only distribute copies if you do so under the same license as you got it under”).

So, if some hypothetical party named “Bob” started a project, they’d hold the copyright. If Bob put the AGPL on that project and also required any contributor to assign copyright on their specific contributions to Bob, Bob would hold the copyright on the entire project code, including all contributions. Someone else could take advantage of the terms of the AGPL allowing derivative works and redistribution to create their own fork (so long as they abided by the conditions(/covenants) in the AGPL), and if they did so, they could omit on their fork any copyright assignment requirement, in which case the fork could end up owned by a mishmash of different copyright holders (making it hard to impossible for the administrator of the fork to do anything tricky like changing what license future versions were under.)

However, on Bob’s original (non-fork) version, if Bob, as the copyright holder, changes the license file to something proprietary, Bob has (arguably?) created a new work that is not the same work as the previous version, and Bob can license that new version under a different license. (I suppose one might be able to argue that changing just the license file isn’t legally enough to make a new version, but the very next time a nontrivial change was made to the codebase, that would qualify as a new version, so it kindof doesn’t matter.) Bob has already licensed previous versions of his non-fork under the AGPL, so Bob can’t really rescind that license already granted on older versions. But new versions could indeed be put under a different license. (Mind you, there are licenses that have specific terms that make them rescindable on old versions. Take for instance the Open Gaming License fiasco that WotC tried to pull not terribly long ago. But I don’t think the AGPL is a license that can be rescinded.)

Since Bob can’t rescind the license on older versions, if Bob made a future version proprietary, the community or any particular party that wanted to could take the last AGPL version of the non-fork and make a fork from there that remained under the AGPL.

The moral of the story is: if you don’t want the copylefted code project you start to be changed to a proprietary license later, don’t do any copyright assignment agreement. The codebase being owned by a diverse mishmash of different copyright holders is a feature, not a bug.

And, as mentioned elsewhere in this post, Immich is owned by a lot of different copyright holders as it has no copyright assignment requirement.

Can you name one other personality with a large following that comes even close to Louis Rossmann in bringing stuff to light and fighting back against enshittification?

Well, there’s Corey Doctorow, of course. He literally wrote the book on Enshittification.

There are definitely more “behind the scenes” folks doing a lot for that particular cause who don’t so much have a following of anywhere near the same size, but nonetheless do fight enshittification in big ways. Bradley Kuhn comes to mind.

Huh. So anyone could maintain a fork or patchset and distribute builds that were feature-for-feature identical to Immich but with no nag screens. Just an interesting thought.

From what I’m seeing, you’re right. If there was a contributor assignment policy (some official policy associated with Immich saying that by submitting a PR, you agree to assign copyright on your code changes go the Immich project), FUTO could change the license on future versions as they wished. But it doesn’t look like there’s any contributor assignment or contributor license agreement on Immich.

To be pedantic, Immich did change from MIT to AGPLv3 a while ago. FUTO could technically scrap the current version, grab the last MIT version of the code, relicense it under their “source-first” license (or any other license they like, pretty much), and declare “this is now the official development version of Immich from which new releases will come.” That would be drastic even for FUTO, though (I don’t think that’s likely any time soon), and the community could then fork the latest AGPLv3 version with a different name and carry on with development.

That is not supporting fascist projects

Literally what I just said in the comment you responded to.

Ah. My mistake. I’ll edit my comment.

Edit: According to another comment in this post, FUTO “took over” Immich. Seems like maybe Immich was AGPLv3 before FUTO got hold of it. Still qualifies as “one of FUTO’s projects”, and your point is still well made, but it does still add a bit of context, and honestly I have to wonder whether future versions of Immich will remain FOSS.

The article I linked in another comment explains more, but Eron Wolf, founder of FUTO, kindof pressured or hoodwinked Louis Rossmann into publicly interviewing Curtis Yarvin who happily refers to himself as a “reactionary fascist” and publicly states that black people are inherently suitable for enslavement.

I don’t know that it’s so much that they support “fascist projects” as much as they go out of their way to be a platform for spreading fascist propaganda, and particularly promoting the fascist Curtis Yarvin.

This is all in reference to this article.

FUTO is an organization that talks a lot of rhetoric about being some bastion of consumer rights in tech, but they’re doing a lot of shitty, shady, and downright evil things. Among them, FUTO has been in the practice of making small grants to FOSS projects (like ffmpeg and musl) and then plastering the FOSS project’s name and logo all over the FUTO site in a way that makes it seem as if FUTO is endorsed by said FOSS projects when that’s not the case at all.

(All this after doing everything in their power with their rhetoric to try to discredit and degrade the entire FOSS community. They wrote an “apology”, but even in the apology, they express their “disdain for OSI approved licenses”. Mind you, none of FUTO’s projects are Open Source most of their projects are proprietary.)

After that article came out just a couple of days ago, apparently they redid their site, I’d have to guess in an effort to address the concern that the way FUTO presented their grant program before implied endorsement by a lot of FOSS projects that didn’t endorse them in any way. I don’t think they’ve done enough, and there are tons of other reasons to think FUTO is evil assholes using consumer rights rhetoric to manipulate people in service to its (fully for-profit) bottom line.

Other concerns in the article include FUTO’s connection to explicit/proud fascists and using their platform to (even coercing Louis Rossmann into) spread fascist propaganda.

And I’ve got plenty more to say about how evil FUTO is.

Doubtful, unfortunately. FUTO seems like the sort of narcissistic assholes who won’t be swayed in any way, shape, or form

There’s no third option between FOSS and proprietary (unless there are licenses that match the Free Software definition but not the Open Source definition or vice versa, I suppose, but I’m not aware of any). All software that is not FOSS is proprietary by definition, whether the source is available or not. It’s not “disingenuous” to call FUTO software proprietary. It’s simply factual.