dasyatidprime

Because many forms of human social norms include a “private but not strongly so” register with neither the friction and psychological barriers of ‘true’ secrecy nor the overt exposure of publicity, and in fact a huge amount of communication takes place in that register, such as most conversation among groups socializing in public spaces.

In the digital world, this hits a wall of “the security mechanisms don't integrate with the nuances of human social understanding directly” and gets split off into “posts that are technically public but not really addressed to the public” and “invite-only areas which would probably be okay being somewhat more open but lack a good way of moderating this without leaving things open all the way”.

And GMAC Bank later became Ally Bank.

HSV and not HSL for that mental model? That's surprising to me—I thought the tint/shade system would lead to wanting to put max-chroma in the middle.

Because if you're modifying a function and you're adding a new variable to hold some intermediary value, you might accidentally pick a variable name that already exists. If you're in the habit of writing := to signal your intent to create a new variable, that error will probably get caught at compile time rather than being a source of subtle bugs where you trample data needed by something else.

The multi-variable case is trickier, especially because (result, err) is such a common pattern and you neither want to impede the very common case of storing a result into a new variable nor to require a profusion of differently-named variables for errors that are very likely to just get folded anyway. So it allows it if any of the variables are new rather than requiring all. But clearing a large chunk of the bug surface is still much better than not.

(This is only my extrapolated reasoning, of course.)

They both contain both elements, but the point is to sharply alter the emphasis. “Software as a Service” is a reasonable description in the salience model of the mainstream software industry, because it is something that you choose and integrate and use like software, except it just so happens to be a service, coming with various costs like subscription fees and very direct vulnerability to upstream changes, and various benefits like installation and maintenance functions being centralized far away where you don't have to directly manage them.

“Service as a Software Substitute” pushes the ‘service’ part to be the most salient. It depicts something which is fundamentally a service, where ‘substitute’ once more emphasizes that you do not, in fact, have the software itself, even though it is taking the role of software. The FSF considers this very important, because they wish users to be able to copy and modify the software they use, and pseudo-distribution purely as a service does not naturally allow for this. If that is not something you care about, then the emphasis will seem strange, yes.

The GPL doesn't require “contributing back”.

You can't just give out binaries without source, but you can choose not to release anything. The thing is, as usage models shift to include a lot of what the FSF very reasonably relabels SaaSS (Service as a Software Substitute), you can release nothing but provide the functionality over the network and reasonably expect people to use it that way. Since you haven't distributed the software itself to anyone, the GPL doesn't put you out of compliance for not showing them your code or letting them distribute it further.

This is the situation the Affero GPL (AGPL) was intended to address—it requires (or tries to require) that source code also be accessible to anyone who is given access to the software over a network. (Imprecisely speaking, that is; do check the license text if you want the specifics.)

That would have much the same cost effect if the server were kept up for it, and wouldn't be effective if the server were taken down.

‘support’ can take the form of financial support, in which case anyone who would be able to pay for SaaSS subscriptions or proprietary software would only have to do something very similar here. The biggest problem there is that the social infrastructure and coordination points don't seem to exist, and that's a place where the fragmentation probably does hit badly, where consolidating users' needs more would let development be much more efficient and coordinated in such a way that crowd financial support would become more economically plausible at least. Of course, there's the free-rider game-theoretic issue, among other things…

While there's a lot of issues between “where are are now” and “broadly end-user-friendly libre software support models being common enough to be recognizable and repeatable”, “sure, so long as you can code” isn't meant to be the first answer. It's just become a cultural laughing-point because a lot of people do ignore the issues, and “you can choose whom you ask to handle the code” requires more work on the environment to be practical. IME, “exclusionary mindset” usually has little to do with it.

The “VCs have conditioned people to never pay for software, so you may have to have massive scale and run on surveillance capitalism to survive” and “user model of technology being vertically integrated service by default” are of course a whole other environmental game which may crush attempts to make this work.

Triplebyte was trying to do this a while ago, weren't they? My last observations from a year and change ago suggested they were running into what might have been signaling equilibrium issues / risk aversion, but I'd be curious if there's other perspectives around.

I tend to term such a device a ‘handset’ for that among other reasons.

Having a workplace sponsor a separate GitHub account for your work there would be reasonable, but is this actually common practice? I certainly wish it were, for the above reason, but I can't speak to the reality.

GitHub forbids multiple free accounts. https://docs.github.com/en/github/site-policy/github-terms-o...

It's strictly dependent on the table definitions you use. A column with no type (the simplest way to use them) will take data in any of the natively supported types—whatever you actually insert in your SQL statement—and will not attempt any translation.¹

If you do add column types to your tables as in traditional databases, then of course you should expect some varying behavior based on that.

I can imagine some potential for issues from people expecting to insert exact decimal numbers and choosing floating-point by accident, perhaps, or using integer types for actually-text numeric ID fields. If you go all-strings on input with no column affinity, then you'll get the same strings on output, though, so that's isomorphic to good-CSV.

¹ https://www.sqlite.org/datatype3.html – §3.1 “If the declared type for a column contains the string "BLOB" or if no type is specified then the column has affinity BLOB.” + §3 “A column with affinity BLOB does not prefer one storage class over another and no attempt is made to coerce data from one storage class into another.”

The most amusing part about TNT is that it's called TNT, it's crafted with gunpowder, and it looks like dynamite, which in real life are three different explosives.

Aka “malignant citogenesis” to the xkcd generation.

I'd worry that translating this to an end-user-relevant concept of security would lead to a lot of scares, though.

Probe all the files in a directory to see which ones are “yours”: “What? Why is it accessing all the files? So suspicious!”

Require a specific name pattern or something: “I never have to remember to do this on the other apps…”

There's a lot of these tradeoffs that in human life are resolved through reference to all sorts of subtle human things that the machine knows not of. We're at this liminal point where “app” software is given a bare form of “agency” from a social perspective as an extension of its developer, but it doesn't have the intelligence to negotiate over it much (and I think that's behind some of the model-simplification pressure that's encouraged heavy vertical integration).

That's very dependent on whether the definition of “good” can be yanked around by what the rest of the industry does, for example by ease of compatibility with other software APIs or by UI and behavior models that are familiar to lower friction for people who have already used something popular.

Anything else that would ‘force’ you to Have an Account somewhere, combined with that somewhere happening to outsource their identity framework to “social” of any kind. Or: the same things that would ‘force’ you to have a social network account to start with, potentially.

To get on my hobbyhorse some: what about a platform that your employer uses to distribute relevant documents and updates which you need for your job, which has become sufficiently well-known and implicitly available that hesitating about the dependency is perceived as bizarre?

Note that the response of “don't deal with them then” runs into market information and churn amplification¹ issues when it is a social assumption that picking up a ‘tool’ (which is actually a relationship with a third party, but where this is close to invisible in the steady state) is essentially a free action which requires no consideration, because the information about “does this employer require me to use this tool” neither propagates efficiently nor stays stable.

Past source: wound up changing the email address on one of my Google accounts recently for exactly that reason. They actually asked me up-front whether I had a GMail account they could use instead, which turned out to be because they use restricted Google Docs for critical material. Not naming them, but in a broad sense, this is one of the more ethical companies I've ever dealt with, by the way.

Future source: I should probably be considering digging into LinkedIn soon despite their past abusive behaviors, because as it turns out, if I want to dig myself out of this hole…

¹ I assume there's a ‘real’ term for what I'm thinking of, but I don't know what it is, so I cobbled that one together out of the most relevant-seeming bits.

This is not an unreasonable point, but I will note that I was partly using “AWS” metonymically to refer to a whole cluster of technology decisions surrounding the current wave of “cloud” everything, and I was also partially referring to the purer “have you been part of a full-scale service deployment with real users” aspect—there's a bigger difference between “operating a service when there's hundreds of thousands of requests from real people hitting you and they really count” and “operating a service for a random who-cares project” than there is between, say, “having a precise understanding of how Ruby behaves in a business-critical backend” and “having a precise understanding of how Ruby behaves on your laptop”. Lacking experience or even just lacking legibility of experience on the former seems to close a lot of doors, and that means you run into the “need experience to get experience” cyclic dependency on a critical subset of what's needed, regardless of how good your other skills are.

Do you have any recommendations on how best to sell yourself from such a position? I have a significant C99-era/close-to-the-assembly fundamentals background, but I see a lot of “if you haven't already AWS'd all the AWS at your last company in production at scale, then there's someone behind you who has”, and that kind of thing has both been a notable lacuna in my experience and something that's notably less fixable independently than e.g. “plow through a React tutorial to familiarize yourself with the basics”. In particular, it seems like in the heavily external-service-integrating style of development, a lot of experience comes in along the lines of “remembering the pitfalls you ran into with particular vendors under particular loads” and the fundamentals don't get you as far. But I hear a lot of conflicting things about this.

I'm someone who's (supposedly) quite good underneath but has stagnated a lot over the last few years (and on and off before that, sadly) as other issues drained away my ability to work. I'm gradually stabilizing things and trying to find the best way to maneuver, and I'm pretty worried that everything's going to pass me by because my experience isn't of the right kind and I'm not legible enough. The people who are getting the jobs with all the traits I want are the ones who got a Real (that is, close to culturally archetypical) Job in 2018 and did their time in the salt mines with the three verifiable contiguous recent years of experience.

If what you say is true, then it's possible what I mostly have is a marketing problem, and it may be that e.g. some of the “actual demonstration of ability” is more readily solvable with “pull some stuff out into public repositories and freshen it up” than I've been imagining.

> we don't legally require every that everyone who posts an image on social media include a written description

Not that it takes too much away from your point, but I've experienced an interesting gap in this example. While not legally required, big chunks of the short-form-text fediverse (Mastodon/Pleroma/…) have had circulating posts recommending descriptive text for image posts, and I'm actually surprised by how many people get into the habit of complying naturally—perhaps because there's also an easily-noticeable slot in the UI for it? Ten or so years ago I remember it being like pulling teeth to explain to some people doing media projects on the Web that this kind of accessibility was important, and now with what seems to be culturally a similar crowd… huh, y'know?

You should look into https://www.w3.org/TR/exi/ if you're interested in binary XML encoding.

People will install 20 social media apps for 20 different psychological contexts that have subtly different needs and can use a UI distinguisher to key their brain in on which set is active.

People won't install another social media app if there isn't sufficient incoming social credit encouraging them to use it and the need falls into an existing category. Something sufficiently different can gain an “I couldn't have done this with” credit to overcome some stickiness with.

… maybe. Sort of. I think.

(I wonder what this means for e.g. Facebook's acquisitions of Instagram and WhatsApp, if it's true.)

It's true, but worse is also better; of reasonable answers, the “right” one is often the one that will be popular enough for people to rearrange the world (sometimes retroactively) to make it right.

When thinking of that as the default, within the culture, “we know what it actually means” is the most salient.

When thinking of that as not the default, in a cultural comparison, “how are analogous-role conversational elements expressed differently” is the most salient. So compared to a culture in which you don't ask that so cheerily, obligatorily, without-‘really’-meaning, yes, it totally is fluffy.

(Speaking as someone who's theoretically American but culturally ambiguous, who's recently been suggested by someone reviewing my social skills in a professional capacity to inject more low-content happy-mood pleasantries into my conversation—they may have actually used the word “American” in the process, and I hardly disagree with the connection. Of course, that's papering over a lot of intra-American variation too.)

And then notice that a Forth is a sort of meta-mini-interpreter. (Whether it's good to use one in any particular case is a different matter, of course.)

Well, I still want to take over the world, but my first few plans got blown out of the water badly enough that I'm having to fall back to “scramble to unbreak my own body and mind before I can try again”, and the forecast is not so good.

(How much of that is a joke is left as an exercise for the reader.)

The techniques are often effective at manipulating you even if you are ‘aware’, unless you are also so very well-aligned internally that the consciously aware part of you can and does counter-affect the subconsciously affected parts strongly and consistently enough. Intelligence and knowledge won't save you. Combining them with ‘enlightenment’ and/or strong detachment might, but that's a much taller order.

In the case of central/popular social/communication software, this is much, much worse, because it can be set up so that taking up the mindset for resistance directly conflicts with taking up a mindset that will be (or be read as) adequately engaged for relationship maintenance and continued connection; the persuasion can use your friends as an attack reflector, getting mixed up in the signal. And no amount of purely individual willpower will help you not get steamrolled by that so long as you remain dependent on the society around you.

(I want to find an analogy to use for the latter—I'm almost tempted to say it compares to historical examples of ideological spirals the same way nuclear explosives compare to conventional ones, but that feels not-quite-right. History pointers appreciated if anyone has 'em.)

Last I checked (which was quite a while ago), Android with multi-user support did in fact assign one Linux UID to each (user, app) tuple! But I don't recall there being a particularly rich privilege model available in practice for the multi-user sharing case, only for isolation. Inter-app intents were handled using Binder IPC underneath; I don't know what use that made of the Linux credentials.

Many server applications handle user separation internally, without reference to the underlying OS, while application separation is much stronger (separate VMs, SELinux, etc.), and desktop platforms have user separation but often-unsandboxed apps, so those are in some ways duals of each other…

I'm not sure what Windows does with UWP and sandboxed apps from the Microsoft Store, but that would be a good place to look.

To add a note to that, caricatured for demonstration: Why would you ever execute on an answer which no one with the right kind of status has told you is okay to execute on?

In the harder domains of STEM? Okay, sure, maybe the answer is self-proving and once your weird innovation works great your detractors will have to eat their words. But subjective things like emotions? You know what happens to people who have some wrong emotion, don't you? Even if your conscious mind doesn't, what's underneath sure might. Highly not recommended to put it that way to yourself raw! And yet.

Emotional processing infrastructure is an important part of society, and its blueprints an important part of culture. I suspect that the more densely packed our sociopsychological world is, the more the equivalent of mental building codes and city utilities are something that has to be negotiated to make life workable. Hindbrain Owners Association, anyone?