For me in Europe, Google Maps coverage quality can be best described by this personal anecdote.
I used to live in Nürnberg, which is in top 10 German cities by population, and where are some major and well-recognized international companies are headquartered.
Nürnberg has a subway system (U-Bahn) since 1980s, and it's significant enough: a few dozen stations over three lines (one is fully driverless, btw).
Google didn't have any representation of U-Banh in Nürnberg till at least 2017.
I don't mean "wasn't supporting it in navigation and routing", I mean " stations weren't even marked and labelled on the general overview maps.
And it's not like they didn't have the into: there was a widely-used user layer which added at least station labels.
They just simply didn't care enough, and had other priorities.
In the meanwhile, the level of detail on OSM covered details as minor as every mailbox not only in Nürnberg, but in every small town around Frankonia (I used to participate in postcrossing and used this a lot from random places).
As a counter argument, here in Sweden I've yet to run into a place where Google maps has failed me when it comes to driving, public transport or address/POI search. OSM on the other hand is missing half the buildings once you get more then 15 km out from major city centers and even in major cities, things like house numbers and addresses is often wrong. The only scenario OSM is better than Google Maps is pedestrian and hiking routes.
France and Spain had a pretty good coverage, Japan cities are decent as well. I've only seen Munchen and Kolhn's most touristic areas but there were decent enough we didn't hit any critical issue. Commerce data and opening hours was abysmal on the other hand.
OSM was pretty good too in France but has different issues: they don't get the same access to up to date commerce data as exposed through local aggregators, and there's just not enough user data to have good heuristics on navigation times.
OK, since we're pitching our SSS implementations here in comments, I welcome everyone to check out BananaSplit, https://bs.parity.io
Not sure about year 2023, but at the time I wrote it for my previous employer there was nothing remotely usable for regular user.
Thus, BananaSplit.
It doesn't allow you to specify many parameters (just the number of shards, and then requires 50%+1 to recover); aimed at printed backups (generates printable full-page QR codes, while asking to copy a decryption phrase to the pages by hand to avoid an "evil printer" attack); and takes the concept of _portable web app_ to its extreme, being a self-contained single html file which requires you to save it locally and open via file:// protocol for it to work.
Disclaimer: while being in use for years, the code had never seen a proper independent code review; there might be bugs, despite me trying to minimize their impact by design, and using only reputable (and pretty minimalistic) JS primitives.
If you want to check out the sources yourself before using, of course those are available under GPL at https://github.com/paritytech/banana_split/
I like how it pushes you towards safety by requiring the code to run offline and off the disk. The resulting UX is not that great (though the explanation of what to do is clear), but it's a good step towards a little bit more security.
On the contrary, I hated it. I just want to see the UI, I don't want to go offline! Give me an "I'm just playing around, I promise!" button I can click on to test it.
Interesting idea, but the restore function doesn't work for me.
Access to fetch at 'file://redacted/Banana%20split.html#/combine' from origin 'null' has been blocked by CORS policy: Cross origin requests are only supported for protocol schemes: http, data, isolated-app, chrome-extension, chrome, https, chrome-untrusted.
Would it be too much to ask for you to open an issue in Github for this?
Things I'd be interested in the most would be details about your environment: browser/version/platform and if this is reproducible in a "fresh" browser profile without any extensions added (or just a list of your extensions, if that's not too privacy-invasive for you).
Mozilla SOPS¹ also supports this, but it's not nearly as user friendly for non-technical folks. Probably one of those solutions you reviewed before creating Banana Split!
We probably shouldn't even bother recommending browser-specific offline modes; I've created an issue in the project's repo to reword that piece by the next release.
Yeah, and I failed to even find an option to open a local HTML file on mobile Firefox (for Android) just now.
That's a shame — things were definitely different in 2019 when I built the initial version; and mobile browsers were definitely a target I had in mind for the tool (especially when it comes to the recovery).
Instead of wrapping the existing tool into a mobile app, I'm thinking about standardizing the QR code format from the tool a bit more — so multiple, more task-specific recovering mobile apps would be possible. (Like, the one in your password manager detecting certain internal text formatting and importing the entries automatically and such).
The idea is that our taxonomy selects a group named "trees", but there's little internal coherence in that group, which probably results in more online hilly wars in the plant-loving communities than we the laypeople can think of.
I don't think this article is referring to physical organisms known as trees (as your article does). It's talking about abstract tree structures, in particular:
1. There is a coherent parent-child relationship
2. Children have only one parent
3. Children have no lateral relationships with other children
4. Relationships are mediated through parents, not e.g. grandparent-child directly.
5. We can coherently differentiate between different children
etc.
All of these are assumptions that greatly simplify analysis and therefore useful in any situations, but almost always fail in some way on closer inspection.
More generally we go to the saying "all models are wrong, some are useful."
I'm 95% sure that the author (who is on HN[1]) is at least referring to the article "There's No Such Thing As A Tree" in his line "There is no such thing as a tree." - regardless of the fact that the article as a whole isn't about trees made of wood.
It's possible that he's additionally making a double entendre about abstract tree structures.
What exactly do you mean by "our taxonomy". There's no plant order, family, or genus called trees. It's just a common plant form. Do you just mean some people's personal mental model?
Does this imply that GPS satellites are a legitimate military target for Russia right now? (I'm pretty sure GPS is used for guidance of drones and missiles pretty heavily there.)
Or for a satellite top be a legit target the communication needs to be two-way? In such case I suspect Türksat satellite should be considered within scope, since Bayraktar TB2S (which reportedly has been used by Ukraine in this war) uses it for SATCOM.
What about other military uses of Starlink, outside of guidance systems? Those are plentiful, and it's hard to see why there would be a drastic difference between guiding a drone and providing communication backbone for military operation coordination in the eyes of the Russians. After all, they are bombing civilian infrastructure just fine, and didn't even care to formally declare this "special military operation" a war...
My point being, I don't think "legitimate target" has to do anything with Russia not shooting down any satellites; it's either lack of capacity or fear of retaliation, both being rather orthogonal to the targets "legitimacy" IMO.
I am damn sure that GPS satellites are a military target! They are even owned by a military organisation. It just so happens that organisation is the most powerful military power in the world. Given that, the next best thing is to jam GPS signals on the ground and I would be very surprised the Russians are not doing it at some scales.
I highly doubt Americans gave the Ukrainians the keys for the special modulation that bypasses usual jamming abilities.
Yeah I agree that SpaceX's concern is unlikely over potential attacks on their satellites. Not only is SpaceX able to put them up faster than Russia could shoot them down, attacking a Starlink satellite would be an act of war against the US regardless.
I think SpaceX's concern is more related to domestic liability. The US's policy has been that they will not provide offensive aid for attacking into Russia. They've been very careful about the range of the systems they send for that reason.
Unrestricted Starlink is not subject to those range limits. So, if the attacks which went deep into Russian territory involved Starlink mounted to drones, it would essentially be in violation of the US's own stated policy about military aid to Ukraine. Thus far this hasn't caused any trouble, but if something were to happen (eg an escalation), SpaceX might get thrown under the bus by the US government since they aren't contracted by the DoD to provide such service.
SpaceX wasn't worried about any of these issues until, mysteriously, "SpaceX" suddenly was coincidental to Elon Musk having his Twitter breakdown when he decided he personally was going to negotiate peace for Ukraine[1] and was promptly rebuffed by the Ukrainian foreign minister.
Then, totally coincidentally[2] suddenly "SpaceX" had very big corporate concerns about how Starlink was used and also wanted more money from the DoD at a higher commercial rate[3] and...
You know, just a whole pile of totally normal coincidences...
>SpaceX wasn't worried about any of these issues until...
If they weren't worried about these issues, why are they very explicitly spelled out in the Starlink terms of service?
The issue came up because Ukraine violated the terms of service by using them as command/control of an explosive drone boat system.
You're just making things up.
Biden administration decided to halt plans to weaponize space started under Trump. This mothballing decision was made in early 2022 and led to said reactions by Elon.
>SpaceX is able to put them up faster than Russia could shoot them down
This doesnt seem correct to me. Building, launching, and positioning replacement satellites to fill gaps in service seems like it would take much more effort and cost than a takedown.
Can you further expand on why you think your claim is accurate?
Russia doesn't have 3500 ASAT missiles and can't build (and launch) them at a rate of ~50 per week (it's estimated that they produce ~50 cruise missiles per month).
Admittedly they don't have to shoot down 3500 sats, probably only the sats in the group servicing Ukraine, but that's still a couple hundred sats and thus, likely more than the Russian ASAT stockpile.
The only two ways I can see to make ASAT logistics work against large constellations like Starlink is if they figure out laser-based ASAT or an in-orbit attack mechanism that shoots multiple cheap missiles/bullets (avoiding expensive separate launch per target). Neither sounds impossible, though.
Yeah it's difficult to envision a truly cost effective ASAT measure, "shooting" bullets is deceptive in that it'd still require carrying the large amount of fuel needed to effectively change orbit (especially for changing inclination) and would be a bit too messy in terms of spread of the resulting debris.
Another interesting and clean approach might be jamming the satellites from orbit. You can't go after all the satellites, but since you can predict which ones will be over when you're doing something important, you can launch vehicles to approach specifically those and jam them at close range.
One hit will create thousands of fragments. How many consecutive hits would be necessary to create severe enough Kessler event to destroy most of the Stralink satellites? Dozens?
Blowing up dozens of satellites would start a chain reaction that if left unchecked will eventually bring down all of them. But I doubt it would happen over a timeframe that's useful for this war. We arguably are already in a Kessler syndrome situation where fragments are created faster than they deorbit, but like any exponential process it starts out incredibly slow.
Space is big. The average breakup seems to create about 300 fragments [1], so if Russia blows up 100 Starlink satellites that's 30,000 bullets that have to randomly hit another 3500 objects spread over an area roughly the area of the surface of the earth. Except that it's worse because space is 3-dimensional and most bullets will spend little time at that altitude, and eventually deorbit. Over a decade it will do a lot of damage, and at some point things will escalate to a point where it's hard to handle. But Russia needs results within weeks or at most months.
No it wouldn't bring down all of them. They are not all on the same level and they can raise and lower their level. Debris analysis would be done and the fleet would be commanded to move strategically.
Causing a chain reaction is significantly more difficult then people think.
As everyone knows, the best possible way to deal with an aggressor is just straight appeasement. The risks are so large that appeasement is the only answer. If only we had historically tried appeasing belligerent foreign powers, think what wars would have been avoided... /sarcasm
Ironically this only further proves the point: US expansionist foreign policy would not have been checked by giving the US what it wanted. There was no version of Iraq complying with the UN which was going to prevent the neo-conservatives from getting their regime change invasion in 2003.
The only preventative measures which would've worked would be diplomatic and economic counter-pressure from the UN and more specifically Europe and other NATO allies, or substantial military build up capable of seriously endangering US conventional forces.
Ukraine would be in a much different military position today if it were a pariah state, as opposed to a nation declaring a strong intent to join both the EU and NATO. Turns out diplomacy, allies and soft-power are actually kind of useful things when you're not the dominant military hegemon.
Yes. The lesson from history is that no appeasement should ever be done. Any war, no matter how small or over how little a thing, MUST ALWAYS BE TOTAL. No war should ever end until one nation enacts genocide on another. Great plan.
So I'm sure you have literally any example of when that worked, historically?
Spoiler: it has never worked. What you are thinking of are diplomatic, economic and military moves and counter-moves or negotiated settlements but the aggressor is always forced to compromise or delayed such that they lose their advantage.
Appeasement - the aggressor making demands and getting them with the idea that they will be then be satisfied - has never worked.
>Does this imply that GPS satellites are a legitimate military target for Russia right now?
GPS is a US military asset and consequently a military target for hostile actors.
The only thing keeping Russia from just blowing them out of the sky is because to do so would be an unconditional declaration and act of war against the USA.
And because they are using it in their planes and within their units. Sure you can argue with Glonass, but if it is working so well, why SU-34s are using Garmin with GPS instead?
>Does this imply that GPS satellites are a legitimate military target for Russia right now?
Of course it does. In the real world, anything is permissible (because there is no global federal government) as long as you have the strength to enforce. In the case of GPS satellite constellation, that is infrastructure owned and operated by the US military, so any attack on those satellites brings Russia directly into conflict with the US and NATO.
>My point being, I don't think "legitimate target" has to do anything with Russia not shooting down any satellites; it's either lack of capacity or fear of retaliation, both being rather orthogonal to the targets "legitimacy" IMO.
Right ... To rephrase: is the US military willing to indemnify Starlink to the same degree if it is attacked by Russia?
You can bet they would retaliate, simply because the game theory aspect of all this implies there needs to be a severe cost associated with damaging US infrastructure. Maybe a huge explosion at some vital oil infrastructure or actually destroying the bridge to Crimea.
>You can bet they would retaliate, simply because the game theory aspect of all this implies there needs to be a severe cost associated with damaging US infrastructure.
It doesn't necessarily imply that that IFF Starlink is providing military assistance to Ukraine, and in response, Russia targets Starlink satellites, that the US will necessarily escalate further. And there is precedence for this. For example, after US assassinated Soleimani, and Iran retaliated against an American base in Iraq (leading to ~100 casualties - though no deaths, and extensive damage to the base itself), America did not further escalate.
> US intelligence, surveillance and reconnaissance (ISR) satellites orbit at about 800 km (500 mi) high and move at 7.5 km/s (4.7 mi/s), so a Chinese Intermediate-range ballistic missile would need to compensate for 1350 km (840 mi) of movement in the three minutes it takes to boost to that altitude. Even if an ISR satellite is knocked out, the US possesses an extensive array of manned and unmanned ISR aircraft that could perform missions at standoff ranges from Chinese land-based air defences, making them somewhat higher priority targets that would consume fewer resources to better engage.
> The Global Positioning System and communications satellites orbit at higher altitudes of 20000 km (12000 mi) and 36000 km (22000 mi) respectively, putting them out of range of solid-fuelled Intercontinental ballistic missiles. Liquid-fuelled space launch vehicles could reach those altitudes, but they are more time-consuming to launch and could be attacked on the ground before being able to launch in rapid succession. The constellation of 30 GPS satellites provides redundancy where at least four satellites can be received in six orbital planes at any one time, so an attacker would need to disable at least six satellites to disrupt the network.
Grimmsc version is a bit weird then, because there is a plausibly-sounding "freundian" interpretation of the most common folklore version, in which a frog turning into a prince after being kissed is a thinly veiled metaphor for fellatio.
That said, in the most common Russian folk version it's *Princess* Frog (Царевна Лягушка), and the metaphor is not working that well once again.
It still has limitations (the biggest one is the requirement for the os&architecture to match between the builder and the deployment target) — but the result is a standalone binary which not only embeds the VM and preloads the app's bytecode, but even "trims" the stdlib to only ship the required functions.
Right, so the moral of the story centers on the target user of the CLI tool. If you're building something for the Elixir community - game on I suppose, though there is still the complexity of build-env per OS/arch.
I wonder where WASM/container enters the discussion.
> Firefly compiles Elixir applications faster and more efficiently than the BEAM can, and introduces WASI targeting to run applications in resource-constrained environments.
Containers are already solved, its trivial to build and boot a mix release - but whether that's appropriate for a CLI tool depends on the complexity of the tool I guess, but not too far from flatpaks etc no?
Fun fact: Russian translation of Dancing men is legitimately solvable, with messages being _in Russian_!
The translators had to butcher the characters' names and twist the deduction logic here and there — but it's still a marvel of localization work (especially given the languages, grammar and even alphabets being very different).
I've wrote https://bs.parity.io/#/ exactly for this; it also tries to avoid the most obvious opsec mistakes, like trusting your printer too much =)
BananaSplit is of very experimental quality, it hasn't been battle-tested and independently reviewed yet.
But at least it's there, it's FOSS, and I encourage you inspecting the code, contributing to it, or re-hosting the HTML.
Judging from the discussion in [1], the "going down this road" already started to happen about half a year ago.
Please notice how everyone in conversation seems to implicitly understand that their customers won't like the new forced telemetry, but still pushing on this, "for customers' own good" (with the only possible exception being made for public sector customers).
For me in Europe, Google Maps coverage quality can be best described by this personal anecdote. I used to live in Nürnberg, which is in top 10 German cities by population, and where are some major and well-recognized international companies are headquartered. Nürnberg has a subway system (U-Bahn) since 1980s, and it's significant enough: a few dozen stations over three lines (one is fully driverless, btw).
Google didn't have any representation of U-Banh in Nürnberg till at least 2017. I don't mean "wasn't supporting it in navigation and routing", I mean " stations weren't even marked and labelled on the general overview maps. And it's not like they didn't have the into: there was a widely-used user layer which added at least station labels. They just simply didn't care enough, and had other priorities.
In the meanwhile, the level of detail on OSM covered details as minor as every mailbox not only in Nürnberg, but in every small town around Frankonia (I used to participate in postcrossing and used this a lot from random places).