• 13 Posts
  • 923 Comments
Joined 9 months ago
cake
Cake day: October 4th, 2023

help-circle
  • since this type of interference is geographically relevant, it clearly has nothing to do with the war.

    So, I’m not sure what the rationale is – it might be worrying about long-range drones, like those light aircraft, flying around Belarus, but I’ve definitely seen some sources say that they believe that the Kaliningrad jamming is a function of the war.

    This article, from late May, has the Finnish government saying that they believe that it’s related to the war.

    https://www.politico.eu/article/gps-jamming-is-a-side-effect-of-russian-military-activity-finnish-transport-agency-says/

    Jamming GPS signals over the Baltic Sea is “most likely” a side effect of Russia’s anti-drone activities, Traficom, the Finnish Transport and Communications Agency, said today.

    “The interference intensified when Ukraine’s drone attacks on Russia’s energy infrastructure began in January 2024,” Traficom said in a press release.

    Estonia also blames Russia for the signal jamming, but the Finnish agency doesn’t agree with the Tallinn government in defining the interference as a hybrid attack.

    “It is possible that the interference observed in aviation currently are most likely a side effect of Russia’s self-protection” that is used “to prevent the navigation and control of drones controlled by GNSS [Global Navigation Satellite System] or mobile frequencies,” Traficom said.


  • Well, as things stand, there’s no right in customary international law to a radio frequency. If Russia wants to produce disruption, Russia basically can. It’s going to annoy countries, but there isn’t really a lot that can be done aside from pressure, sanctions and such. Hell, even if there were such a right, it’d ultimately need to be backed up by ability and willingness to use force at some level, by someone. I’m not sure that countries are willing to wage war over radio broadcasts, no matter how disruptive.

    In theory, a country could probably run a pretty powerful broadcast off a ship. So I don’t think that Kaliningrad is even all that special here – Russia could be doing the same even if it didn’t have a little enclave of territory and didn’t mind annoying whoever is using the signal.

    My guess is that Russia will stop when the war ends. That is, my assumption is that Russia isn’t willing to do peacetime jamming for the purpose of just being obnoxious.

    If it’s enough of a problem, it’d probably be possible to set up navigation systems that are more-limited, don’t have the degree of military utility of something like GPS. Maybe reactivate LORAN, say.

    In the case of transatlantic aircraft, I’d assume that those normally have INS navigation systems; while limited in accuracy, those should be usable as a backup to GPS in most roles.



  • Aside from broken controllers, which I don’t think can reasonably count, the Atari 2600 joystick.

    One button, a lot of resistance to push on the stick.

    After that, an elderly Logitech gamepad from the 1990s that had a D-pad that rolled diagonal way too easily. IIRC it had a screw-in mini-joystick that could attach to the center of the D-pad. Don’t remember the model. White case, attached directly to a joystick/MIDI port.

    After that, I think the NES controller. I have no idea why people like those or actually buy recreations. Yes, nostalgia, but the ergonomics on it were terrible. Hard buttons, sharper corners on the D-pad than is the norm today, and a squared-off controller made the thing downright uncomfortable to use for long periods of time.









  • Right now when updates get applied to the NAS, if it gets powered off during the update window that would be really bad and inconvenient require manual intervention.

    You sure? I mean, sure, it’s possible; there are devices out there that can’t deal with power loss during update. But others can: they’ll typically have space for two firmware versions, write out the new version into the inactive slot, and only when the new version is committed to persistent storage, atomically activate it.

    Last device I worked on functioned that way.

    you might lose data in flight if you’re not careful.

    That’s the responsibility of the application if they rely on the data to be persistent at some point; they need to be written to deal with the fact that there may be in-flight data that doesn’t make it to the disk if they intend to take other actions that depend on that fact; they’ll need to call fsync() or whatever their OS has if they expect the data to be on-drive.

    Normally, there will always a period where some data being written out is partial: the write() could complete after handing the data off to the OS’s buffer cache. The local drive could complete because data’s in its cache. The app could perform multiple write() calls, and the first could have completed without the second. With a NAS, the window might be a little bit longer than it otherwise would be, but something like a DBMS will do the fsync(); at any point, it’d be hypothetically possible for the OS to crash or power loss or something to happen.

    The real problem, that I need an nas for, is not the loss of some data, it’s when the storms hit and there’s flooding, the power can go up and down and cycle quite rapidly. And that’s really bad for sensitive hardware like hard disks. So I want the NAS to shut off when the power starts getting bad, and not turn on for a really long time but still turn on automatically when things stabilize

    Like I said in the above comment, you’ll get that even without a clean shutdown; you’ll actually get a bit more time if you don’t do a clean shutdown.

    Because this device runs a bunch of VMs and containers

    Ah, okay, it’s not just a file server? Fair enough – then that brings the case #2 back up again, which I didn’t expect to apply to the NAS itself.



  • I’m assuming that your goal here is automatic shutdown when the UPS battery gets low so you don’t actually have the NAS see unexpected power loss.

    This isn’t an answer to your question, but stepping back and getting a big-picture view: do you actually need a clean, automatic shutdown on your Synology server if the power goes out?

    I’d assume that the filesystems that the things are set up to run are power-loss safe.

    I’d also assume that there isn’t server-side state that needs to be cleanly flushed prior to power loss.

    Historically, UPSes providing a clean shutdown were important on personal computers for two reasons:

    • Some filesystems couldn’t deal with power loss, could produce a corrupted filesystem. FAT, for example, or HFS on the Mac. That’s not much of an issue today, and I can’t imagine that a Synology NAS would be doing that unless you’re explicitly choosing to use an old filesystem.

    • Some applications maintain state and when told to shut down, will dump it to disk. So maybe someone’s writing a document in Microsoft Word and hasn’t saved it for a long time, a few minutes will provide them time to save it (or the application to do an auto-save). Auto-save usually partially-mitigates this. I don’t have a Synology system, but AFAIK, they don’t run anything like that.

    Like, I’d think that the NAS could probably survive a power loss just fine, even with an unclean shutdown.

    If you have an attached desktop machine, maybe case #2 would apply, but I’d think that hooking the desktop up to the UPS and having it do a clean shutdown would address the issue – I mean, the NAS can’t force apps on computers using the NAS to dump state out to the NAS, so hooking the NAS up that way won’t solve case #2 for any attached computers.

    If all you want is more time before the NAS goes down uncleanly, you can just leave the USB and RS-232 connection out of the picture and let the UPS run until the battery is exhausted and then have the NAS go down uncleanly. Hell, that’d be preferable to an automated shutdown, as you’d get a bit more runtime before the thing goes down.


  • and to remote in.

    This is the approach I use with laptops domestically, and I think that there’s something to be said for it. Like, the laptop itself doesn’t store important information. A remote server does. The laptop is just a thin client. If the laptop gets lost or stolen – which I’ve had happen – I revoke the credentials. No important information is lost, and no important information is exposed.

    Whole-disk laptop encryption has improved things too from an exposure standpoint (albeit not a loss standpoint), though I don’t use it myself (don’t want to spend any battery life on it). I assume that smartphones have some form of reasonably-secure storage hardware, but I don’t know if it involves encryption.

    What I found irritating – and this is years back now – was an employer who didn’t care if I took a laptop in or out or what information I stored on it (as long as it was a work system), but who refused to provide remote access to the network, so I couldn’t just keep the important information on the work network. I mean, I get if they want to have some sort of isolated DMZ and require an externally-accessible server to live there, not provide VPN access in to the general network, but not having the ability to have remote network access to work systems at all is just incredibly obnoxious.

    I think that some of it is that Windows is not phenomenal to use remotely. Yeah, there are solutions, but they aren’t great if you’re on a high-latency, low-reliability, or low-bandwidth link. I try to use console Linux for as much of my stuff as possible. That whole ecosystem was designed around thin-client, remote use.


  • Your age 30 is fine. Age is always an excuse, but mostly not true.

    It’s fine for single-player shooters, which are less demanding, but speaking as someone who has packed on some decades, your reaction time definitely becomes a noticeable factor over the years for competitive multiplayer games. I definitely can’t play competitive twitch shooters nearly as well as when I was 18, which is about when your reaction time is at its best.

    That being said, there are shooters where twitch time is less-critical or roles or play-styles that focus less on it.

    And I don’t see how someone couldn’t learn to play with a dual-stick or trackpad (or trackball, for that matter), which is what I think OP is talking about. I haven’t had any problems picking up new input methods…that just takes time. Took time to learn when I was 18, too.


  • I mean, twin stick gamepad or to lesser extent touchpad just isn’t going to be as good as a mouse for an FPS. A good mouse player will beat a good touchpad or gamepad player.

    And the problem with the Deck is that it has a PC game library, and a lot of those are designed with a mouse in mind. Console FPSes usually adjust the game difficulty so that playing with twin sticks are practical. Enemies give you more time to slowly turn around without inflicting enormous amounts of damage. Auto-aim assist is common. Ranges are shorter. Stuff like that.

    If this is a single-player game – which it sounds like you’re playing – you can reduce the difficulty to compensate for the input mechanism.

    There’s an input mechanism that some people developed for twin-stick gyro controllers called Flick Stick, which someone else mentioned; Steam Input supports this. The mouse is still going to win, but it’s an improvement over traditional pure-stick input.

    There’s also some input mechanism which I think was different from the “Flick Stick” approach – though maybe I’m wrong and misremembering, didn’t have an interest in exploring it – that IIRC someone put together using Steam Input. The way it worked, as I recall, was that one could tap the thumbstick in a direction and it’d immediately do a 90 degree turn. The idea was to provide for a rapid turn while keeping sensitivity low enough to still permit for accurate aiming. But I’m not able to find the thing with Kagi in a few searches, and it’s not impossible that I’m misremembering…this was only a single video that I’m thinking of.

    I don’t think that there’s any trick to learning this, just playing games and picking it up over time. I mean, I was atrocious at using a keyboard+mouse when I first started doing it, and ditto with twin-stick FPSes.

    You could also attach a keyboard and mouse, though I think that that kind of eliminates the point of the Deck, at least as long as one also has a PC to play on – it might make sense for someone who just uses a Deck and a phone.

    is there an easy FPS game where I don’t have to move or shoot too fast

    Play games that are designed for consoles or which have a gamepad mode, rather than a keyboard+mouse PC game. They’ll be tuned for controller limitations. Like, can you play Halo comfortably with the Deck? That was designed for a gamepad originally, and it’s available on Steam (though I’d note that it requires a Microsoft account, which you may-or-may-not be willing to do).

    https://old.reddit.com/r/truegaming/comments/8f7oyr/the_core_reasons_thumbsticks_are_inaccurate/

    This also talks about some limitations of thumbstick aiming (if you’re using thumbsticks and not trackpads). It might be possible to tweak some of these, like sensitivity or dead zone, but I’d assume that for a given game, the developers have already chosen pretty reasonable defaults.


  • For those who haven’t played the series, VATS is an alternate aiming mode where one can pause (or in later games in the 3d series, greatly slow) the game, select a certain number of targets depending upon available action points, and then have all those shots taken in rapid succession, with the game aiming.

    I’d say that VATS is kind of a “path” than a purely alternate input method in those games; you need to make a VATS-oriented build, though it’s true that it makes it possible to play the game with minimal FPS elements. Like, in Fallout: New Vegas, VATS provides major benefits close-up. While VATS is active, there’s enormous damage reduction applied to your character, IIRC 90%, so for short periods of time, they have enormous damage output and little risk. They can also turn rapidly and target multiple enemies, probably better than a player manually-playing could. At close ranges, VATS is just superior.

    But VATS suffers severe accuracy penalties at range. Whether-or-not a target is moving doesn’t affect VATS accuracy, but range does a lot, whereas with manual aiming, whether-or-not a target is moving makes a big difference and range doesn’t matter much. As a result, VATS isn’t great for sniping, which is also an aspect of the game. You can do it (especially, oddly-enough, with pistols, in Fallout 4, where the Concentrated Fire perk lets later shots in a flurry of pistol shots at range be very accurate.

    In Fallout 76, VATS provides such dramatic damage benefits that I’d say that it’s impractical to play a non-VATS build – VATS is required to get damage up to a reasonable level later in the game.



  • The planning board’s decision was based on health concerns due to the possible negative environmental impact of telecommunication on the residents, especially the children studying at the school who could potentially be exposed to electromagnetic radiation. The town felt the residents would be ‘unsafe’ due to radio frequencies and rejected the company’s notion of building the tower on the land.

    I mean, I think that the planning board is idiotic, but I don’t see why T-Mobile cares enough to fight it. If they don’t build it, okay. It looks like the school in question is right in the middle of town. Then Wanaque is going to have crummy cell coverage. Let them have bad cell coverage and build a tower somewhere else. It’s not like this is the world’s only place that could use better cell coverage. The main people who benefit from the coverage are Wanaque residents. Sure, okay, there’s some secondary benefit to travelers, but if we get to the point that all the dead zones that travelers pass through out there are covered, then cell providers can go worry about places that are determined not to have have cell coverage.

    If I were cell companies, I’d just get together with the rest of the industry and start publishing a coverage score for cities for cell coverage. Put it online in some accessible database format, so that when places like city-data.com put up data on a city, they also show that the city has poor cell coverage and that would-be residents are aware of the fact.