Adventures in walkie-talkie apps
Aug. 24th, 2020 12:03 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
brin_bellwayApps that think the distinguishing feature of a walkie-talkie, as opposed to a cell phone or VoIP line--the thing that makes you say "what we need right now is a walkie-talkie"--is "push to talk" rather than "infrastructureless": no, go away.
(honestly, I'm not actually firmly attached to being able to "talk" *at all*. it's a nice bonus, might come in handy occasionally, but text-only would be fine.)
Briar: Aimed at dissidents getting their Internet shut down by the government, so the level of security is overkill for my needs and sometimes comes at the cost of features more useful to me (like having multiple devices under the same account). Doesn't seem to have multihop even in group chats, *severely* limiting the range (can't even communicate with someone two floors down in the same house). However, can make use of Internet if it's available to both parties. Has a Linux client in beta, which is nice: if I have to choose one OS for offline texting to work on, I'll pick Android for its portability, but being able to type on my laptop keyboard would certainly be good.
(Also, drains the battery pretty hard, at least in Bluetooth mode. I'm lucky to get 24 hours of standby time out of it, when usually I can get nearly a week of standby. Admittedly this probably says more about how many of my smartphone's capabilities I normally keep turned off than it does about Briar, and someone who was going to have Bluetooth on anyway might not find the difference significant.)
Serval: Great idea, well aligned with my needs and desires, but it's still in alpha after nearly a decade of development. Apparently it's not completely abandoned, but progress is very, *very* slow. I have subscribed to their blog and will be following them with interest.
Firechat: People talk wistfully about it, but nobody seems to know where to find a reputable APK. It's not on Google Play anymore, and it was never on F-Droid.
Bridgefy: ...okay, I guess *technically* an offline-messaging app that *requires Internet during the installation process* is not 100% useless, but it's *far* less useful than it could be, and I am deeply uncomfortable with the fragility implications of "during a disaster, the mesh is irreplaceable and inextensible". As I've said previously, I generally consider it a dealbreaker in *any* software to require Internet during the installation process, unless the whole purpose of the software is to interact with the Internet. I might grudgingly use it if there were zero other options, but I'll pick almost anything over a cloud-dependent version.
Also their data-collection policies felt kind of shady to me, certainly by comparison to the utopian non-profits making most of the other stuff on this list.
IPFS: Despite the talk about "diversely resilient networks that enable persistent availability — with or without Internet backbone connectivity", from what I can tell IPFS in its current state is *not* distributed, merely decentralised. That is to say, you can talk to your peers rather than going through a server, but you must talk to said peers over the Internet. Distributedness is #4 on the IPFS team's long-term to-do list: for scale, note that in 2019 they attempted one item off the to-do list and didn't manage to complete it by the end of the year. It was a big project and I'm not blaming them, but I do figure the completion of item #4 is going to be a while.
I like how they think, though: their planned Internet fails gracefully not just in the face of loss of connectivity, but in the face of linkrot as well. I have installed a node on my laptop, which I expect to be very occasionally useful at first and gradually more useful over time. (They say file-sharing works well right now: I might make use of that.) I have also subscribed to their newsletter.
(There are two IPFS apps for Android, one of which crashes immediately on startup and one of which requires Android 8+. That's the first app I've encountered to require an Android version later than my current 7. If *I* can't afford an Android version later than 7, it's going to be a while until people living in shacks on little Pacific islands--Serval's initial target audience--can do it.)
Berty: In alpha--actually, maybe not quite even that--but development seems to be more active than on Serval. Very similar to Briar, but without the ability to use Internet if available. Unclear whether it's going to have multihop, the only way it could be better than Briar. (well, okay, not the *only* way, I guess there's *something* to be said for iOS and Windows compatibility) I have subscribed to their newsletter.
goTenna: Why do I need a whole other piece of hardware--that costs as much as a smartphone itself! each!--when there are so many transmitters already packed into an ordinary smartphone? None of those--nor the *combination* of them--were good enough? *Really*?
(And excuses about "longer range" are going to have to be pretty damn good excuses, given that for the kind of money they're charging for a two-pack I could buy...*checks*...six cheap smartphones to use as mesh nodes, *and* I'd be able to recruit Android-using neighbours into the mesh just by offering them an APK to sideload.
Actually, no, no excuses about "longer range" are good enough. Dedicated long-range mesh hardware should *complement* smartphone-based mesh nodes, not replace them.)
Also, buggy as hell apparently.
---
Iron star for Most Usable As-Is: Briar. Plan to keep it Internet-connected as much as possible, but unlike other Internet-based texting apps it fails *somewhat* gracefully in the Internet's absence.
Mithril star for Closest Match Between Their Aspirations and Mine: IPFS, with an honourable mention to Serval.
---
(edit: update on Bridgefy)
(honestly, I'm not actually firmly attached to being able to "talk" *at all*. it's a nice bonus, might come in handy occasionally, but text-only would be fine.)
Briar: Aimed at dissidents getting their Internet shut down by the government, so the level of security is overkill for my needs and sometimes comes at the cost of features more useful to me (like having multiple devices under the same account). Doesn't seem to have multihop even in group chats, *severely* limiting the range (can't even communicate with someone two floors down in the same house). However, can make use of Internet if it's available to both parties. Has a Linux client in beta, which is nice: if I have to choose one OS for offline texting to work on, I'll pick Android for its portability, but being able to type on my laptop keyboard would certainly be good.
(Also, drains the battery pretty hard, at least in Bluetooth mode. I'm lucky to get 24 hours of standby time out of it, when usually I can get nearly a week of standby. Admittedly this probably says more about how many of my smartphone's capabilities I normally keep turned off than it does about Briar, and someone who was going to have Bluetooth on anyway might not find the difference significant.)
Serval: Great idea, well aligned with my needs and desires, but it's still in alpha after nearly a decade of development. Apparently it's not completely abandoned, but progress is very, *very* slow. I have subscribed to their blog and will be following them with interest.
Firechat: People talk wistfully about it, but nobody seems to know where to find a reputable APK. It's not on Google Play anymore, and it was never on F-Droid.
Bridgefy: ...okay, I guess *technically* an offline-messaging app that *requires Internet during the installation process* is not 100% useless, but it's *far* less useful than it could be, and I am deeply uncomfortable with the fragility implications of "during a disaster, the mesh is irreplaceable and inextensible". As I've said previously, I generally consider it a dealbreaker in *any* software to require Internet during the installation process, unless the whole purpose of the software is to interact with the Internet. I might grudgingly use it if there were zero other options, but I'll pick almost anything over a cloud-dependent version.
Also their data-collection policies felt kind of shady to me, certainly by comparison to the utopian non-profits making most of the other stuff on this list.
IPFS: Despite the talk about "diversely resilient networks that enable persistent availability — with or without Internet backbone connectivity", from what I can tell IPFS in its current state is *not* distributed, merely decentralised. That is to say, you can talk to your peers rather than going through a server, but you must talk to said peers over the Internet. Distributedness is #4 on the IPFS team's long-term to-do list: for scale, note that in 2019 they attempted one item off the to-do list and didn't manage to complete it by the end of the year. It was a big project and I'm not blaming them, but I do figure the completion of item #4 is going to be a while.
I like how they think, though: their planned Internet fails gracefully not just in the face of loss of connectivity, but in the face of linkrot as well. I have installed a node on my laptop, which I expect to be very occasionally useful at first and gradually more useful over time. (They say file-sharing works well right now: I might make use of that.) I have also subscribed to their newsletter.
(There are two IPFS apps for Android, one of which crashes immediately on startup and one of which requires Android 8+. That's the first app I've encountered to require an Android version later than my current 7. If *I* can't afford an Android version later than 7, it's going to be a while until people living in shacks on little Pacific islands--Serval's initial target audience--can do it.)
Berty: In alpha--actually, maybe not quite even that--but development seems to be more active than on Serval. Very similar to Briar, but without the ability to use Internet if available. Unclear whether it's going to have multihop, the only way it could be better than Briar. (well, okay, not the *only* way, I guess there's *something* to be said for iOS and Windows compatibility) I have subscribed to their newsletter.
goTenna: Why do I need a whole other piece of hardware--that costs as much as a smartphone itself! each!--when there are so many transmitters already packed into an ordinary smartphone? None of those--nor the *combination* of them--were good enough? *Really*?
(And excuses about "longer range" are going to have to be pretty damn good excuses, given that for the kind of money they're charging for a two-pack I could buy...*checks*...six cheap smartphones to use as mesh nodes, *and* I'd be able to recruit Android-using neighbours into the mesh just by offering them an APK to sideload.
Actually, no, no excuses about "longer range" are good enough. Dedicated long-range mesh hardware should *complement* smartphone-based mesh nodes, not replace them.)
Also, buggy as hell apparently.
---
Iron star for Most Usable As-Is: Briar. Plan to keep it Internet-connected as much as possible, but unlike other Internet-based texting apps it fails *somewhat* gracefully in the Internet's absence.
Mithril star for Closest Match Between Their Aspirations and Mine: IPFS, with an honourable mention to Serval.
---
(edit: update on Bridgefy)
