Snowden’s Favorite Cloud Service Now Has a Group Chat App
motherboard.vice.comI always cringe when I come across this type of Snowden branding. Can't put my finger on the reason.
Perhaps it's because it reeks of limited hangout (trusted insider reveals a bunch of info, and then everyone uses his brief software mentions to convince everyone what tools to use), even if he isn't actually one?
Vice has me very cautious these days in general, as well: https://tobacco.ucsf.edu/disney-fox-hearst-time-warner-tied-...
From my perspective, it's like branding something with names like Stalin, Dahmer or Ames. I really wish more people would maturely reflect on Snowden's misunderstanding; pride; failure to avail himself of legal means to correct his misunderstanding; and ultimate treason. He has caused extraordinarily grave damage to his nation's security and — since that nation is the sole guarantor of global security — to the security of mankind. From my perspective, he should be despised at worst, pitied at best.
I'd love to hear more about his misunderstanding, please continue.
I'm starting to look through the code right now. A scroll through their open source page indicates they're using libsodium, which is a good sign that they avoided most of the foot-cannons alluded to on this page: https://gist.github.com/tqbf/be58d2d39690c3b366ad
The SpiderOak folks seem to know what they're doing. But if you don't trust them, here's a completely open secure communications tool:
https://github.com/Spark-Innovations/SC4
It's based on TweetNaCl, libsodium's smaller sibling. It can run standalone, i.e. no server required.
No group chat yet, but that's a relatively straightforward extension. I have two-party chat working in private beta. (Chat does require a server, though.)
> The SpiderOak folks seem to know what they're doing.
I'm inclined to agree with that assessment.
> But if you don't trust them
Trust, but verify.
IIRC they were discussing hiring a friend of mine (whose knowledge in software security is something I respect greatly).
I wonder why they don't use golang.org/x/crypto/nacl instead of bindings to C library (libsodium-go). All primitives in it are implemented in assembly (for amd64 arch), so speed shouldn't be a concern.
Great question!
We really wanted to use the pure go implementation, but it doesn't cover as much functionality. Detached signatures for example.
Ah, yes, there are no signatures in x/crypto/nacl. There's https://github.com/agl/ed25519, however it will be slower, as there's no assembly.
Edit: oops, looks like signatures were added 2 months ago (https://github.com/golang/crypto/blob/master/ed25519/). (I guess it's the same agl's code). There is a detached version (Sign/Verify).
Is Semaphor focused on group chat (Slack) or are there plans to support private communications (text, audio, video) like Wire? Unlike Signal, Wire allows registration with only an email address (via http://app.wire.com) and does not force you to upload your contacts to their server.
Article: http://arstechnica.com/business/2016/03/go-ahead-make-some-f...
Security: https://www.documentcloud.org/documents/2756350-Wire-Securit...
How's the UX of wire?
I'm dying for an app that eventually does three things:
1. Secure. A few of them exist
2. UX. I love Telegram, shame it doesn't fit item #1
3. Temporary. I actually really love Snapchat's ephemeral images and/or messaging. Telegram does a good job at this with auto destructing messages.. but it saves images on the system[1], and i don't trust it removing from the cloud in a timely manner. And of course, Telegram fails #1.. making temporary communication all the more troublesome.
[1]: This may be limited to certain systems, such as Android. But yea, don't save an image of an important document in a "secret" chat on Android.. it saves it to your filesystem.
> How's the UX of wire?
No UX complaints here. I don't think it has ephemeral messages, but you can delete per-contact history of end-to-end encrypted msgs. There's a neat doodle/whiteboard feature. The surprising killer feature of Wire has been the quality of audio, which may be due to their hiring of former Skype engineers, http://www.wired.com/2015/08/wire-declares-war-on-terrible-c...:
"You know that vague hissing noise that’s omnipresent during your phone calls? It’s called “comfort noise,” and it’s totally artificial. It’s placed there so that when you hear those brief moments of silence between each speaker’s vocalizations, you don’t think the call has disconnected. In Wire, there’s none of that; it’s actually a little jarring, not hearing anything at all when no one’s speaking ... a Wire group call is also set up in a sort of virtual space. Sound comes through the app in stereo—you really need a headset to experience it—and the app’s post-processing is able to delay it by a few milliseconds in one ear or the other. The effect is that even though you’re all on a call, it will always sound like Mary’s sitting on your left, Mike is right in front of you, and Stephanie is a few seats over to the right. Your brain doesn’t have to re-identify a voice every time it starts speaking."
1. End-to-end using Proteus which is inspired by Axolotl (now Signal). Whitepaper available (wire.com/privacy), independent security review underway). Right now only crypto/comms part open source but there will be more news in this.
2. It's similar. Visually distinct the UX is close enough to what people are used to from other IM apps, I would say.
3. Not. We've experimented with this internally but so far have not decided to release ephemeral aspect to public. Not enough demand. As someone else commented - you can delete content from your devices (syned across if you're logged in from more than one), but content will remain on other people's devices.
> 3. Not. We've experimented with this internally but so far have not decided to release ephemeral aspect to public. Not enough demand. As someone else commented - you can delete content from your devices (syned across if you're logged in from more than one), but content will remain on other people's devices.
You may not see demand, but the userbase for an app like Wire or Signal is itself very small and with almost no demand (compared to just having a messaging app with emojis, GIFs and stickers, for which there is a huge demand - encrypted or not). :)
I use this feature in Telegram called secret chat with the self-destruct timer set to a specific duration to exchange sensitive information like banking information or other things that I don't want to remain for long.
(By trusting Telegram's encryption for secret chats) This makes it quite simple to exchange information which otherwise would involve asking people to delete specific messages or wondering if someone really deleted it or just forgot to do it because of laziness, forgetfulness or just being busy (not attributing malice to the other party here, since the information can be captured for posterity through many other means).
I hope you would consider this feature even if you see "not enough demand", because for those who need it, it's a real blessing to have it!
Are decrypted messages/assets stored in plaintext in the local device cache, or subject to local device backups? Can the local device cache be manually flushed, made subject to OS-level data protection policy, or eliminated entirely?
-Messages stored locally use the device encryption/OS-level protection
-Not included in backups
-Can be deleted either manually (per message/per chat) or altogether by uninstalling
Your "temporary" criteria is not yet available in the 1.0.x version of Semaphor, but the very near term roadmap includes retention policy per channel/conversation, with settings like "N days" and/or "N messages". Then Semaphor uses ephemeral key rotation to match the retention policy (i.e. both the messages and ephemeral keys are destroyed as the retention expires.)
Just a friendly reminder: Snapchat is not actually ephemeral. They save everything they just stop showing it to users.
I honestly figure as much with any of these temporary-messengers. Snapchat, Telegram's Secret feature, etc. Really sucks.
Source code: https://spideroak.com/solutions/semaphor/source
Good luck to them.
After helping make a slack-style client-side encrypted productivity app (https://balboa.io) that has been on life-support for the last 2 months, I have a lot of respect for people that attempt to make this space more secure.
It's not easy.
A few lessons learned that may be useful to others:
1. Most businesses and consumers are ok with their data being available to companies like Slack and Google because they trust these companies. They feel that regardless of reality, their data is safer with Google or Slack because those companies have a lot more to lose if they fail.
2. The SME space for productivity apps is pretty much the same as the consumer space. You're going to be competing with Google. SMEs are actually really cheap and scrappy: they don’t spend money on non-bottom line affecting stuff. If you want to sell security to them, it has to be essentially free
3. Reputation is more important than (or at least AS important as) your technical chops. You're asking people to trust you. You can show that you are competent by demonstrating a mastery of the technology, but that may not be enough to show that you are also trustworthy.
SpiderOak describes this as "Collaboration and messaging for teams." It seems like it's for (for-profit) business teams and not for personal "teams" or groups or non-profits/social groups. I guess the pricing model mentioned in the article is to get businesses that use Slack or Hipchat or other system.
I don't like SpiderOak's pricing models in general because of how it seems to oversell and upsell services. For personal teams/groups, there are free services like Telegram (awesome user experience that keeps improving at a fast pace but poorer homegrown crypto with normal messages stored in plaintext on the servers) and Signal (great crypto but awful user experience, slow and buggy app and slow and unreliable message delivery).
Thanks for taking a look and for your feedback! Do you have any recommendations for us for alternative pricing?
One of the things we tried to accommodate is that teams can be paid for by the individual members instead of one entity having to foot the whole bill. This was one of the common complaints we saw about Slack, where large communities enjoy using it but had no way to pay for better service.
You can use Semaphor for free just like Slack, with limited historical content retention.
Also, for what it's worth, I use Signal daily for personal messaging, and my own experience with it has been great. We think of Semaphor and the team/business context as as having pretty different requirements (and therefore somewhat different underlying crypto structures) than individual messaging. The biggest differences are about message retention and what happens when you want to invite a new member to an established conversation.
Sorry, when I wrote my comment, I could not find any information from your website about the pricing of Semaphor and what the tiers provide (it just says "plans starting from $6"). The article didn't really explain the plans and what they provide either.
It seems like the personal plan pricing is high, considering that "personal" use as such may not be high volume or high storage in general for such an application. But that's just a thought without any information, and as such, not useful. It may be easier to judge it after knowing what it provides in every tier.
My comment on pricing was based on the pricing model you have for the backup service, which is highly nonlinear with a very high jump between tiers to push people to buy into a larger one. The plans at 30GB and 1TB remind me of Dreamhost and oversold plans (not that it's wrong business wise, but it doesn't seem fair from the customer's point of view).
Downloadable here: https://spideroak.com/opendownload
Bummer, they say Android, but the download isn't available.
Sorry! Ironically (since we're using Google's Golang) Android has been the hardest platform for us to finish. We've got Mac, iOS, Linux, and Windows now, and are running Android internally but it's not release ready yet. Should have it within a few weeks. It is likely to require a newish Android version though. Thanks for taking a look!