The key is protocol, as far as I understand it. XMPP OMEMO encryption isn’t as robust as Signals’.
That’s why Molly works a little better on GOS than the Signal app.
–
Welcome to the forum btw!
Conversations is an XMPP client, and XMPP is by default an end-to-end plaintext protocol. The context in most of these messenger discussions is secure instant messaging, where people cannot unintentionally send unencrypted messages (exposed to the server operator).
Signal is reputed for having a battery-hungry WebSocket-based notification implementation (used instead of Firebase Cloud Messaging if Google Mobile Services are not present at the time of install). They don’t see it as a resource priority because non-GMS Android users are a minority.
Some in the community get around it by using Molly (a hardened Signal service client) which implements more efficient notifications via WebSockets, as well as supporting UnifiedPush and FCM via an open source reimplementation of the Google FCM support library.
2 Likes
If the threat model isn’t “high” (you have to decide this for yourself…do not just take advice from strangers on the internet…do your research) then Conversations does offer an option to encrypt chats and may be an OK solution for some people. It is harder to use than signal and less secure though so I don’t see a reason to use it as a messenger unless you want to use it with a VOIP number.
BUT Conversations actually has a built-in unified push distributor, which means you can solve the Signal battery drain issue by switching to Molly-FOSS and enabling Unified Push through the Molly-FOSS notifications options and also in the Conversations unified push options.
I also turned off “Allow background usage” in Molly-FOSS. I still get my notifications, just sometimes a little delayed. My phone lasts almost twice as long with this setup. I recommend using Molly-FOSS anyway, and if you aren’t going to use Conversations, there are other dedicated unified push distributors like NextPush.
I tried Orbot on GOS a few days ago. Uninstalled after an hour of not being able to get a connection. Pretty sure it’s an Orbot issues as Tor browser seldom had a problem connecting over several days of testing.
It’s way more nuanced than this. I’m self-hosting mollysocket, ntfy server and using ntfy as my UnifiedPush client app.
The UnifiedPush notifications will be slower than FCM. There’s a chance that your mollysocket will bug out like I ran into a race condition ( WebSocket to Signal frequently closed with code 4409 ("Connected elsewhere") → no pushes until manual connection ping · Issue #102 · mollyim/mollysocket · GitHub ) that sometimes caused my notifications to be delayed by ~1 hour until I figured it out (had to unlink everything from signal first). If you’re not self-hosting your own mollysocket + ntfy you might get rate limited or throttled by the public services.
Here’s the personal Nitter feed of the SimpleX founder and main dev: https://nitter.net/epoberezkin
Here is a pasted copy of a retweet from 19 December 2025:
Our Peer-Reviewed Reanalysis of the Henry Ford Vaxxed Vs. Unvaxxed Study Finds Vaccinated Children FAR Sicker Across ALL 22 Chronic Diseases
1. Neurodevelopmental disorders:
+1254%
2. Autoimmune disease:
3. Cancer:
4. Autism:
5. Motor disability:
6. Speech disorder:
7. Mental health disorders:
8. Asthma:
9. Developmental delay:
10. Atopic disease:
11. Seizure disorder:
12. Food allergy:
13. Neurological disorder:
14. Any chronic condition:
Observed only in vaccinated children (zero cases in unvaccinated):
15. ADHD
16. Diabetes
17. Brain dysfunction
18. Behavioral disability
19. Learning disability
20. Intellectual disability
21. Tics
22. Other psychological disability
Bottom line:
22 out of 22 chronic disease categories were worse in vaccinated children. ZERO exceptions.
The CDC’s 81-dose hyper-vaccination schedule is driving the modern epidemics of chronic disease and autism.
Don’t know how relevant this is to the security of SimpleX itself, I work in a medical clinic and let me tell you doctors and surgeons have all sorts of wacky beliefs, general physicians give shitty damaging advice on nutrition, nutritionists give shitty damaging advice on dental issues and so on, shouldn’t we go to hospitals then?
Just because someone has a weird take on something (that’s not their main field btw) doesn’t mean they are generally untrustworthy or unreliable…
2 Likes
It’s time to grow up.
Deniability, Forward Secrecy are things which can be discussed on the forum. In the Court it does not matter. 
Have you heard about Manning and Assange?
They used xmpp (jabber)+OTR, chat logs were used in the Court. This is a real life.
According to a Jabber chat log, Manning asked a person called Nathaniel Frank – alleged to be Assange – whether he was any good at cracking a password hash. Manning sent Assange a hexadecimal string that she had found on her computer network.
No later than January 2010, Manning repeatedly used an online chat service,
Jabber.ccc.de, to chat with ASSANGE, who used multiple monikers attributable to him. 2 - Criminal No. l:18-cr-111 (CMH)
2 Likes
What make this relevant to me, personally, is that an encrypted messenger is a piece of software I put an exceptional amount of trust in and that feels a lot harder to do when the lead dev dabbles in all kinds of conspiracy theories.
Your mileage may vary, you do you etc. … as long as there are other messengers available I prefer those where the dev doesn’t publicly partake in such behaviour. 
1 Like
Wrt deniability, you’re not wrong. E.g. if your ex-partner sends threatening messages, courts aren’t interested in deniability, screenshots are usually enough. Note that I didn’t drag out deniability for comparison, SimpleX did.
As for forward secrecy, it still matters. You said
Yeah obviously, if you keep logs of your chats and your device is seized, it’s not going to help. But if you’d recall Lavabit, here is Ladar Levison, the CEO of the company admitting he failed to enable forward secret TLS
As for the initial lack of perfect forward secrecy; that was my mistake. In my rush to leave town I only confirmed that the server supported perfect forward secrecy. I didn’t realize the server wasn’t configured so clients would use those ciphers default.
The agencies came after Snowden’s email by requesting access to the non-forward secret RSA keys. They already had the data collected from the backbone. They just couldn’t decrypt it without the key. Hacking user’s endpoint and/or server to exfiltrate the long term RSA keys is much less noisy and it allows passive, retrospective decryption of messages collected from the backbone.
k
What we discuss here is cryptographic protocols in their appropriate context. You could argue message encryption doesn’t matter when in real life one can just look at the endpoint. You could argue The XKCD §538, you’re just not adding any new insight or being helpful.
Cryptographic protocols can and should be compared for their security properties. When there is a better way to do something (like MACs over digital signatures), that’s called a best practice, and the written rule is you don’t deviate from best practices unless you know exactly why you’re deviating (non-repudiation in say, a business context). Looking at the protocol design also shows if the author is familiar with the field, and these smell tests are the easiest way to filter out snake oil security products.
I get your point is the real life context matters and I couldn’t agree more. For forward secrecy to be most useful, you shouldn’t store logs. Cwtch doesn’t store message logs by default. Neither does TFC. TFC encrypts all logs the user opts in, with Argon2id+XChaCha20, and stores them in the exfiltration secure environment.
2 Likes
You do you, but people literally put their lives on the hands of surgeons who believe that you can treat tooth infections with just antibiotics, that’s not just factually wrong, its dangerous advice, yet they do great things in their area of expertise.
1 Like
I’d like to ask which area of expertise is that? Looking at SimpleX CEOs LinkedIn, it looks like the guy has a degree from London School of Economics. Sure theres a cert for ethical hacking which isn’t bad. It’s important to have adversarial mindset when designing security systems. But still, IMO this stuff is best left to computer scientists.
I strongly believe that all of this unhinged twitter posting from a personal account is completely irrelevant. The only thing that matters is the product and that the creator doesn’t start portraying views that are contrary to its purpose. There’s a wise saying to “never meet your heroes” and that applies here too.
Who cares if a crazy mofo goes on an anti-vax tirade on twitter? Is his software still good? That’s all that matters.
The creator of Obtanium is also an unhinged mofo that started posting politicized opinions on his project’s github README. Do I care? No. Obtanium is good enough for me and his personal opinions are irrelevant, so I’ll keep using it.
2 Likes
I will disagree, the creator does reflect the project, Exhibit A would be GrapheneOS, It is influencing other users too and it is not limited to them, and therefore encourages that behavior, this is not something you should have ever and this is one of those exactly why I dislike being political or at least not involve my projects on it
3 Likes
Before commenting anything else I should say that I haven’t personally researched him or simplex that deeply, however I do know something.
We should probably avoid these 3 views:
- Your epistemic competence on one area always determines your epistemic competence in everything.
- The only way a person can be an expert (or proficient in something) is by having an academic degree on that thing.
These are widely acknowledged to be wrong (or at least overly simplifications of deeper issues) especially on computer sciences where practical experience is so overwhelmingly important, and that’s what I can comment on, we can’t use these arguments against anyone’s credibility because they are simply crappy arguments.
At the end of the day what matters is that the software was designed properly and that it does what it claims, if those conditions why wouldn’t you use it? Other reasons would probably be ideological.
2 Likes
That sounds ideological (but still a valid response), my concern with your approach would be that if you consider these attitudes from Simplex CEO (not by any means the sole developer) bad enough to not use it, then what is your stance on Firefox? Android? Apple devices? Any billionaire owned enterprise? Voting?
There are a lot of crazy wackos out there who are the owners of something we use, if a product/service is worth it depends on your personal circumstances i think.
2 Likes
It’s a big world with big players driving their politics through their works. Three guesses why Elon “overpaid” for Twitter. He now has reach to feed his shit to the screen of 400 million people. What you support is what you enable to shape the world. The thing is, you might have aligned interests wrt privacy, but when you feed the post-factual system and its advocators, don’t be surprised when you’re the next meal. It won’t happen through SimpleX. Probably. But what could go wrong supporting wrong people.
Obviously not. But usually people who are passionate about something attend university to learn from the best in their field.
Yeah I’m sure there’s plenty of great armchair heart surgeons and structural engineers. Cryptography isn’t a field where you can fake it till you make it either.
It’s a good heuristic usually, and it’s a good heuristic here considering the SimpleX CEO’s stubborn infantile attitude towards communicating the project’s threat model that deviates from the marketing, and treating the truth as a fine print hidden deep in the documentation, and then expanding to the world of web3.
Citation needed. Every run-of-the-mill charlatan of this post-factual age peddling their pseudoscience says the same thing.
The makers and doers attitude can create new stuff but you’re not going to beat 80 years of work by the academic community on your own. We stand on the shoulders of giants. You can build your own knowledge base from scratch but it’s a long, long, climb.
Well that’s just it, it doesn’t.
1 Like
While I want to nitpick some things you said I think we agree in more aspects than it seems 
so please don’t get too hung up over my words and try to see what i mean in case I don’t explain myself correctly.
Should say this again, I’m not defending SimpleX, I’m criticising some arguments here, and it seems you’ve given some independent reasons to distrust Simplex on your post, I will be checking them out.
First nitpick: “people who are passionate… attend university to learn”.
Assuming you say that because of the simplex CEO anti vaxx takes, please understand you can’t possibly attend university for every topic that’s interesting to you, and in some cases (like philosophy or cooking) you just don’t need “formal” education to learn what you need. You can’t expect everyone, even otherwise smart people to deeply research everything they say (even if that would be optimal).
Now, if you mean that he should go to college and study computer sciences then maybe you’re right, but he clearly isn’t the only one working in Simplex and I don’t see why he couldn’t be a valuable asset to the team with the qualifications he has (in theory).
Second Nitpick: “armchair heart surgeons” What matters in education is that you have the theoretical knowledge and that you get enough practice in a similar environment to real life. As a medical professional I tell you in theory one could get that without a university (regulations wouldn’t let you practice though), In fact IIRC, the first dentistry universities in south america basically let you study on your own and gave you some tests to evaluate your capabilities.
In any case, it’s unclear why any of this even matters since the topic in question is software development, something that clearly can be learned in theory and practice without a university, a lot of the giants we stand on have material outside of universities now more than ever, courses, YouTube videos, books, etc.
To give you a citation, Google seems to agree that someone without an academic degree can be proficient since they hire some of them :
I don’t think I need to give a citation for my point 1 since you seem to agree .
Finally, I need to say I actually agree with a lot of your other criticisms of simplex, and ultimately if you have independent evidence SimpleX or it’s CEO are not up to the task that’s what matters, the only thing I can speak on is that you can’t strictly rely on his vaccination takes to dismiss Simplex (which based on your posts you don’t seem to do anyways), but hey where I live there is no anti Vax cultural movement so maybe I just don’t see the cultural implications.
2 Likes
While Signal has deniability and Session does not, cryptographically, neither have plausible deniability in practice.
Session made this argument wrt deniability when, unveiling the first version of their protocol in 2020, justifying why they sign messages with the sender’s long-term identity key. They cited the problem that courts and media may treat messages as authentic evidence even when there is no cryptographic authenticity.
Practicality
Deniability in the Signal protocol is a cryptographic property which prevents a cryptographic proof from being provided that a person signed a message with a particular key. In practice, cryptographic deniability is often disregarded when it comes to court cases or media reporting. For example, cryptographic deniability was used unsuccessfully as a defense in a court case involving the communications of Chelsea Manning and Adrian Lamo. Instead, courts often rely on screenshots of conversations from a dishonest chat partner or seized devices to establish the real world identities of the chat participants.
For plausible deniability to exist, the official clients of Signal, Session, SimpleX etc could add something like the ability to edit anyone’s messages locally (without leaving an edit log of course) and not just one’s own messages globally. This is not my idea; Session proposed this feature in the same article but AFAIK never implemented it.
On the loss of cryptographic deniability
…
Instead of designing a cryptographic protection, Session will add the ability to edit other users’ messages locally, thus providing a way to completely forge conversations. Since signatures are deleted after messages are received, there will be no way to prove whether a screenshot of a conversation is real or edited, diminishing the value of screenshots as evidence.
The anti‑vax retweet was just one example. He posts daily and most of his tweets are hot takes. I picked that particular tweet because I didn’t want to turn this thread into a political debate and I figured it wouldn’t attract any defenders here.
If you’re concerned about local cultural effects, you’ll probably come across something unsettling if you scroll through his feed. I recommend checking the link so you can see the full context for yourself.
Looking at the company itself, SimpleX’s reliance on VC funding and its founder’s claim that venture capital is the only way to bring privacy to the masses raises red flags. Sure, privacy needs funding, but VC money brings with it perverse incentives and pressure. Rather than own up to that and explain the safeguards put in place to mitigate those issues, he goes off on tangents explaining how it’s actually a good thing.
He also argued that Proton moving to nonprofit governance “robbed the founder and other people of millions of dollars (if not billions) of shareholder value.” That profit‑centric focus is worrisome, especially given that his own company still lacks a stable revenue stream. What compromises might he make to protect that “shareholder value”, particularly when venture capital is breathing down his neck?
2 Likes