3 ms·

> you need to set it running on a remote server So someone else's computer then? If this is running on a server, who's running the server where the file's bein

> you need to set it running on a remote server So someone else's computer then? If this is running on a server, who's running the server where the file's being loaded to? Should I only install this on pwned machines, for you know, safety?

keepamovin 3y ago
i don’t understand, what do you mean? if you can form your criticism into more coherent points and less generalized angst and snark it will be more helpful. i trust you have something to say, please do say it well.
- NoZebra120vClip 3y ago
  Your #1 feature for this is that it opens and parses files without having to download them to a computer that you're using, which I assume is a safety measure first and foremost? Or is it a matter of space or time to transfer or something? At any rate, now I need to install your service on some other server, presumably one I control, and now your feature evaporates in smoke, because this app on this server is now downloading the files and parsing them anyway, which I could've just done in the first place without a middleman. So I really don't see the point in your "download-free" architecture here if we're depending on a server in the middle.
  - keepamovin 3y ago
    Oh, OK. Sorry. Thanks for your reply, and I assumed you were familiar with the concept of defense in depth and isolation for security. I apologize because sometimes working so deeply in this I incorrectly assume it's just obvious for people. Somehow bizarrely I thought that if someone was bothering to comment on this little niche guy I posted then they are probably interested in that sort of thing, haha! I admit it can be a paradoxical thing to get one's mind around at first, and also it's not a panacea. I'll try to explain but please bear in mind I'm not the best to actually explain this stuff in the correct jargon, you may need to go online and search for these terms to understand more. I'll try to explain this from the ground up. Sorry if I'm waffling! :) The Core Idea: Security by Isolation If you've heard of Secure Email Gateways (SEGs) that isolate potentially harmful email attachments, the concept here is similar. We're focusing on isolation as a primary security feature—keeping potentially harmful files at arm's length from your system. Ask yourself, would you feel safer opening a sketchy document in Google Docs than downloading it directly onto your machine? This works in a similar vein but adds more layers of security. How It's Done: More Secure than You Think Like Advanced Threat Protection (ATP) services, our server acts as an isolation chamber, safely converting documents into high-quality images. This disarms malicious code that might be embedded in the files. Parsing to convert to images is a far less common code-path for exploits than explotis targeting viewers and editors, but even if an exploit chain did poke through the parse path, it still only manages to compromise the limited privilege user account on the remote machine. For additional protection, you can even run it inside docker. How is this more secure than you think? Because you're compounding the difficulties for exploit creators. Instead of getting RCE from a PDF and stealing your ssh keys on your local machine, they need RCE from a PDF converter, then privilege escalation on the remote machine, then maybe break out of the docker container, then they still need to compromise your local machine, so they need to compromise the web server's content and deliver the exploit to you. It's about reducing attack surface, and increasing exploit chain length. It doesn't make it impossible, it just makes it more difficult, and therefore it's more secure. Also, because it's self-hosted, you have control over the environment, making it even safer than third-party cloud services. Finally, because it's open-source you can even modify parts of it that are not to your liking. Dislike the simple API_KEY based auth? No problem, replace it with WebAuthn or SSO and you're good to go! :) Your Server, Your Rules One of the best parts about our solution being self-hosted is that you have complete control over the server. It's a bit like having your own personal security guard rather than relying on a communal one. If you want a more tailored, commercial experience, we offer licenses that provide additional features and support. Get in touch at sales@dosyago.com to learn more. Trade-offs and Weaknesses While this system adds a layer of security, it's not foolproof—no system is. You need to trust the server you're running this on and the network security needs to be robust. Yet, these are trade-offs for gaining an isolation layer that serves as a protective buffer between you and potential threats. Why This Matters: Defense in Depth and Zero Trust In cybersecurity, Defense in Depth and Zero Trust aren't just buzzwords. They're frameworks that have proven effective. Our product fits neatly into these paradigms by offering multiple layers of security and subscribing to a 'never trust, always verify' model. The Bottom Line Consider this as an extra roll-cage when you're navigating the wild open plains of the internet. It's another option in your security toolkit—a precautionary measure that sits between you and the unknown. Curious about leveling up your security for not just documents but all web content? We've got you covered! BrowserBox already has Chai integrated and amplifies the security umbrella to cover your entire online footprint. Check it out: https://github.com/dosyago/BrowserBoxPro https://github.com/dosyago/BrowserBoxPro Does this address your concerns? I may have misunderstood your criticism. You may have understood all of this but still had some criticism--if so, then I've very curious to hear it! Fire away! :)

keepamovin 3y ago

i don’t understand, what do you mean? if you can form your criticism into more coherent points and less generalized angst and snark it will be more helpful. i trust you have something to say, please do say it well.

NoZebra120vClip 3y ago
Your #1 feature for this is that it opens and parses files without having to download them to a computer that you're using, which I assume is a safety measure first and foremost? Or is it a matter of space or time to transfer or something? At any rate, now I need to install your service on some other server, presumably one I control, and now your feature evaporates in smoke, because this app on this server is now downloading the files and parsing them anyway, which I could've just done in the first place without a middleman. So I really don't see the point in your "download-free" architecture here if we're depending on a server in the middle.
- keepamovin 3y ago
  Oh, OK. Sorry. Thanks for your reply, and I assumed you were familiar with the concept of defense in depth and isolation for security. I apologize because sometimes working so deeply in this I incorrectly assume it's just obvious for people. Somehow bizarrely I thought that if someone was bothering to comment on this little niche guy I posted then they are probably interested in that sort of thing, haha! I admit it can be a paradoxical thing to get one's mind around at first, and also it's not a panacea. I'll try to explain but please bear in mind I'm not the best to actually explain this stuff in the correct jargon, you may need to go online and search for these terms to understand more. I'll try to explain this from the ground up. Sorry if I'm waffling! :) The Core Idea: Security by Isolation If you've heard of Secure Email Gateways (SEGs) that isolate potentially harmful email attachments, the concept here is similar. We're focusing on isolation as a primary security feature—keeping potentially harmful files at arm's length from your system. Ask yourself, would you feel safer opening a sketchy document in Google Docs than downloading it directly onto your machine? This works in a similar vein but adds more layers of security. How It's Done: More Secure than You Think Like Advanced Threat Protection (ATP) services, our server acts as an isolation chamber, safely converting documents into high-quality images. This disarms malicious code that might be embedded in the files. Parsing to convert to images is a far less common code-path for exploits than explotis targeting viewers and editors, but even if an exploit chain did poke through the parse path, it still only manages to compromise the limited privilege user account on the remote machine. For additional protection, you can even run it inside docker. How is this more secure than you think? Because you're compounding the difficulties for exploit creators. Instead of getting RCE from a PDF and stealing your ssh keys on your local machine, they need RCE from a PDF converter, then privilege escalation on the remote machine, then maybe break out of the docker container, then they still need to compromise your local machine, so they need to compromise the web server's content and deliver the exploit to you. It's about reducing attack surface, and increasing exploit chain length. It doesn't make it impossible, it just makes it more difficult, and therefore it's more secure. Also, because it's self-hosted, you have control over the environment, making it even safer than third-party cloud services. Finally, because it's open-source you can even modify parts of it that are not to your liking. Dislike the simple API_KEY based auth? No problem, replace it with WebAuthn or SSO and you're good to go! :) Your Server, Your Rules One of the best parts about our solution being self-hosted is that you have complete control over the server. It's a bit like having your own personal security guard rather than relying on a communal one. If you want a more tailored, commercial experience, we offer licenses that provide additional features and support. Get in touch at sales@dosyago.com to learn more. Trade-offs and Weaknesses While this system adds a layer of security, it's not foolproof—no system is. You need to trust the server you're running this on and the network security needs to be robust. Yet, these are trade-offs for gaining an isolation layer that serves as a protective buffer between you and potential threats. Why This Matters: Defense in Depth and Zero Trust In cybersecurity, Defense in Depth and Zero Trust aren't just buzzwords. They're frameworks that have proven effective. Our product fits neatly into these paradigms by offering multiple layers of security and subscribing to a 'never trust, always verify' model. The Bottom Line Consider this as an extra roll-cage when you're navigating the wild open plains of the internet. It's another option in your security toolkit—a precautionary measure that sits between you and the unknown. Curious about leveling up your security for not just documents but all web content? We've got you covered! BrowserBox already has Chai integrated and amplifies the security umbrella to cover your entire online footprint. Check it out: https://github.com/dosyago/BrowserBoxPro https://github.com/dosyago/BrowserBoxPro Does this address your concerns? I may have misunderstood your criticism. You may have understood all of this but still had some criticism--if so, then I've very curious to hear it! Fire away! :)