so i notice there hasn't been any activity on the git repo since last year just wondering what the status is? i would be willing and im sure others as well to donate money for your time. ive tried Steve's windows version and it does work in wine but im not sure how to get it to authenticate sites on the linux side.
progress on linux version?
- Thread starter markboston35
- Start date
- Watchers 13
- Status
Even The Big G has said that, if he were starting this today, he would probably have written the reference/POC client as a browser extension rather than a Windoze app. Of course, he would have written it in WebAssembly. ;-) So, it seems like, until and unless SQRL steps out of the HTTP realm, or if you want to point your web cam at another computer to log into something, nobody needs a native desktop client.
Just FYI, the remote authentication mode using QR-Codes is much less secure than Same Device mode & should probably not be highlighted for general use.
yeah this is also what i ended up doing after generating my sqrl id on my windows box. everything is moving to the cloud pretty much who would of saw that in 2013? not really a developer but i dabble how hard would it be to make a sqrl pwa?
that seems to be the direction of the world maybe Steve or someone should focus on that.
that seems to be the direction of the world maybe Steve or someone should focus on that.
Same device vs. same process space, though. Out-of-band ought to be more secure in a separate memory space. I'd appreciate the million-dollar ask message being very isolated from the Chrome bug surface. Maybe in a different cgroup, grsecurity hardened, or more robust measures.
Kubuntu user here, Windows free in my house for 5+ years now. Currently running the Windows SQRL client in Wine on Kubuntu, the Chrome and Firefox extensions, as well as the Android client. All that being said, whilst there is not an absolute necessity, it would be nice to have a native Linux client.
I think for adoption to be successful, the Linux user community has to be a priority; a huge proportion of the people actually responsible for security in institutions and corporations are Linux users, and they're unlikely to take seriously a security initiative that doesn't have full native support for Linux. These are often the people making recommendations and setting policies for organizations that take security seriously. If Linux is neglected, I don't think SQRL will take hold.
Do you have a github or similar source repo link for that?A Native SQRL Client for Linux
SQRL Clients - Google Drive
drive.google.com
It's OK if you want it to remain closed source but that would tend to count me out of trying it.
Do you have a github or similar source repo link for that?
[/QUTE]
No.
Okay, but you're missing out on one the greatest programs ever written.
This is a good point event though iim one of the people you talk about and don't use Linux a majority of security people do.
The point i'd like to raise here is that rather than build a native client for Linux we should probably be trying to build it into the OS itself as @Steve documented and has commented on before this would result in the best security for SQRL possible and allow for more protection of the authentication than a native client could provide. How we go about getting the Linux foundations to look at this I don't know but that would be my suggestion for this OS as its clearly the easiest to convince to do this. Although we could try to UNIX and get both Linux and Mac in one go.
I agree that a Linux CLI client should be a big priority. Being able to create and re-key identities from the Linux CLI is essential in my opinion and would open up a lot of possibilities for this promising technology.
It would be very helpful for any Linux compatible code to be posted on GitHub to enable testing and further development. I would be interested in contributing if possible. It seems a lot of the SQRL code on GitHub is incomplete and out of date. It seems the browser extension is the best reference implementation I've found.
It would be very helpful for any Linux compatible code to be posted on GitHub to enable testing and further development. I would be interested in contributing if possible. It seems a lot of the SQRL code on GitHub is incomplete and out of date. It seems the browser extension is the best reference implementation I've found.
Not least of all it could be bundled into something like Damn Small or Puppy Linux & put on a CD or thumb drive so that you can pretty safely generate a new identity & encrypt your Identity Unlock Key in a safe environment.
I was thinking about this in a different thread where I wrote about booting from a USB drive to create the ID in safety. If you had a CLI client, which could easily be written in Python or something like that, you could have a special boot .iso that's a command line Linux with a Python script that runs the identity creation. You could be assured that there wasn't any malware spying on you, and it also takes care of a problem we discussed on the newsgroup years ago about the possibility of a high processor load at the time of creation compromising the security of the Rescue Code.
The USB Drive could be made from a raspberrypi Zero for prototyping. I am going to try and see if I can help with the libpam module, as I have worked with PAM modules in the past. However I have hit the same compile problems as described previously (on the android app forum), which suggests that the compile environment is different, assuming that the code did compile originally.
Here's the source for a CLI program written in GO to create a basic SQRL Identity File (rename it to sqrlid.go). It can be compiled for Windows, Linux, and may other OSes (look for GOOS and GOARCH here: https://golang.org/doc/install/source#environment)
Tried it on Termux. Rescue code generated and prompt for password okay. However, the identity file could not be found in the tmp directory.
Identity file:
/data/data/com.termux/files/usr/tmp/go-build027305743/b001/exe/sqrlid.sqrl
What am I missing here?
I don't have an Android device to try this on. Instead I downloaded and installed the BlueStacks emulator and then Termux. I ran the 32-bit Linux version of sqrlid under Termux and it created a sqrlid.sqrl file without any problem.
To be sure it didn't create a file for you, run sqrlid again. It will warn you if the identity file already exists.
Last edited:
- Status