We have tested on Tor Browser (which uses Firefox as its base), the Chromium derivatives, WebKit browsers, various mobile browsers, etc.
BTW, failure to post via the QR box, with the confirmed-correct captcha, just happened again attempting this post the first time. Trying a second time...
This does not provide sufficient detail to let us know exactly what you see going wrong and when. This description of it "failing" the first time but not the second suggests that you are encountering the behavior we described above:
>Note that if the board has per-post captcha active and you don't have a block bypass when you post, you will be challenged first for a block bypass and then your post's CAPTCHA will be rejected (since you already solved it to get your block bypass) and you'll have to solve the new post CAPTCHA. This is not a bug, but rather something LynxChan does normally because... well, you know the rest. We're note sure if the behavior has changed in newer versions of LynxChan.
Until we receive a sufficiently detailed description of what is experienced, when, and how, there's nothing we can do.
>you only have to look over the overboard to see it was a common complaint across several boards, so it likely wasn't anything board-specific.
Yes. We did not think it was board-specific, but rather was a problem with the server's copy of ffmpeg, as we describe above.
>Care to let us know what version of the software you're using here on this site BTW?
You can always see the LynxChan version number in use by looking at the bottom of the site's front page.
>While you're on the line, are you ever going to consider implementing a warrant canary of some sort here?
No, we are not. Warrant canaries are useful only in certain legal systems and operational circumstances, and only serve to provide a false sense of security otherwise. If you are worried about such things, you should assume that all of Anon.cafe's traffic is being monitored and traced in real time. Anon.cafe and all other webring sites are slow enough that it is trivial to connect any given post to any given source IP by traffic monitoring and timing alone. Assume that our server is compromised and we're run by whatever spook house scares you the most.