The paste "speed" is too fast, so characters are not being pasted as shown on the paste window. It's affecting changing passwords with passwd before installing openssh #1249
Comments
|
PS. Just tested a couple of times and it seems it's mostly affecting the first character, so the first character is always lowercase... If I use Apple it will do apple, if I use Elevator it will do elevator, if I use Supercalifragilisticexpialidocious it will do supercalifragilisticexpialidocious... |
|
It seems this is the same problem: #858 |
|
I didn't test on any mobile device, only on Linux servers connected to the PiKVM directly (no multiport KVM in-between). Don't know what's causing the issue but definitely need addressing urgently if you're able, pretty please. This can create massive headache when administering remote locations and the password isn't accepted because of errors. But don't stress Maxim, you have a life to live too. I think you're amazing for taking such good care of this project :-) Thanks so much! |
|
I've come back to this problem again and I can't find any way to reproduce it. Is it only with linux hosts? |
I'm sysadmin for only one windows server, the rest are all debian linux, and it happened on debian. The root password I was setting during setup was massive as you can see below. Not sure if this is a buffer issue, is there any error correction when sending keypresses over the virtual usb keyboard to the host? I also have a problem where if I press Shift key to capitalize a work, sometimes the virtual keyboard from the UI will show that the Shift button remained pressed down even though I am physically not pressing it anymore on the physical keyboard. Seems like PiKVM is not polling the real keyboard properly if it's thinking a key is depressed when it's not. |
|
Did you try other browser? |
|
For the shift thing both chrome and firefox. For the password thing just chrome. Also something to add, on the shift button thing it normally happens when I am typing an acronym like AT&T |
Due to business requirements I'm having to change some systems passwords to a very large string...
So to test this, I used Bitwarden to generate a password with 64 words (using the word function, instead of character). This means I get a password around 500 characters long (remember this is a test).
I paste the password into the Text > paste box in the PiKVM, then I make sure the layout is en-gb and hit paste... I can see that the password that gets pasted is not the same as was originally provided.
Why is this a big issue? Because on the terminal you can't see what you're typing... so you paste, change the password using passwd command, but next time it won't logon because the paste from the PiKVM won't be the same, there's always a small variation.
Here's an example I just generated, original text:
And here is what the PiKVM pastes into the application:
Can see right from the first character it's already wrong, with lowercase "sacrament" instead of "Sacrament". Obviously passwords will be case sensitive...
It would be amazing if this could be looking into because it can have terrible ramifications - as in losing access to remote system because the password seemed to change correctly but since it was invisible it was impossible to confirm, so only way to confirm is logging off the user, but when logging back on the password isn't accepted... for something of that length it's impossible to even guess, the whole point of a difficult password ;-) but of course the user needs to know it haha.
The text was updated successfully, but these errors were encountered: