Trouble with virtual terminals when supporting a SCO app
Hi all, I've got an issue where I am supporting an older closed-source application written for SCO, and I am attempting to run it under linux.
It runs amazingly fast under linux vs. SCO using the ICBS packages. We're talking in the order of 300-400% faster.
We connect to the system via telnet or ssh. No serial terminals are involved.
This application 'identifies' users within the program via their tty # (ie tty123, tty54), therefore, a single user can login to this program correctly, but once a second user telnets to the machine, the software believes the user is already connected as it only identifies the user by the controlling tty that telnetd is running on, and is not aware of virtual ttys (ie, pts/0 pts/1).
Is there a way to force users to be mapped to a tty (like cu did in Solaris) that our application can understand? Printing is not an issue, as we use IP-based printing vs print back to tty. a hacked telnetd perhaps? A different terminal controller than mingetty? Any ideas are appreciated.
If this is not the correct forum for this question, or if you are aware of a more approriate venue for this, please let me know.