Canvas size changes on disconnect

[opicron]

Sorry for the many issue notifications. I updated to the latest version and notice that the canvas resizes on disconnect and we lose half of the screen output. Is this correct? It seems to double the canvas font width somewhere.

Visit https://bbs.theforze.eu/ and wait for 10 seconds for timeout.

Edit: update, when scrollback is selected before timeout/disconnect the screen isnt rescaled. Nor when exiting scrollback mode. I believe it has something to do with the resolution/screensize resetting to 80x25 chars.

The following elements get their height and width reset to a smaller size then initially set:

fTelnetClientContainer
fTelnetFocusWarning
fTelnetStatusBar

This happens with no custom css, and with my custom css.

Connected:

Disconnected:

[rickparrish]

On your site when it first connects it's requesting an 8x14 font, but then after disconnecting I see a request for an 8x16 font, so that's likely what is causing the issue. The question is what is making it request a new font upon disconnecting...

I don't see this happening with the default settings on the embed wizard. In that case it uses an 8x16 font to start with, and then no extra request is made at the time of disconnect:

EDIT: Scatch that, I do see the behaviour you're describing on the embed wizard if I set it to 132x37 before connecting. So yeah looks like it is resetting to 80x25 upon disconnection. I'll look to see what's causing that.

EDIT 2: During the disconnection process the ANSI sequence ESC[8;25;80t is received, which resets the window to 80x25, so that explains why fTelnet is doing what it's doing. So either Mystic or fTelnetProxy is sending that, and at quick glance I don't see anywhere that fTelnetProxy sends it, so I think it's probably Mystic.

(An undocumented fTelnet feature is that you can add #ftelnetdbug=1 to the url to view all the socket i/o, so for example https://bbs.theforze.eu/#ftelnetdebug=1 would let you see all the ansi sequences that are being sent and received. Non-printable characters are printed as ~code, eg an escape character is ascii character 27 so gets printed as ~27)

[opicron]

Oh my, I remember reading you implemented a request to support that Ansi code. I didnt put 1+1 together.

So sorry to mention this as a issue while its caused by external factor.

Thanks for the additonal information, it will help any future debugging.

Edit: donation made for your continued efforts!