This problem is reproduced easily with PuTTY, the popular
SSH client for Windows. PuTTY is started in a remote
desktop session, of course. The "text select" mouse cursor
over a PuTTY window is black and thus hardly visible on the
window's default black background if using rdesktop. The
cursor is white, as it should be, if using MSTSC. Moreover,
the mouse cursor becomes black or white if reconnecting to
the same session with rdesktop or MSTSC, respectively.
In general, rdesktop always seems to use black color for
the "text select" mouse pointer while MSTSC, as well as the
regular Windows desktop, does something like logical NOT
over the underlying pixels to get the pointer image on the
screen. The latter approach ensures easily that the pointer
is visible over any solid filling or pattern.
Versions used to reproduce the bug:
MS Windows Server 2003 SP1
rdesktop 1.4.1
PuTTY 0.58
Logged In: YES
user_id=1588262
I am running into this problem as well. I use a black background in Visual Basic,
and the text cursor (the I-beam) is black as well, making it hard to select text or
position the mouse on a line.
Some sort of logical NOT makes sense here.
Logged In: YES
user_id=1588262
Forgot to mention the versions I'm using:
MS Windows 2000 Professional
rdesktop 1.4.1 (compiled on Mac OS X 10.4.7)
Visual Basic 6
Logged In: YES
user_id=645231
Originator: NO
I'd like to bump this bug. I've recently started using Putty over rdesktop sessions, and the black on black I-beam cursor is killing my eyes. Windows XP here, although I'm sure it's a fairly generic issue.
This isn't really an rdesktop bug but a limitation of X. (That's the short answer - but there may be a workaround)
For performance, rdesktop is mapping windows cursors to X cursors. RDP servers transmit the actual bitmaps of the cursors (i.e. it doesn't say "switch to I-beam cursor, it sends the I-beam cursor image over the network). Rdesktop creates an X cursor from that data.
The problem is that windows is capable of using "xor" pixels which change based on the pixels underneath them. X is not capable of this. So you can't simply map a windows cursor onto an X cursor.
There may be a workaround which would be to hide the X cursor and just simulate a windows cursor right on the main surface. I don't know what the performance impact would be (might need to be an option on the cmd line)
A short term workaround might be to find a windows I-beam cursor that isn't all black (or all white either, probably), or use putty with a non-black background or something.
Warren,
I really appreciate that this issue is remembered and understood. Indeed, it is deeper and more of an architectural one than I would anticipate back when I submitted it.
Thank you!
Yar
I found a workaround here (submitted by David Simmons) : http://osdir.com/ml/network.rdesktop.user/2005-03/msg00012.html
The workaround is to change the "Text Select" pointer on the Windows machine by going to Control Panel->Mouse->Pointers. I changed it to the beam_r.cur image, which is a beefier version of the I-beam bitmap which does render nicely in rdesktop.