When we have the HTML5 client ready, we should phase out the Java applet SSHVnc.
r29231 - removed the Java client from the docs r29232 - removed .jar and .tar.gz files and cleaned up makefiles r29241 - cleanup of tlclient.cgi, removed: .no.sshvnc, clienttype and javaclient image r29242 - removed /vsmagent/default_geometry parameter
r29246 - removed network-NAT-javaclient.png from makefile r29247 - removed autotest for default_geometry
r29249 - updated tl-setup
r29271 - renamed browser_client directories to web_integration
r29273 - updated references to web_integration sections in the tag
http://www.cendio.com/resources/docs/platforms/selinux.xhtml has a note a workaround to let tlclient.cgi connect to vsmserver. Is this still required now that the Java client is gone?
> http://www.cendio.com/resources/docs/platforms/selinux.xhtml has a note a > workaround to let tlclient.cgi connect to vsmserver. Is this still required now > that the Java client is gone? r29279
The same note was present in the tag, it was removed in r29280.
(In reply to comment #1) > r29231 - removed the Java client from the docs > r29232 - removed .jar and .tar.gz files and cleaned up makefiles > r29241 - cleanup of tlclient.cgi, removed: .no.sshvnc, clienttype and > javaclient image I think it would be good if the clienttype parameter is still recognized, but with an error/info message printed if you try to use clienttype=java. Otherwise, customers using such URLs will be silently redirected to the native client. Reopening for discussion.
(In reply to comment #9) > I think it would be good if the clienttype parameter is still recognized, but > with an error/info message printed if you try to use clienttype=java. > Otherwise, customers using such URLs will be silently redirected to the native > client. Reopening for discussion. Fixed in r29295. A small bug with the javascript code was fixed in r29296. Redirect to https was broken, fixed in r29297. After internal discussions we decided to keep default_geometry, the removal the parameter was reverted in r29294.
(In reply to comment #4) > r29271 - renamed browser_client directories to web_integration This will likely break customers apache configuration unless they run tl-setup. We should consider writing a note regarding this in the release notes.
(In reply to comment #11) > (In reply to comment #4) > > r29271 - renamed browser_client directories to web_integration > > This will likely break customers apache configuration unless they run tl-setup. > We should consider writing a note regarding this in the release notes. No problem, tl-setup is required after installation/upgrade. Services will not start until it is executed.
Minor problem: The TAG still talks about "the common CGI script". I believe we should remove "common" since tlclient.cgi is exclusively used for the native client now. Also, this section was removed entirely: - <para> - When using the Java Browser Client, usernames and passwords are limited - to ASCII characters (ISO 646). Some PAM modules and authentication - mechanisms are case sensitive, while others are not. Usernames in the - ThinLinc client are case sensitive by default, however this behaviour - can be changed. See - <varname>LOWERCASE_LOGIN_NAME</varname> - in <xref linkend="clientconf" /> for details. - </para> Only the first sentence should be removed; the rest still applies.
(In reply to comment #13) > Minor problem: The TAG still talks about "the common CGI script". I believe we > should remove "common" since tlclient.cgi is exclusively used for the native > client now. > > Also, this section was removed entirely: > > - <para> > - When using the Java Browser Client, usernames and passwords are > limited > - to ASCII characters (ISO 646). Some PAM modules and authentication > - mechanisms are case sensitive, while others are not. Usernames in the > - ThinLinc client are case sensitive by default, however this behaviour > - can be changed. See > - <varname>LOWERCASE_LOGIN_NAME</varname> > - in <xref linkend="clientconf" /> for details. > - </para> > > Only the first sentence should be removed; the rest still applies. Fixed in r29403 and r29404
Ok now.
