/*
* TODO:
*
- * - Copy-and-paste from the Event Log.
- *
- * - Remainder of the context menu:
- *
- * - New Session and Duplicate Session (perhaps in pterm, in fact?!)
- * + Duplicate Session will be fun, since we must work out
- * how to pass the config data through.
- * + In fact this should be easier on Unix, since fork() is
- * available so we need not even exec (this also saves us
- * the trouble of scrabbling around trying to find our own
- * binary). Possible scenario: respond to Duplicate
- * Session by forking. Parent continues as before; child
- * unceremoniously frees all extant resources (backend,
- * terminal, ldisc, frontend etc) and then _longjmps_ (I
- * kid you not) back to a point in pt_main() which causes
- * it to go back round to the point of opening a new
- * terminal window and a new backend.
- * + A tricky bit here is how to free everything without
- * also _destroying_ things - calling GTK to free up
- * existing widgets is liable to send destroy messages to
- * the X server, which won't go down too well with the
- * parent process. exec() is a much cleaner solution to
- * this bit, but requires us to invent some ghastly IPC as
- * we did in Windows PuTTY.
- * + Arrgh! Also, this won't work in pterm since we'll
- * already have dropped privileges by this point, so we
- * can't get another pty. Sigh. Looks like exec has to be
- * the way forward then :-/
- *
- * - Saved Sessions submenu (not in pterm of course)
- *
- * - Change Settings
- * + we must also implement mid-session reconfig in pterm.c.
- * + This will require some work. We have to throw the new
- * config at the log module, the ldisc, the terminal, and
- * the backend; that's the easy bit. But within pterm.c
- * itself we must also:
- * - redo the colour palette if necessary
- * * might be nice to move this over into terminal.c.
- * That way we could check which palette entries in
- * cfg have actually been _changed_ during
- * reconfiguration, and only update those ones in
- * the currently visible palette. Also it'd save
- * some of this hassle in the next port.
- * - enable/disable/move the scroll bar if necessary
- * - change the window title if necessary
- * - reinitialise the fonts
- * - resize the window if necessary (may be required
- * either by terminal size change or font size change
- * or both)
- * - redraw everything, just to be safe.
- * + In particular, among the above chaos, we must look into
- * how the choice of font affects the choice of codepage
- * since the Unix default is to derive the latter from the
- * former.
- *
- * - Copy All to Clipboard (for what that's worth)
+ * - Go through all the config options and ensure they can all be
+ * configured and reconfigured properly.
+ * + icon title appears to be a non-option on Unix.
+ * + Why the hell did I faff about disabling two of the vtmode
+ * options? The rest aren't used either in pterm! Work out
+ * whether they should be, and how they can be.
+ * + Refresh in the codepage combo is badly broken.
+ * + `Don't translate line drawing chars' ?? What is this crap?
+ * It does nothing at all, and where's the option to paste as
+ * lqqqk? What was I smoking?
+ *
+ * - Better control of the individual config box features.
+ * + SSH packet logging shouldn't be mentioned in pterm, and in
+ * fact not PuTTYtel either.
+ * + Keepalives, and the Connection panel in general, shouldn't
+ * crop up in pterm. (And perhaps also not mid-session in
+ * rlogin and raw?)
*/
/*
int cfgbox(Config *cfg)
{
- extern int do_config_box(const char *title, Config *cfg);
- return do_config_box("PuTTY Configuration", cfg);
+ return do_config_box("PuTTY Configuration", cfg, 0);
}
static int got_host = 0;
-const int use_event_log = 1;
+const int use_event_log = 1, new_session = 1, saved_sessions = 1;
int process_nonoption_arg(char *arg, Config *cfg)
{