(NOT Subscribed To List)

On some platforms (such as Freerunner, with 2.8" screen) it is a common occurrence for the screen orientation to change portrait<->landscape, and also for size to be too small in one dimension for comfortable visual layout. (IE, vertical_split is awkward when portrait-mode console is <50 characters wide) Also, on most platforms it's quite reasonable to presume that the window dimensions are subject to change, and that those changes can quite easily involve severe limitation in one dimension.

But if you have selected 'vertical split' for example, and then change the X*Y proportions of the window sufficiently, things can get too cramped. So I propose 'auto_split' option - if selected by user, then split will be automatically horizontal or vertical, depending on whether height*2>COLS. (a somewhat arbitrary proportion I've found to work well in my tests so far - essentially, with most fonts, if the window is square or wider it chooses Vertical, if narrower it chooses Horizontal) It's a simple patch (against 4.6.1), entailing a few additions and a few alterations to layout.c, and adding 'extern int auto_split;' to layout.h and '{ "auto_split", &auto_split },' to layout[] in setup.c. It may not be correct, but so far it 'works for me'. (it gets quite frustrating after a while using mc on a handheld, when switching from portrait to landscape means you need to pull down Options->layout, change horizontal|vertical split, & save - you don't necessarily change orientation frequently, but mc could be invoked while in either mode, and it would be very helpful if it were able to adapt) The only potential bug I've seen is that numerically-specified panel sizes get bungled if 'auto split' is selected in the Layout dialog while 'equal split' is not set. (when changing from vertical to horizontal in the layout dialog with 'equal split' not set, it forces reversion to 50%/50%, however for an 'auto split' scenario it really isn't sensible to talk about absolute split dimensions - but when it comes down to it, it doesn't really make much sense to have column/row counts for manual unequal split to begin with, when window dimensions and fonts are subject to unpredictable changes)

j

Joel Newkirk
[EMAIL PROTECTED]

78a79,81
> /* Set if the panel split horiz vs vert choice is handled automatically based on terminal dimensions */
> int auto_split=0;
>
123a127
> static int _auto_split;
170a175
>     { N_("&Auto split"),       &auto_split,   0 },
183a189,196
>     if (_auto_split)
>     {
>     _horizontal_split=0;
>     if (height*2<COLS)
>     {
>         _horizontal_split=1;
>     }
>     }
217c230
<     dlg_move (layout_dlg, 6, 6);
---
>     dlg_move (layout_dlg, 7, 6);
219c232
<     dlg_move (layout_dlg, 6, 18);
---
>     dlg_move (layout_dlg, 7, 18);
295a309
>     _auto_split = check_options [9].widget->state & C_BOOL;
370c384
<     for (i = 0; i <= 8; i++) {
---
>     for (i = 0; i <= 9; i++) {
425,426c439,440
<     add_widget (layout_dlg, groupbox_new (4, 2, first_width, 6, title1));
<     add_widget (layout_dlg, groupbox_new (4, 8, first_width, 4, title2));
---
>     add_widget (layout_dlg, groupbox_new (4, 2, first_width, 7, title1));
>     add_widget (layout_dlg, groupbox_new (4, 9, first_width, 4, title2));
467a482,483
>
>
469c485
<     button_new (6, 15, B_2RIGHT, NARROW_BUTTON, "&>", b2right_cback);
---
>     button_new (7, 15, B_2RIGHT, NARROW_BUTTON, "&>", b2right_cback);
472c488
<     button_new (6, 9, B_2LEFT, NARROW_BUTTON, "&<", b2left_cback);
---
>     button_new (7, 9, B_2LEFT, NARROW_BUTTON, "&<", b2left_cback);
474c490
<     check_options[6].widget = check_new (5, 6, XTRACT (6));
---
>     check_options[6].widget = check_new (6, 6, XTRACT (6));
481a498,499
>     check_options[9].widget = check_new (5, 6, XTRACT (9));
>     add_widget (layout_dlg, check_options[9].widget);
528a547,554
>     if (auto_split)
>     {
>     horizontal_split=0;
>     if (height*2>COLS)
>     {
>         horizontal_split=1;
>     }
>     }

_______________________________________________
Mc-devel mailing list
http://mail.gnome.org/mailman/listinfo/mc-devel

Reply via email to